the

catalan

by Alex Raetsky & Maxim Chetverik EVERYMAN CHESS Gloucester Publishers pic

www.everymanchess.com

First published in 2004 by Gloucester Publishers plc (formerly Everyman Publishers plc), Northburgh House, 10 Northburgh Street, London EC1V OAT Copyright © 2004 Alexander Raetsky and Maxim Chetverik The right of Alexander Raetsky and Maxim Chetverik to be identified as the authors of this work has been asserted in accordance with the Copyrights, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without prior permission of the publisher.

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN 1 85744 346 2 Distributed in North America by The Globe Pequot Press, P.O Box 480, 246 Goose Lane, Guilford, CT 06437-0480.

All other sales enquiries should be directed to Everyman Chess, Northburgh House, 10 Northburgh Street, London EC1V OAT tel: 020 7253 7887 fax: 020 7490 3708 email: [email protected] website:

www.everymanchess.com

Everyman is the ed trade mark of Random House Inc. and i� used in this work under license from Random House Inc.

EVERYMAN CHESS SERIES (formerly Cadogan Chess) Chief advisor: Garry Kasparov Commissioning editor: Byron Jacobs Typeset and edited by First Rank Publishing, Brighton. Cover design by Horatio Monteverde. Production by Navigator Guides. Printed and bound in United States by Versa Press Inc.

CONTENTS

I

Bibliography

4

Introduction

5

1

d4 d5 2 c4 e6 3 tiJf3 tiJf6 4 93

1 Closed Catalan: 4 ... .i.e7 5 .i.g2 0-0

7

2 4 ... c6 5 �g2 �d6

38

3 4 ... dxc4 5 'iVa4+

44

4 4 ... dxc4 5 .¥l.g2 .¥l.e7 6 0-0 0-0

58

5 4 ... dxc4 5 �g2 a6

96

6 4 ... dxc4 5 .¥l.g2 tiJc6

124

7 4 ... dxc4 5 .i.g2 c5

151

8 4 ... dxc4 5 .i.g2 b5

165

9 4 ... dxc4 5 .i.g2 c6

172

1 0 4 ... dxc4 5 .i.g2 .i.b4+

176

1 1 4 ... dxc4 5 .i.g2 tiJbd7

181

1 2 4. . .dxc4 5 .i.g2 .¥l.d7

185

Index of Complete Games

190

BIBLIOGRA PHY

Books Katalonskoe Nachalo, J.Neistadt (Moscow, 1986) Beating the Flank Openings, v'Kotronias (Batsford, London 1996) Winning with the Catalan, A.Dunnington (Batsford, London 1997) The Queen's Gambit and Catalan for Black, L.Janjgava (Gambit, London 2000) Die Katalanische Eroeffnung, A.Raetsky, M. Chetverik (Kania, 2001) Periodicals: New In Chess Yearbook 1-67 Chess Informant 1-86 Shakhmatny bulletin, USSR/Russia. Shakhma!J v SSSR Database: Chess Assistant: Huge base 2003

Acknowledgements We would like to thank Esben Lund and Jacob Aagaard for moral and technical dur­ ing our work with this book, as well as our families for their forgiveness for our not so secret Catalan lover. This book is dedicated to the Catalan bishop. Thank you for everything!

INTRODUCTION

1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 93 Between the authors there exists almost a lifetime of dedication to the Catalan Open­ ing. This positionally oriented opening fits our classical style perfectly and has brought us much pleasure in our chess careers. There­ fore it is quite natural for us to write a book about this opening, and we have already done so in German. We should point out here that, rather than a translation, this is aimed at being a complementary work, with slightly different points of gravity. In some positions, where style is the main factor, we have chosen slightly different recommenda­ tions compared to the earlier book. Addi­ tionally, this edition is more up-to-date. We have decided to use the move order 1 d4 dS 2 c4 e6 3 4J O 4Jf6 4 g3 as a Catalan standard in these pages because in our ex­ perience this is more likely to be seen in prac­ tice than 1 d4 4Jf6 2 c4 e6 3 g3 dS 4 4J O, both leading to the same position:

(see following diagram)

This is the starting position of the Cata­ lan. White could have played 4 ..tg2 instead of 4 4J O, but this will make a difference only in very few lines, mainly in Chapter 3 with S 'iVa4+ after 4...dxc4. The lines with 1 d4 4J f6 2 c4 e6 3 g3 cS and 1 d4 4J f6 2 c4 e6 3 g3 ..tb4+ will not be

I found in this book. After 3...cS White has a choice between 4 4J O, going into the Eng­ lish Opening, or 4 dS, transposing to the Modern Benoni. After 3 .....t b4+ the game will almost always transpose to the Bogo­ Indian Defence, where White would have played 3 4J O instead of 3 g3.

The layout of the chapters in this book has been set out mainly according to importance, in a descending order. But we have also taken into consideration the natural branches of the move order. Therefore we have included slightly obscure lines with 4...c6 and S.....td6 in the Closed Catalan and 4...dxc4 S 'iVa4+ as Chapters 2 and 3 respectively. You will find that some of the games in this book are those of the authors. Obviously 5

Th e Catalan

we are proud of our nicest wins but the main reason for their inclusion is that we feel we understand these games quite well. Not sur­ prisingly, our losses also feature. One of Alexander's most brilliant victo­ ries, against a strong Hungarian Grandmas­ ter, did not find its way into the first draft of the book, but our friend Jacob Aagaard in­ sisted that we include it, so here it is. It should serve as a warning of the power of the Catalan ... Raetsky-Varga Zurich 2000 1 d4 d5 2 c4 dxc4 3 ttJf3 c6 4 g3 b5 5 .tg2 .tb7 6 0-0 e6 7 a4 a6 8 b3 cxb3 9 �xb3 ttJf6 1 0 J:td 1 ttJbd7 1 1 ttJe5 �b6 1 2 ttJc3 ttJxe5 1 3 dxe5 ttJd5 13...ttJg4!? is also possible. White has com­ pensation but it is not clear how much. The following lengthy - and by no means forced - fantasy line illustrates relevant themes: 14 ttJe4 ttJxe5 15 .i.f4 ttJc4 16 l:tac1 ttJa5 17 �c3 .l:!.d8 18 .i.e3 l:txdH 19 l:txdl c5! 20 .i.xc5 �c7 21 axb5 .i.xe4 22 .i.xe4 .i.xc5 23 l:tc1 rJ;;e7 24 �xc5+ �xc5 25 l:txc5 axb5 26 l:txb5 ttJc4 and Black maintains the balance. 1 4.txd5 ! ? White parts with his good bishop t o fur­ ther the collective cause - namely the exploi­ tation of Black's tardy kingside development. 1 4 . . . exd5 14...cxd5 does not work because after 15 Jte3 �c7 16 axb5 .i.e7 17 bxa6 .i.xa6 18 �a4+ �d7 19 'iVxd7+ rJ;;xd7 20 ttJa4 White wins the exchange. 1 5 axb5 axb5 1 6 .te3 �d8? After this Black is, surprisingly, completely outplayed. The problem is the advance e2-e4. After the accurate 16...Jtc5 17 .i.xc5 'iVxc5 18 l:txa8+ Jtxa8 19 e4 d4! 20 ttJe2 d3! 21 'iVxd3 0-0 22 'iVd6 'iVc2 Black is doing fine. 1 7 .ta7 ! Wilc7 1 8 e4! b4 Black cannot avoid the deadly opening of 6

the centre. After 18...dxe4 19 ttJxe4 �xe5 20 �d3 �d5 21 'iVe3 �e6 22 �6! the end is fast approaching. 1 9 exd5! bxc3 20 d6 �d7 2 1 .tc5! l!!.xa 1 21...l:tb8 22 l:tdbl g6 23 l:ta7 .i.g7 24 l:txb7 l:txb7 25 �xb7 'iVxb7 26 l:txb7 0-0 27 f4 c2 28 .i.e3 l:td8 29 l:tb2 and Black has a hopeless task - 29 ...f6 30 e6! l:txd6 31 l:tb8+ and White wins. 22 ':xa 1 �d8 23 Wilxc3!? Simple and safe, but 23 Jtb6+ rJ;;c8 24 Jtc7! is another way to end the game. 23 . . . h5 24 l:ra7 �c8 25 'iVa5 �b8 26 .tb6! ? This does win, but a more direct route is 26 l:txb7+! �xb7 (26 ...rJ;;xb7 27 �a7+ rJ;;c8 28 �a8 mate) 27 �d8+ �c8 28 .i.a7+ rJ;;b7 29 �6+ rJ;;a8 30 d7! etc. 26 . . .txd6 . 27 exd6 �xd6 28 J:ta8 + !

Winning a piece. 28 . . .txa8 . 29 Wila7 + �c8 30 Wilxa8+ �d7 31 'iVxh8 c5 32 'iVd8 + �c6 33 Wilxd6 + �xd6 34 �f 1 �d5 35 �e2 g5 36 h3 g4 37 h4 1 -0 The Catalan is a dangerous weapon for the positional player, and often we find our­ selves enjoying a lasting advantage in the rniddlegame and endgame due to the strength of the Catalan bishop, or the weak­ ness of c6 or c5 and so on. We hope you will enjoy many such advantages, and that you find this book useful.

I

CHA PTER ONE Closed Catalan: 4 .,te7 5 .,tg2 0-0 .

.

.

1 d4 d5 2 c4 e6 3 lUf3 lUf6 4 g3 i.. e 7 5 �g2 0-0 Black decides to address development (first the kingside, then the queenside - in real classical fashion) and not undertake any untimely action in the centre. This quite logi­ cal approach was always popular against the Catalan. In fact even before the birth of the opening we can find games with this set-up, one example being Blackburne-Baird, Leipzig 1884, where the game started 1 d4 dS 2 ttJf3 ttJf6 3 g3 e6 4 i..g2 Ae7 S 0-0 0-0 6 c4 c6 7 ttJc3 - a rather modern move order. After 1 d4 dS 2 c4 e6 3 ttJf3 ttJf6 4 g3 i..e7 S i..g2 0-0 the most common move is 6 0-0. Occasionally White plays 6 �c2 to reach a position from the semi-open Catalan, in­ tending to secure an advantage by postpon­ ing castling. However, after 6 �c2 the d4pawn becomes weaker and Black can chal­ lenge it immediately with 6 ... cS. Black then has a good game after 7 0-0 cxd4 8 ttJxd4 ttJc6 9 ttJxc6 bxc6, or 8 ... eS!? (Game 1) . After 6 0-0 we get the basic closed Catalan position (from hereon referred to as the CC) . The CC is very similar to other closed openings. Now 6 ... b6 transposes to the Queen's Indian Defence, 6 ...cS to the Tar­ rasch Defence and 6 ... ttJe4 followed by . . f7f5 to the Stonewall variation of the Dutch .

Defence. Obviously we cannot examine these openings here, but we do pay attention to the possible transpositions.

The arrangement of the theory in the CC can be awkward due to the number of move orders. For instance after 1 d4 dS 2 c4 e6 3 ttJf3 ttJf6 4 g3 i..b4+ S i..d2 i..e7 6 j,g2 0-0 7 0-0 ttJbd7 8 i..f4 ttJhS 9 j,c1 ttJhf6 we have a normal position in the CC with 6 0-0 ttJbd7. However, we will use the most popu­ lar routes. Since the game Botvinnik-LRabinovich, St. Petersburg 1938 Black has used the plan with ... c7-c6 and ... b7-bS. This hits the c4pawn and forces White to define the situa­ tion in the centre. In the case of an exchange on bS and the opening of the c-ftle, the most 7

Th e C a t a l a n

White can achieve i s a slightly better position because the g2-bishop lacks bite. On the other hand, c4-c5 notably restricts Black's freedom, and after e2-e4 (and sometimes even e4-e5) White has the better prospects. The main response to 6 0-0 is ...ttJbd7. The plan with 6...c6, 7...b6 and 8.....tb7 (Game 2) has gone out of fashion because Black has serious problems after 8 ttJe5. Here Black can easily go wrong, for example 8...ttJfd7 9 cxd5 cxd5 10 ttJxd5! or 8.....tb7 9 e4 ttJbd7? 10 ttJxc6! - in both cases White should win. In the main line, with 8.....tb7 9 e4 dxc4 10 ttJxc4 i..a6 11 b3, White has the initiative in the centre, while in the case of 6...c6 7 �c2 b6 8 ttJbd2 ..tb7 9 e4 ttJa6 White has more space. The main gambit continuation in the CC is 6...ttJbd7 7 ttJc3 dxc4 8 e4 (Game 3). There are various possibilities after 7 ttJc3 c6, which we examine in Games 4-6: 8 �3 (Game 4), 8 iVd3 (Game 5) and 8 b3 without �c2 (Game 6). Game 7 is devoted to rare continuations in the 7 �c2 line. We have paid attention to Zagoryansky's plan with 7...c6 8 .l:td1 b6 9 a4!? with the idea of a4-a5. Here Black must play with some accuracy in order to avoid trouble; Black seems to equalise after 9 .....ta6 10 b3 �c8 11 as c5 12 axb6 �xb6 13 �a2 ..tb7 14 ..ta3 �fe8, as in Tratar-Atalik. The main possibilities after 7 �c2 c6 are examined in Games 8-15. 8 ..tf4 (Games 8 & 9) allows Black to engineer a Stonewall posi­ tion (8...ttJh5 and 9 ...f5). In our opinion this appears to be the best option, but note that after 8...ttJe4 9 ttJc3 (or 9 ttJbd2, 9 ttJfd2) the direct 9 ... f5 is no good in view of 10 ttJxe4 fxe4 11 ttJd2. White's main plan in the CC is the ad­ vance e2-e4, when ...d5xe4 is not a popular reaction in modern international practice, being played mostly in games among ama­ teurs. After ...c6-c5 and ...c5xd4 (or d4xc5) a prototypical pawn structure arises: 8

White has the pawn maJonty on the queenside, which provides a modest but sta­ ble advantage, especially in the endgame. In these positions it is not easy for Black to find counterplay and very often Black is forced to fight long and hard for a draw. After 7 �c2 c6 8 ttJbd2 b6 9 e4 ..ta6 or 9 ... ..tb7 White can grab space with e4-e5, and in such a situation White has kingside attack­ ing chances (Game 12). Occasionally after e2-e4 the sides manoeuvre their pieces with­ out altering the pawn structure, as in Game 13, where play continues 7 �c2 c6 8 ttJbd2 b6 9 b3 ..tb7 10 e4 .l:tc8 11 i..b2 'iVc7 12 �ad1 .l:tfd8. Again, thanks to his space ad­ vantage, White's chances are slightly better. Generally after e2-e4 Black prepares for and then advances ...c6-c5. The modern pawn structure in the CC arises after the exchange e4xd5 ... e6xd5:

C lo s e d C a t a la n : 4

With this confrontational pawn structure the current placement of the pieces and other tactical nuances are more important than strategic considerations. Black develops his light-squared bishop either to b7 (Game 13) or - more often - on a6 (Games 14 & 15). In the line starting with 7 'it'c2 c6 8 ttJbd2 b6 9 b3 iLa6 White has a choice between 13 'ili'fS, 13 iLb2 and 13 ttJf1. The move order 10 iLb2 l:!.c8 11 e4 cS 12 exdS exdS 13 'ir'fS is analysed in Game 15.

Game 1

.

.

.

iL e 7 5 iL g 2 0 - 0

ttJeS 1 5 "iV f4 ttJh5 1 6 'it'e3 iLc5 1 7 b4! iLxb4 18 0-0 Black did not have enough compensa­ tion for the exchange in C.Horvath-yu Sha­ oteng, Budapest 2000) 13 'if'h4 (13 'it'f4 e5! 14 'iih 4 ttJc6! 15 iLxa8 ttJd4! gives Black a very strong attack) 13. .. ttJd5 14 'it'e4 iLc6! 15 iLxc6 ttJxc6 with strong play for the pawn. Even worse here is 12 ttJc3?! l:!.c8 13 ttJb3 'ir'hS 14 'iif4 iLc6 15 0-0 h6 16 e4 ttJbd7 17 f3 as! 18 g4 'ili'eS 19 "iVg3 "iVxg3 20 hxg3 ttJe5 and Black's prospects were preferable in Korchnoi-Short, Lucerne 1989. 7 . . . cxd4 8 lL'lxd4

H . Olafsson-E . Gelier

Reyifyavik 1986 1 d4 d5 2 c4 e6 3 lL'lf3 lL'lf6 4 g3 iLe7 5 iLg2 0-0 6 �c2 The downside to this is Black's reply. 6 . . . c5! After 6 ... dxc4 7 'it'xc4 a6 there is a major difference between this and the Semi-Open variation in that White is not forced to castle yet, and can achieve a slight advantage with 8 iLf4!, e.g. 8 ... ttJdS 9 ttJc3 bS 10 "iVd3 iLb7 11 ttJxdS iLxdS 12 0-0 ttJd7 13 l:tfdl cS 14 e4 iLb7 15 dS exdS 16 exdS iLf6 17 'it'c2 .l:!.e8 18 h4 h6 19 ttJd2 'i!i'b6 20 ttJe4, when White was a shade better in Dorfman-Bansch, Lvov 1984. 7 0-0 White needs to be careful even here. He could easily get himself into trouble: K.Grigorian-Vyzhmanavin, Pinsk 1986 con­ tinued 7 cxdS?! cxd4 8 ttJxd4 ttJxdS 9 'iVb3 ttJc6 10 ttJxc6 bxc6 11 e4 ttJb4 12 0-0 cS!, when White was behind in development and was therefore not suited by the open charac­ ter of the position. After 13 ttJa3 ttJc6 14 iLe3 l:!.b8 15 "iVc3 ttJd4 16 'lithl iLa6 17 �fdl eS 18 b3 'it'c8 19 "iVaS .l:!.b4 Black had a slight edge. There is also 7 dxcS 'it'aS+ 8 ttJc3 dxc4 9 ttJd2 'it'xcs 10 ttJa4 "iVaS 11 "iVxc4 iLd7, e.g. 12 iLxb7 iLbS! (an important zwischenzug as after 12 ... iLxa4? 13 iLxa8 ttJbd7 14 iLg2

8 . lL'lc6 Interesting is 8 ... e5!? 9 ttJfS d4 10 ttJxe7+ 'iixe7 11 iLgs ttJc6 12 ttJd2 h6 13 i.xf6 'it'xf6 14 c5 iLfS 15 'it'a4 as in Zilberstein­ A.Geller, Belgorod 1989. Now it is possible for Black to continue I s .. :iVg6 16 ttJc4 i.e4 17 ttJd6 iLxg2 18 'litxg2 'iVe6 with reasonable chances. It is not clear that the d6-knight is well placed. 9 lL'lxc6 bxc6 1 0 b3 White has also tried 10 l:td l 'iib 6 11 ttJc3 iLa6 12 iLe3 'i!Vb7 13 cxd5 cxd5 14 iLd4 but without much success. After 14.J:tac8 15 'i!Vb3 iLc5 16 'ili'xb 7 iLxb 7 17 e3 iLxd4 18 .uxd4 .l:!.cs 19 ltad1 as 20 iLfl iLc6 21 h3 l:tb8 a draw was agreed in Gulko-Khalifman, Lucerne 1997. 1 0 . . a5 This looks like a luxury that offers Black nothing in of development. Serious . .

.

9

Th e C a t a l a n

consideration should b e given t o the con­ tinuation 10 ... �a6 1 1 �b2 .l::!.c 8 12 e3 iVb6 1 3 ttJd2.l::!. fd8 14.l::!. fd1 ttJd7 1 5 l:tac1 �b7 1 6 ttJ f3 c5, when 1 7 ttJe5? ttJxe5 1 8 �xe5 d4! 19 exd4 �xg2 20 'itxg2 f6 21 �f4 cxd4 22 'iVe4 e5 23 �d2 �a3 24.l::!.b 1 as was excel­ lent for Black in Gulko-Karpov, Thessaloniki 1 988. Instead Karpov gives 1 7 cxd5 exd5 1 8 'ti'fS ttJf6 1 9 ttJg5 with chances for both sides. 1 1 .i.b2 After 1 1 ttJc3 iLa6 12 .l:!.d1 'ti'c7 13 ttJa4 dxc4 14 bxc4 .l:[ab8 1 5 iLd2 e5 1 6 nab 1 .l::!.xb 1 1 7 .l:txb 1 ttJd7 Black - in control of the c5-square - was doing fine in Gulko­ Hjartarsson, Iinares 1 989. 1 1 . . . .i.a6 An interesting alternative here is 1 1 ...a4!? 12 ttJd2 a3 13 �d4 (1 3 �c3?? d4 and the bishop is lost) 1 3. .. dxc4 1 4 e3 cxb3 1 5 ttJxb3 ttJd5 1 6.l::!. fd 1 'ifc 7 1 7 e4 ttJf6 1 8 'ti'c3 �a6 1 9 ttJc5 .l::!. fd8 20 .l:tab 1 when White had compensation for the pawn in Kengis-Gild. Garcia, Manila 1 992 (whether there is more than enough is not completely clear) . 1 2 4Jd2 4Jd7 1 3 J:rfd 1 .i.f6 1 4 J:rac 1 J:rc8 1 5 4Jf3

1 5 . . :ii'e 7 White is already a little better. After 1 5 ... iLxb2 1 6 'ti'xb2 'iib 6 1 7 e4 ttJc5 1 8 iVd4! dxc4 1 9 ttJe5.l::!. fd8 20 ttJxc4.l::!.xd4 21 ttJxb6 l:.cd8 22.l::!.xd4.l::!.xd4 23 �fl i.xfl 24 'itxfl ttJxe4 25 l':txc6 the endgame was slightly 10

unpleasant for Black in Kengis-Klovans, Riga 1 987. 1 6 .i.xf6 4Jxf6 After 1 6 ...'ti'xf6 1 7 'iVd2! 'iVd8 1 8 e4! White is better according to H.Olafsson. 1 7 4Je5 .i.b 7 1 8 �d2 �a3 1 9 c5 .i.a6 White has secured the queenside and is now free to start action on the kingside. 20 g4!

20 . . . J:rc7 21 g 5 4Jd7 22 4Jxd7 J:rxd7 23 e4 d4 24 e5 .i.b5 25 .i.e4 d3 26 h4 a4 26 .. .'ilib4 looks safer, when after 27 'ii'xb4 axb4 28 f3 f6 29 gxf6 gxf6 30 exf6 .l:txf6 3 1 'it f2 .l:!.f4 3 2.l::!.g 1+ 'it f7 33.l::!.g4 e 5 3 4 'it>e3 h 5 35 l':tg5 l':txh4 3 6 .l::!.c g1 White has nothing special. 27 h5! axb3 28 axb3 'ilt'xb3 29 h6 g6?! An improvement is 29 ....l::!. fd8 30 hxg7 'itxg7 3 1 'iY'f4 d2 32 'iY'f6+ 'itg8 33 .l::!.x d2! 'ti'a3! (not 33 ... l':txd2? 34 i.xh7+! �xh7 35 'ti'xf7+ 'It>h8 36 g6 etc.) 34 .l:tb 1 'it'a4 and Black still has sufficient counterplay. 30 'ii'f4 f5 Black loses in the event of 30 ....l:[fd8 3 1 'tWf6 'it>f8 32 �xg6!, when White comes crashing through: 32 ... 'ite8 33 'iVh8+
C lo s e d C a t a la n : 4

i.xd3 and White is completely winning. He is threatening 41 f7 and after 40 ... b4 there is even 41 'iYg7+ lIxg7 42 hxg7+ �g8 43 .Jtc4 mate. 31 . . . l:td4 32 f3 iLc4? 32 ... l:td7 puts up some resistance, al­ though 33 'ifg5 'ifb2 34 �h1 ! still gives White a strong attack. 33 l:tb 1 1 -0 After 33 ...'ilVc2 there comes 34 f7! etc.

Game 2 Pigusov-Goldin

Irkutsk 1986 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 iLe7 5 iLg2 0-0 6 0-0 c6 Black has some inventive alternatives: 6 ...ltJa6 7 ltJc3 c5 8 cxd5 exd5 is a Tar­ rasch with the knight on a6 instead of c6, a feature that sees Black without any hold on d4. Rausis-Ponomariov, Enghien les Bains 1 999 went 9 dxc5 ltJxc5 1 0 i.e3 i.g4 1 1 h3 .Jte6 12 ltJd4 J:tc8 1 3 nc1 a6 14 l:tc2 b5 1 5 a3 'ii'd 7 1 6 'it>h2 lIfd8 1 7 lId2 'iib 7 1 8 'ilVb1 ltJce4 1 9 ltJxe4 ltJxe4 20 lld3 .Jtd7 21 l:tfd 1 with a slight edge for White. 6 ... ltJc6 7 'i¥c2 ltJb4 8 'i¥b3 dxc4 9 'ifxc4 b5!? is ambitious but weakening. Then 1 0 � 3 a s 1 1 ltJc3 J:tb8 1 2 .Jtf4 a4 1 3 'ifd1 .Jtb7 1 4 nc1 ltJbd5 1 5 ltJxd5 ltJxd5 was an edge for White in Kengis-Hertneck, Ger­ many 2000. Here 9 ... b6 looks solid but is not too exciting for Black: 10 ltJc3 i.a6 1 1 'tIYb3 c5 1 2 dxc5 .Jtxc5 1 3 ltJe5 l:.c8 1 4 .Jtg5 'ilVd4 1 5 i.f4 ltJbd5 1 6 e3 'iVb4 17 l:tfd1 ltJxf4 1 8 gxf4 l:tc7 1 9 ltJc6 'iYc4 20 'ifc2 .Jtxe3?! (20 ... ltJg4!? with an unclear game is better, with the following tactical justification: 2 1 b3?! ltJxe3 2 2 fxe3 'i¥xf4 and Black has good counterplay; 21 ltJd4 improves) 21 fxe3.l!Ixc6 22 l:td4 'ilVc5 23 SLxc6 'ifxc6 24 'ifg2 'ifc7 25 lIad1 .Jtb7 26 ltJb5 'iVb8 27 'iVd2 and White was on his way to winning in Rausis-Kulaots, Sweden 2000. 7 ttJc3

.

.

.

iL e 7 5 iL g 2 0 - 0

I n the event o f 7 ltJbd2 Black should transpose to the main lines with 7 ...ltJbd7. Instead Grivas-Papaioannou, Portaria 1 996 continued with the unnecessary 7 ... b5, and after 8 cxb5 cxb5 9 ltJb3 as 10 .Jtf4 ltJa6 1 1 ltJe5 .Jtb7 1 2 l:tc1 'iVb6 1 3 .Jtg5 lIfd8 1 4 e3 h6 1 5 .Jtxf6 .Jtxf6 1 6 f4 White had a slight lead. 7 'iVc2 can of course easily transpose to the main lines featured later in this chapter, but White's immediate action in the centre is of independent significance, e.g. 7 ... b6 8 ltJbd2 i.b7 9 e4 ltJa6 1 0 a3 c5 1 1 e5 ltJd7 1 2 cxd5 i.xd5 1 3 'iVd3 �c8 1 4 ltJe4 cxd4 1 5 ltJeg5 f5 1 6 exf6 ltJxf6 1 7 ltJxd4 SLc4 1 8 'ifd1 nd8 1 9 .Jtxa8 'ifxa8 2 0 'ifB 'it'xB 21 ltJgxB .Jtxfl 22 �xfl and White had the more comfortable ending in Karpov-Tal, Brussels 1 987. An interesting alternative here is 1 1 ...ltJe4 1 2 ltJxe4 dxe4 1 3 ltJd2 cxd4 1 4 ltJxe4 'ifc7 1 5 .Jtf4 ltJc5, when Damljanovic­ Popchev, Yugoslavia 1 997 saw the over the top 1 6 ltJf6+?! i.xf6?! (preferable is 1 6 ... gxf6 1 7 exf6 .Jtd6 1 8 "it'd1 �h8 with the better game for Black; it is difficult to see how White's pieces should succeed in attacking the black king) 1 7 exf6 'ifd8 1 8 .Jtxb7 ltJxb7 1 9 fxg7 l:.e8 20 lIfe1 f6, and now 21 'iVe4 ltJc5 22 'iVB would have maintained White's lead . 7 . . . b6 Black is slightly worse - in traditional Catalan fashion - after 7 ... dxc4 8 ltJe5 ltJbd7 9 ltJxc4, e.g. 9 ... ltJb6 1 0 ltJe5 .td7 1 1 e4 c5 12 i.e3 cxd4 1 3 i.xd4 i.e8 14 a4 ltJfd7 1 5 ltJxd7 ltJxd7 1 6 e5! ltJc5 1 7 b4! ltJb3 1 8 'it'xb3 'ifxd4 1 9 b 5 and White had the freer game in Raetsky-Lemanczyk, Dusseldorf 2001 . 8 ttJe5 iLb7 Black can run into trouble with 8 ... ltJfd7?, when 9 cxd5 cxd5 1 0 ltJxd5! exd5 1 1 .i.xd5 ltJxe5 1 2 dxe5 .Jth3 1 3 i.xa8 .txfl 14 'i'xfl ltJd7 1 5 .Jtd5 ltJxe5 1 6 e4 left White a clear pawn up in Chetverik-Laihonen, Nagyka­ nizsa 1 993. 1 1

Th e C a t a l a n

8 ... .i.a6 9 cxd5 (9 'it'a4!? b5 1 0 cxb5 cxb5 1 1 ttJxb5 'iie 8 12 ttJc3 'it'xa4 1 3 ttJxa4 ..txe2 14 �el .i.b5 1 5 ttJc5 l:te8 1 6 i.e3 ttJfd7 1 7 a4 ttJxe5 1 8 axb5 ttJc4 1 9 b 4 produced an edge for White in Tunik-Dautov, Moscow 1 989) 9 ... cxd5 1 0 .i.f4 ttJfd7 (after 1 0 ... .i.b7 1 1 nel ttJh5 12 .i.d2 ttJf6 1 3 ttJb5 ttJe8 1 4 'it'c2 f6 1 5 ttJd3 ttJc6 1 6 ttJf4 'it'd7 1 7 e4! favoured White in Smejkal-Bc)nsch, 1 998) 1 1 ]::te l (1 1 ttJxd5?! exd5 1 2 .i.xd5 ttJxe5 1 3 dxe5? does not work this time due to 1 3 ... .i.xe2!, when Black wins material) 1 1 ...ttJxe5 1 2 .i.xe5 and White is a shade better.

9 e4! dxc4 Black loses a pawn after 9 ... ttJbd7? 1 0 ttJxc6! .i.xc6 1 1 exd5 exd5 1 2 cxd5 .i.b7, when White has 13 d6! .i.xg2 14 dxe7 'it'xe7 1 5 'it>xg2 l:tfd8 1 6 'it'f3 l:tac8 1 7 .l:tdl 'it'b4 1 8 .l:tb 1 h 6 1 9 .i.f4 'iia5 20 d5, which was close to winning in Poluliahov-Dragomaretsky, Sochi 1 996. Black was also rather quickly in a bad way in Korchnoi-Yusupov, Rotterdam 1 988, finding himself clearly worse after 9 ... ttJa6 1 0 .i.f4 dxc4 1 1 ttJxc4 b5 1 2 ttJe3 ttJd7 13 ttJg4 b4 14 ttJa4 ttJb6 1 5 ..te3 ttJc4 1 6 .l:tel ttJxe3 1 7 ttJxe3 .l:tc8 1 8 e5! etc. 1 0 ttJxc4 i.a6 The weakening of the queenside with 1 0 ... b5 seems risky. For example 1 1 ttJe3 b4 12 e5 bxc3 1 3 exf6 .i.xf6 1 4 bxc3 ttJd7 1 5 i.. a3 .l:te8 1 6 .l:tbl ttJb6 1 7 ];tel 'YWc7 1 8 ttJg4 i.e7 1 9 .i.xe 7 .uxe 7 20 ttJe5 led to a clear 12

advantage for White in Ivanchuk-Dautov, Tashkent 1 987. 1 1 b3 b5 1 2 ttJe3 Also possible is 12 ttJe5, after which the thematic 12 ... c5 results in superior prospects for White following 1 3 d5 b4 14 ttJe2 'YWc7 1 5 .i.b2 lId8 1 6 lIe1 . Instead 12 ... b4 looks a little suspicious (and greedy), as was demon­ strated in Dizdar-limbach, Hamburg 1 993, which went 13 ttJe2 .i.xe2 14 'YWxe2 'YWxd4 1 5 i.b2 'it'b6 1 6 l:!ael .l:td8 1 7 .tIc4!? with more than sufficient compensation. 'The subse­ quent 17 ... a5 1 8 'it'c2 .l:ta6 19 ttJxf7!? 'itxf7 20 e5 ttJd5 21 'it'xh 7, l:trn 22 l:tg4 'it>e8 23 .tIxg7 'YWd8 24 .i.h3 c5 25 l:tdl left White in the driving seat. 1 2 . . . b4 1 3 ttJe2 ttJbd7 I t is dangerous to take the pawn 1 3. .. ..txe2 1 4 'it'xe2 'it'xd4 1 5 ..tb2 'it'b6 1 6 ttJc4 'it'b5 1 7 .i.xf6 gxf6 1 8 .:tad 1 a s 1 9 .tIfel a4 20 e5 and White had a strong attack in Smyslov-Guimard, Groningen 1 946. 14 �e 1 �c8 1 5 i.b2 'i'a5 1 6 ttJf4 g6 16 ... c5 17 e5 ttJe8 18 d5 and Black is in trouble. 1 7 ttJd3 �fd8 1 8 ttJc4 'i'h5 1 9 'i'c2 Even the ending after 1 9 'iix h5 ttJxh5 20 a3!? offers White the better prospects. 1 9 . . . 'ilfb5 20 a3

20 . . . ttJd5 ! ? Black has t o try something. 2 0. . .bxa3 2 1 .i.xa3 c5 2 2 d 5 gives White a clear lead. 21 axb4!

Clo s e d C a t a la n : 4... � e 7 5 Ji.g 2 0 - 0

A strong yet logical move, allowing White to retains his structural advantage. 21 exdS?! cxdS 22 ttJxb4 dxc4, on the other hand, is not clear. 21 . . . lbxb4 22 lbxb4 .Jtxb4 23 �ed 1 lbb8 24 �a4 "ilfb7 25 �da 1 .Jtf8 26 "ilfc3 .Jtg7 27 e5!' The weakness of the d6-square is clearly felt. 27 . . . .Jtxc4 27 ... .JtfS loses to 28 ttJd6! .Jtxd6 29 exd6 J:[xd6 30 dS, when the dark squares look like a Swiss cheese. 28 bxc4 �c7 29 .Jta3 .Jtf8 30 c5 30 .JtxfS cj;;x fS 31 J:[b4 'iVc8 32 .l:!.ab1 ttJd7 33 cS should also win 3 0 . . . �cd7 3 1 .Jtc 1 .Jte7 32 h 4 �c8 33 .Jte4 f5 34 exf6 .Jtxf6 3 5 �b1 "ilfc7 36 .Jtf4 "ilfd8 37 "ilfb3 �xd4 38 1:rxd4 1 -0 .

r------_

Game 3 Topalov-Kramnik

Unares 1997 1 d4 d5 2 c4 e6 3 lbf3 lbf6 4 g3 .Jte7 5 .Jtg2 0-0 6 0-0 ttJbd7 7 ttJc3 ! ? This gambit line i s n o longer s o 'hot' thanks in part to what now develops. 7 . . . dxc4!

The key to refuting a gambit is often in accepting the pawn(s) . 8 e4 c6 One of several possibilities at this junc-

ture. 8 . ..lt:Jb6 9 'ili'e2 cS 10 dxcS .JtxcS 1 1 ttJeS .Jtd7 1 2 .l:.dl 'iVe7 1 3 .JtgS gives White a slight edge, this being transfonned to a dan­ gerous attack in Gretarsson-Bjornsson, Reyk­ javik 1 999 after 1 3. .. h6 1 4 .Jtxf6 gxf6 I S ttJg4 cj;;g7 1 6 eS! f5 1 7 ttJf6 .Jtc6 1 8 g4 .l:.fd8 1 9 .Jtxc6 bxc6 2 0 cj;; h 1 . 8. . .c S 9 d S exdS 1 0 exdS ttJb6 1 1 ttJeS .Jtd6 12 f4 l:te8 13 a4 as (13. .. .JtxeS 14 fxeS nxeS I S .Jtf4 gives White a strong attack for the pawns; the lead in development is the most important factor here) 1 4 ttJbS .Jtf5 I S J:[e1 ttJg4 1 6 ttJxg4 J:[xe 1 + 1 7 'iVx e 1 .Jtxg4 1 8 .Jtd2 'ili'e7 1 9 'ili'xe7 .Jtxe7 20 d6 .Jtf6 21 .Jtxb7 .l:.b8 22 .Jtc6 .Jtxb2 23 l:tel was an edge for White in Petursson-Dutreeuw, San Bernadino 1 989 . 8 ... a6!? is an interesting option, Black try­ ing to hang on to the pawn. 9 a4 J:[b8 1 0 as bS 1 1 axb6 cxb6 1 2 .Jtf4 lIa8 1 3 ttJd2 bS 1 4 ttJxbS ttJb6 IS .Jtc7 'ili'd7 16 .Jtxb6 'i¥xbS 17 .JtaS is good for White according to Volzhin, so Kramnik-Lputian, Wijk aan Zee 2000 went l 1 ...ttJxb6 1 2 'iVe2 .Jtb7 1 3 .l:!.dl .Jtb4 14 ttJeS as! (making way for ... .Jta6 in case of ttJxc4 and improving on 14 ...'i¥e7, which gives White the chance to win a tempo and claim an advantage with I S .JtgS! h6 1 6 .Jtxf6 'iWxf6 1 7 ttJxc4 ttJxc4 1 8 'ii'x c4 'ili'e7 1 9 eS .Jtxg2 20 cj;;xg2, as in Volzhin-Neverov, Du­ bai 2001) I S ttJa2 na8 1 6 .JtgS?! (1 6 .Jtf4!? looks like an improvement) 16 ... .Jte7 and now 1 7 ttJxc4? .Jta6 1 8 b3 a4 1 9 .Jtxf6 axb3 20 'ili'g4 .Jtxf6 21 ttJxb6 cxb6 22 eS .Jte7 resulted in a clear advantage for Black. Kramnik gives the lesser evil 1 7 ttJc3, when the strongest is 1 7 ... ttJfd7! 1 8 ttJxd7 'ii'xd7 1 9 .Jte3 a4 with the better game for Black. 9 a4 e 5 ! ? A n interesting, fresh try. Older games contribute the following knowledge: 9 ... aS 1 0 'ili'e2 ttJb6 1 1 J:[dl .Jtb4 1 2 ttJeS 'ili'e7 1 3 .Jte3 .Jtd7 14 ttJxc4 ttJxc4 I S 'ili'xc4 and now Black went wrong in Botvinnik­ Em.Lasker, Moscow 1 936 with I S ... bS?! 1 6 13

Th e C a t a l a n

�e2 l:Iab8?!, when 1 7 axbS cxbS 1 8 eS ttJe8 19 dS! proved rather inconvenient for the defender. After the improvement l S ... eS 1 6 dxeS ttJg4 1 7 i.d4 .ie6 1 8 �e2 .ics White has 1 9 ttJdS!? cxdS 20 exdS .ixd4 21 .l:!.xd4 .ifS 22 l:Iel , with sufficient compensation for the piece to claim a lead. 9 ... b6 10 'iVe2 and now 1O ... .i.a6 1 1 Udl 'ifc8 12 dS!? exdS 13 exdS .l:!.e8 14 dxc6 ttJcS 1 5 ttJbS ttJd3 1 6 .u.xd3! cxd3 17 'iYxd3 fur­ nished White with compensation for the exchange in I.Zaitsev-Lukacs, Dubna 1 979, while 10 ... aS 1 1 .l::!.d l i.a6 12 .i.f4 .l:!.e8 1 3 ttJd2 ttJ f8 1 4 ttJxc4 ttJg6 1 5 .i.e3 should be better for White, which was the case in Ivan­ chuk-Dreev, Tallinn 1 986 after l S ... ttJdS!? 1 6 e S .l:!.c8 1 7 .l:!.ac1 ttJxe3 1 8 fxd b S 1 9 ttJa3 'iYb6 20 i.f1 etc. 1 0 dxe5 ttJg4 1 1 .ltf4 'ii'a 5 1 2 e6 After 12 'iYd4 .l::[d 8 13 'ii'xc4 ttJdxeS 1 4 ttJxeS ttJxeS 1 5 'iYb3 b 6 1 6 ttJdS cxdS 1 7 i.xeS i.a6 1 8 .l::t fel d4 Black has the better chances. 1 2 . . . fxe6 12 ... ttJdeS 13 ttJxeS ttJxeS 14 exf7+ J:!.xf7 15 ttJdS cxdS 1 6 i.xeS dxe4 17 i.c3 'it'a6 1 8 .ixe4 i. fS 1 9 i.xfS l:txfS 20 'i¥g4 .l:i. f7 even­ tually led to equality in Ignatiev-Yagupov, St. Petersburg 1 998. 1 3 'ii'e 2 Later Topalov tried to improve on his play in the main game, still without proving an advantage - Topalov-Kasparov, Sarajevo 2000 went 1 3 eS ttJdxeS 14 ttJxeS ttJxeS 1 5 'ifbs i.d6! 1 6 ttJe4 1:I.fS 1 7 'iYe8+ .l:!. f8 1 8 'iWhs ::tfS and a draw was agreed. 1 3 . . . ttJge5 1 4 ttJd4 ttJd3! The e6-pawn is of lesser importance than the placement of the pieces. 1 5 ttJxe6 %:tf6 1 6 .ltc 7 1 6 ttJc7? loses to 1 6 ... ttJxf4 1 7 'iYxc4+ �h8 1 8 ttJxa8 ttJeS 1 9 'iYb3 ttJxg2 20 'iitxg2 ttJf3 21 Ufdl �hS and Black has a winning attack. 1 6 . . . 'ii'b4 All of this was - oddly enough - played 14

again later the same year. Perhaps White did not know of this game? 1 7 ttJd4 After 17 ttJf4 'ifxb2 1 8 'ifxb2 ttJxb2 19 eS Uh6, as in Maherramzade-Kacheishvili, Baku 1 997, White has compensation after 20 %:tab 1 ttJd3 21 ttJxd3 cxd3 22 ttJe4, but better is 1 7 ... ttJ7eS! with a strong position for Black. 1 7 . . . .ltc5 1 8 ttJc2 'iWxb2 1 9 ttJd 1 'ii'b 3 The black queen will survive, leaving White's minor pieces looking very odd in­ deed. Topalov clearly must have overlooked something. 20 ttJde3 ttJ7e5 2 1 h3 .lte6 22 Wh2 22 i.. x eS ttJxeS 23 1:I.ab 1 'ifxa4 24 .l:!.xb7 'i¥a6 gives Black a clear advantage. 22 . . . %:th6 23 %:tab 1 ? ! 2 3 f4 ttJg4+ 2 4 ttJxg4 .i.xg4 2 5 'iYxg4 'iVxc2 26 h4 improves, although Black is ahead after 26 ... .l:!.f8. 23 . . . .ltg4!

A key tactic. 24 ttJxg4 24 'iid2 .l:txh3+! 25 i.xh3 ttJf3+ wins for Black, while 24 f3 loses to 24 ... ttJxf3+! 25 i.xf3 i.xh3 with a deadly attack after 26 ttJg2 'ifc3! etc. 24 . . . ttJxg4+ 25 'iWxg4 'iWxc2 26 'iWg5 26 f4 runs into 26 ... ttJf2! 27 'iff3 l:Ixh3+ 28 .i.xh3 ttJg4+ 29 �hl 'iWh2 mate. 26 . . . ttJxf2! 27 %:txf2 27 'it'xcs again meets with 27 ... .l:!.xh3+!, now with a twist after 28 'it>g1 Uh1+! 29

C lo s e d C a t a la n : 4

.txhl ttJh3 mate. 27 . . :i'xf2 28 �f1 'i'd4 29 ii.. e S 'i'd7 30 'i'c 1 'i'e6 31 ii.. x g7 Wxg7 32 'i'gS + �g6 33 'iWxcs 'i'd6 0-1

Game 4 Romanishin-Ribli

Polanica Zdrrj 1993 1 d4 dS 2 c4 e6 3 lLlf3 lLlf6 4 g3 ii.. e 7 S ii.. g 2 0-0 6 0-0 lLlbd7 7 lLlc3 c6 8 'iWb3

8 . . . b6 Black has two main alternatives: 8 ...'iVb6 appears inadequate. White was a litde better in Korchnoi-S.Polgar, Munich 2000 after 9 .l:.d 1 'it'xb3 10 axb3 b6 1 1 e4 i.b 7 1 2 e5 ttJe8 1 3 .tg5 .txg5 1 4 ttJxg5 h6 1 5 ttJh3 ttJc 7 1 6 ttJf4 a5 1 7 h4 i.a6 1 8 l:tac1 .l:.ac8 1 9 cxd5 cxd5 20 J:td2. But 8 ... ttJb6!? looks sound. Then 9 c5 ttJbd7 10 �c2 b6 1 1 b4 a5 12 b5 i.b7 1 3 cxb6 cxb5 1 4 ttJxb5 �xb6 1 5 .l:.bl .l:.fc8 1 6 �d3 i.a6 1 7 a4 i.b4 1 8 ..td2 ..txb5 1 9 axb5 :c4 gave Black the initiative in Martynov­ Sherbakov, Taby 1 99 1 . This leaves 9 cxd5 cxd5, when an interesting idea is 10 l:tdl !? with the point that after 1 0 ... .td7 1 1 ttJe5 .l:.c8 12 i.f4 White is a lime better. Instead Chetverik-o.Kozlov, Pecs 1 996 went 1 0 ttJe5 ttJfd7! (the knight o n b 6 i s the better placed of the two) 1 1 e4 ttJxe5 12 dxe5 d4 1 3 .I:tdl i.c5 1 4 ttJb5 .id7 1 5 a4 a6 and now White should have played 1 6 ttJd6 ..txd6 1 7

.

.

.

ii.. e 7 5 ii.. g 2 0 - 0

exd6 e 5 1 8 a 5 ttJc8 1 9 'flVxb 7 ttJxd6 2 0 �d5, with chances for both sides. 9 ii.. f4 Also good is 9 i.g5 i.b7 1 0 .l:.fdl a5 1 1 cxd5 cxd5 1 2 t!.ac1 with the more comfort­ able game for White. After the subsequent 1 2 ... ttJe4 1 3 .ixe7 �xe7 1 4 ttJd2 ttJef6 1 5 e3 �fc8 1 6 a3 .ia6 1 7 ttJbs lhc1 1 8 .l:.xc1 l:!.c8 1 9 .l:.xc8+ i.xc8 20 �c2 .ia6 21 .ifl White still had an edge in Tukmakov-Chernin, Lvov 1 990. 9 . . . ii.. b 7 Black has a couple of alternatives. 9 ... .ia6 1 0 cxdS cxdS 1 1 ttJbS! i.xbS 1 2 �xbS gives White a lasting edge with the two bishops, e.g. 1 2 ... a6 1 3 �c6 l:!.c8 1 4 'ii'a4 .l:.a8 1 5 .l:.fc1 bS 1 6 'ii'c 2 l:!.c8 1 7 i.c7 'ii'e 8 1 8 'ii'd2 ttJb8 19 ttJeS, Markowski-Filippov, Koszalin 1 999. 9 ... a5!? 10 .l:.fd 1 .ia6 1 1 cxdS ttJxdS! sees Black keep his options open, 12 ttJeS ttJxeS 1 3 .ixeS a4 1 4 'iVc2 a3 offering good coun­ terplay according to Kasimdzanov. 1 0 cxdS

White has also tried 10 .l:.ac1 dxc4 1 1 'iVxc4 ttJdS 1 2 ttJeS, when the natural 1 2 ... ttJxeS results in an unclear game after 1 3 i.xeS .tgS 1 4 e 3 i.xe3!? 1 5 fxe3 ttJxe3 1 6 'iie 2 ttJxfl 1 7 '.txfl f6 1 8 .if4 'tid7 (Ulibin & Tiviakov) . In Ulibin-Tiviakov, Kastellanata 1 998 Black opted for the dubious 12 ... bS?! but soon suffered: 1 3 'iVd3 ttJxe5 14 ..txe5 'it'b6 1 5 ttJe4 f6 1 6 .if4 .l:.ad8 17 .id2 .ic8 1 8 ttJcS as 19 .l:.fd 1 'it>h8 20 'iVa3! b4 21 �a4 15

Th e C a t a l a n

'iWb5 2 2 'it'xb5 cxb5 2 3 lLlb3 a 4 2 4 lLla5 �d7 25 lLlc6 �xc6 26 ':'xc6 and life was not easy for the defender. 1 0 . . . cxdS 1 O ... exd5 1 1 l:tad1 lteS 1 2 e4!? lLlxe4 1 3 lLlxe4 dxe4 1 4 lLle5 lLlxe5 1 5 dxe5 'WcS 1 6 Ji.xe4 Ji. f8 1 7 .l:.d2 c 5 1 S Ji.xb7 'Wxb 7 1 9 l:tfd1 1:te7 2 0 h4 l:[cS 21 'it'd5 'iVxd5 22 l:.xd5 was a litde better for White in Rashkovsky­ Goloshchapov, Ekaterinburg 1 999 in view of the domination of the d-ftle. 10 ... lLlxd5 1 1 lLlxd5 exd5 12 nael c5 1 3 l::t fd1 ncS 1 4 h4 cxd4 was played i n Chet­ verik-Hajnal, Gyula 1 999. Now 1 5 lLlxd4 lLlc5 1 6 'it'f3 gives White a slight plus due to the superior pieces and the isolated d-pawn. If possible the queen will go to g4, and in the event of 1 6 ... 'ii'd 7 White can consider 1 7 �h2! with the idea o f 1 S �h3. 1 1 l:tac 1 1 1 lLlb5!? is interesting, after which 1 1 ...a6?! is risky thanks to 1 2 lLld6!, when White gets the two bishops. Tiviakov gives 1 1 ...lLleS 1 2 lIael a6 1 3 lLlc3 with an even game, or 1 3 lLlc7 g5 1 4 lLlxaS gxf4 1 5 lLlxb6 lLlxb6 1 6 gxf4 with unclear play. 1 1 . . . lLlhS After 1 1 ...lLleS White achieves a slight edge with 1 2l::tc2 lLld6 1 3l::t fel lLlc4 1 4 lLld2 lLlxd2 1 5 �xd2, when an invasion down the c-ftle is imminent. 1 2 .lteS f6

1 3 .ltc7 ! ? 16

A very exciting piece sacrifice. White gets two pawns and free access to the queenside in return for the bishop. Meanwhile the knight remains stuck on h5. The alternative is 1 3 �f4!? lLlxf4 1 4 gxf4 rs 1 5 lLlb5 l:tf6 1 6 :c7 l:[bS 1 7 :tfel a6 1 S lLla7! with an attack on the queenside (Romanishin). 1 3 . . . ii'xc7 1 4 lLlxdS ii'd6 White gets an extra possibility after 1 4... 'it'dS, when 1 5 lLlxe7+ transposes and 1 5 lLlc7!? offers a hell o f a ride: 1 5 ... lLlc5 1 6 l:txc5! bxc5 1 7 'Wxb7 l:!.bS 1 S 'Wxa7 'it'cS! 1 9 lLld5 �d6 20 'it'e7! exd5 2 1 'it'xd6 and White will probably have the better chances in the endgame after 2 1 ...l:tdS 22 'i¥xc5 l:txb2 23 'it'xcs ktxcS 24 e3 as all the pawns are on one flank and Black has problems with the knight on h5. 1 S lLlxe7 + ii'xe7 1 6 l:tc7 .ltdS 1 7 ii'bS l:tfd8 17 .. .l:1adS 1 S ':xa7 'iVd6 19 .l:.el , with .l:i.cc7 to follow, also looks dangerous for Black. 1 8 e4 .ltxe4 Worth considering was 1 S ... a6 19 'We2 �xa2! (1 9 ... 'it'd6 20 :tfel Ji.xa2 21 e5! looks dangerous) 20 .l:ta1 Ji.b3 21 'it'e3 'i¥b4 22 'it'c3 'it'xc3 23 ':xc3 e5 24 ':xb3 exd4 with approximately even chances in the endgame, although White might argue otherwise. 1 9 ii'xhS ii'd6 19 ... l1dcS is strongly met with 20 lLle5! �xc7 2 1 Ji.xe4 fxe5 22 'ir'xh7+ 'iii' f8 23 �xaS exd4 24 lId1 and Black's king is walking on thin ice. 20 l:tfc 1 a6 21 lLlh4 .ltxg2 22 lLlxg2 eS?! It was necessary to take the pawn with 22 ... 'it'xd4, and after 23 lLlf4 play 23 ...'iVe4, when White has compensation. Perhaps 24 .l:i.1 c4!? is the best chance to fight for the initiative. 23 lLlh4 g6? 23 ... exd4 24 lLlrs 'it'e5 puts up more resis­ tance, although White has 25 lLlxg7! �xg7 26 .l:i.xd7+l::tx d7 27 'iVg4+ 'it>hS 2S 'iixd7 with a near decisive lead.

C lo s e d C a t a la n : 4 .

. .

i. e 7 5 i.g2 0 - 0

i.f4 lbc5 1 4 'it'xd8 l:[fxd8 1 5 h 3 lbh6 1 6 l:tfd 1 i.xc4 1 7 lbd4 l:tac8 1 8 i.xh6 gxh6 1 9 lb fS i.f8 20 lbd6 i.xd6 2 1 exd6 lbd3 22 ttd2 l:txd6 23l::tad1 with a clear advantage to White. Here the brutal 1 3 ...g5!? is more ap­ propriate, e.g. 1 4 i.xg5 (14 l'!ad 1 I?) 1 4 ... i.xg5 1 5 lbxg5 'iixg5 1 6 ii'xd7 lbxe5 1 7 'iY fS 'iYg7 with an unclear situation i n which White has interests on the kingside while Black is looking to the centre and to the queenside. 9 . . . i.a6 1 0 b3 J:!.cS Black could also choose to develop the other rook. An interesting choice is 1Q ... ii'c8 Stripping away the defences. 1 1 i.f4 l:td8 1 2 e4 b5 1 3 cxb5 cxb5 1 4 :tac1 24 . . . hxg6 25 �xg6 + 'ithS b4 1 5 lbxd5! (forced) 1 5 ... exd5 (1 5 ... 'iYxc1 1 6 After 25 ... �f8 26 I:t l c6 ii'e7 27 �xd7! the lbxe 7+ � f8 1 7 l1xc1 i.xd3 1 8 lbc6 l:!.dc8 1 9 last line of defence falls. i.d6+ 'it>e8 20 lbd2 leaves White with good 26 dxe5 1 -0 compensation for the exchange; Black will ,....--- have difficulties freeing himself) 1 6 l:!.xc8 Game 5 i.xd3 1 7 l:ha8 lha8 1 8 lhd3 lbxe4 (1 8 ... dxe4 1 9 lbel .l:!.e8 20 l:te3 favours Gleizerov-Barua White) 19 lbe5 lbdf6 20 lbc6 i.f8, UlibinKo/kala 2002 ----------------.. Vaulin, Bydgoszcz 2000, and now 21 i.b8!? gives White a modest plus. 1 d4 d5 2 c4 e6 3 tt:lf3 tt:lf6 4 g3 i.e7 5 1 1 e4 i.g2 0-0 6 0-0 tt:lbd7 7 tt:lc3 c6 S �d3 A patent of the Russian Grandmaster. S . . . b6 9 J:!.d 1 This seems to give White the best chances for an edge. Now the queen will not be un­ protected after e2-e4. Alternatives: 9 cxd5 cxd5 10 i.f4 a6 1 1 a4 i.b7 1 2 l:!.fc1 l:!.c8 1 3 lbd2 lbh5 1 4 i.e3 i.d6 1 5 f3 lbhf6 1 6 i.f2 'iYe7 1 7 i.h3 nfd8 1 8 .:tel lbb8 1 9 lIac1 lbc6 resulted in a balanced game in Ivanchuk-A.Petrosian, Lvov 1 988. 9 e4 i.a6 10 b3 dxc4 1 1 bxc4 e5!? and now 1 2 .l:i.dl seems to be rather hannless. One game continued 12 ...'iYc7 13 dxe5 lbxe5 1 1 . . . c5 14 lbxe5 ii'xe5 1 5 i.f4 ii'e6 1 6 lbd5!? cxd5 1 1 ...ii'c7 is best met with 1 2 i.f4 iVb7 13 17 exd5 ii'd7 1 8 d6 lIad8 19 'iic2 i.b7 20 e5 lbe8 1 4 a4 iVb8 15 i.gs i.xgS 1 6 lbxgS dxe7 'iYxe7 21 l:!.e1 �d7 22 l::ta dl �c8 23 g6 1 7 g4 with bright prospects on the king­ l:!.xd8 �xd8 24 i.xb7 'iYxb7 25 i.g5 ii'c6 26 side, Yevseev-Azarov, Minsk 2000. i.xf6 and a draw was agreed in Flear-Dizdar, 1 2 exd5 exd5 1 3 dxc5 Montpellier 1 998. Instead Vaganian-Mueller, 13 i.b2 dxc4 14 bxc4 cxd4 15 lbb5 lbc5 1 995 continued 1 2 dxe5 lbg4 1 3

24 tt:lxg6!

17

Th e Ca t a l a n

1 6 'iYxd4 'iWxd4 1 7 tLlfxd4 i.xb5 IS cxb5 tLla4 19 tLlf5 l:tfeS 20 i.d4 iLc5 21 l:tac1 gave White a fantastic advantage in Roman­ ishin-Ciric, Dortmund 1 976. As in so many positions in the Catalan, Black has a big problem with the c6-square. 1 3 . . . dxc4 1 4 bxc4

1 4 . . Jbc5? ! Preferable IS 1 4...tLlxc5 1 5 'ii'f l 'iVeS (1 5 ...tLlfd7!?) 1 6 tLlb5 tLlce4 1 7 tLlfd4 i.c5 I S iLb2, when I S ...l:tdS should limit White to a slight advantage. Worse is I S ...'iVe5?! 1 9 'ii'e2 iLxb5 20 cxb5 l:tcdS 21 tLlc6 l:txd1+ 22 l:txdl 'iVf5 23 i.d4, when Black was strug­ gling a little in Gleizerov-Perez, Malaga 2001 . 1 5 ltJb5 �xb5 1 6 cxb5 �d5 1 7 ltJd4 ltJe5 After 17 ... tLlc5 IS 'iVc4 tLlfe4 1 9 i.e3 i.f6 20 iLO! White stands clearly better according to Gleizerov. The main idea behind iLo is, of course, to protect the rook. Now tLlf5 and tLlc6 are threats, and Black will probably have to give up his second bishop. 1 8 'ific3 IS 'iVe2? would be a slip as after IS ... i.c5! ! Black assumes the initiative. 1 8 . . . �c5 1 9 �b2! Forced. 19 i.e3? tLlfg4! 20 i.xd5 'ii'x d5 21 'irb3 'iVe4 22 'irbl 'iNaS gives Black a sensational attack, and 1 9 i.xd5? 'iVxd5 20 iLe3 iLxd4 21 iLxd4 tLlf3+ 22 Wfl tLlxh2+ 23 Wg 1 tLlf3+ 24 Wfl tLlxd4 25 'iVxd4 'ifh 1+ 26 �e2 l:teS+ looks deadly. 1 9 . . . �d6 20 �d2 'ifid7 18

20 ... tLld5 fails to 21 i.xd5! l:txd5 22 tLlc6! l:txd2 23 tLlxdS iLd4 (only move) 24 'iVxd2 tLlf3+ 25 Wg2 tLlxd2 26 iLxd4 IhdS 27 l:tc1 with a wonderful ending. 21 �ad 1

2 1 . . . 'ilfg4 2 1 ...tLld5? 22 .txd5! l:txd5 23 tLlb3 and White wins material. 21 ...l:tdS 22 tLlb3! is also very uncomfortable. 22 h3! This leads by force to a highly favourable ending. 22 . . . �xd4 23 hxg4 �xc3 24 �xc3 �xd2 25 �xd2 ltJexg4 26 �c6 h5 27 f3 ltJe3 28 �xf6 gxf6 29 �d7 White has all the chances. 29 . . . 'it>g7 30 �xa7 �d8 31 �d7 ! The minor piece endgame i s winning, and the attack on f7 cannot be allowed. 3 1 . . . �xd7 32 �xd7 ltJc4 33 'it>f2 'it>f8 34 �h3! The h5-pawn is the target of this bishop manoeuvre. 34 . . . 'it>e7 35 �f1 ltJa3 36 f4 'it>d6 37 �e2 'it>c5 38 �xh5 ltJxb5 39 �xf7 'it>d4 40 'it>f3 ltJd6 4 1 �b3 f5 42 g4 White does not hesitate in creating a deci­ sive ed pawn. An important point to notice here is that as is a light square, so White does not risk having a pawn + bishop versus king scenario in which he cannot win. 42 .. .fxg4+ 43 'it>xg4 b5 44 'it>g5 b4 45 'it>g6 ltJe4

Clo s e d C a t a la n : 4

.

.

.

ii.. e 7 5 ii..g2 0 - 0

Black also loses after 45 ... 'it>e4 46 fS! ttJxfS 47 1i.c2+ etc. 46 ii.. c 2 ttJc5 47 f5 'it'e5 48 f6 ttJd7 49 f7 'it'd6 50 'it'g7 'it'e7 51 ii..f 5 1 -0

1 5 .. .'�Jfd7! is good, when 1 6 ttJxf7!? 'it>xf7 1 7 exd5 i s unclear. The alternatives are worse, e.g. 1 5 .. d4?! 16 e5! dxc3 17 'i!fc2! with a very dangerous position for Black. And 1 5 ... h6?! 16 ttJxf7! 'it>xf7 17 e5 'it>g8 (1 7 ... ttJe4 18 e6+! Game 6 'it>xe6 1 9 ttJxe4 is not something Black will Korchnoi-Nogueiras survive, an important line being 1 9 ... dxe4 20 lWe2! i.d5 21 i.xe4 'i£tf7 22 'Wh5+ 'it>g8 23 Moscow 1994 '----------------.. lIxd5! and Black can resign as mate is com­ 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 ii.. e 7 5 ing) 1 8 exf6 1i.xf6 1 9 1i.a3 was a clear lead for White in Tisdall-Chernuschevich, Halls­ ii.. g 2 0-0 6 0-0 ttJbd7 7 ttJc3 c6 8 b3 b6 dall 1 996. 9 ii.. b 2 1 0 a4! ? A highly original approach, gaining space on the queenside. The alternative is lO ttJd2, when Black has two main moves. Ehlvest-Nogueiras, New Delhi 2000 went lO ... b5 1 1 c5 (1 1 cxb5 cxb5 12 a3 lIc8 1 3 b4 ttJb6 14 ttJb3 ttJc4 1 5 1i.c1 ttJd7 1 6 lIa2 i.b7 1 7 ttJxb5 a6 1 8 ttJc3 was enough for an edge in Korchnoi-Lputian, Sarajevo 1 998) l 1 ...b4 1 2 ttJa4 1i.b5 1 3 lIe 1 �c7 14 'iVc2 e5 1 5 e4!? dxe4 1 6 ttJxe4 exd4 1 7 i.xd4 lIfe8 1 8 lIac1 J:tad8 1 9 ttJb2 and White's extra territory afforded him more room for manoeuvre and therefore the superior prospects. 9 . . . ii.. a 6 lO ... lIc8 1 1 e4 dxe4 (l 1 ...c5 12 exd5 cxd4 Black's main alternative is, not surpris­ 13 dxe6 fxe6 14 ttJce4 slightly favoured ingly, 9 ... 1i.b7, when White has tried a few White in Tregubov-Fomichenko, Krasnodar moves: 1 999) 1 2 ttJcxe4 (1 2 ttJdxe4?! b5! 1 3 lIel 10 "ifc2 might be best, transposing to 7 bxc4 1 4 'iVe2 'i¥a5 was pleasant for Black in 'it'c2 c6 8 b3 b6 9 ii.b2 ii.b7 l O ttJc3 (see Cifuentes-Gomez, Terrasa 1 995) 1 2 ...ttJxe4 Game 1 0) . 1 3 ttJxe4 ttJf6 14 'YWc2 ttJxe4 1 5 1i.xe4 h6 1 6 1 0 l:!.c1 ttJe4!? 1 1 ttJxe4 dxe4 1 2 ttJe5 fS Ihdl 1i.f6 1 7 'it'e2 lIe8 1 8 'ii f3 1i.b7 1 9 h4 1 3 'it'c2 ttJxe5 1 4 dxe5 c5 1 5 l:!.cdl "ifc7 was l:!.c7 20 l:!.fel and White was slightly better in a draw in Szabo-Sky, Palma de Mallorca Kaidanov-Cifuentes, Andorra 1 99 1 . 1 969, probably because the position offers 1 0 . . . dxc4 little promise of progress for either player. Black is not forced to take the pawn but it Stojanovic-Jovcic, Tivat 1 995 went l O seems to be the best option. After lO ... lIc8 ttJe5 ttJxe5 1 1 dxe5 ttJd7 1 2 cxd5 cxd5 1 3 e4 d4!? (a logical temporary pawn sacrifice, de­ 1 1 ttJd2 l:!.e8 12 e4 ttJf8 13 l:!.el ii.b7 14 as bxa5 1 5 c5 'iVc7 1 6 e5 ttJ6d7 17 ttJe2 ii.a6 1 8 signed to close out the g2-bishop) 14 'iVxd4 1i.c3 fS 1 9 h4 ii.b5 20 1i.xa5 'i!fb 7 2 1 ttJf4 i.c5 1 5 "it'dl ttJxe5 1 6 'it'h5 ttJg6 1 7 l:!.fdl ttJb8 22 'iic 2 White had the advantage in 'iVe 7 1 8 ttJa4 ii.a3 19 ii.xa3 'iVxa3 with an­ Poluliahov-Akhmadeev, St. Petersburg 1 999. other draw. 1 1 bxc4 ii.. x c4 1 2 ttJd2 ii.. a 6 1 3 ii..x c6 lO 'it'd3 c5 1 1 cxd5 exd5 1 2 .l:!.adl l:!.e8 1 3 .l:!.c8 1 4 ii.. g 2 ttJb8! dxc5 bxc5 1 4 ttJg5 ttJb6 1 5 e4!? and now 19

Th e Ca t a la n

The knight is transferred to a square with a future, at the same time vacating a square for the queen - and, in turn, the king's rook. Black should not be worse here. 1 5 lLlb5 'tWd7 1 6 'tWb 1 ! ? �xb5? ! After this White always has the advantage of bishop versus knight. 1 6 ... ltJc6!? looks like an improvement. After 1 7 ltJc4 ltJxd4 1 8 ltJe5 ltJxe2+ 1 9 'lith 1 'iVd2 20 ltJc6 .l:!.xc6!? 2 1 �xc6 ltJg4 Black has strong compensation for the exchange in the form of a direct at­ tack on the enemy king. White is forced to play 22 'iWe4 'iWxb2 23 'iWxg4, with chances for both sides. 1 7 axb5 'ii'x b5 1 8 l:txa 7

1 9 �c 1 lLlc6 Forced. After 19 ... �d6 20 .if3 l:txc1 21 .ixe2 l:txb 1 22 l:txb 1 ltJc6 23 l:ta4 White is close to winning. 20 �xc6 l:txc6 21 l:txe7 l:tfc8 22 lLlb3 h5 23 l:ta7? ! White should play 2 3 �e3!? ltJd5 2 4 l:ta7 J:k2 25 l:tel "it'd3 26 ltJal ltJxe3 27 ltJxc2 ltJxc2 28 .l:i.c1 l:!.c3 29 .l:.b7 with decent chances of success. 23 . . . h4 Black needs to create some insecurity around his opponent's king, but this is not the only way. After 23 ... .l:!.c2 24 ..ia3 ltJg4 25 'iVel (25 'it'dl 'iVxdl 26 lhdl lhf2 27 l:td2 l:tf3! and Black is still fighting) 25 ... "it'xel 26 l:txe 1 J::tx f2 27 J::tc 1 J::tc c2! 28 .l:!.xc2 llxc2 there are still some drawing chances. 24 �f4 lLld5 25 .!:te 1 'tWg4 26 'tWd 1 'tWxd 1 27 l:txd 1 l:tc3 28 l:tb 1 l:td3 28 ... ltJxf4 29 gxf4 l:tf3 30 l:tb7 .l:!.c6 31 llc1 l:td6 also offered stubborn resistance. 29 l:ta2 l:tcc3 30 l:tab2 h3? ! 30 ... ltJxf4 31 gxf4 l:!.f3 was necessary ac­ cording to Nogueiras. 31 �f 1 lLlxf4 32 gxf4 l:tf3? A blunder. After 32 ... 'lith7 33 'litel 'it>g6 White still needs to solve some technical problems before he can write 1 -0 on the score sheet. 33 lLlc 1 ! 1 -0

Game l Chiburdanidze-Vaganian

Biel 1994

1 8 . . . 'tWxe2! ? Essentially a piece sacrifice. 1 8 ... .id6 1 9 e3 ltJc6 20 l:ta2 gives White a modest but enduring advantage. The ed pawn IS more a weakness than a strength. 20

1 d4 d5 2 c4 e6 3 lLlf3 lLlf6 4 g3 �e7 5 �g2 0-0 6 0-0 lLlbd7 7 'ii'c 2 c6 This slow build-up is the main line, and the way the position should be played. How­ ever, there are two notable alternatives. 7 ... c5 8 cxd5 ltJxd5 9 ltJc3 ltJxc3 (9 ...liJb4 10 iVb3 b6 1 1 lldl ltJc6 12 d5! favours White, and in Adianto-Dao Thien Hai, Is­ tanbul 2000 the subsequent continuation 1 2 ... ltJa5 1 3 �c2 exd5 1 4 ltJxd5 ..ib7 15

C lo s e d C a t a l a n : 4

�h3 lLlb8 1 6 lLlf6+ �xf6 1 7 l::tx d8 �xd8 1 8 � f5 h6 1 9 b4!? cxb4 20 �b2 lLlbc6 2 1 'ii'e 4 lLle7 22 �h7+ 'it>h8 23 'iWg4 f6 24 �d3 lLlac6 25 'ii'e 4 f5 26 'iWe6 saw Black come under heavy attack) lO bxc3 cxd4 1 1 cxd4 lLlb6 1 2 �f4 lLld5 1 3 l:!.ac1 �d7 1 4 lLle5 �b5 1 5 'iWb2 �a6 1 6 l:lfdl with a healthy develop­ ment lead for White in Allen-Soloveychik, Canberra 1 999. There followed 1 6 ...l:tc8 1 7 �xc8 'ii'x c8 1 8 ::tcl 'ii'd 8 1 9 �d2 �d6 20 lLlc4 �c7 21 e4 lLlb6 22 lLle3, and White's stronger centre and more attractively posted pieces combined to form an advantage. 7 ... b6 weakens the c6-square and gives White the advantage after 8 cxd5!, when Black has two unpleasant recaptures. In the case of 8 ...lLlxd5 9 lLlc3 �b7 lO lLlxd5 exd5 1 1 .l:tdl lLlf6 1 2 lLle5 White has a slight but stable advantage. The knight would like to go to f4 and exert pressure on d5. Karpov­ Spas sky, Riga 1 975 continued 1 2 ... c5 1 3 dxc5 �xc5 1 4 lLld3 �d6 1 5 �f4l::te 8 16 e3 lLle4 17 �xd6 'iVxd6 1 8 lLlf4 l:tac8 1 9 'ii'a4, when White's advantage had increased. 8 ... exd5 9 'iWc6 Si.a6 lO lLlc3 l:tc8 (l O ... ne8?! 1 1 lLle5 J::tb 8 12 'iVa4 �b7 1 3 lLlxd7 lLlxd7 1 4 lLlxd5 and White was a pawn up for nothing in Raetsky-Renner, Apolda 1 993) 1 1 'ii'a4 lLlb8 12 �h3 'iVe8 1 3 'iWc2! (1 3 �xc8 'iWxa4 14 lLlxa4 �xe2 1 5 l:tel �xf3 1 6 Si.b7 is less clear) 1 3. .. l:td8 1 4 Si.f4 �d6 1 5 �e5 and Black will have consider­ able problems getting his pieces into the game. 8 .l:!.d 1 b6 Black has little chance of equalising with 8 ... b5. Then White has no use for the c-ftle, so best is 9 c5, when Rashkovsky-Farago, Sochi 1 980 went 9 ... lLle4 1 0 a4 �a6 1 1 axb5 Si.xb5 1 2 lLla3 a6 1 3 lLle1 f5 1 4 lLld3 �f6 1 5 lLlf4 IIe8 1 6 f3 lLlg5 1 7 h4 lLlfl 1 8 e4 g6 1 9 e 5 �g7 20 �d2 'ii'e7 2 1 lLlxb5 axb5 22 lLld3 with an advantage for White, who had more space and access to the f4-square (which might prove problematic for Black on e6) . 8 ... lLle4 9 b3 f5 i s a n interesting idea. Of

.

.

.

� e 7 5 �g2 0 - 0

course Black must accept the exchange of the dark-squared bishops. Ara Minasian­ Lputian, Yerevan 2000 continued 10 �a3 �xa3 I 1 lLlxa3 f4 1 2 lLlbi g5!? 1 3 lLlc3 g4 1 4 lLlh4 lLlg5 ( 1 4. . .lLlxf2!? 1 5 l::tf l fxg3 1 6 hxg3 lLlh3+ 1 7 'it>h2 followed by e2-e4 looks promising for White) 1 5 e4 lLlf6 1 6 exd5 exd5 1 7 cxd5 cxd5 with chances for both sides. In this kind of position Black obviously hopes that he will achieve something on the kingside, and he has reasonable chances of doing so. White, for his part, will be looking to quickly invade the queenside. 9 a4! ?

White i s putting immediate pressure on the queenside. Consequently Black should now be careful how he develops. 9 �a6 Inferior is 9 ... �b7?! due to 10 as!, with the threat of a5-a6. In Zagorjansky-Bonch­ Osmolovsky, Moscow 1 949 White emerged with the better game after 1 0 ... bxa5 1 1 c5 lLle4 1 2 lLlbd2 f5 1 3 lLlb3 �a6 1 4 lLlxa5 �b5 1 5 b4 'ii'e 8 1 6 lLle5l::t c 8 1 7 f3 lLlef6 1 8 lLlb7. 1 0 b3 .l:!.c8! Black needs to build up his ... c6-c5 push with some care. After lO ... c5 1 1 lLlc3 dxc4 1 2 bxc4 �xc4 1 3 dxc5 �xc5 1 4 lLle5 lLlxe5!? 1 5 l:txd8 .l:taxd8 1 6 �b2 there i s compensation for the queen, but not much. It seems to us that White still has slightly superior long­ term prospects because Black has no obvious . . .

21

Th e C a t a l a n

point of attack o n which t o concentrate. Note that here 1 1 .J::k8?! is too late com­ pared with the main game as now White has 12 lLlb5! dxc4 1 3 bxc4 cxd4 14 lLlfxd4 and Black has the usual problems with the c6square, e.g. 14 ... lLld5?! 1 5 'iib 3 lLlc5 1 6 'itb1 lLlf6 1 7 .1i.f4 'iii e8 1 8 lLlxa7 lId8 1 9 lLlac6 and White was on his way to the full point in Pigusov-Tavadian, Irkutsk 1 983. 1 1 a5 1 1 lLlc3 throws a pawn away. 1 1 . . . c5! Black is finally fully prepared. 1 2 axb6 'Wii' x b6 1 3 'Wii' a 2

1 3 . . . �b7 The strange looking 1 3. .. lLlb8 14 .1i.f4 .1i.d6 was played in Korchnoi-Spraggett, Wijk aan Zee 1 985. Now White could try 15 b4!? .1i.xf4 16 bxc5 'iic 7 17 gxf4 .1i.xc4 18 'it'd2 when the control over e5 and the weak a­ pawn, together with his own protected ed pawn, secure a lead. 1 4 �a3 tLie4 Also possible is 14 ... l:tfe8 1 5 lLlbd2 cxd4 1 6 .1i.xe7 lIxe7 17 'iixa7 e5, as in Tratar­ Atalik, Nova Gorica 1 999. Black should never get into trouble here, and Atalik earned approximate equality after 1 8 'ii'x b6 lLlxb6 1 9 lIa7 lIcc7 2 0 c5!? lIxc5 21 lLlxe5 lLlc8 22 lLld3 lIc3 23 lIa4 lIxe2 24 ltxd4 g5 25 lLlb4 lLle7 26 i.f1 .t!.e6 27 lId3 lIc5. 1 5 e3 �f6 1 6 �b2 dxc4 1 7 bxc4 z:.b8 1 8 tLic3 22

On 1 8 lLlbd2 Black has the interesting 1 8 ... lLlxf2!?, leading to equality after 19 c,t>xf2 .1i.xf3 20 'it>xf3 'iVxb2 21 'iix b2 lIxb2 22 ltxa7 lId8. 1 8 . . . �a8 1 8 ... lLlxf2?? now loses to 1 9 lLla4! etc. 1 9 tLixe4 19 lhb 1 might well leave White ahead, Black needing to improve on 1 9 ...'iVxb2!? 20 'it'xb2! (20 lIxb2 lLlxc3 21 'iVa 1 lIxb2 22 'ilixb2 lLlxd1 23 'iVd2 lLlxe3 24 fxe3 lId8 gives Black good counterplay) 20 ... .l:i.xb2 21 lLlxe4 .l:i.c2 22 lLlxf6+ lLlxf6 23 dxc5 .l:i.xc4 24 l::t dc1 , when the c-pawn is annoying for Black. 1 9 . . . �xe4 20 �c3 20 lLle5 leads to equality - 20 ... i.xe5 21 dxe5 .1i.xg2 22 'it>xg2 �7+ 23 'it>g1 lLlb6 and the knight is no worse than the bishop. 20 . . . 'Wii' b 7 Also possible is 20 ... cxd4!? 21 exd4 'iVc6 22 lLle5 lLlxe5 23 i.xe4 lLlrn 24 i.xf3 'iix f3 with equality. 21 'Wii' x a7 21 lLle1 .1i.xg2 22 lLlxg2 cxd4 23 exd4 'iVb3 is fine for Black. 21 . . . �xf3 22 �xf3 'Wii' x f3 23 'ilfxd7 h5 24 'Wii' c 7 z:.b3 25 �e 1 cxd4 26 exd4 g5!7

Black is trying to generate something in­ teresting. 27 z:.db 1 27 l::t ab 1 ? .1i.xd4! is not what White wants. 27 . . . �xd4 28 z:.xb3 'iYxb3 29 z:.a5 .if6

C lo s e d C a t a la n : 4

30 lXb5 tt'd 1 3 1 tt'a5 tt'f3 32 iLd2 lXa8 33lXxg 5 + ! White correctly goes for the draw before he gets mated. 33 . . . iLxg5 34 'ii'x g5 + 'itt h 7 35 'ii'h 6+ 'itt g 8 36 'ii'g 5 + Y. - Y.

Game 8 Kengis-M . G u revich

Jurmala 1985 1 d4 d5 2 c4 e6 3 tLif3 tLif6 4 g3 iLe7 5 iLg2 0-0 6 0-0 tLibd7 7 ifc2 c6 8 iLf4 ! ?

.

.

.

iL e 7 5 iL g 2 0 - 0

�e8 1 7 f3 hxg3 1 8 hxg3 'ii'g 6 1 9 l2Je2 �d7 20 l2Jf4 'it'g5 21 '.t>f2 \itJf7 22 l:th1 llh8 with a balanced game. 9 �d2! fS to b4 g5 1 1 a4 g4 1 2 l2Je 1 l2Jb6 1 3 l2Ja3 dxc4 14 e3 e5 1 5 dxe5 l2Jxa4!? 1 6 l2Jxc4 i s a shade preferable for White, while Black was in trouble in Kurajica-Flear, Las Palmas 1 993 after 1 5 ... .te6?! 1 6 as l2Jd5 1 7 l2Jxc4 l2Jxb4 1 8 .1Lxb4 .txb4 1 9 l2Jd3 .te 7 20 l:tfd1 etc. 9 tLibd2 Harmless is 9 cxd5 cxd5, e.g. to l:tc1 �b7 1 1 l2Jc3 l:!.c8 1 2 "it'd 1 l2Je8 1 3 l:tc2 g5 1 4 �d2 l2Jd6 1 5 l:tac 1 f6 1 6 l2Je 1 fS 1 7 l2Jd3 l:te8 1 8 l2Jb 1 l:txc2 1 9 .l:txc2 'itb8, Andersson-Bareev, Ubeda 1 997. After the second pair of rooks is exchanged Black is by no means worse. 9 . . iLb7 1 0 e4 .

This development of the bishop does not guarantee an advantage but is nonetheless a sound and practical alternative to b2-b3, which can be seen in Games to-IS. 8 . . . b6 8 ... a5 9 l2Jbd2 l2Jh5 to �e3 a4 1 1 l:tfd 1 �a5 1 2 �g5 f6 1 3 �e3 .1Ld6 1 4 l2Jel fS 1 5 l2Jdf3 'it'a6 1 6 .l:tac1 gives White a slight edge. Khalifman-LSokolov, Rogatska Slatina 1 99 1 saw this increase i n value after 1 6 ... 'ii'x c4?! 1 7 'ii'x c4 dxc4 1 8 l:txc4 l2Jb6 1 9 l:tcc1 a 3 2 0 b3 l2Jd5 21 �d2 l2Jb4 22 .1Lxb4 .1Lxb4 23 l2Jd3 .1Le7 24 l2Jfe5, with a dominating position. In the event of 8 ... l2Jh5, little is achieved by assuming that this is a normal Stonewall and playing 9 �c 1 . Bareev-Balashov, Elista 1 996 went 9 ... fS (9 ... l2Jhf6!? with a repetition is also sensible, but perhaps 10 b3 was the ide� 10 b3 g5 1 1 e3 g4 1 2 l2Je5 l2Jxe5 1 3 dxe5 l2Jg7 1 4 l2Jc3 h 5 1 5 �b2 h4 1 6 l:tad1

1 0 . . . !!.c8 The immediate to ... c5!? is possible, e.g. 1 1 exd5 exd5 1 2 l:tfe 1 l:tc8 1 3 'ii'fS !? dxc4 1 4 l2Jxc4 g6 1 5 �h3, which was seen in D.Gurevich-Browne, Key West 1 994. Now Black should have played 1 5 ... cxd4 1 6 l2Jd6 �xd6 17 �xd6 l:te8 1 8 l2Jxd4 ..ixg2 1 9 "it'xg2 l2Jc5 with equality (Browne) . 1 1 lXad 1 White also achieves very little from closing the position. After 1 1 e5 l2Je8 12 h4 h6 1 3 cxd5 cxd5 1 4 �d3 �c7 1 5 �e3 .ta6 1 6 l:t fb l 'iVc2!? 1 7 l2Jel "ij' fS 1 8 l2Jdf3 l2Jc7 1 9 l:tc1 l:tfd8 20 l:tc2 l2Je8 2 1 .l:tac1 l2J rn the position was complicated but no worse for 23

Th e C a t a l a n

Black i n Stohl-San Segundo, Moscow 1 994. 1 1 . . . c5 1 2 exd5 exd5 1 3 'i!t'b 1 1 3 �fe l l!e8 1 4 �f5 g6 1 5 'ii'h 3 dxc4 1 6 ttJxc4 cxd4 also gives Black a good game, particularly after 1 7 ttJce5?! ttJxe5 1 8 i.xe5 d3 19 'ir'h6 i.c5 20 h3 i.xf3 21 i.xf3 l:txe5! 22 l:!.xe5 'ir'd4 as in Kaidanov-Vaganian, Armenia 1 994, when Black had a strong attack on the dark squares as well as a dangerous ed pawn. All in all this is way too much for an exchange. 1 3 . . .l::t e8 Also fine is 1 3. .. dxc4 14 ttJxc4 i.e4 1 5 'iii'ai cxd4 1 6 ttJd6 i.xd6 1 7 1i.xd6 I:te8 1 8 .l:txd4 with roughly even chances (Kengis) . 1 4 b3 dxc4 1 5 tZlxc4 �e4 1 6 'i!t'b2 b5 16 ...ttJd5 17 i.d6 1i.f6, with unclear play, is a sound alternative. 1 7 tZlce5 c4

An alternative is 1 7 ... 'iib 6!? 1 8 .l:tfe l 'itb7 1 9 ttJh4 i.xg2 20 ttJxg2 ttJxe5 21 1i.xe5 cxd4 22 i.xd4 with a draw. 1 8 tZlxf7 ! White enters the spirit of wild tactics. Af­ ter 1 8 bxc4 bxc4 1 9 ttJxf7 Black has the at­ tractive alternative 1 9 .. .'ii b 6! 20 1i'xb6 axb6 21 ttJd6 i.xd6 22 1i.xd6 c3 etc. 1 8 . . . c3 1 9 'i!t'e2 'i!t'b6 19 ... i.xf3?? 20 'ir'e6! and White wins due to the threat of ttJh6+ and smothered mate. 20 tZl7e5 �a3 21 �h3 �a8 21 ...c2!? 22 i.xd7 cxd l 'iV 23 I:txdl i.xf3 24 'iVxf3 ttJxd7 25 ttJxd7 'ir'c6 26 'i¥xc6 l::.x c6 24

2 7 d 5 I:tc2 2 8 d 6 with unclear play was an interesting idea. But Black is happy with the half point. 22 'i!t'c2 �e4 23 'i!t'e2 Y:z - Y:z

Game 9 Dautov-Tiviakov

Venlo 2000 1 d4 d5 2 c4 e6 3 tZlf3 tZlf6 4 g3 �e7 5 �g2 0-0 6 0-0 tZlbd7 7 'i!t'c2 c6 8 �f4 b6 9 l:td 1

9 . . . �b7 Here this is sounder than 9 ... i.a6, as now White would have different options: 10 b3 l:!.c8 1 1 ttJc3 dxc4 12 ttJd2 b5!? 1 3 bxc4 bxc4 1 4 'iVa4 i.b5 1 5 ttJxb5 cxb5 1 6 'iVxb5 'itb6 1 7 'ir'xb6 ttJxb6 presented Black with interesting counterplay in Delchev­ Beliavsky, Moscow 200 1 . 1 0 ttJe5! uses the posting o f the bishop on f4 creatively. After 1O .. .'�c8 1 1 ttJc3 Black has tried a couple of moves. 1 1 ...�7 12 b3 l:.ac8 1 3 e4 h6 14 ttJxd7 'ir'xd7 1 5 e5 ttJe8 1 6 1i.f1 ttJc7 1 7 a4 1i.b7 1 8 'ilVe2 favoured White in Burmakin-Graf, Ohrid 2001 . It is not clear where Black will find genuine counterplay; a lack of breathing space is a problem. Wojtkiewicz-Zubarev, Moscow 2002 went 1 1 ....l1.xc4 12 ttJxc4 dxc4 1 3 ttJe4 h5 14 ttJd6 i.xd6 1 5 i.xd6 l::.d 8 1 6 a4 with an advantage for White. The two bishops are clearly stronger than the knights.

C lo s e d Ca t a la n : 4

1 0 4:le3 dxe4 Also possible is lO ... .l:!.c8. Then after 1 1 ttJeS ttJhS 1 2 3l.d2 ttJhf6 1 3 e4 cS Black has good counterplay. In Fridman-Yusupov, Essen 2002 a draw was agreed after 1 4 exdS cxd4 1 5 ttJc6 3l.xc6 1 6 dxc6 ttJeS 1 7 ttJbS ttJxc6 1 8 �f4 'iVd7 19 J:tac1 3l.cS 20 �xc6 lhc6 21 3l.eS 'iVe7 22 3l.xd4. Now the strongest approach appears to be the slow 1 1 b3, e.g. l 1 ...ttJhS 12 3l.c1 f5 1 3 �b2 3l.d6 1 4 e 3 ttJhf6 1 5 ttJe2 ttJe4 1 6 ttJf4 'iWe7 1 7 ttJeS! ttJxeS 1 8 dxeS �b8 1 9 a4 with a pleasant game for White in Marin­ Pogorelov, Andorra 1 994. 1 1 4:ld2 4:lh5 1 2 4:lxe4 4:lxf4 1 3 gxf4 "fie 7 1 4 e3 J:.aeS 1 5 J:.ab 1 In Lautier-Tiviakov, Mondariz 2000 White played 1 5 l:tac1 ttJf6 1 6 a3 ttJdS 1 7 ttJeS �d6 1 8 'iVa4 ttJxc3 1 9 l::tx c3 �xeS 20 fxeS f6 21 f4 �a8 22 b4 'iVf7 23 'iVc2 'iVhs 24 l:td2 fxeS 25 dxeS and achieved lasting pressure. 1 5 . . . 4:lf6 1 6 b4 4:ld5 1 7 a3 f5! ? 1 7 . . .ttJxc3 1 8 'it'xc3 c S 1 9 dxcS �xg2 20 'it'xg2 bxcS 21 bS gives White a long lasting advantage due to the strong knight and the potential ed pawn, which will soon come to b6. 1 S 4:le5 ii.. d 6 1 9 l:tbe 1 "ile7 20 "ila4 a6 21 "ilb3

2 1 . . . g5! Black needs to create counterplay, other­ wise White will have all the fun. 22 fxg5

.

.

.

ii.. e 7 5 ii.. g 2 0 - 0

22 ttJxdS cxdS 2 3 'iWd3 g4 with equality. 22 . . . ii.. x e5 23 dxe5 "fixg 5 24 �h 1 �hS 25 f4 "ilh5 26 4:lxd5 exd5 27 "ilb2 l:tgS Black might solve his problems both eas­ ier and quicker with 27 ... bS!? 28 'iVf2 lIc4, when White has to search for an advantage. 2S l:tg 1 l:txe 1 29 l:txe 1 l:teS 30 l:txeS + ii.. x eS 3 1 h3 3 1 'it'd4 'iVe2 32 h3 'it'bs 33 a4 'iVc6 and Black maintains equality. 3 1 . . . "ileS 32 "ile2 "ild7 33 "ilf2 "fidS 34 �h2 ii.. d 7 35 ii..f 3 a5 36 "fie 1 �gS 37 ii.. h 5 �fS 3S ii.. f 3 ii.. e S 39 bxa5 bxa5 40 "ile3 Y2 - Yo

Game 1 0 Kozul -lputian

Lucerne 1997 1 d4 d 5 2 e4 e6 3 4:lf3 4:lf6 4 g3 ii.. e 7 5 ii.. g 2 0-0 6 0-0 4:lbd7 7 "ile2 e6 S b3

The main system in the Closed Catalan. White is in no hurry and slowly builds up his forces. S . . . b6 The same goes for Black - the queen's bishop needs to find a route into the game. However, there are interesting alternatives. 8 ... ttJe4 9 l:tdl (9 ttJfd2 ttJxd2 lO ttJxd2 f5 1 1 �b2 'it'e8 1 2 ttJf3 'it'hS 1 3 'it'c 1 ttJf6 1 4 �a3 gave White a slight plus i n Paunovic­ Piankov, Burgos 2001) 9 ... b6 l O ttJe l ttJd6 1 1 ttJd2 �b7 1 2 �b2 l:tc8 1 3 ttJd3 cS 1 4 25

Th e Ca t a l a n

dxc5 bxc5 1 5 cxd5 exd5 1 6 ttJf4 ttJf6 1 7 .l:tac1 d 4 1 8 ii.xb 7 ttJxb7 1 9 ttJc4 ttJd5 20 ttJxd5 'it'xd5 21 e3 ii.f6 22 e4 'ii'e 6 23 f4 and White had the better game in Poluliahov­ Telnov, Tomsk 200 1 . 8 ... b5 9 c5! (opening the c-ftle makes little sense with the queen on c2) 9 ... ttJe4 (this seems to be too time consuming) 1 0 ii.b2 as 1 1 ttJe5 ttJxe5 12 ii.xe4 fS (1 2 ... dxe4 1 3 dxe5 e3!? 14 fxe3 is perhaps a slight advantage for White; the knight has a freer game than the bishop on c8) 1 3 ii.g2 ttJg4 1 4 ttJd2 f4 1 5 h3 ttJh6 1 6 ttJf3 fxg3 1 7 fxg3 ttJfS 1 8 ttJe5! and White is better. V.Mikhaievsky-Rechlis, Israel 1 999 saw White continue with the weaker 1 5 .l:tae 1 ?!, when 1 5 ... fxg3 1 6 hxg3 'ii'e 8 1 7 ii.f3 'ifh5 18 ii.xg4 'ii'xg4 19 ttJf3 'iVh3 20 ttJe5 .l:tfS 21 g4 l:txe5 22 dxe5 'ii'xg4+ led to a draw. Instead of 9 ... ttJe4 the encounter A.Petrosian-Bischoff, Gennany 1 999 went 9 ... a5 10 ttJbd2 b4 1 1 e4 ii.a6 12 l:f.el ttJe8 1 3 e 5 ttJc7 1 4 h4 and White had the superior prospects. There followed 1 4 ...ttJb5 1 5 ii.b2 lte8 1 6 ttJf1 'iib 8 1 7 ttJe3 'ilka7 1 8 ttJg5 ttJfB 1 9 f4 ii.c8 20 ii.f1 h6 21 ttJf3 ii.d7 22 fS with a menacing attack. 9 .i.b2 9 ttJc3 is premature in view of 9 ... ii.a6, when the c-pawn is genuinely hanging. After 1 0 ttJd2 b5 1 1 l:f.dl l:tb8 1 2 e3 bxc4 1 3 bxc4 'ii'a 5 14 ttJb3 'ikc7 1 5 c5 e5 1 6 e4 dxe4 1 7 ttJxe4 ttJd5 1 8 dxe5 ttJxe5 1 9 ii.b2 .l:tfe8 20 ii.d4 ii.fB Black had solved his opening problems in Bernard-Sosonko, 1 999. 9 .i.a6 Black needs to place the bishop on a6 if he wants to see .l:tfd 1 . After 9 ... ii.b7 10 ttJc3 l:f.c8 White has two attractive alternatives to the standard l l lIfd 1 . 1 1 lIadl 'ii'c 7 1 2 e 4 ttJxe4 1 3 ttJxe4 dxe4 1 4 'iVxe4 c5 1 5 d5 ii.f6 (1 5 ... ttJf6 1 6 1Vc2 exd5 1 7 ii.e5! 'ii'd 8 1 8 ttJg5 g6 1 9 h4 gave White a strong attack in Geller-Ciric, Ober­ hausen 1 961) 1 6 'it"c2 exd5 1 7 cxd5 and now Hausner-Kotan, Prague 2000 went 1 7 .. .'ii d 6 1 8 ii.c1 b5 19 ii.f4 'i!ib6 20 l:f.fe l l:f.ce8 21 d6

l:txe1+ 22 l:txe1 h6 23 h4 l:td8 24 h5 with an advantage to White thanks to the d-pawn and the better king. Black's lot was much worse in Smejkal-Padevski, Smederevska Palanka 1 9 7 1 , when 1 7 ... ii.xb2 soon got him into trouble: 1 8 'ii'xb2 l:tcd8 1 9 d6 'iib 8 20 ttJh4! g6?! 21 l:tfel f6 22 ii.xb7 'it'xb7 23 l:te7 'ii'c 6 24 'ii'e 2 ttJe5 25 l:tc7! 'ii'a 8 26 f4 ttJf7 27 'ii'e 7 1 -0. Vokac-Hrtanek, Karvina 2001 saw White play the immediate 1 1 e4 dxe4 12 ttJxe4. Then 12 ... c5 1 3 ttJxf6+ ii.xf6 14 ttJg5 ii.xg5 1 5 ii.xb7 l:tc7 1 6 ii.e4! g6 1 7 d5 exd5 1 8 cxd5 ii.f6 1 9 d6! is a good illustration of the kind of problems Black can quickly run into, Black avoiding trouble with 1 2 ... ttJxe4 1 3 1Vxe4 ttJf6 1 4 'ii'c2 c 5 1 5 dxc5 ii.e4 (1 5 ... ii.xc5?? 1 6 ttJg5! and Black can resign) 1 6 'iIfc3 ii.xc5 1 7 I:tfe 1 ii.a8 1 8 ttJe5 ii.xg2 1 9 c;to>xg2 'ii'e 7 20 'ilff3 when White had a slight plus. Again it is the c6-square that is Black's main problem here. 1 0 J:!.d 1 J:!.c8

. . .

26

1 1 lLlc3 ? ! The stronger 1 1 ttJbd2 leads t o play along the lines of Game 1 4. 1 1 . . . "fic7 An alternative is l 1 ...dxc4!? 12 ttJe5 cxb3 1 3 axb3 ttJb8, when Lputian prefers Black. Perhaps this is a bit optimistic, but Black is clearly okay. 1 2 e4 It might be time for 12 cxd5 cxd5 1 3 l:f.ac1

C lo s e d C a t a la n : 4 . . � e 7 5 � g 2 0 - 0 .

'iVb8 with complete equality. 1 2 . . . dxe4 1 3 'ilfe2 b5 1 4 bxe4 b4!

White clearly has some problems here. 1 5 ttJb 1 ? ! Necessary is 1 5 ttJa4 when, after 1 5 ... 'iVa5 1 6 a3!? b3 1 7 ttJc3 .l::t fd8, it is difficult to say who is better and why. 1 5 . . . ttJb6 1 6 ttJbd2 ttJa4 1 7 .l:!.ab 1 e5! White's centre collapses. Black's control over the dark squares affords him the lead. 1 8 J:rde 1 J:rfd8 1 9 dxe5 ttJxb2 19 ... ttJd7 20 .ll. a l ttJdxc5 favours Black. 20 J:rxb2 'ilfxe5 21 .l:!.be2 ttJd7 22 ttJb3 'ilfe7 23 ttJfd4 ttJe5 24 �f 1 .l:!.d7 25 ttJb5 'it'b8 26 'ilfe3 J:red8 27 'it>g2 h6 27 ... ttJxc4!? 28 .ixc4 .ll. x b5 29 .ll. x b5 'ii'x b5 30 .l::t c 7 a6 is stronger, when White has very little compensation for the pawn. 28 ttJ5d4 .l:!.xd4!

Black completely annihilates the enemy's

defence o f the light squares. 29 ttJxd4 ii.e5 30 J:rd 1 'iWd6 The accurate move. After 30 .. :iVb6 White has 31 ttJxe6!!, escaping with a draw (31 .l::tc d2?! .ll. x c4 leaves White in big trouble), e.g. 3 1 ....i.xe3 (3 1 ....l::tx dl 32 ttJxc5 is far from clear) 32 .l::t xd8+ �h7 33 ttJf8+ �g8 34 ttJe6+! with a draw by repetition. 31 J:red2 'it'b6 32 f4 ttJxe4 33 �xe4 ii.xe4 Black has definite compensation for the exchange, with two strong bishops against a poorly placed rook and a pinned knight. 34 'it>h3 as 35 e5 ii.a6 36 'iWe4 ii.b7 37 'ilfe3 a4 38 f5?! White is getting desperate. A better try is 38 g4!? a3 39 f5 with counterplay. 38 . . . exf5 39 e6 ii.e4 40 exf7 + 'it>xf7 4 1 'it'e2 g 5 ! White's king i s i n much more trouble than its opposite number. 42 g4 ii.xd4 43 'ilfe4+ ii.d5 44 'ilfxd4 fxg4+ 0-1

Game 1 1 Beliavsky-M itkov

Panormo 200 1 1 d4 d5 2 e4 e6 3 ttJf3 ttJf6 4 g3 ii.e7 5 ii.g2 0-0 6 0-0 ttJbd7 7 'iWe2 e6 8 b3 b6 9 J:rd 1 ii.b 7 1 0 ttJe3

1 0 . . .J:re8 Also possible is 1O ... c5!?, e.g. 1 1 cxd5 exd5 27

Th e C a t a la n

1 2 dxc5 �xc5 1 3 �b2 l:tc8 1 4 e 3 l:te8 1 5 'ii'f5 'it'e 7 1 6 l:!.ac1 tiJ f8 1 7 tiJd4 g6 1 8 it'g5 �xd4 1 9 exd4 �g7 20 �f1 h6 21 'ii'd2 tiJ8h7 and Black was no worse in Salov­ Bauer, Enghiens-Ies-Bains 1 999. 1 1 e4 dxe4 Following up ".�b7 with 1 1 ...�a6 looks illogical, and after 1 2 �f4 dxc4 1 3 tiJd2 c5 1 4 d 5 White i s able to generate an initiative. Korchnoi-M.Gurevich, Antwerp 1 995 con­ tinued 1 4".e5 1 5 �g5, when Korchnoi gives 1 5".tiJxd5 1 6 tiJxd5 �xg5 1 7 tiJf3! �b5 1 8 �h3 .ltc6 1 9 �xd7 .ltxd7 20 tiJxe5 �e6 21 tiJxc4 as a shade preferable for White. In Odendahl-Tiviakov, Dieren 2001 Black elected to capture on d5, but 1 4".exd5 1 5 e5 caused problems: 1 5".d4 (1 5".tiJg4 1 6 tiJxd5 tiJgxe5 1 7 tiJxc4 �xc4 1 8 bxc4 gives White good control over the light squares and am­ ple compensation) 1 6 exf6 �xf6 1 7 tiJce4 c3 1 8 tiJc4 �e7 19 tiJxc3! g5 (19".dxc3? 20 �h3 �b5 21 tiJe5 and White wins) 20 'ii' f5 ! gxf4 2 1 tiJd5 l:te8 22 �e4 and Black was under pressure. l 1 ...c5 12 exd5 exd5 1 3 �b2 dxc4 14 d5! gives White the advantage. The point is that 1 4".cxb3 1 5 axb3 �d6 opens the way for 1 6 l:txa7. Sulava-Ferretti, Verona 1 998 went 1 4".il.. d 6 1 5 bxc4 tiJe5 1 6 tiJe4 l:te8 1 7 tiJxd6 tiJxf3+ 1 8 �xf3 'ii'xd6 19 it'd3 Itcd8 20 a4 'iVf8 21 a5 11d6 22 axb6 axb6 23 .l:tel .l:txeH 24 l:he l , although White had all the chances. 1 2 lZJgS White wants to establish a knight on e4. Another possibility is 12 tiJxe4, when 1 2".tiJxe4 1 3 'ii'x e4 'ii'c 7 14 �f4 �d6 1 5 �xd6 it'xd6 1 6 c5! �c7 1 7 l:tac1 :fd8 1 8 'iVel �a6 1 9 a4 �b8 20 b4 bxc5 2 1 dxc5 tiJf6 22 'iVe5 llxdH 23 .l:!.xdl it'c8 24 tiJd4 was very nice for White Jo.Horvath­ Krizsany, Paks 1 996. A year later Black im­ proved in Cvitan-Borgo, Porto San Giorgio 1 997, 1 2".c5 1 3 tiJxf6+ �xf6 1 4 tiJg5 �xg5 1 5 �xb7 llc7 1 6 dxc5 �xc1 1 7 J::taxc 1 bxc5 1 8 il..g2 it'e 7 19 l:[d2 tiJf6 20 l:tcd 1 g6 21 'ii'c 3 h5 22 h3 l:!.fc8 23 f4 .l:.d7 24 llxd7 28

tiJxd7 2 5 .lt f3 providing White with a mod­ est lead. In this position Black has problems getting the knight to d4, which is the only good square available. Nevertheless Black should be able to hold with accurate play. 1 2 . . . h6 1 2".c5 1 3 dxc5 �xc5 1 4 tiJgxe4 tiJxe4 1 5 tiJxe4 favours White. 1 3 lZJgxe4 lZJxe4 1 4 lZJxe4 lZJf6

1 S cS! A thematic solution. The d5-square is not so important. Instead White has control over d6 and Black's queenside is under a bind. Less troubling for Black is 1 5 �e3 l:!.c7 1 6 IId2 lld7 1 7 tiJxf6+ �xf6 1 8 :adl it'e7 1 9 c 5 l:tfd8 20 a4 e 5 2 1 dxe5 l:txd2 22 l:txd2 'iVxe5 23 h4 l:[xd2 24 it'xd2 bxc5 25 'ifa5, Sulava-Inkiov, Nice 200 1 , with the smallest of edges for White - if anything. 1 S " .lZJdS 1 5".tiJxe4 16 .ltxe4 bxc5 17 dxc5 'ifc7 1 8 �b2 .l:!.cd8 1 9 a3 gives White a slight but enduring advantage in view of the c6-pawn and, in tum, the bishop on b 7. 1 6 a3 as 1 7 i.f1 1kc7 1 8 i.d2 J:!.fd8 1 9 b4 axb4 20 i.xb4 J:!.a8?! 20".tiJxb4!? is necessary, when 21 axb4 l:ta8 lirnits White's lead. 21 cxb6! Now White occupies the c5-square and starts pushing his a-pawn - a perfect reward for his game-plan. 2 1 " .'ii'x b6 22 i.xe7 lZJxe7 23 lZJcs i.c8

C lo s e d C a t a la n : 4

.

.

.

iL e 7 5 iL g 2 0 - 0

quick central expansion with e2-e4 while simultaneously monitoring the c4-pawn. S . . . b6 8 ... bS 9 cS should be to White's advantage, this version being even better than with 8 b3, which is quite useless here. Black has tried a couple of responses. 9 .. :iVc7 to ttJb3 eS?! 1 1 ttJxeS ttJxeS 12 il.. f4 ttJfg4 13 e4! dxe4 14 il.. x e4 f6 (14 ... f5 1 5 il..g2 il.. f6 1 6 1Iaei i s nice for White) 1 5 dxeS fxeS 1 6 il.. x h7+ \t?h8 1 7 il.. d2 il.. e 6 was the course of Botvinnik-I.Rabinovich, Leningrad 1 938, when 1 8 f3 ttJf6 1 9 il.. f5 il.. c 4 20 i.d3 il.. e 6 21 il.. a S would have gained White a free pawn. In the event of 9 ... aS 1 0 e4 Black should play to... dxe4 rather than drop back to e8. For example in Damljanovic-Semkov, Saint John 1 988 Black was too cramped after 1 0 ... ttJe8 1 1 eS, creating breathing space with l 1 ...f6 1 2 exf6 gxf6 1 3 lIel ttJg7 failing to improve his chances: 1 4 a4 b4 1 5 ttJb3 l:te8 1 6 i.h6 ttJrn 1 7 ttJh4 'ii'd7 1 8 l:te3 i.d8 1 9 l:tael 'iV f7 2 0 il.. xg7 it'xg7 21 f4 i.d7 22 il.. h 3 and White was doing very well. Return­ ing to 1 0 ... dxe4, Gulko-Campora, Biel 1 987 Game 12 continued 1 1 ttJxe4 ttJxe4 1 2 'iVxe4 l:!.a6 1 3 Comas Fabrego-Nogueiras l:!. e 1 l:!.e8 1 4 h 4 ttJf6 1 5 'it'c2 ttJdS 1 6 a4 b4 1 7 ttJeS 'iVc7?! 1 8 il.. e 4 g6 1 9 hS i.f6 20 hxg6 Havana 1999 '-------.. hxg6 21 il.. xg6! fxg6 22 'iVxg6+ wrn 23 l:te4 \t?e7?! (23... i.xeS was the last try) 24 i.gS 1 d4 d5 2 e4 e6 3 ttJf3 ttJf6 4 g3 iLe7 5 lIrn 25 l:!.f4! and White won. Gulko gives iLg2 0-0 6 0-0 ttJbd7 7 fie2 e6 S ttJbd2 1 7 ... i.f6 as the improvement in an anyway difficult position. He believes that after 1 8 il.. f1 il.. x eS 1 9 l:txeS White has a small but enduring advantage. It is often the case in the Catalan that the enduring nature of White's positive stock is the more significant factor from a practical point of view. 9 e4 iLb7 This seems wrong here. After 9 ...i.a6 to eS ttJe8 1 1 l:!.e1 ':c8 1 2 i.f1 ttJc7 1 3 b3 dxc4 1 4 ttJxc4 ttJdS 1 5 'ilVe4! White was better in Cifuentes-Flear, Polanica Zdroj 1 992, which continued I S ... bS 1 6 ttJe3 l:!.e8 1 7 h4 ttJrn 1 8 ttJxdS! cxdS 1 9 'iVg4 l:tc6 20 hS f5 21 exf6 I n this game White elects t o execute a

24 a4 White is close to being strategically win­ rung. 24 . . .11a5 24 ... ttJdS 25 l:tdbl 'it'c7 26 ttJb3 i.a6 27 i.xa6 l:!.xa6 28 ttJcS l:!.a7 29 as! and the a­ pawn makes progress. The main point is that 29 .. .l:has 30 l::tx aS 'ii'x aS 31 ttJb7 'iVc7 32 ttJxd8 'ii'x d8 33 it'xc6 is decisive. 25 ttJb3 l::ta d5 26 a5 fia7 27 a6 iLd7 2S iLe4 As we can see, Black should still have tried to get rid of the bishops. 2S . . . l::td 6 29 ttJe5 iLeS 30 fib2 ttJd5 Beliavsky gives 30 ... ttJf5 3 1 ttJb7! il.. x b7 32 axb 7 'it'bs 33 l:taS and White wins. 31 l::td b1 ! This time White should refrain from 3 1 ttJb7? a s 3 1 ...il..x b7 3 2 axb7 'it'b6 3 3 .l:tdbl 'ili'xb2 34 .l:txb2 l:!.b8 is less clear. 31 . . . ttJe7 32 fibS ! The blockade of the pawn is removed, and with it the last line of defence. 32 . . . fixbS 33 l::t x bS l::tx d4 34 a7 l::tx e4 35 ttJa6! ttJxa6 36 l::t x a6 l::te 1 + 37 'it>g2 1 -0

29

Th e Ca t a l a n

..Il.xf6 2 2 ..Il.d2 'iNd6 2 3 a 3 'it'd7 2 4 ..Il.d3 and Black's future was getting darker thanks to the lack of play and White's presence in the centre and prospects of a kingside offensive. 9 ... dxe4 10 lLlxe4 ..Il.b7 is solid but ive. Black needs to pay special attention to the possibility of the clamping c4-c5, as in the previous main game. Raetsky-Whiteley, Cap­ pelle la Grande 1 997 went 1 1 �d 1 ttJxe4 1 2 'ii'x e4 ttJf6 1 3 'ii'c2 �c8 1 4 ..Il.f4 ..Il.d6 1 5 .i.xd6 'ii'x d6 1 6 c 5 it'c7 1 7 lIac1 �fd8 1 8 ttJe5 ttJd7 1 9 cxb6 'it'xb6 20 ttJc4 it'c7 21 �a4 and White was on top. In Krasenkow­ Azmaiparashvili, Groningen 1 997 White must have been in the driving seat after 1 1 ...'it'c8 1 2 lLlxf6+ lLlxf6 1 3 c5, but the sub­ sequent 1 3. .. lLld5 14 ..Il.g5 f6 1 5 ..Il.d2 bxc5 1 6 dxc5 e 5 1 7 b4 a s 1 8 bxa5 .i.d8 1 9 lLlh4 f5 20 ttJf3 ..Il.f6 21 �abl e4 22 ttJg5 .i.a6 23 �b6!? was unclear... 10 eS lLle8 1 1 b3 1 1 cxd5 is an alternative. Then 1 1 ...cxd5 is a somewhat ive option - where is the knight going from e8? Sargissian-Rivas, Ubeda 2001 continued 12 �el l:tc8 1 3 it'a4 as 1 4 lLln .i.a6 1 5 h4 .i.d3 16 .i.g5 h6 1 7 ..Il.xe7 'iVxe7 1 8 �ac1 ttJc7 1 9 �e3 ..Il.b5 20 'it'dl lLla6 21 �ec3 �xc3 22 �xc3 ttJdb8 23 g4 g6 24 'iid2 �g7 25 'ii'f4 and White had more territory, chances to break through on the kingside and control of the only open fIle. Black does better to follow the example set in Buhmann-Vaganian, 2000, when l 1 ...exd5!? 12 �e1 ttJc7 1 3 ttJn lLle6 14 lLle3 g6 1 5 h4 c5 generated suffIcient counterplay, 1 6 ttJg4 cxd4 1 7 ..Il.h6 �e8 1 8 'it'd2 lLldc5 1 9 ttJxd4 it'd7 20 ttJxe6 lLlxe6 resulting in a double-edged game. 1 1 . . . lLlc7 After 1 1 ...�c8 12 ..Il.b2 c5 13 �ac1 cxd4 14 ttJxd4 Black was struggling in Hertneck­ Stangl, Austria 1 997. There followed 1 4 ...'it'c7 1 5 :fe 1 'itb8 1 6 l\ib 1 lLlc7 1 7 a3 'Yia8 1 8 b4 dxc4 1 9 ..Il.xb7 'iixb 7 20 lLlxc4 ttJd5 21 'iid 3 �fd8 22 ttJb5 and White dominated proceedings. 30

1 2 ..tb2 c S 1 3 �fe 1 'iYb8 1 4 dxcS lLlxcs 14 ... ..Il.xc5 1 5 ttJg5 g6 1 6 .l:tac1 favours White. Black has no active plan and ... dxc4 exposes both d6 and f6 to the white knights. 1 S lLld4 as! ? Perhaps 1 5 . . .dxc4!? makes life easier for Black, e.g. 1 6 ttJc6 �e8 1 7 ttJxe7+ 'Yixe7 1 8 ..Il.xb7 lLlxb 7 1 9 lLlxc4 and Black stands no worse. 1 6 a3 �c8 1 7 b4 axb4 1 8 axb4 lLld7 1 9 bS �xa 1 20 ..txa 1 lLlcs Also possible is 20 ... ..Il.c5 21 'iib 3 dxc4 22 lLlxc4 .i.xg2 23 �xg2 lLld5 24 lLlc6 when the knights are of equal worth. Black will proba­ bly play ...lLld7-b8 very soon in order to ad­ dress White's outpost. 21 'i!Vb 1 dxc4 22 ..txb7 'i!Vxb7 23 lLlxc4 lLldS 24 lLlc6

24 . . . �a8 2S 'i!Vc2 ..tf8 26 ..td4 White would be worse after 26 ttJd6?! ..Il.xd6 27 exd6 'it'd7 28 ..Il.e5 f6 29 ..Il.d4 'i¥xd6 because the bishop is not very effec­ tive and the knight on c6 is basically out of play. 26 . . . lLld7 27 �a1 �xa 1 + 28 ..txa 1 ..tcS %-%

Game 13 Raetsky-Landenbergue

Scuol 200 1 1 d4 dS 2 c4 e6 3 lLlf3 lLlf6 4 g3 ..te7 S ..tg2 0-0 6 0-0 lLlbd7 7 'ii'c 2 c6 8 lLlbd2

C lo s e d C a t a la n : 4

b6 9 b3 iLb 7 1 0 e4 White needs to get going. After 10 i.b2 c5 l 1 l:1ael l:tc8 12 'iVbl dxc4 1 3 ttJxc4 cxd4 14 ttJxd4 ii.xg2 1 5 'it>xg2 a6 16 ttJe3 l:txel 1 7 l:txel 'ifa8+ 1 8 f3 .l:tc8 Black had solved all his problems in Chiburdanidze-P.Nicolic, Linares 1 988. 10 .. J:tc8 1 1 iLb2 "*c7 l 1 ...dxe4 1 2 ttJxe4 transposes to 8 b3 b6 9 i.b2 i.b7 1 0 ttJc3 lIc8 1 1 e4 dxe4 1 2 ttJxe4, covered in Game 1 0, note to Black's 9th move. l 1 ...c5 1 2 exd5 exd5 1 3 dxc5 dxc4 1 4 b4!? was the interesting course of Lindgren­ Rasmussen, Budapest 2000, when Black went astray with 1 4 ... i.a6?! 1 5 c6! c3 (1 5 ... 1:txc6 1 6 ttJd4 I:!.c7 1 7 b5! i s the problem) 1 6 'ifxc3 ttJb8 17 ttJd4! i.xfl 1 8 ttJxfl and White had more than enough compensation for the exchange - 1 8 ...l:te8 1 9 ttJe3 ttJa6? 20 lId I ! i.xb4 2 1 �c4 b 5 22 ttJxb5 'iVb6 2 3 i.xf6 gxf6 24 ttJd5 proved decisive. Black should play 1 4 ... bxc5 1 5 b5 'it'c7 1 6 ttJxc4, although White has a very pleasant structural advan­ tage. The alternative 1 3 ... ttJxc5 1 4 ttJg5 g6 1 5 'i'c3!? h6 1 6 ttJh3 seems to help White. In­ deed in Wilhelmi-Lau, 2000 White emerged from 13 ... i.xc5 14 l:tadl J:te8 1 5 'iffS 'iIi'c7 1 6 ttJd4 a6 1 7 cxd5 g6 1 8 'iVbl ttJxd5 1 9 ttJe4 ii.f8 20 lHe 1 b5 21 a3 b4 22 ttel with the better game. Instead of the trade on c5 White has an interesting idea in 1 3 l:tfdl dxc4 1 4 ttJxc4 b5 1 5 d5!?, when Wojtkiewicz-Shabalov, USA 2000 continued with multiple exchanges: 1 5 ...ttJxd5 1 6 ttJce5 ttJ7f6 1 7 ttJg5 h6 1 8 ttJexf7 l:1xf7 1 9 ttJxf7 �xf7 20 ii.xf6 i.xf6 21 l:txd5 ii.xd5 22 l:1dl ii.d4 23 'it'fS+ �g8 24 i.xd5+ and a draw was agreed as White had failed to prove an advantage. 1 2 l::ta d 1 l::tf d8 1 3 l::tfe 1 "*b8 1 4 "*b 1 An interesting alternative was the ambi­ tious 14 h4 h6 1 5 e5 ttJe8 1 6 c5!? bxc5 1 7 i.a3 ttJc7 1 8 dxc5 as 1 9 g4 'it'a7 20 g5 ttJb5 21 i.b2 'it'xc5 22 'ti'd3 ..ta6 23 gxh6 ttJc7 24 'iVb 1 gxh6 25 ttJd4 where White has com-

.

.

.

iL e 7 5 iLg2 0 - 0

pensation for the pawn through his active play, Vaulin-Dobrovolsky, Harkany, 1 994. 1 4 . . . h6 1 5 a3 dxe4 1 6 tL\xe4 c5 1 7 d 5 ! ? exd5 1 8 tL\c3! ? l::te 8 1 9 tL\xd5 tL\xd5 20 cxd5 l::t c d8

21 l::t x e7 ! ? Entertaining stuff. The idea i s to enhance the power of the dark-squared bishop. With 21 'ikfS ..tf6 22 .l:txe8+ .l:txe8 23 i.el White guarantees a modest plus. 21 . . . l::tx e7 22 tL\h4 l::te 2! Black has to be very careful here. The fol­ lowing variations help demonstrate what lurks in the shadows. 22 ...l:tee8?! 23 ttJfS f6 (23 ... ttJe5 24 f4 looks dangerous) 24 ttJxh6+! gxh6 25 1Wg6+ �f8 (25 ... �h8 26 'it'xh6+ �g8 27 'it'g6+ �h8 28 ..th3! 'it'd6 29 ii.xd7 and White wins) 26 d6! i.xg2 27 ..tel ! and Black is savagely mated. 22 .. . f6!? is preferable, when 23 ttJfS .l::[ f7 offers Black solid defensive prospects on the kingside. An interesting approach is 24 "iVe4!? ttJf8 25 'ii'g4 �h7 26 'ifh5 'ii'c 7 27 i.e4 with a strong attack, while there is also 24 llel , e.g. 24 ...ttJ f8 2 5 ttJe7+ �h8 (25 ...l:1xe7? loses instantly to 26 .l:txe7 i.xd5 27 i.xf6! gxf6 28 'ii' fS and Black is soon mated) 26 ttJg6+ ttJxg6 (after 26 ...�g8 White is not obliged to go for the repetition) 27 'ili'xg6 I:!.e8 (27 ...'it>g8 28 i.e4 looks dangerous, although after 28 ... �f8 29 'iih7 .l:!.e7 30 i.xf6 gxf6 3 1 'ifh8+ �f7 3 2 'i!Vh7+ White has only a draw) 28 .l:1e6!? 31

Th e C a t a l a n

The surpnsmg point. After 27 ':'xe 1 .l:!.xeH 28 1t.fl 'itd6! the position is less clear, but now the king can no longer escape via e8. 27 . . . �c8 27 .. :iWd6 28 'ii'g4+ 'it>fB 29 'ii'g7 mate (e8 is not available) . 28 'iiVh S 'iiVd 6 29 l:txe 1 l:txe 1 30 'iiVx h6 'iiVx f6 31 'iiVx f6 �h3 3 1 ...1t.a6 32 d6 .i.xfl 33 h4 also wins smoothly for White. 32 'iiVg S + 1 -0

Game 14 28 ... l'.:tfe7 ? ? loses t o 2 9 1t.e4 'it>g8 3 0 l:txf6! in view of 30 ....:.xe4 31 l:tfB+ 'it>xfB 32 'ii'xg7 mate, which leaves 28 .. .lhe6 (28 ... :ee7!?) 29 dxe6 lIe7 30 1t.xf6! 'iVd6! (30 ...gxf6? 3 1 'ikxf6+ l:tg7 3 2 e 7 and White wins) 3 1 1t.xb7 (31 1t.xe7?? 'ii'd H 32 1t.fl 'itf3 and suddenly it is Black who wins) 3 1 ...ltxb7 32 e7 'it'dH 33 'it>g2 'iVd5+ 34 f3 'iVd2+ 35 'it>h3 l:he7 36 1t.xe7 11i'd7+ with a drawn ending. 23 lLlfS lLlf6? This allows a combination that is not too difficult to find. Consequently Black should play 23 .. .l::txb 2 24 'iVxb2 'it'e5 25 'ii'x e5 tLlxe5 26 l:el or 23 ... tLle5!? 24 f4 l:Ixg2+! 25 'it>xg2 tLlg6 26 'ii'e 4!? (26 1t.xg7? l:!.xd5 27 llxd5 'ii'd 8! and Black wins) 26 ... 'ii'a 8 27 tLle3, in both cases with a slightl edge for White. 24 lLlxh6 + ! gxh6 2S �xf6 l:tde8 26 'iiVf S l:re 1 +

27 �f 1 ! ! 32

Marin-Berescu

Iasi 1999 1 d4 dS 2 c4 e6 3 lLlf3 lLlf6 4 g3 �e7 S �g2 0-0 6 0-0 lLlbd7 7 't!Vc2 c6 8 lLlbd2 b6 9 b3 �a6 1 0 l:td 1 1 0 e4 .l:!.c8 1 1 l:te1 is a different approach, against which Black should have sufficient resources. 1 1 ...c5 1 2 exd5 exd5 1 3 1t.a3 l:te8 1 4 cxd5 tLlxd5 1 5 'iff5 tLl7 f6 1 6 dxc5 1t.xc5 17 1t.xc5 l:txc5 simply equalised in Tosic­ Kosic, Trebinje 2001 , but 1 6 ... g6!? 1 7 'ii'h 3 1t.xc5 1 8 1t.xc5 l:txc5 19 .l::tx e8+ 'ii'x e8 20 ]:tel 'ii'd 8 21 'ii'h4 tLlc3!? is unclear according to Kosic. Black seems fme in this position. Also possible is 14 ... cxd4!? 1 5 1t.xe7 lIxc2 1 6 1t.xd8 I:txd8 1 7 :ec1 l:tdc8 when Black, again, looks no worse. Perhaps White could consider 1 8 tLlc4!? .l:txcH 1 9 .l:!.xc 1 b5 20 tLlxd4 bxc4 21 tLlc6 'it>fB 22 bxc4 with a complex struggle ahead. 1 0 . . . l:tc8 10 ... b5 led to a slight advantage for White in Filippov�liang Chong, Shenjang 1 999 after 1 1 c5 b4 12 e4!? tLlxe4 1 3 tLlxe4 dxe4 1 4 'iVxe4 tLlxc5 1 5 'ii'c 2 tLld7 1 6 tLle5 tLlxe5 1 7 dxe5 'ii'c 7 1 8 'ii'x c6 'ii'x c6 1 9 1t.xc6 .l:!.ac8 20 1t.e4 1t.b5 21 1t.e3. 1O ... c5 1 1 e4 dxc4 12 tLlxc4 cxd4 13 tLlxd4 .l:!.c8 1 4 e5 tLld5 1 5 'ii'b 2! secures White a small plus. Black, as always, has problems with the light squares. Poluljahov-Feigin, Kakhovka 1 997 went 1 5 ... 1t.xc4 1 6 bxc4

C lo s e d C a t a la n : 4

l:!.xc4 1 7 .1i.xd5! exd5 1 8 lbf5 .1i.b4 1 9 l:txd5 .1i.c3 20 'ii'e2 'ii'c 7 21 .1i.b2 .1i.xb2 22 'iVxb2 lbc5 23 lbd6 'ii'c 6 24 J::tadl and White was firmly in control. 1 1 e4 c5 1 1 ...'ii'c 7 1 2 e5 lbe8 1 3 lbfl 'itb8 1 4 .1i.g5 .1i.xg5 1 5 lbxg5 g6 1 6 f4! is clearly better for White according to Beliavsky. 1 2 exd5 exd5

.

.

.

iL e 7 5 iL g 2 0 - 0

l:txd8 �cxd8 24 .1i.g5 lIb8 25 'ii'd l :£1+ 26 'iit h3 .1i.xa 1 27 'iVxa 1 .1i.f1+ 28 'iitg4 .1i.e2+ 29 'iit h4 l:txh2+ 30 .1i.h3 l:tb5! with a balanced game. In Delchev-L.B.Hansen, Istanbul, 2000 Black chose the alternative capture: ( 1 3 lbfl) 1 3. .. dxc4 1 4 d5 lbe8 15 bxc4 lbd6 16 lbe3 l:!.e8 1 7 .1i.fl b5 1 8 ltbl bxc4 1 9 'ii'a4 .1i.b7 20 .1i.d2 l:ta8 21 .1i.a5 lbb6 22 'ii'c 2 .1i.c8 23 lbe5 and White was doing well. Stohl prefers 1 4 ... cxb3 1 5 axb3 .1i.xfl 1 6 .1i.xfl 'ii'c 7, al­ though he still likes White after 1 7 .1i.f4 .1i.d6 1 8 .1i.xd6 'ii'x d6 1 9 �xa7 .I:ta8 (1 9 ... lbxd5? 20 .1i.c4 lb7f6 21 lbg5 g6 22 .1i.xd5 lbxd5 23 lbe4 'itb8 24 .l:!.d7 lbb4 25 'iVc3 and White wins) 20 l:txa8 .I:txa8 21 lbh4 g6 22 lbg2. 1 3 cxd4 Black's alternatives have had mixed for­ tunes. 1 3. .. dxc4 14 d5! cxb3 15 axb3 i.e2 16 .I:tel .1i.b5 17 lbh4 l:!.e8 18 l:txa7 .1i.d6 19 �aal lbe5 20 lbc4 lbxc4 21 bxc4 .1i.d7 22 lbe left White slightly better in Cvitan-Godena, Gen­ eve 1 996. After 13 ... b5?! 14 cxd5 c4 White used to respond with 1 5 bxc4, but stronger is 1 5 lbe5!, when 1 5 ...lbxd5? 1 6 lbxfl! 'ii'a 5 does not work for Black (as it normally does) be­ cause the b5-pawn obstructs the queen. In­ stead 16 ... lbb4 17 lbxd8 lbxc2 1 8 lbc6 leads to a clear advantage for White. This leaves the lesser evil 1 5 ... lbb6! 1 6 lbc6 'ii'c 7 1 7 bxc4 bxc4 1 8 lbe4 lbfxd5 1 9 lbxe7+ 'ii'x e7 20 lbc5 when White's outlook is the brighter, with strong knight on c5. Ernst-Aagaard, Groningen 1 998 continued 1 4 ... .1i.b7 1 5 lbg5 h6 1 6 d6 .1i.xg2 1 7 dxe7 'iVxe7 1 8 'iitxg2 hxg5 1 9 'ii'd 3 c4 20 bxc4 bxc4 21 lbxc4 'ii'b4 22 lIdc1 .I:txc4 23 'ii'x c4 'ii'xb2 24 .l:tabl 'it'd2 25 l:tc2 'iVa5 26 .l:.b5 'iVel 27 'iVe2 'iVaI and a draw was agreed. However, White has an improvement in 1 5 'it'd3! c4 1 6 bxc4 bxc4 1 7 lbxc4 .1i.xd5 1 8 lbfd2 with advantage. It would appear that 13 ... b5 is inadequate. 1 3. .. l:te8 1 4 'iVfS g6 1 5 'it'h3 .1i.fS (1 5 ... b5?! is poor due to 16 lbg5! with the threat of . . .

This is an important position for the Cc. White has more than one option. 1 3 iLb2

1 3 'ii'fS g6 (13. ..l:te8 14 .1i. b2 transposes to 1 3 .1i.b2 lIe8 14 'ii'fS) 14 'ii'h 3 1:[c7 1 5 .1i.b2 .1i.c8 1 6 'it'h6 lbg4 1 7 'ii'f4 lbdf6 1 8 lbe5 cxd4 19 .1i.xd4 .1i.d6 20 cxd5 lbh5 21 'ii'e4 .1i.xe5 22 .1i.xe5 l:te7 was unclear in Beshu­ kov-Fomichenko, Krasnodar 1 999. Particularly interesting is Sulava's 1 3 lbfl !?, when Sulava-Atalik, Cappelle la Grande 2000 continued 1 3. .. cxd4 1 4 lbxd4 b5 (14 ... .1i.c5?! is too optimistic - 1 5 lbe3 dxc4 1 6 lbc6 sees White generate an initia­ tive, Luther offering an example in the line 1 6 ... 'iVc7 1 7 lbxc4 lbg4? 1 8 .1i.f4! lbx£1 1 9 'ii'x £1 .1i.x£1+ 2 0 'iitx £1 etc.) 1 5 lbe3 bxc4 1 6 lbxd5 lbxd5 1 7 .1i.xd5 cxb3 1 8 'ii'x b3 lbc5?! 19 'ii'e .1i.f6 20 .1i.a3 'iVa5 21 l:tab l ! .1i.xd4 22 Uxd4 .1i.e2 23 'iix e2 'iVxa3 24 'iVe7! 'iit h 8 25 i.xfl with a poor position for Black. Stohl offers the improvement 1 8 ... .1i.c5!, with the following instructive line: 1 9 lbe6!? .1i.x£1+ 20 'iitg2 fxe6 21 .1i.xe6+ 'iit h 8 22 lhd7 .1i.d4! 23

33

Th e C a t a l a n I

.Jtxd5, the subsequent 1 6 ... ttJf8 1 7 dxc5 bxc4 1 8 bxc4 .Jtxc5 1 9 cxd5 .Jtxf2+ 20 'it>h 1 .l:!.c2 21 .Jtxf6 'ii'xf6 22 ttJde4 being good for White) 1 6 cxd5 .te2 1 7 l:te1 ttJxd5 1 8 l:tac1 with a complicated struggle ahead. 1 4 lZJxd4 b5 1 5 �f5 bxc4 1 6 bxc4

1 6 . . . g6! An important move. Black has been through a lot of suffering in this position. 1 6 ... 'iVb6?! 1 7 .l:!.ab 1 dxc4 1 8 ttJe4! g6 1 9 ttJxf6+ ttJxf6 2 0 'i!t'f3 lUe 8 2 1 ttJc6 'iYxc6 22 'iYxc6 l:txc6 23 .txc6 l:tc8 24 .Jtxf6 .Jtxf6 25 .Jte4 offers White good winning chances. 16 ... .Jtxc4?! 17 ttJxc4 .l:!.xc4 18 .txd5 g6 1 9 'iVg5 ttJxd5 20 'ii'x d5 ttJb6 21 ttJf5! gx f5 22 'it'e5 .Jtf6 23 'ii'xf6 'iVxf6 24 .Jtxf6 was seen in Nielsen-Danielsen, Randers 1 996, White eventually converting this favourable ending. 1 6 ... ttJb6?! 1 7 cxd5 ttJa4 (1 7 ... ttJbxd5 is met with 1 8 ttJe6! fxe6 19 'iVxe6+ 'it>h8 20 'iNxa6 and White is a pawn up for nothing) 1 8 l:Iab1 ttJxb2 1 9 l:txb2 .tc5 20 ttJc6 't'kd6 21 ttJe4 ttJxe4 22 .Jtxe4 g6 23 'ii f3 f5 24 .td3 .txd3 25 l:txd3 f4 26 g4 ::tce8 27 .l:!.e2 and White was on his way to the full point in Sorokin-Hoffman, Villa Martelli 1 997. 1 7 if'g5?! TIlls proves to be too risky as Black easily obtains compensation for the exchange. Su­ lava proposes the following line as a possible improvement: 1 7 'ii'h 3 dxc4 1 8 .Jtc3 ttJc5 1 9 ttJc6 l:txc6! 2 0 1i.xc6 ttJd3 21 ttJe4 'itb6! 22 'iYg2! ttJg4! 23 .l:!.ab 1 'ii'x c6 24 ttJf6+ 'it'xf6 25 34

.Jtxf6 .Jtxf6 with an assessment of 'unclear' as usual, further tests are required. 1 7 . . . dxc4 1 8 lZJc6 l:txc6 1 9 �xc6 'iYb6 20 �xd7 if'xb2 21 l:tab 1 'ifa3 21 ...'iVd4 is even stronger according to Marin. After 22 ttJf3 'iVxf2+ 23 'it>xf2 ttJe4+ 24 'it>g2 ttJxg5 25 ttJxg5 .txg5 26 l:td6 .tc8 Black has all the chances, while 22 'iVa5 .Jtc5 23 'ii'xa6 'iVxf2+ 24 'it>h 1 ttJe4! leaves White's king in serious trouble. 22 �b5 lZJe4? A tactical error. Black would have been fine after 22 ... c3! 23 ttJc4 'iic 5 24 'iWxc5 .txc5 25 l:tdc1 .Jtxb5 26 l:txb5 ttJe4 27 ttJe3 l:td8, retaining the ed pawn and with excellent compensation for the exchange. 23 lZJxc4 'iYf3 24 'iYe3 ii'xe3 25 lZJxe3 �xb5 26 lZJd5! �c5 27 l:txb5 �xf2+ 27 ...ttJxf2 28 l:tc1 .Jtd4 29 l:tc4 and wins. 28 Wg2 f5? After the more precise 28 ... h5 White still remains well on top after 29 ::tb7 .Jtc5 30 ttJe7+ Wg7 31 ttJc6. 29 l:tb7 l:tf7 30 lZJe7 + Wg7 31 .l:!.dd7 l:tf8 32 1ZJg8 + ! 1 -0

Game 15 Raetsky-Gattenloehner

Winterthur 2002 1 d4 d5 2 c4 e6 3 lZJf3 lZJf6 4 g3 �e7 5 �g2 0-0 6 0-0 lZJbd7 7 if'c2 c6 8 lZJbd2 b6 9 b3 �a6 1 0 �b2 l:tc8 1 1 e4 c5

Clo s e d C a t a la n : 4

In Kasparov-Huss, Zurich (simul) 1 987, 1 1 ...'iVc7 12 l:tfel 'iYb8 13 l!tac1 l::t fd8 1 4 'iYb 1 h 6 1 5 a 3 .i. b7 1 6 e5 ltJe8 1 7 cxd5! cxd5 18 .tfl earned White an advantage, albeit nothing spectacular. However, this soon snowballed, and after 1 8 ...l:txc1 1 9 l:txc1 l:tc8 20 b4 l:txc1 21 .i.xc1 �c8 22 b5 �c3 23 a4 g5?! 24 h3 ltJfS 25 .i.b2 �c8 26 .i.d3 White was clearly better. 1 2 exd5 exd5 1 3 'ii'f 5!? 1 3 l:!.fd 1 transposes to the previous game. An alternative is 1 3 .l:!.fe l , when Lobron­ Knaak, 1 998 continued 1 3 ... cxd4 14 ltJxd4 b5 1 5 .l:tadl .i.c5 1 6 ltJf5 'iVb6 1 7 ltJe 7+ .txe 7 1 8 l:txe7 bxc4 1 9 bxc4 dxc4 20 ltJe4 c3 21 .i.xc3 ltJxe4 22 .i.xe4 ltJf6 23 .i.f3 'tWc5 24 l:te3 lUe8 25 'iVd2, with a slight edge for White. 1 3 . . . 96 1 4 'ii'h 3

1 4 . . . cxd4 Black plans a knight manoeuvre. 14 ... h5 15 l:tfe1 cxd4 1 6 ltJxd4 ltJc5 was played in Raetsky-D.Frolov, Smolensk 2000, but this looks rather risky for Black. There followed 1 7 ltJf5! gxf5 1 8 .l:!.xe 7! 'fixe 7 1 9 'iVh4! Itc6 (19 ... ltJcd7 20 'fig5+ 'it>h7 21 �xf5+ 'it>h6 22 'fif4+ 'it>g6 23 ltJf3 ltJe4 24 l:tel f5 25 .i.h3 and Black is in for a rough ride) 20 �g5+ 'it>h7 21 'iVxf5+ 'it>g8 22 'i'g5+ 'it>h7 and a draw was agreed, but 23 cxd5 l:td6 24 b4! gives White a strong attack. In Rustemov-Nikolenko, Moscow 1 999 Black followed 14 ... l:tc7 1 5 dxc5 with the

.

.

.

ii.. e 7 5 ii..g2 0 - 0

faulty 1 5 ... bxc5?, finding himself a pawn down for nothing after 1 6 ltJg5 d4 1 7 ltJe6! fxe6 18 �xe6+ 'it>g7 19 'tWxa6. Gelfand­ Sorokin, Sochi 1 986 went 1 5 ...ltJxc5 1 6 'it'h6 lIe8 17 .i.e5 l:td7 1 8 .th3 ltJe6 1 9 l:tfe 1 dxc4 20 .i.xe6 fxe6 21 ltJe4 .i.b7 22 ltJfg5 .i.xe4 23 ltJxe4 cxb3 24 axb3 and White had a fonnidable attack. 1 5 ltJxd4 ttJc5 1 6 �ad 1 ttJd3 Kochiev-Shaposhnikov, St. Petersburg 1 996 is worth a look: 1 6 ... l:Ie8 1 7 ltJ2f3 ltJce4 1 8 ltJe5 'iVc7 1 9 ltJe6! 't\Vd6 (1 9 ... fxe6 20 'iVxe6+ 'it>g7 21 'i¥f7+ 'it>h8 [21 ...�h6 22 ltJg4+!] 22 .i.xe4 J:lfS 23 'iVxd5! ltJxd5 24 ltJf7+ �g8 25 ltJh6 mate) 20 ltJxf7! 'it>xf7 2 1 .txe4 ltJxe4 2 2 l:!.xd5 'i¥xd5 2 3 cxd5 and White was winning. 1 6 ... h5 1 7 ltJ2f3 ltJce4 1 8 ltJe5 �e8 1 9 .i.xe4 was the course o f Razuvaev-Lputian, Frunze 1 979, when 19 ... dxe4?! met with 20 ltJf5! gxf5 21 'iVxf5 lIc5 22 b4 l:!.c7 23 lId6! .tc8 24 'iVg5+ �h7 25 ltJd7! �xd7 26 lhd7 .txd7 27 .txf6 .txf6 28 'iVxf6 l:!.xc4 29 'iVe7 1 -0. Razuvaev gives 1 9 ...ltJxe4 20 l:tfe l with an attack for White. 1 7 ii.. a 1 �c5

1 8 ii.. x d5 ! ! A surprising move. Usually White does not surrender this bishop if he can help it, yet here he does not even bother to recapture on d5. In fact this was home preparation. In a previous game 1 8 f4? was played, but with little success. After 1 8 ... dxc4! 1 9 ltJc6 l:txc6! 35

Th e C a t a l a n

2 0 i.xc6 it'c7 21 it'g2? (21 i.f3 c3 2 2 tbe4 tbxe4 23 i.xe4 c2 is better) 21 ...i.c5+ 22 'it>hl tbg4 Black was winning in Raetsky­ Filippov, Smolensk 2000. The contrasting 1 8 i.h I !? l:te8 1 9 a3 i.fB 20 b4 ':c7 21 tbb5 also looks nice for White, but is of less relevance considering the effec­ tiveness of the text. 1 8 . . . ttJxdS 1 9 ttJe4! The key idea. White simply wants to re­ gain the knight for free. 1 9 . . . ttJf6 19 ... tb5b4 20 iVh6! l:te5 21 tbfS! gxfS 22 i.xe5 f6 23 tbg5! and Black is mated, while 1 9 ... .l:!.KC4 20 bxc4 i.xc4 21 'ii'h 6! tbe5 22 tbfS! works in the same way. 20 ttJxf6 + ? ! A n inaccuracy. The game could have been decided with 20 tbxc5! tbxc5 (20 ... bxc5 21 tbc6 it'd6 22 tbxe7+ it'xe7 23 :xd3 and Black must resign) 21 tbc6 'iNc7 22 tbxe7+ 'ii'x e7 23 'ii'h 4!, when White wins as follows: 23 ...'it>g7 24 nfel tbe6 25 .l:!.d7! it'xd7 26 it'xf6+ 'it>h6 27 it'h4 mate. 20 . . ..bf6 2 1 .l:[xd3 'ii'a 8 Black does not have any compensation on the light squares. After 2 1 ...i.b7 22 l:tfd1 it'e7 23 i.c3 l:th5 24 it'n it'c5 25 tbb5 i.xc3 26 tbxc3 White is a pawn up. 22 ttJe6 ! ?

22 . . .l::t h S Only move. After 22.. .fxe6 23 'iVxe6+ <j;g7 24 lId7+ White wins. 36

2 3 'ii'g 4 Also strong is 23 tbxfB!? l:[xh3 24 i.xf6 it'c6 25 l:td8 'it'xf6 26 tbe6+ 'iVxd8 27 tbxd8, with good winning chances in the endgame. 23 . . . .i.c8 24 .i.xf6 fxe6 24... i.xe6 25 it'f3 it'xf3 26 .uxf3 gives White excellent winning chances in the shape of a pawn and a powerfully posted bishop. 2S 'ii'd 4 eS

26 'ii'd 6! Not 26 i.xe5? i.h3, when White is close to being worse. 26 . . . .i.h3 27 f3! e4 28 l:td4

28 . . . e3? A blunder. I f 28 ... :f7 Black is struggling, but while 29 g4 l:lc5 30 lIfd1 exf3 31 "i'e6 £1+ 32 'it>x£1 .l:!.c8 33 'it>g3 'iVg2+ 34 'iith4 favours White, there remains work to do. 29 'ii'e 7 1 -0

Clo s e d C a t a la n : 4 . . . iL e 7 5 iLg2 0 - 0

Summary The evaluation of the gambit continuation 6 . ..lbbd7 7 ttJc3 dxc4 8 e4 (Game 3) has changed during the last decade. Earlier Black defended with 8 ... ttJb6, 8 ... c6 9 a4 as or 9 ... b6, but White has a strong initiative in the centre in all of these lines. Now Black has discovered 8 ... c6 9 a4 e5!? 1 0 dxe5 ttJg4 1 1 i.f4 'ifa5. Another modern idea is action on the queenside with 8 ... a6 9 a4 l:[b8 1 0 as b5, but this variation needs more practice before a safe conclusion can be made. After 7 ttJc3 c6 White seldom develops his queen on b3, often being forced to exchange on d5, when Black usually equalises quite comfortably. It seems to us that 8 'iVd3 has more going for it than 8 'ifb3. However, after 8 ... b6 White should not hurry with 9 e4 because 9 ... i.a6 1 0 b3 dxc4 1 1 bxc4 e 5 i s nice for Black. O n the other hand, recent victories b y G M Gleizerov have shown that after 9 .l:!.d1 i.a6 10 b3 1:[c8 1 1 e4 Black has problems. In the case of the move order 7 ttJc3 c6 8 b3 b6 9 i.b2 i.a6 (Game 6), then 10 ttJd2, with the standard idea of e2-e4, deserves attention. Indeed White's chances look preferable. An­ other interesting idea is 10 a4, although with 10 ... dxc4 1 1 bxc4 i.xc4 12 ttJd2 i.a6 13 i.xc6 Black can equalise with accurate defence. After 7 'ifc2 c6 8 i.f4 b6 the standard plan 9 ttJbd2 i.b7 10 e4 (Game 8) does not give White any advantage as the bishop is not well placed on f4 after ... c6-c5. Perhaps White should play 9 l:td1 i.a6 10 ttJe5, or 9 ... i.b7 1 0 ttJc3 dxc4 1 1 ttJd2 (Game 9) . In the event of e2-e4 the reply ... d5xe4 is unpopular. If Black is late with the freeing advance ... c6-c5 White can himself push with c4-c5 (Game 1 1), which is in fact quite effective. The advance e4-e5 is still relevant (Game 1 2). After 7 'iVc2 c6 8 ttJbd2 b6 9 e4 i.a6 this plan promises White the better chances, although in the case of 9 ... i.b 7 1 0 e5 ttJe8 1 1 b3 ttJc7 12 i.b2 c5 the bishop is well placed on b7 and the position is about equal. After 1 1 cxd5 cxd5 White is slighdy better but instead of 1 1 ... cxd5 the alternative 1 1 ... exd5 looks more promising for Black, whose intention is to play ...ttJe8-c7-e6 and prepare counterplay involving ... c6-c5. In the CC the most popular position arises after 7 'ifc2 c6 8 ttJbd2 b6 9 b3 i.a6: Recent practice has demonstrated that 10 lId1 does not lead to an advantage for White. In the complex struggle Black has sufficient resources with which to oppose White's activity in the centre and on the kingside (Game 1 4) . On the other hand, after 10 i.b2 ltc8 1 1 e4 c5 1 2 exd5 exd5 1 3 'iWf5!? White has a genuine initiative, and we d o not see how Black can equalise. 1 d4 d5 2 c4 e6 3 tiJf3 lbf6 4 g3 iLe7 5 iLg2 0-0 6 0-0 6 'ilVc2 c5 - Game 1 6 . . . lbbd7 6 ... c6 7 ttJc3 b6 8 ttJe5 - Game 2 7 'ifc2 7 ttJc3 7 ... dxc4 - Game 3 7 ... c6: 8 'iWb3 - Game 4; 8 'iVd3 - Game 5; 8 b3 - Game 6 7 . . . c6 S lbbd2 8 .l:[d 1 b6 9 a4 - Game 7 8 i.f4 b6: 9 ttJbd2 Game 8; 9 lId1 - Game 9 8 b3 b6: 9 i.b2 i.a6 1 0 l:td1l::tc 8 1 1 ttJc3 Game 10; 9 l:td 1 i.b7 1 0 ttJc3 S . . . b6 9 e4 - Game 12 9 b3 iLa6 9 ... i.b7 - Game 13 1 0 l:td 1 - Game 14 10 i.b2 - Game 15 -

-

-

Game 1 1

CHA PTER TWO 4

.

.

.

I

c6 5 i.,g2 i.,d6

1 d 4 d 5 2 c 4 e 6 3 ttJf3 ttJf6 4 g 3 c 6 5 �g2 �d6 This system resembles the Closed Catalan (4 ... �e7 5 �g2 0-0 6 0-0 ltJbd7). However, in contrast to the positions with ... i.. e 7, Black can concentrate his efforts on the major plan of expansion with ... e6-e5. This approach enjoyed some popularity in the 1 920s, when Catalan theory was in its infancy. Particularly noteworthy is the encounter Reti-Bogoljubov (Game 1 6) when, after transposition, a key position was reached in which Black at­ tempted in vain to steer the game to Dutch Stonewall territory. After a correct move order such a transposition is indeed possible, but the Dutch Defence is outside the scope of this book. �_After 6 0-0 ltJbd7 Wh.ite occasionally sac­ n�es the c4-pawn with 7 ltJc3, when his initi �� more or less compensates the mod­ est inve�ent. However, most of the time Wh.ite prefe�� avoid this possibility, opting instead to contiritJ�J b3, 7 'iVc2 or 7 ltJbd2. The 'tabia' occur5'a.(ter 6 0-0 ltJbd7 7 'iVc2 0-0 8 ll'lbd2, and sint<:, Bogoljubov-Joss, . Zurich 1 934 it has been kndwn that the im­ mediate advance 8 ... e5 is premature here. After 9 cxd5 ltJxd5?! 10 ltJc4 Black h.as seri­ ous problems (panov-Makogonov - see the notes to Game 1 7), while after the stronger 38

9 . . .cxd5 1 0 dxe5 ltJxe5 1 1 ltJxe5 i.. x e5 1 2 ltJf3 the minuses of the isolated pawn are not quite compensated by Black's freer develop­ ment. Therefore Black's best strategy is to prepare ... e6-e5 with 8 ...'iVe7 or 8 ....l:!.e8 (often these moves simply transpose) .

Game 1 6 Reti-Bogoljubov

New York 1924 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 c6 5 �g2 �d6 6 0-0 ttJbd7

7 b3 The alternative is 7 ltJc3 dxc4 8 'iWc2 (du­ bious is 8 e4?! e5 9 dxe5 ltJxe5 10 ltJxe5 .1Lxe5 1 1 'iVxd8+ 'it>xd8 12 .l:!.d H �c7 Chet-

4

verik-Jurcisin, Presov, 1 999. White doesn't have full compensation for the pawn deficit in this endgame.) 8 ... 0-0 9 l:!.dl 'fie7 10 i.g5 h6 1 1 .ixf6 ttJxf6 1 2 ttJd2 e5, as in Lputian­ Arencibia, Biel 1 993, which continued 1 3 d5!? cxd5 1 4 ttJxd5 ttJxd5 1 5 i.xd5 c 3 1 6 'ifxc3 .ig4 1 7 ttJc4 .l:!.ac8 1 8 'fie3 i.b8 1 9 'ife4 with a slight edge for White (more ac­ tive). 7 . . . 0-0 Early play in the centre with 7 ... b6 8 .ib2 c5?! is not to be recommended. In Chetverik­ Hajek, Bratislava 2000 White won after 9 cxd5 exd5 1 0 dxc5 bxc5? 1 1 e4! i.e7 1 2 e5 .ia6 13 exf6, so Black should prefer to ... ttJxc5 1 1 ttJc3 i.b7 12 b4, when White has a useful lead in development in this iso­ lated d-pawn position. 8 il.. b 2

\

8 . . . .l:!.e8 Black can transpose to the Dutch Defence with 8 ...ttJe4 9 ttJbd2 fS. 9 ct:Jbd2 White can prevent ... e6-e5 with 9 ttJe5 'iVc7 to f4 b6 1 1 ttJd2 .ib7 12 e4, as in Euwe-Davidson, Amsterdam 1 924, when White took action in the centre and was re­ warded with a pull after 1 2 ... dxe4 1 3 ttJxe4 ttJxe4 1 4 .ixe4 ttJf6 1 5 i.g2 c5 1 6 d5 exd5 17 cxd5. 9 . . . tUe4? ! Lokvenc-Stolz, Haque 1 928 went 9 ... e5 to cxd5 cxd5 1 1 dxe5 ttJxe5 1 2 ttJxe5 i.xe5 1 3

.

.

.

c 6 5 il.. g 2 il.. d 6

.ixe5 lhe5 1 4 ttJ O lIe8 1 5 ttJd4 i.e6 1 6 'iVd3 'ii'd7 1 7 l:tac1 l:tac8 1 8 "ifb5 and Black's isolated pawn left him a little worse. Black can also try 9 ... b6 to 'iVc2 .ib7 1 1 e4 e5 with a mix of systems. Samisch-Rellstab, Swine­ munde 1 930 continued 1 2 dxe5 ttJxe5 1 3 l:tad l ?! 'iVc7 1 4 exd5 cxd5 1 5 ttJd4 a6 1 6 ttJfS i.fS 17 ttJO ttJed7 with chances for both sides. However, 1 3 ttJxe5 i.xe5 1 4 .ixe5 .l:txe5 1 5 f4 .l:te8 1 6 e5 ttJg4 17 'it'd3 im­ proves, gaining space and thus putting White on top. 1 0 tUxe4 dxe4 1 1 tUe5 f5 1 2 f3! White opens up the position while Black's queenside remains undeveloped. 1 2 . . . exf3 1 2 ...ttJxe5 has also been tried. Koller­ Wiechmann, 1 994 went 1 3 dxe5 .ic5+ 14 �hl exO 1 5 .l:!.xO 'fie7 16 .l:td3 (Black cannot liberate the queen's bishop) 1 6 ... i.b6 17 'iVd2 l:[d8 1 8 .l:tdl l:lxd3 1 9 'ii'x d3 with a huge advantage for White. 1 3 il.. x f3 "iic 7 1 4 tUxd 7 ! il.. x d7 1 5 e4 e5 15 ... fxe4 16 i.xe4 g6 17 'fid3 cJi;g7 18 c5! .ifS 19 d5+ e5 20 d6 'iid 8 21 'fic4 is given by Tartakover, Black's position being critical. Black should play 1 5 ... .ifS 1 6 e5 l:tad8 in order to facilitate development, although White still has the initiative after the move 1 7 'iVe2. 1 6 c5 il..f 8 1 7 'iWc2 exd4?! 1 7 ... fxe4?! 18 i.xe4 g6 1 9 i.xg6! and White wins. 1 7 ... f4!? is interesting, keeping the position closed, 1 8 gxf4 exf4 1 9 'it>h 1 leaving White a little better thanks to his presence in the centre and more breathing space for his forces. 1 8 exf5 .l:!.ad8 1 9 il.. h 5! .l:!.e5 20 il..xd4 .l:!.xf5? ! Black can also try 2 0. . .lId5, when 21 'iic4 �h8 22 i.g4 sees White retain the extra material and the initiative, while 2 1 ...i.e6? loses to 22 fxe6 .l:i.xd4 23 .l:!.xfS+! �xfS 24 illIt-! lbd5 il.. x f5 22 "iix f5 .l:!.xd4 23 .l:!.f 1 ! .l:!.d8 Alekhine gives 23 .. .'iVe7 24 .if7+ cJi;h8 25 39

Th e C a t a l a n

.idS! 'iVf6 2 6 'it'c8 and wins. 24 i.f7 + �h8 2S i.e8! 1 -0

A famous final move.

Game 1 7 Panov-M . M a kogonov

Kiev 1938 1 d4 dS 2 c4 e6 3 ttJf3 ttJf6 4 g3 c6 S i.g2 -td6 6 0-0 ttJbd7 7 'ii'c 2 0-0 8 ttJbd2 eS?! Too early. Another try is 8 ... b6 9 e4 (after the greedy 9 cxd5?! cxd5 1 0 'iVc6 'iVc7 1 1 'ii'x a8 .i.a6 1 2 'iVxfB+ liJxfB White has two rooks for the queen but is behind in devel­ opment) 9 ... dxe4 10 liJxe4 liJxe4 1 1 'iVxe4. Raetsky-J . Gunnarsson, Hafnarfjordur 1 996 continued 1 1 ....i.b7 12 .i.f4 liJf6 13 'ii'e 3 c5 14 lladl .i.xf4 15 'ii'xf4 'iVe7 16 .l:!.fel cxd4 1 7 1:txd4 lIac8 1 8 liJe5 .i.xg2 1 9 �xg2 .l:!.fd8 20 .l:!.edl l:i.xd4 21 'iVxd4, White possessing the d-flle and a queenside pawn majority. Black's prospects failed to improve after 2 1 . ..h6 22 a3 'iVb7+ 23 f3 b5 24 c5 liJd5 25 b4 f6 26 liJd3 'ii'a6 27 liJf4. 8 ...'iVe7 presents White with a few choices: 9 !:tel b6 10 e4 liJxe4 1 1 liJxe4 dxe4 1 2 Ihe4 .i.b7?! 1 3 .i.g5 'ii'e 8 1 4 lIael 'iVc8 1 5 l:th4! g6 was seen in A.Petrosian-Skembris, Dortmund 1 992. Petrosian then recom­ mends 1 6 .i.h6! .l:te8 17 liJg5 with a menac­ ing attack for White. Perhaps 9 e4 is premature compared with 40

the previous line. The natural 9 ... dxe4 1 0 liJxe4 liJxe4 1 1 'iVxe4 e 5 1 2 dxe5 liJxe5 1 3 liJg5 liJg6 (1 3 ...g6!? i s an interesting option) 14 'ii'c 2 h6 1 5 liJe4 .i.f5 1 6 .l:tel lUe8 1 7 .i.d2l::tad8 1 8 'iVb3 .i.e5 1 9 .i.b4 'iVc7 earned Black a decent game in Smyslov-Sehner, Farum 1 986. 9 b3 is the quiet approach. Indeed in Ra­ zuvaev-Skatchkov, St. Petersburg 1 997 this policy resulted in a steady lead for White after 9 ... b6 1 0 .i.b2 .i.b7 1 1 e4 dxe4 1 2 liJxe4 liJxe4 1 3 'ii'xe4 l:tad8 1 4 !:tadl liJf6 1 5 'iVc2 thanks to Black's development problems. These were compounded after the subse­ quent 1 5 ... c5? 1 6 dxc5 .i.xf3 ( 1 6 ... .i.xc5? 1 7 liJg5) 1 7 cxd6 .i.xd 1 1 8 'ii'xh 7+! liJxh7 1 9 dxe7 .i.e2 2 0 exfB'ii' + liJxfB 2 1 l:t a l 1:td2 22 .i.c3 l::!c 2 23 .i.e5 liJd7 24 .i.e4 J:td2 25 .i.c3 and White was close to winning. 9 cxdS ttJxdS? ! Or 9 ... cxd5 1 0 dxe5 liJxe5 1 1 liJxe5 .i.xe5 1 2 liJf3 �th an edge for White. Rogers­ Handoko, Djakarta 1 993 continued 1 2 ... .id6 1 3 .l:!.d 1 .l:!.e8 1 4 .i.g5 .i.e6 1 5 liJd4 .ie5 16 liJxe6 fxe6 17 e4 h6 18 .i.xf6 'iVxf6 19 exd5 .l:!.ac8 20 'iVd3 .i.xb2 21 .l:!.abl exd5 22 .i.xd5+ �h8 23 .ixb7 J:!.cd8 24 'iVf3 and, although White had won a pawn, the oppo\ site coloured bishops offered Black chances of saving the game. 1 0 ttJc4 'ii'e 7 1 1 e4 Also good is 1 1 liJxd6 'iVxd6 1 2 l:[dl 'i'e7 1 3 e4 liJb4 1 4 'ifb3 liJa6 1 5 .i.g5 when White has the bishop pair and an active posi­ tion. In Bogoljubov-Joss, ZUrich 1 934 this proved significant after 1 5 ... 'ife8 1 6 dxe5 liJxe5 17 �d8 liJxf3+ 1 8 .i.xf3 'iVe5 1 9 lhfB+ �xfB 2 0 .i.f4 'iVe6 21 'ii'c 3. 1 1 ttJSb6 Panov gives l 1 .. .liJb4 12 'iib 3! exd4? 1 3 .i.g5! and Black has problems. 1 2 ttJxd6 'ii'x d6 1 3 l:td 1 White had a healthy initiative for the pawn in Filipovic-J erbic, Pula 1 999 after 1 3 .i.e3 exd4 14 .i.xd4 .l:!.e8 1 5 .l:tadl 'ii'e6 16 b3 'ifxe4 1 7 'ifb2. There followed 1 7 .. .f6 1 8 . . .

4

.l:!.del 'iVg6 1 9 lbh4 'iVf7 20 lbfS 'ii' f8 21 f4 lbd5 22 'it'f2 g6 23 J::tx eS 'it'xeS 24 .l:!.el 'it'f8 25 i.xd5+ cxd5 26 lbe7+ h8 22 ... fxe5? is poor in view of 23 lbg5, while 22 ... lbxe5 loses: 23 i.xe5 fxe5 24 lbg5 'iig6 25 'iixg6 hxg6 26 .l:!.dS etc. After 22 ...'iVg6 23 'i'xg6 hxg6 24 i.xe6+ l::tx e6 25 i.a3 Black's cramped position will soon fall apart. 23 l:tg4! 4Jdf8 Or 23 ... lbxe5 24 lbxe5 fxe5 25 i.xe5 l:tgS 26 l:th4 and wins. 24 exf6 'it'xf6 25 4Je5 b5 26 i.a2 c5 27 i.d5 i.a6 28 l:th4 g5 29 l:th6 'iWg7 30 i.xf8 4Jxf8 31 4Jf7 + 1 -0

Game 1 8 Orsag-Haba

Tumov 1996 1 d4 d5 2 c4 e6 3 4Jf3 4Jf6 4 g3 c6 5 i.g2 i.d6 6 0-0 4Jbd7 7 'iWc2 0-0 8 4Jbd2 l:te8

9 b3

.

.

.

c 6 5 i. g 2 i. d 6

9 e4 is good enough only for equaliry. For example Pirc-Bogoljubov, Bad Harzburg 1 935 went 9 ... dxe4 1 0 lbxe4 lbxe4 1 1 'iVxe4 e5 1 2 i.g5 f6 1 3 i.d2 exd4 14 'iVxd4 lbe5 1 5 i.c3 c5 1 6 'iVd5+ i.e6 1 7 'iVxb7 ttbS l S 'iVe4 lbxc4 1 9 b3 lbb6 20 J::ta el 'iVc7, and Black had solved his opening problems. Alternatively 9 J::td l 'iVe7 1 0 e4 lbxe4 1 1 lbxe4 dxe4 1 2 'ii'xe4 e5 1 3 i.g5! f6 1 4 i.d exd4 1 5 'iVxe 7 i.xe 7 16 i.xd4 i.c5 17 .l:!.e 1 saw White maintain a modest lead in devel­ opment in Ricardi-Panno, Florida 2001 . An ending advantage resulted after 1 7 ...l::tx e 1+ I S J::txe l i.xd4 1 9 lbxd4 lbe5 20 b3 i.g4 2 1 h3 i.d7 22 f4 lbf7 23 c5
Th e C a t a l a n

1 4 �xeS �xeS 1 S ttJf3 �e7. Andersson­ Cifuentes, Benidorm 1 994 continued 1 6 ttJd4 .l:!.c8 17 �2 J::te c7 1 8 l:tac1 h6 (Cifuentes gives 1 8 ... hS!?) 19 e3 a6 20 l:tfe 1 'iid7 21 f3 SLe6 22 'it>f2 'iVd6 23 l:txc7 'iVxc7 24 l:le2. With �c2 White trades more pieces in order to step closer to an endgame that is favour­ able due to the isolated d-pawn. 1 3 !:tfe 1 .ltg4 1 4 a3 l:te8 1 5 'ii'd 3 ttJe4 1 6 b4 Interesting is 1 6 'YWxdS ttJxf2!? 1 7 ttJgS ttJeS 1 8 i..x eS 'iVxgS 1 9 �xd6 'iVe3 with a highly complicated position in which Black has a strong attack for the piece. 1 6 . . . .ltf8 1 7 e3 'ii'b 6? ! Better is 1 7...'YWd6!? 1 8 �ac 1 i.. f5 . 1 8 !:tae 1 White can grab the pawn with 1 8 'iVxdS!? lIcd8 19 'YWc4! as the g4-bishop is hanging. 1 8 . . !:ted8 1 9 .ltd4 ttJxd4 20 ttJxd4 ttJf6 21 1tJ2b3 .

White is slightly better due to the control of cS. 21 . . . a6 22 h3 .lth5 Black enters the following complications as the more normal 22 ... i..e 6 23 ttJcS leads to a small but enduring advantage for White. 23 f4 h6 24 f5 'ilVd6 ! 25 g4 .ltxg4 26 hxg4 ttJxg4 27 ttJf3 'ii'g 3 Black has attacking possibilities for the piece, but probably not enough. 28 'ii'd 2 .ltd6 29 !:te2 ttJe5 30 ttJxe5 !:txe5 31 ttJd4 !:tde8 42

Or 3 1 ...�e4 32 'it>f1 :g4 33 WiVe1 and White consolidates . 32 'ii'e 1 'ii'g 4 33 !:td 1 !:te4 Black can win one more pawn but at the cost of exchanging all the heavy pieces: 33 ... l:!.xe3 34 �xe3 l:txe3 3S 'it'xe3 'YWxd1+ 36 'it>f2 and the three pawns are no match for the minor piece, Black losing his attacking potential. 34 'ii'f 2? ! 34 �d3 secures White the better game. 34 . . . 'ilVg5 Or 34... �xe3 3S 'ii' f3 !? �xf3 36 �xe8+ �h7 37 ttJxf3 "it'xfS with a complex situa­ tion. 35 f6 g6 36 'it>f 1 .ltg3 37 'ii'f 3 l:txe3 38 !:txe3? Poor. White should go for 38 'iVxdS 'iVg4 39 �f3 'YWh3+ with equal chances. 38 . . .!:txe3 39 'ii'x d5 'ii'x f6 + 40 ttJf3 Not 40 SLf3? �eS 41 'it>g2 SLxd4 42 l:txd4 l:txf3 and Black wins. 40 . . . 'ii'b 2 4 1 'ii'd 8+ 'it>h7 42 'ii'd 2 'ii'x a3

Black now has more than adequate com­ pensation the piece. 43 ttJd4 !:te7 44 .ltd5 .ltd6 45 'ii'f 2 'ilVh3 + 4 6 .lt g 2 'ii'g 4 47 J:rd3?! 47 i.. f3 'it'f4 48 bS offers better prospects for White than the game continuation. 47 . . . .ltxb4 48 .lth3 'ilVe4 49 l:tf3 'ilVe 1 + 50 'it>g2 'ii'x f2+ 5 1 !:txf2 .lte3 52 !:te2 !:te7 53 !:te2 b5 54 'it>f3 b4 55 .ltf1 a5 56 ttJb5 !:te5 0-1

4 . . . c 6 5 iL g 2 iL d 6

Summary These examples indicate that in the ... .td6 system it is not necessary for White to hurry with e2-e4 as this serves only to justify Black's own advance of the e-pawn, which appears to prom­ ise excellent equalising chances (for example 6 0-0 tiJbd7 7 'iic 2 0-0 8 tiJbd2 .l:te8 9 e4 dxe4 1 0 tiJxe4 tiJxe4 1 1 'iie 4 e 5 with further simplifications that cancel out any White edge). A more promising strategy for White is to complete queenside development with b2-b3 and .tb2. After 9 b3 e5 10 cxd5 cxd5 1 1 dxe5 tiJxe5 1 2 .tb2 we again have a position with an isolated pawn, albeit an acceptable version for Black, who enjoys the better development and, in prac­ tice, tends to achieve equality. 1 d4 d5 2 c4 e6 3 lLIf3 lLIf6 4 g3 c6 5 iLg2 iLd6 (D) 6 0-0 lLIbd7 7 'ifc2 7 b3 (D) - Game 16 7 0-0 8 lLIbd2 (D) 8 ... e5 - Game 1 7; 8 ... .l:te8 - Game 18 ...

5 . iLd6 . .

7 b3

8 1L1bd2

43

CHA PTER THREE

I

4 . . . dxc4 5 'iVa4 +

1 d4 d5 2 c4 e6 3 lLlf3 lLlf6 4 93 dxc4 5 'iVa4+ After 4 ... dxc4 White can regain the pawn back immediately. This continuation was popular during the infancy of the Catalan. In particular world champions Capablanca, Alekhine (his two nice victories are shown in Games 22 & 25), Botvinnik and Smyslov played this line, but much has happened since. When White regains the pawn with the aid of the queen check he is also assisting Black's development. Moreover, the queen is subject to attack on c4 (mainly from ... b7b5). These days 5 'i!Va4 is seen less often than 5 ..tg2, and is usually played by those looking for the kind of edge which brings with it little chance of losing. Not surprisingly the main adherent of 5 'iVa4+ is GM Ulf Andersson, a peace loving man with superb endgame tech­ nique. 1 d4 It:if6 2 c4 e6 3 g3 d5 4 ..tg2 dxc4 5 'i!Va4+ is the standard move order for this line from a theoretical perspective, but in reality It:if3 is seen more often, and the variations where White tries to play without It:igl -f3 are hardly relevant - 5 'i!Va4+ ..td7 6 �xc4 ..tc6 7 f3, for example, can never give White an advantage. After 1 d4 d5 2 c4 e6 3 It:if3 It:if6 4 g3 dxc4 5 'it'a4+ Black does have the extra 44

possibility o f 5 . . .�d7 6 'i!Vxc4 'i!Vc6, but after 7 It:ibd2 'i!Vxc4 8 It:ixc4 White has a stable advantage. In the event of 5 ...lt:ic6 6 'it'xc4 'iWd5 the position is acceptable for Black, but only in the line with It:if3, of course, as otherwise the queen would be hanging. More attention should be paid to 5 ... c6 6 'i!Vxc4 b5, with a position that most often arises from the Slav after 1 d4 d5 2 c4 c6 3 It:if3 It:if6 4 'i!Vc2 dxc4 5 'ii'xc4 e6 6 g3 b5. Now after 7 'i!Vc2 ..tb7 8 .i.g2 It:ibd7 Black is trying to achieve ... c6-c5 as quickly as possible. If White allows this (Game 1 8) Black is afforded decent chances of equality, so preference should be for 9 It:ie5, preventing the immediate 9 ... c5 (Game 20) . After 5 'i!Va4+ it is logical to play 5 ... .i.d7 6 'iVxc4 ..tc6, after which Black develops his pieces (Game 21) or plays ... ..tc6-d5 (with tempo), preparing the freeing advance ... c7c5 (Game 22). Nevertheless, the main de­ fence against 5 �a4+ is 5 ... lt:ibd7. It is im­ possible to clearly classify the variations based on concrete moves as there are so many transpositions. But what is possible is to pick out the different strategies. Insuffi­ cient for equality is the plan with ... e6-e5 (after 6 ..tg2 c6 7 'iVxc4 ..td6 8 0-0 e5). And quite ive is 6 ... ..te7.

4 . . . dx c 4 5 'il a 4 +

One idea for White is to hold back the capture on c4 and instead continue develop­ ment with 6 �g2 (Game 23), although 6 "i'xc4 is the main continuation. Then 6 ... a6 (Games 24-26) is designed to harass White's queen with the help of ... b7-b5 and, in so doing, solve Black's main strategic problem the development of the queen's bishop. White quickly drops back to c2 in Game 24 to prevent ... b7-b5, while in Games 25 & 26 Black is allowed to complete his main plan and push the b-pawn.

Black plays 5 . . . c6 r---------------..

Game 19 Mednis-Prie

Cannes 2000 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 'i'a4+ c6 5 ...'iir'd7 6 'ii'x c4 'iVc6 7 lLlbd2 'it'xc4 8 lLlxc4 i..b4+ 9 .id2 .ixd2+ 1 0 lLlfxd2! lLlc6 1 1 e3 lLlb4 12 'It>e2 .id7 13 �g2 .ic6 14 f3 lef� White in control of the centre in Botvinnik­ Vidmar, Groningen 1 946. There followed 14 ... lLld7 1 5 a3 lLld5 1 6 e4 lLl5b6 1 7 lLla5 i..b5+ 1 8 'It>e3 0-0-0 1 9 l:thc1 lLlb8 20 b3 i..d7 21 .if1 lLlc6 22 lLlxc6 iLxc6 23 a4 iLe8 24 as lLla8 25 a6 b6 26 b4 with a big edge. After 5 ... lLlc6 White should avoid 6 lLle5? "i'd5 7 l:tgl b5, when he stands clearly worse. Meanwhile, 6 iLg2 transposes to 5 �g2 lLlc6 6 'ii'a4. This leaves 6 'iWxc4 'iVd5 7 'iWd3, e.g. 7 ... lLlb4 8 'it'dl 'iWe4!? 9 lLla3 iLd7 1 0 .ig2 i..c 6 1 1 0-0 h6 1 2 �f4 lLlbd5 1 3 .id2 �xa3 14 bxa3 0-0 1 5 e3 l:tfd8 as in Kobalija­ Dokuchaev, Majkop 1 998, when White's weakened queenside pawns and Black's activ­ ity contributed to a balanced game. Another sensible line is 7 ... 'iWfS 8 'iVxfS exfS 9 a3 �e6 10 e3 lLla5 1 1 lLlbd2 lLlb3 1 2 lLlxb3 �xb3 1 3 i.d3 g6 and Black was doing fine i n Kura­ jica-Zvjaginsev, Pula 1 997. This status quo continued after the subsequent 14 .id2 as 1 5 e2 �d6 1 6 %:thc1 �d5 1 7 lLle5 c 6 1 8 lLlc4

�c 7 1 9 b4 axb4 20 axb4 'It>d7. 6 'ilxc4 b5 7 'ilc2 Abramovic-Maksimenko, Novi Sad 2000 went 7 'iWd3 lLlbd7 8 �g2 �b7 9 0-0 a6 1 0 a4 c 5 1 1 axb5 �e4 1 2 'iWdl axb5 1 3 k[xa8 'it'xa8 1 4 lLla3 b4 1 5 lLlc4 cxd4 1 6 �f4 'iWa6 1 7 b3 �c5 1 8 lLlxd4 �xg2 1 9 'It>xg2 0-0, Black completing development. 7 . . . iLb7 8 iLg2 Also possible is 8 lLlbd2 lLlbd7 9 e4 iLe7 10 a3 c5!? 1 1 �xb5 0-0 with compensation according to M.Gurevich. In this line White can try to prevent the thematic ... c6-c5, a plan adopted in the encounter Gheorghiu­ R.Bagirov, Bern 200 1 : 9 lLlb3 'iWb6 1 0 �g2 as!? 1 1 .ig5 a 4 1 2 lLlbd2 c5 1 3 �xf6 lLlxf6 14 0-0 l:tc8 15 l:tac1 �e7 16 dxc5. Now 1 6...�xc5! 1 7 'ifb 1 0-0 would have given Black the better chances due to the bishop pair and space on the queens ide. 8 . . ttJbd7 .

9 0-0 9 a4 c5!? 10 axb5 cxd4 1 1 0-0 �c5 1 2 �g5 .l:tc8 1 3 lLlbd2 0-0 1 4 'iWd3 h 6 1 5 .ixf6 'iWxf6 1 6 lLlg5 'iVxg5 1 7 �xb7 l:tc7 was Timman-Lautier, Malmo 1 999, with chances even after 1 8 lLlf3 lLle5 19 lLlxe5 'it'xe5. After 9 lLlc3 c5 White must avoid 10 lLlxb5? 'iVa5+ 1 1 lLlc3 cxd4, when Black is winning. White fared better in Zagorskis­ Vaisser, Lyon 1 994: 1 0 0-0 b4 1 1 lLla4 l:tc8 1 2 lLlxc5 lLlxc5 1 3 dxc5 .ixc5 14 'it'a4+ �c6 1 5 'it'a6 0-0 1 6 �g5 i.d5 1 7 l:tac1 'iVb6 1 8 45

Th e Ca t a l a n

'it'xb6 axb6 1 9 ..Itxf6 gxf6 20 b3 with an equal endgame. In Magerramov-Doroshkievich, Podolsk 1 992 White produced an interesting pawn sacrifice aimed at generating an initiative: 9 ... a6 10 0-0 cS 1 1 ..ItgS l:.c8 1 2 dS!?, al­ though after 1 2 ... exdS 1 3 4Jh4 h6 1 4 ..Itxf6 4Jxf6 1 5 l:.ad 1 l:.c7 1 6 4JxdS 4JxdS 1 7 ..ItxdS ..ItxdS 1 8 e4 ..Ite7 19 l:.xdS �a8 20 4Jf5 0-0 21 l:.fdl l:.e8 Black was okay. 9 . . . c5 1 0 �e3 Another option is 10 a4 b4 1 1 ..ItgS l:.c8 12 dxcS ..Itxcs 13 �d3 0-0 14 4Jbd2 with a balanced game, a situation that is preferable from White's side of the board than 1 1 as?! l:.c8 12 dxcS �xcS 13 'it'dl 0-0 14 4Jbd2 �e7 1 5 4Jb3 l:.fd8 1 6 'ilVel ..Itd6 17 a6 ..ItdS, when the advance of the a-pawn proved unsuccessful in A.Kulikov-Najer, St. Peters­ burg 2000. Black is better. 1 0 .. J�c8 1 1 dxc5 �xc5 1 2 �xc5 lLixc5 1 3 lLic3 b4 1 4 lLib5 'i!fb6 1 5 lLibd4 0-0 1 6 a3 a5 Better is 16 ... .u.fd8, e.g. 17 axb4? 4Jce4 with an initiative for Black. 1 7 axb4 axb4 1 8 'i'c4 �xf3 1 8 ... .i.dS 19 'iib s 'it'xbS 20 4JxbS 4Jb3 21 l:.a4 l:.c2 yields nothing for Black. 1 9 lLixf3 lLice4 20 'i'b3 20 �d4 it'xd4 21 4Jxd4 eS 22 4Jf3 .l:!.c2 23 4JxeS l:.xb2 is equal. 20 . . JUd8 21 e3 h6 22 h3 lLic5 23 'ii'a 2 lLife4 24 l::ta d 1 l::tx d1 25 l::t x d 1

46

25 . . . lLixf2! ? An interesting piece sacrifice. 26 'it>xf2 lLie4+ 27 'it>g 1 'ii'x e3 + 28 'it>h2 l::t c 2 29 'ii'a 8+ 'it>h7 30 l::tf 1 l::t x g2+ Black forces the draw. 3 1 'it>xg2 'iVe2 + 32 'it>g 1 'iVe3 + 33 'it>g2 'ife2 + Y. - Y2

Came 20 Raetsky-Sveshni kov

Kolontaevo 1994 1 d4 d5 2 c4 e6 3 lLif3 lLif6 4 g3 dxc4 5 'i'a4+ c6 6 'iVxc4 b5 7 'iVc2 �b7 8 �g2 lLibd7 9 lLie5

9 . . . 'ii'b 6 Another option is 9 ... 4JxeS 10 dxeS 4JdS! 1 1 0-0 ..Ite7 12 l:.dl �8 13 4Jc3 0-0 14 4JxdS cxdS 1 5 e4 l:.c8 16 'ife2 dxe4 17 ..Itxe4 ..Itxe4 1 8 �xe4 l:.c4! with a good game for Black in Volzhin-Ibragimov, Elista 2001, although 19 �f3 �e8 20 ..Ite3 .u.c7 was agreed drawn. Note that it is important for Black to find (or be prepared with) 1 O ... 4JdS! as this is superior to 10 ... 4Jd7 1 1 0-0 'ifb6 1 2 a4, which featured i n two Raetsky games. After 1 2 ... a6 1 3 axbS axbS 14 ..Ite3 'irc7 1 5 l:.xa8+ ..Itxa8 1 6 4Jc3 White had the initiative in Raetsky-Klingelhoefer, Giessen 1 994, and the further 1 6 ... �8 1 7 l:.al ..Ite7 1 8 l:.a7 ..Itd8? 1 9 4JxbS 0-0 20 l:.xd7 cxbS? 21 iLa7 saw White win . Raetsky-Volzhin, Hastings 1 992/93 also favoured White after 1 2 ... c5 1 3

4 . . . dx c 4 5 'iW a 4 +

.ixb7 'iVxb7 1 4 axb5 ttJxe5 1 5 ttJc3 iJ.. e 7 1 6 .if4 ttJg6 1 7 'iVe4 'iix e4 1 8 ttJxe4 etc. 10 i.e3 10 0-0 'ifxd4 (10 ... c5!? is interesting) 1 1 ttJxd7 'iVxd7 1 2 .l:!.dl 'it'c8 1 3 ttJc3 a6 1 4 iJ..g5 with compensation for the pawn. White is well developed and ready for action. 10 . . . c5 1 1 Q\xd7 Q\xd7 1 2 i.xb7 'iWxb7 1 3 0-0

1 3 . . . cxd4? ! 1 3 ... 1:tc8 has also been played. Raetsky­ Rausch, Biel 1 997 continued 14 l:[c1 iJ.. e 7 1 5 dxc5 ttJxc5 1 6 b4 iJ.. f6 1 7 ttJc3 ttJd7 1 8 'iVe4 'i'a6 1 9 i.d4 0-0 with equality, while 17 ... ttJa4 1 8 'it'e4 'ifa8! (1 8 .. .'ika6? 19 ttJxb5! .l:!.xc1+ 20 .l:!.xc1 'ii'x b5 21 .l:!.c8+ iJ.. d 8 22 .l:!.xd8+!) 19 ttJxb5! leads to a forced draw after 1 9 .. .'it'xe4 20 l:txc8+ 'it>e7 21 l:tc7+ 'it>ffi. Here 14 ...c4?! risks giving White a free hand in the centre. Raetsky-Tregubov, Krasnodar 1 995 went 1 5 a4 a6 1 6 axb5 axb5 1 7 ttJc3 .ib4 1 8 d5 exd5?! (1 8 ... i.xc3 19 dxe6 fxe6 20 'iVxc3 0-0 and White has only a small advantage) 1 9 .l:ta7 �c6 and White now changed gear and emerged with a clear ad­ vantage after 20 .l:!.ca1 ! iJ.. c 5 21 l:t l a6 ttJb6 22 .ixc5 'iix c5 23 1:tb7. 14 i.xd4 .l:!.c8 After 14 ... e5 1 5 iJ.. e 3 i.e7 16 l:tdl 0-0 1 7 'it' fS White has an action position. 1 5 Q\c3! b4 Again the push in the centre with 1 5 ... e5 is insufficient for equality - 16 i.e3 i.b4 1 7

'i'e4 'iixe4 1 8 ttJxe4 fS 1 9 a3 iJ.. e 7 20 ttJc3 etc. 1 6 'iWe4 'i!fxe4 Avoiding the exchange with 16 ... 'it'a6 gives White the better chances after 17 ttJd5 'tIi'c6 1 8 .l:!.ac1 'it'xc1 19 .l:txc1 l:!.xc1+ 20 'iit>g2. 1 7 Q\xe4 f5 1 7 ... .l:tc4?! runs into 1 8 i.xg 7! .l:tg8 1 9 iJ.. x ffi .l:!.xe4 20 iJ.. d 6 l::t xe2 21 l::t ac1 and White is well on top. 1 8 Q\g5 e5! 1 9 i.xa7 i.e7 20 i.e3 h6 2 1 Q\f3 Wf7 Completing development is the aim of Black's pawn sacrifice, resulting in an active game for Black. 22 .l:!.fc 1 g5 23 b3 f4 24 i.d2 We6 25 .l:!.xc8 .l:!.xc8 26 .l:!.c 1 .l:!.a8 27 .l:!.c2 27 l:!.c6+ 'iit>d 5 28 .l:txh6 g4 29 l:!.h7 'it>e6 and again Black has compensation, winning back one of the invested pawns with an ag­ gressive stance. 27 . . . Q\f6 28 Q\e 1 28 gxf4 gxf4 29 e3 promises White a small edge. 28 . . . Q\d5 29 e4

29 . . .fxe3 30 i.xe3 e4 3 1 i.d4 i.f6 32 .l:!.c6 + Wd7 33 .l:!.c4 33 l:txf6 ttJxf6 34 .i.xf6 l:txa2 and, with a rook versus two minor pieces, Black has definite compensation. His king and rook are active and 'W'hite's pieces lack co-ordination. 33 . . . .l:!.xa2 34 i.xf6 Q\xf6 35 Q\c2 l:!.b2 36 Q\xb4 Yo - Yo 47

Th e C a t a la n

Black plays 5 . . . iLd7

Game 2! Vila-Sky

Castrop-Rauxel !990 1 d4 d5 2 c4 eS 3 ttJf3 ttJfS 4 g3 dxc4 5 �a4+ iLd7 S �xc4

s . . . iLcs 6 ... c5 7 dxc5 i.c6 transposes to 5 i.g2 c5 6 'ii'a4+ i.d7 7 �xc4 i.c6 8 dxc5 (see the notes to Game 76). 7 iLg2 ttJbd7 Gleizerov-Akhmadeev, Kstovo 1 997 went 7 ... i.e7 8 ttJc3 0-0 9 0-0 a6 1 0 �d3 b5 1 1 i.g5 ttJbd7 1 2 a3 i.b7 1 3 b4 h6 1 4 i.xf6 ttJxf6 1 5 ttJe5 i.xg2 1 6 �xg2 with the finn grip on c5 earning White the pleasant pros­ pects. 8 ttJc3 ttJbS Also possible is 8 ... i.e7 9 0-0 0-0 10 'iVd3 ttJd5 1 1 e4 ttJxc3 1 2 bxc3 fS (12 ... ttJc5 1 3 ir"e3 ttJxe4 i s best avoided i n view o f 1 4 ttJe5!, regaining the pawn and keeping a plus) 1 3 exfS exfS 14 ttJel i.xg2 1 5 ttJxg2 ttJb6 1 6 a4 a s 1 7 l:te 1 tt f7 with approximate equality in Zaichik-Shabalov, Philadelphia 2000. 9 �d3 iLb4 1 0 iLg5 Petrov-Alekhine, Buenos Aires 1 939 saw considerable trading during the next dozen or so moves: 10 0-0 0-0 1 1 l:tdl h6 12 ttJe5 i.xg2 1 3 �xg2 'it'e7 14 ttJe4 l:tad8 1 5 ttJxf6+ 'it'xf6 1 6 'it'f3 'it'xf3+ 1 7 ttJxf3 ttJc4 1 8 b3 48

ttJa3 1 9 i.b2 f6 20 ttJe1 ttJb5, and after 2 1 l:tac1 i.xe 1 22 Ihe 1 ttJxd4 23 i.xd4 l:txd4 24 l:txc7 l:tf7 the situation cleared and ended in an even ending. 1 0 . . . ttJa4 Forcing the exchange of the bishop with 10 ... h6 1 1 i.xf6 'it'xf6 seems okay for Black. Indeed in Sosonko-Ligterink, Hilversum 1 987 Black had a solid position with possi­ bilities of healthy counterplay in the centre after 1 2 0-0 0-0 1 3 ttJe4 �e7 14 l:tac1 ttJd7 1 5 a3 i.d6 1 6 ttJc3 i.xf3 1 7 i.xf3 c6 1 8 ttJe4 i.c7 1 9 b 4 i.b6 20 l:tfdl l:tfd8. 1 1 J:tb 1

1 1 . . . �d5?! 1 1 ...h6 is once again a sensible option, with 1 2 i.xf6 'it'xf6 1 3 0-0 ttJxc3 1 4 bxc3 i.a5 giving White central control, while the bishop pair helps Black's . quest for equality. 1 2 iLxfS iLb5 The recapture 1 2 ...gxf6 invites White to generate an initiative after 1 3 0-0 i.xc3 14 bxc3 ir"a5 15 c4, although this might be pref­ erable to the text. 1 3 �c2 �xa2 1 4 0-0 ttJxc3 1 5 iLxg7 �g8 1 5 ... ttJxbl 1 6 'it'xc7! .l:!.g8 1 7 ttJe5! l:txg7 1 8 �xb7 a6 1 9 'it'xa8+ �e7 20 i.c6 and White's attack against the bare king is very dangerous. 1 S J:ta 1 ttJxe2+ 1 7 'it>h 1 �d5 1 8 �xh7 O-O-O? 1 8 ... l:txg7 19 "iixg7 i.c6 20 h4 is clearly

4 . . . dx c 4 5 "ii a 4 +

better for White, but Black is holding on. 19 J:txa7

White has a decisive lead. 1 9 . . . c6 20 J:tfa 1 20 .i.e5 is clear-cut. 20 . . . "iif 5 21 "iix f5 exf5 22 i.f6 J:tdeS 23 i.e5 f4 24 i.h3 + J:te6 25 J:taS + 1 -0

Game 22 Alekhine-Rabar

Munich 1942 1 d4 d5 2 c4 e6 3 lLlf3 lLlf6 4 g3 dxc4 5 "iia4+ i.d7 6 "iix c4 i.c6 7 i.g2 i.d5

S "iid 3 Other queen moves have been tried here. After 8 'fia4+ .i.c6 9 'fidl e5!? 10 0-0 exd4 1 1 'iixd4 'iixd4 12 ttJxd4 .i.xg2 1 3 'it>xg2 .i.c5 the smoke begins to clear and Black faces no problems, e.g. 14 ttJb3 ttJa6 1 5 ttJc3

.i.b6 1 6 .i.g5 ttJd7 1 7 .l:!adl h6 1 8 .i.f4 0-0-0 with a level game. 8 'it'c2 ttJc6 (8 . . . .i.e4 9 'fid 1 transposes to 8 �d3 .i.e4 9 'iWdl) 9 'fia4 .i.b4+ 1 0 ttJc3 ttJe4 1 1 0-0 saw White sacrifice a pawn for the initiative in S.Larsen-Thesing, Soro 1 982. There followed 1 1 .. ..i.xc3 12 bxc3 ttJxc3 1 3 'it'c2 ttJe4 1 4 .i.a3 ttJd6 1 5 .i.xd6 'it'xd6 1 6 e4 ttJb4 1 7 �e2 .i.c6 1 8 d 5 .i.d7 1 9 a3 ttJa6 20 e5 'it'e7 21 d6 cxd6 22 exd6 'it'xd6 23 ttJe5 and, despite being two pawns down, White's active play furnished him ample compensa­ tion. S . . . c5 Andersson-Ille scas, Pamplona 1 997/98 is the way to play this line for Black: 8 ... .i.e4 9 'it'd 1 c5 1 0 ttJc3 .i.c6 1 1 0-0 ttJbd7 1 2 .i.e3 .i.e7 13 dxc5 (typical Andersson) 1 3. .. .i.xc5 14 .i.xc5 ttJxc5 1 5 'fixd8+ llxd8 16 lIac1 'it>e7 1 7 b4 ttJcd7 and the ending was level, although White was probably happy anyway! 9 lLlc3 i.c6 1 0 0-0 lLlbd7 1 1 J:td 1 The more direct approach with 1 1 e4 promises White an edge, l 1 ...cxd4 1 2 ttJxd4 ttJe5 1 3 'iWd 1 'i!ib6 14 ttJxc6 ttJxc6 1 5 'fia4 giving White the bishop pair and the more active forces. 1 1 . . . cxd4 With 1 1 ...'iVb6 Black retains the tension in the centre for a while. Gereben-Szabo, Bu­ dapest 1 948 continued 12 e4 cxd4 13 ttJxd4 .i.c5 14 .i.e3 ttJe5 1 5 'fie2 ttJfg4 1 6 ttJxc6 .i.xe3 17 ttJxe5 .i.xf2+ 1 8 'it>h 1 ttJxe5 1 9 ttJa4 'it'a5 2 0 'it'xf2 'it'xa4 2 1 'it'c5 and Black had a pawn but needed to address develop­ ment, White's activity maintaining the bal­ ance. 1 2 lLlxd4 i.xg2 1 3 Wxg2 i.e7 1 4 "iif 3! The queen takes over the bishop's role on the long diagonal. 1 4 . . . "iib 6? 14 . . .'ir'c8 also favours White after 15 ttJb3 ttJe5 1 6 'fif4 ttJc6 1 7 .i.e3, although this improves on the game for Black. Now White sacrifices material to generate an initiative. 1 5 i.e3! 0-0 49

Th e Ca t a l a n

15 .. :iVxb2? 1 6 ltJcb5 highlights White's strategy. 1 5 ... ltJe5 16 ltJdb5 is also problem­ atic for Black. 1 6 lDfS !ii.. c s The continuation 16 .. :ii'd8 17 ltJxe7+ 'iVxe7 1 8 'iVxb7 .l:tfb8 19 'ilc7 .l:txb2 20 ..td4 is given by Alekhine as being better for White. 1 7 lDa4 'iWaS 1 8 lDxcS lDxcS

6 . . . a6 Not the only possibility. After 6 ... c6 7 'it'xc4 ..td6 8 0-0 e5 the break in the centre is not entirely satisfying for Black, e.g. 9 ltJc3 0-0 1 0 .l:td1 'ike7 1 1 'ikb3 exd4 1 2 ltJxd4 i.c5 13 h3 ltJe5 14 ..tg5 h6 15 ..txf6 'iIxf6 1 6 ltJe4 'it'e7 1 7 ltJxc5 'iix c5 1 8 .l:tac1 and White has a small advantage due to his lead in de­ velopment. 6 ... ..te7 7 'it'xc4 0-0 8 0-0 a6 9 �c2 c5 1 0 .l:td1 .l:ta7 1 1 e4 b6 1 2 ltJc3 .tb7 1 9 lDxg7! A nice, traditional wrecking of the king­ 1 3 d5 exd5 1 4 e5 ltJe8 1 5 ltJxd5 saw White seize the initiative in Lautier-Pinter, side from one of histoty's greatest attacking players. 1 993, when Black had to surrender his light­ 1 9 . . .'�xg7 squared bishop - 1 5 ... ..txd5 1 6 .l:txd5 ltJc7 1 7 .l:td3 ltJe6 1 8 b 3 'ilb8 1 9 i.b 2 .l:td8 2 0 .l:tad 1 19 ... ltJce4 20 b4! 'iie 5 21 ..tf4 'itb5 22 a4 (White really wants his bishop on the a1-h8 b5 21 h4 etc. 7 lDc3 !ii.. e 7 diagonal!) 22 .. .'ihb4 23 i.e5 is nice for White. Again Black has other options: 20 !ii.. d 4! lDce4 21 'ilixe4 'tiffS 22 'ilixfS 7 ... c5 8 0-0 .l:tb8 9 i.f4 b5 10 'ild1 .l:tb6 1 1 d5 (Black keeps his pawn but now White exfS 23 ':!:!'ac 1 ':!:!'fe8 24 .:!:!.c7 ! ':!:!'xe2 2S attacks in the centre) 1 1 ...exd5 1 2 ltJxd5 ':!:!'xb7 'itig6 26 .bf6 'itixf6 27 .:!:!.d6 + ! 1 -0 In view of the continuation 27 ...�g7 28 ltJxd5 1 3 �xd5 ..tb7 1 4 �d2 ii.e7 1 5 a4 with compensation for the pawn in Speel­ .l:tdd7 .l:!.f8 29 '>itG .l:tc2 30 .l:tdc7 .l:td2 3 1 'it>e3. man-Andersson, Hastings 1 980/8 1 . There followed 1 5 ... ltJf6 1 6 'ilxd8+ i.xd8 1 7 .l:tfd1 ltJd5 1 8 axb5 ltJxf4 1 9 gxf4 .l:txb5 20 ltJe5 Black plays S . . . lDbd7 ....-------------. --. i.xg2 21 '>itxg2 i.c7 (21 ...l1xb2? is poor in Game 23 view of 22 .uxd8+!) 22 ltJxc4 .txf4 23 .l:txa6 0-0 and Black had finally castled. White won Andersson -A . Sokolov the pawn back but the game was equal. Also Bar 1997 '-------... possible here is 8 ... cxd4 9 ltJxd4 'iWb6 10 ltJc2 i.e7 1 1 'it'xc4 0-0, when Black ad­ 1 d4 dS 2 c4 e6 3 lDf3 lDf6 4 g3 dxc4 S dressed development and achieved a good 'ilia4+ lDbd7 6 !ii.. g 2 50

4 . . . dx c 4 5 'W a 4 +

position after 1 2 b4 ltJe5 1 3 'iVb3 .i.d7 1 4 .i.e3 'it'c 7 1 5 l:Iac1 .i.c6 1 6 .i.xc6 ltJxc6 1 7 .i.f4 'iVb6 i n Lindberg-Jakovenko, Oropesa 2001 . 7 ... .l:Ib8 looks sensible. 8 'iixc4 b5 9 'iYd3 .i.b7 1 0 0-0 c5 1 1 .i.f4 .l:i.c8 1 2 dxc5 .i.xc5 1 3 Itadl 0-0 1 4 ltJe5 and White was trying to generate an initiative In Andersson­ Kasparov, Belgrade (6) 1 985, which went 14 ...1Lxg2 1 5 'It>xg2 ltJxe5 1 6 .i.xe5 .i.e7 1 7 'it'f3 'it'a s 1 8 it'b7 .l:i.fe8 1 9 a3 b 4 (securing a level game) 20 .i.xf6 gxf6 21 axb4 'it'xb4 22 Wixb4 .i.xb4 23 ltJe4 'It>g7 24 ltJd6 with a draw. Returning to the 1 2th move, the more sophisticated 1 2 ... b4 has also been tried. In Mochalov-Sveshnikov, Minsk 2000 compli­ cations soon led to the arrival of the ending phase after 1 3 ltJa4 'it'a5 1 4 ltJb6 ltJxc5 1 5 ltJc4 it'b5 1 6 ltJd6+ .i.xd6 1 7 �xb5+ axb5 1 8 .i.xd6 ltJfe4 1 9 1Lxc5 .l:i.xc5, when Black's activity was sufficient to maintain the bal­ ance: 20 .i:tfc1 'It>e7 21 .l:i.xc5 ltJxc5 22 .l:i.c1 .l:!.c8 23 ltJe5 ltJa4 24 .uxc8 1Lxc8 etc. 8 lLle5

8 . . . 0-0 Another Andersson game went 8 .. J::tb 8 9 ltJxd7 (or 9 'iYxc4 c5 1 0 ltJxd7 'iYxd7 1 1 .i.f4 b5 1 2 'iVd3 c4 1 3 'iYc2 .i:tb6 with counterplay for Black) 9 ...'iVxd7 10 'iYxc4 b5 1 1 'iVd3 1Lb7 1 2 .i.xb7 ':'xb7 1 3 Wio .l:i.b6 1 4 0-0 'it'c6 1 5 'ii'xc6+ l::txc 6 and Black was well developed in Andersson-Hubner, Tilburg 1 98 1 , with White getting no chance to ex-

ploit the c-file or the often weakened c5square. Consequently 16 .i.g5 l:Ic4! 17 e3 c5 1 8 1Lxf6 gxf6 19 dxc5 0-0 20 l:tac1 l:!.d8 21 ltJb 1 .l:i.xc5 22 .l:i.xc5 .i.xc5 23 .l:i.c1 .i.b4 left Black with the superior ending . 9 lLlxd7 9 ltJxc4 has also been played. Black was never in danger in the opening in Poluljahov­ Zakharevich, St. Petersburg 200 1 , which went 9 ... c5 10 dxc5 1Lxc5 1 1 0-0 .l:i.b8 1 2 'it'dl 'ii'c 7 1 3 1Lf4 e5 1 4 .i.e3 b5 1 5 .i.xc5 ltJxc5 1 6 ltJe3 .i.b7 1 7 1Lxb7 .l:i.xb7 1 8 .l:i.c1 'ii'd 8 etc. 9 . . . .i.xd7 1 0 'Wxc4 b5 1 1 'Wb3 b4 1 2 lLle4 .i.b5 1 3 lLlxf6 + .i.xf6 1 4 .i.e3 Also possible is the greedy 14 .i.xa8 'ifxa8 1 5 0, although in the following lines Black obtains healthy compensation: 1 5 ... 1Lxd4 1 6 .i.e3 (1 6 'ii'xb4 c 5 1 7 'iVb3 e 5 1 8 a4 .i.d7 is fine for Black as White still needs to com­ plete development) 1 6 ... e5! 1 7 .l:i.c1 (White must be careful here, e.g. 1 7 .i.xd4 exd4 1 8 'ii'xb4? .l:i.e8 1 9 0-0-0 c5! with the better chances for Black thanks to his busier pieces and White's insecure king) 1 7 ... c5 18 .i.xd4 exd4 1 9 .i:txc5 .l:i.e8 20 .l:i.xb5 (White returns the exchange and secures equality) 20 ... axb5 21 'it'f2 'iVa7! 22 .i:te1 �c5 etc. (A.Sokolov) 1 4 . . J:tb8 1 5 l:td 1 c5!

Black temporarily sacrifices a pawn in or­ der to activate his forces. Note that White's king is still in the centre. 1 6 dxc5 'Wa5 1 7 l:td2 .i.a4 1 8 'Wc4 .i.c3! 51

Th e Ca t a l a n

Black continues t o play with purpose. 1 9 O-O! i.b5 20 'tib3 20 'ii'g4? is poor in view of 20 ....i.xd2 21 .i.xd2 'ii'xa2 and White has lost a pawn for no compensation. 20 . . . i.xd2 20 ... .ta4 21 'it'c4 .i.b5 with a repetition. 2 1 i.xd2 i.xe2 22 l:!.e 1 The ed pawn on the c-ftle gives White only a little compensation. 22 . . . it'b5 22 ... 'ii'c7 23 c6 as is more accurate, with an edge for Black. 23 e6 a5 24 i.e3 a4 25 it'e2 i.d3 26 'tid 1 i.e4 27 b3 axb3 28 axb3 i.e2 29 'tie2 l:!.be8 30 h3 Now it is difficult for Black to proceed. 30 . . . e5 31 'tib2 f6 32 ..t>h2 i.h5 33 l:!.e5 %-%

After 3 3... 'it'd3 3 4 'iVxd2 'il'xd2 3 5 .txd2 .i.f7 36 .txb4 .txb3 37 c7 Wf7 38 .tb7 .te6 the position is balanced.

Game 24 H Gbner-Smyslov

TilbufJ!, 1982 1 d4 d 5 2 e4 e6 3 liJf3 liJf6 4 g3 dxe4 5 'tia4+ liJbd7 6 it'xe4

6 . . . a6 The immediate break in the centre with 6 ... c5 is also possible, bringing the game to a new junction. For example in Karpov52

Korchnoi, Moscow (8) 1 974, Black devel­ oped fluidly and had no problems after 7 .i.g2 b6 8 0-0 .i.b7 9 l:tdl a6 1 0 dxc5 .txc5 1 1 b4 .te7 12 .tb2 b5 13 'ii'd4 l:tc8 1 4 lLlbd2 0-0. Also possible here i s 8 lLle5 cxd4 9 lLlxf7! 'it>xf7 10 .txa8 lLlc5 when White wins the exchange but lags behind in devel­ opment, and must face Black's counterplay. 7 dxc5 .txc5 8 .tg2 a6 (8 ...'il'c7 9 lLla3!? might favour White) 9 'ifb3 l:!.a7 1 0 a4 and now the direct 1 O ... b5 1 1 axb5 'it'b6 1 2 0-0 'ilVxb5 1 3 'ilVxb5 axb5 1 4 l:txa7 .i.xa7 1 5 lLlc3 keeps White ahead in the development race and therefore earns him an edge. Chetverik­ Vujosevic, Gyongyos 1 997 instead continued 1 0 ... b6 1 1 0-0 .i.b7 1 2 lLlc3 'iVa8, when Black should be fine in the hedgehog position that results from 1 3 lLlh4 .i.xg2 14 lLlxg2 0-0 1 5 .i.e3 etc. 7 'ilfe2 e5 8 i.g2 b6 Grivas-Kourkounakis, Athens 1 996 went 8 ... cxd4 9 lLlxd4 .ll.b4+ 10 .ll. d2 'it'e7 1 1 0-0 0-0 12 J:td 1 lLlb6 1 3 .i.xb4 'il'xb4 14 lLlc3 and Black had problems with queenside de­ velopment. 9 liJe5 G.Agzamov-Zaid, USSR 1 984 demon­ strated an interesting approach to the posi­ tion: 9 dxc5 .i.xc5 10 lLlg5!? llb8 1 1 lLlc3 .i.b7 12 0-0 .i.xg2 1 3 Wxg2 'ilVc8 14 l:tdl 'itb 7+ 1 5 f3 0-0 1 6 lLlce4 lLlxe4 17 lLlxe4 .i.e7 1 8 lLld6 (after this exchange White has the superior minor piece) 1 8 ... .i.xd6 1 9 1Ixd6 l:tbc8 20 'it'dl and White was better. 9 . . . liJd5 1 0 liJe3 The aggressive 10 lLlc6!? 'iVc7 1 1 e4 is in­ teresting. Zaitchik-A.Ivanov, Beltsy 1 977 continued 1 1 ...'ii'x c6 1 2 exd5 exd5 1 3 lLlc3 lLlf6 1 4 .i.g5 cxd4 1 5 .i.xf6 dxc3 1 6 1l.xc3 and Black once again experienced develop­ ment difficulties. There followed 1 6 ... .i.e6 17 'it'e2 'iWb5 18 'ii'e 5 �c8 1 9 a4 'ii'c 4 20 .i.fl 'iie4+ 21 'ii'xe4 dxe4 22 .i.xa6, White win­ ning back the pawn with advantage. 1 0 . . . i.b7 1O ...lLlxe5 1 1 lLlxd5 exd5 12 dxe5 .i.e7 13

4 . . . dx c 4 5 'Wi a 4 +

0-0 0-0 1 4 f4 and White has an initiative on the kingside. 1 1 liJxd5 exd5 1 2 0-0 !il.. e 7 1 3 J:td 1 0-0 1 4 'Wif5 14 �f4 l:!.c8 1 5 ..wfS is more precise. 14 . . . liJxe5 1 5 dxe5 J:ta7 !

28 . . . J:txa2 29 J:tac 1 J:t2a4 30 'ii"c 5 'Wixc5 3 1 J:txc5 J:ta 1 32 J:tcc 1 32 l:[xa 1 l:!.xa 1 33 'it>f1 d3 34 l:[d5 e4 35 f3 �xb2 36 fxe4 iLc3 is poor for White. 32 . . . J:txc 1 33 J:txc 1 e4! 34 '1t>f 1 J:ta2 35 J:tb 1 '1t>f7 36 !il.. b4 J:ta4 37 !il.. d 2 '1t>e6 38 h4 '1t>d5 Black has control of the centre and supe­ rior forces. 39 '1t>e 1 J:ta2 40 h 5 d3 4 1 !il.. c 1 !il.. d 4 42 g4 b4 43 g5 43 .td2 �xb2 44 �xb4 doesn't work for White in view of 44 .. �d4 45 .td2 .txf2+. 43 . . . 1:.a8 The threat of ... J:Ih8 forces White into a hopeless rook ending. 44 !il.. e 3 !il.. x e3 45 fxe3 J:th8 46 J:ta 1 J:txh5 47 J:ta7 g6 48 J:ta6 J:th2 49 J:txg6 J:te2+ ,?O '1t>d 1 J:txe3 51 J:tg8 J:tg3 52 '1t>d2 '1t>c4 53 b3 + '1t>xb3 54 g6 J:tg2+ 55 '1t>e3 d2 56 J:td8 '1t>c2 57 J:tc8 + '1t>d 1 0-1 .

1 6 e6?! White goes for the direct attack in the cen­ tre. Other tries don't promise an advantage, e.g. 1 6 'iVd3 'iVc7 1 7 �xd5 �xd5 1 8 'it'xd5 l:!.d8 1 9 'iib 3 l:txd 1+ 20 'it'xd 1 'iVxe5 with an equal game, or Smyslov's 1 6 e3 g6 1 7 'iVf4 f6 1 8 e6 'it'c8, when Black frees himself success­ fully. 1 6 . . . d4 1 7 !il.. x b7 J:txb7 1 8 e3 !il.. f 6 1 9 exd4 �e7 ! 20 !il.. e 3 20 exf7+ l:1fxf7 21 ..wd3 'it'xd4! 22 'iVb3 c4 hands over the initiative to Black. 20 . . .fxe6 21 'Wig4 h5! 22 'Wixh5 This is better than 22 'iVe4 cxd4 23 �xd4 l:!.d7 24 'it'xe6+ 'it>h8 25 'ii'xd7 'ii'x d7 26 .i.xf6 'it'e6 27 .td4 when White suffers on the light squares. 22 . . .cxd4 23 'Wie2 'ii"d 5! The centralization of the queen and Black's centre are more important features here than the a-pawn. 24 'ii"x a6 b5 25 'Wia5 e5 26 !il.. d 2 J:ta8 27 'Wib4 J:tea7 28 !il.. e 1 White should prefer the more adventur­ ous 28 l:1ac1 l:txa2 29 l:!.c5 'iWf3 30 l:!.dc1 ..wd3! 31 l:!.xb5 �h7, although Black retains the initiative.

Game 25 Alekhine-Junge

Prague 1942 1 d4 d 5 2 c4 e6 3 liJf3 liJf6 4 g3 dxc4 5 'ii"a 4+ liJbd7 6 'ii"x c4 a6 7 !il.. g 2

7 . . . b5 Also possible is 7 ... c5 8 dxc5 .txc5 9 'it'd3 b5!?, as played in Nogueiras-Marjanovic, Sarajevo 1 985, which continued 10 4Je5 4Jd5 1 1 4Jxd7 �xd7 12 4Jc3 4Jb4 13 'ii'b l �c6 53

Th e Ca t a l a n

1 4 0-0 Jixg2 1 5 �xg2 0-0 1 6 a 4 �d5 1 7 l:t d 1 �xc3! (from here the rest is practically forced) 1 8 l:txd8 �xb 1 1 9 lIxa8 lha8 20 axb5 �a3! 21 bxa3 axb5 22 l:tb1 Jixa3 23 Jixa3 l:txa3 and a draw was agreed. S 'iWc6 8 'ii'c 2?! is less ambitious. 8 ... c5 9 0-0 Jib7 10 �c3 l:tc8 1 1 dxc5 Jixc5 12 �3 �6 looked fine for Black in Skatchkov-Lastin, Nizhnij Novgorod 1 999. After 1 3 a4 b4 1 4 a 5 Wia7 1 5 �a4 i.d5 1 6 'it'd1 'iVb7 1 7 �e1 Jia7 18 i.xd5 �xd5 19 �d3 h5 this unex­ pected pawn advance enabled Black to create an initiative: 20 �f4 �5f6 21 �b6 i.xb6 22 axb6 h4 etc. White delayed castling in Grabaczyk­ Krasenkow, Lubniewice 1 998, so after 9 a4 Jib7 1 0 dxc5 i.xc5 1 1 axb5 axb5 1 2 l:txa8 'ii'x a8 1 3 �c3 0-0 Black was the first to complete development, seizing the initiative after 1 4 �xb5 i.e4 1 5 'it'd1 �e5 1 6 0-0 l:td8 1 7 'iVe1 ? �xf}+ 1 8 Jixf3 Jixf3 1 9 exf3 'ii'x f3 20 �c3 l:td3! 21 Jif4 �g4 22 �e4 �xf2! 23 �xf2 l:td1 ! with a decisive lead. S . . . .l:!.bS 8 ... .l:!.a7 is a bit awkward. 9 'it'c2 Jib7 1 0 0-0 c 5 1 1 a4 �a8 1 2 axb5 axb5 1 3 l:txa7 'iVxa7 14 �a3 is given as slightly better for White by Neistadt as Black still needs to ad­ dress his development. Pogorelov-Korneev, Albacete 2000 went instead 9 i.f4 i.b7 1 0 �c2 c 5 1 1 dxc5 Jixc5 1 2 0-0 0-0 1 3 �bd2 �d5 1 4 �b3 �xf4 1 5 gxf4 Jib6 1 6 l:tfd1 'iVe7 1 7 l:tac1 h6 1 8 �bd4 .l:!.e8 1 9 e3 �b8 with an unclear position in which Black has the bishop pair but White the more active pieces. 9 0-0 9 i.g5 i.b7 10 i.xf6 Jib4+ 1 1 �bd2 i.xd2+ 1 2 Wxd2 gxf6 1 3 'it'c3 c5 and White's king was stuck in the centre in Reschke­ Meijers, 2000. After 14 l:thc1 c4 1 5 We1 �b6 1 6 'iVd2 0-0 1 7 'iWf4 Wh8 1 8 'iVh4 l:tg8 1 9 Wfl f5 20 'ii'x d8 l:tgxd8 21 �e5 l:txd4 22 �xf7+ <J;;g7 23 Jixb7 'it>xf7 24 i.f3 l::td2 25 l::t ab 1 l::t b d8 Black entered the ending 54

the happier of the two. 9 . . . ii.b7 1 0 'iWc2 c5

1 1 a4 1 1 �c3 I:tc8 transposes to 8 'iWc2 c5 9 0-0 Jib 7 1 0 �c3 l:tc8. 1 1 . . . ii.xf3?! The decision to grab material is risky, lead­ ing Keres to suggest 1 1 ...l:tc8 1 2 'i¥b3 'iVb6 1 3 �a3 Jic6 1 4 �e5 �xe5 1 5 dxe5 4Jd5 with counterplay. 1 2 ii.xf3 cxd4 1 3 axb5 axb5 1 4 .l:!.d 1

1 4 . . . 'iWb6 Other continuations are good for White, e.g. 1 4 ... Jic5 1 5 i.f4 e5 16 Jixe5 4Jxe5 1 7 'ii'x c5, o r 1 4. . .e 5 1 5 e3! when the opening of the position is to White's benefit. 1 5 ttJd2 e5 In the event of 1 5 ... 4Je5 1 6 �b3 4Jxf3+ 1 7 exf3 White is in the driving seat. If Black tries to hold on to the d4-pawn he can get

4 . . . dx c 4 5 'iVa 4 +

into serious trouble, e.g. 1 7 ... l:td8 1 8 4Jxd4! .tIxd4 1 9 l:ta8+ <j;; e 7 20 .lte3 and White is winning. 1 5 ... .ltc5 1 6 4Jb3 presents White with compensation. 16 ttJb3 ttJc5?! Better is 16 ... �e7 1 7 e3 dxe3 18 �xe3 'iVe6, when 19 4Ja5! earns White compensa­ tion. 1 7 ttJxc5 Jtxc5?

23 . . J�hc8 24 e4! 'iWb3 25 11a 1 ! b4 26 .!:taS + �b5 27 .!:ta5 + �cS 28 'iVc5+ �d7 29 .!:ta7 + 1 -0 Black is mated.

Game 26 Cu . H ansen-Van Wely

Istanbul 2000

_

Inviting a brilliant reply. After the more circumspect 1 7 ...'iNxc5 1 8 .ltc6+ 4Jd7 1 9 'i'xc5 .ltxc5 20 l:ta5 White has only a modest edge. 1 8 naS ! ! 'iVxa6 1 9 'ii'x c5 'ilVe6 20 Jtc6 + 0Jd7 Alekhine gives 20 ... <j;;d 8 21 �d2 b4 22 .tIal ! 4Jd7 23 �xd7 'ii'xd7 24 l:ta7 and White is wmrnng. 21 Jtxd 7 + �xd7 22 'iVa7 + �cS 23 Jtd2!

Black's king is stranded.

1 d4 d5 2 c4 e6 3 ttJf3 ttJfS 4 g3 dxc4 5 'iVa4+ ttJbd7 S 'iWxc4 as 7 Jtg2 b5 8 'iVc6 .!:tb8 9 Jtf4 ttJd5 011- Morozevich, Groningen 1 997 went 9 ... i.b4+!? to 4Jbd2 (to i.d2 i.c5!? is inter­ esting, 1 1 dxc5 .ltb7 1 2 4Je5 .ltxc6 1 3 tDxc6 "iVc8 1 4 4Ja7 introducing a draw by repetition) to . . i.b7 1 1 'it'xc7 'iVxc7 12 i.xc7 .tIc8 13 i.f4 l:tc2 when, for the pawn sacrifice, Black has an active position. After 14 0-0 4Jd5 1 5 tDb3! l:txb2 1 6 i.d2 tDc3 17 i.xc3 i.xc3 1 8 l:tfc 1 b4 Black was doing well as the game moved into an ending, with the bishop pair and a dangerous queenside majority. Morozcvich evaluates 1 4 a3 i.xd2+ 1 5 i.xd2 l:txb2 as unclear. 1 0 Jtg5 .

1 0 . . . Jte7 Again 1O ... i.b4+?! is possible. After 1 1 tDbd2 4Je7 1 2 .ltxe7 .ltxe7 1 3 l:tc1 White controls the c-file and has the better chances, but this is a lesser evil for Black than the eccentric 12 .. .'itxe7?!, when Smyslov­ S.Polgar, London 1 996 left White better after 55

Th e C a t a l a n

1 3 'it'c2 iLb7 1 4 a 3 iLxd2+ 1 5 'iixd2 f6 1 6 l:. c 1 nc8 1 7 'iVb4+ c;t> e 8 1 8 0-0. 1 1 i.xe7 'i'xe7 1 2 ttJc3 i.b7 12 ... liJb4?! leads to wild complications: 1 3 'it'xc7! liJc2+ 1 4 c;t>dl ! liJxal 1 5 liJe5! 'iVd8! 1 6 'it'xd8+ c;t>xd8 1 7 liJc6+ (better than 1 7 liJxf7+ c;t>e7 1 8 liJxh8 b4 1 9 liJa4 b 3 and Black is doing fine) 1 7 ...c;t>c7 1 8 liJxb8 and now Monin-Vul, Kecskemet 1 992 went 1 8 ... c;t>xb8 1 9 c;t>d2 liJb6 20 l::tx al b4 21 liJdl l:td8, when Black gets his pawn back after 22 e3 e5 but walks into the clever 23 b3 exd4 24 e4 (Vul) with the better prospects for White, whose knight will find a good outpost on d3. Illescas-Onischuk, Wijk aan Zee 1 997 went 1 8 ... liJxb8 1 9 c;t>d2 :d8 20 e3 e5, when Illes­ cas proposes 21 :tc1 with advantage to White, who keeps the extra pawn. Also interesting is 1 2 ...'iVb4!? 1 3 0-0 liJxc3 1 4 bxc3 'it'd6 1 5 'it'xd6 cxd6, when the end­ ing is okay for Black. Atalik-Li Wenliang, Beijing 1 996 was agreed drawn after 1 6 a4 bxa4 1 7 l:txa4 c;t>e7 1 8 l:1fa 1 l::tb 2 1 9 e3 iLb7 20 liJe 1 iLxg2 21 c;t>xg2 l::t c 8 22 l:txa6 .l:!.xc3 23 .l:!.6a2 .l:!.xa2 24 lha2. 1 3 ttJxd5 i.xc6 1 4 ttJxe7

1 5 l:tc 1 1 5 0-0 gives Black the opportunity to launch the c-pawn. Wojtkiewicz-Kaidanov, New York 1 993 is a good example, when

56

1 5 ... iLxf3 1 6 iLxf3 c 5 1 7 dxc5 liJxc5 1 8 .l:i.fc 1 slightly favours White. The bishop i s better than the knight, but Black's king is active near the centre. There followed 1 8 ....l':thc8 1 9 .l:!.c2 liJd7 20 l::t ac1 liJb6 2 1 .l:!.xc8 .l:!.xc8 22 .l:i.xc8 liJxc8 23 iLb7 liJd6 24 iLxa6 c;t>d7 25 b3 c;t>c6 26 a4 c;t>b6 27 iLxb5 liJxb5 28 axb5 c;t>xb5 29 c;t>g2 c;t>b4 30 �f3 c;t>xb3 with a draw. 1 5 " .i.d5 ! ? Black also has 1 5 .. J:tb6 1 6 0-0 .uc8, e.g. 1 7 liJd2 iLxg2 1 8 c;t>xg2 c5 1 9 dxc5 �xc5 20 liJb3 l::t x c1 21 i:txc 1 , when the knight is bet­ ter on b3. Thanks to his control of the c-ftle White has the slightly better prospects. Gle­ izerov-Serper, Moscow 1 992 proved less attractive for White after 1 7 liJg5 iLxg2 1 8 c;t>xg2 .l:i.d6! ' 1 9 e3?! when, i f he had time for liJe4, he would enjoy a solid advantage. However, 1 9 ... c5 20 dxc5 l:td5! is enough for a small lead for Black. 1 6 l:txc7 l:thc8 1 7 l:ta7 1 7 l:txc8 l::tx c8 1 8 c;t>d2 liJf6 and Black controls the light squares in the centre and has compensation for the pawn. 1 7 " .11a8 Also possible is 1 7...l::t c 2!? 1 8 0-0 f6, again with compensation. 1 8 l:txa8 i.xa8 1 9 e3 J:td8 23 iLxa8 l:txa8 24 .l:!.c1 the idea is to generate activity but Black defends successfully with 24 ... liJd5+. 20" .i.xg2 21 ttJxg2 l:td8 22 e3 ttJe4+ 23 d2 l:tal 28 a3 liJd5 with compensation. 27 " .ttJd5 28 e4 fxe4 29 fxe4 ttJf6 30 ttJd3 l:tc4 31 ttJe5 l:tc2 + 32
4 . . . dx c 4 5 � a 4 +

Summary After 5 ... c6 6 'it'xc4 b5 7 i*'c2 �b7 8 �g2 lLibd7 the continuation 9 lLie5 is the most interest­ ing available to White, albeit one that offers nothing more than equality. Black has 9 ... lLixe5 1 0 dxe5 lLid5!, closing the dangerous h l -a8 diagonal, but perhaps 9 ...'iVb6 i s even more precise, attacking the d4-pawn and renewing the threat of ... c6-c5. There is also another defence that confl11TI s 5 'ifa4+ lacks punch. After 5 ... �d7 6 'iVxc4 .ic6 7 �g2 the game Andersson-Illescas is perhaps the most exact ment: 7 ... �d5 8 'l'd3 �e4 9 'iVdl c5 1 0 lLic3 �c6 etc. Black can, step by step, equalize in the 5 ... lLibd7 variation. As is demonstrated in Game 24, Black should not fear 6 'iVxc4 a6 7 'iVc2 c5 8 �g2. Actually this is the position from the variation 5 ... c6 6 'it'xc4 b5 7 'iVc2 �b7 8 �g2 lLibd7 with the extra tempo ... a7-a6 (as Black will play ... c6-c5 later in that line, but here do it in one go with ... c7-c5), which is therefore a definite improvement on a position considered fine for Black. It is interesting that the same posi­ tion can occur after the main continuation 8 'i'c6 :tb8 (after 9 0-0 �b7 10 'iVc2 c5) . For Black it is enough to play ... :b8-c8, which guarantees him an equal game. Therefore White most often aims to use the position of the queen on c6 for an attack on the weakened c7-square with 9 �f4. Black can sacrifice a pawn in order to activate his pieces with 9 ... �b4 (Morozevich). In Game 26, Hansen-Van Wely, we see that in the main line after 9 ...lLid5 10 iLg5 �e7 1 1 �xe7 'i'xe7 the chances are approximately even. 1 d4 d5 2 c4 eS 3 liJf3 liJfS 4 g3 dxc4 5 �a4+ liJbd7 5 ... c6 6 'it'xc4 b5 7 'iVc2 �b7 8 �g2 lLibd7 (DJ 9 0-0 - Game 19; 9 lLie5 - Game 20 5 ... �d7 6 'i'xc4 6 ... c5 7 dxc5 �c6 8 �g2 - see Game 76 (Chapter 7) 6...�c6 7 �g2 (DJ 7 ... lLibd7 - Game 2 1 ; 7...�d5 - Game 22 S �xc4 6 �g2 a6 7 lLic3 - Game 23 S . . . aS 7 .i.g2 7 'iVc2 - Game 24 7 . . . b5 8 �cS � b8 (DJ 9 0-0 - Game 25; 9 �f4 - Game 26

8. . . liJbd7

7 .i.g2

8. . . �b8 57

CHA PTER FOUR

I

The Semi-Open Catalan : 4 . . . dxc4 5 i.. g 2 i.. e 7 6 0-0 0-0

1 d4 d5 2 c4 e6 3 tDf3 tDf6 4 g3 dxc4 5 .i.g2 .i.e 7 6 0-0 0-0 In this chapter we consider the lines where Black simply completes development and refrains from defending the extra pawn on c4. This approach is the so-called Semi-Open Catalan (from now on the sq and is one of the most popular systems. Here, as a rule, the game is less sharp than strategic in nature. The position after 6 0-0 0-0 arises in the main via the move order 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 iLe7 5 iLg2 0-0 6 0-0 dxc4. But all roads lead to Rome and, once you are there, you look at the city rather than the route behind you. Now 7 ttJbd2 allows Black to defend his extra pawn with 7".b5, which is not desirable for White, of course. 7 a4 prevents ".b7-b5 but in itself is not a useful developing move. The gambit continuation 7 ttJc3 was quite popular at the end of the eighties and during the beginning of the nineties, but over the last few years this interest has cooled down. After 7".ttJc6 8 e3 it is not good for Black to play 8".ttJd7 (Game 27). Black should prefer 8".l:tb8 with the intention of ".b 7 -b5, or 8".iLd6 with ".e6-e5. Practice has proved that Black has good counter-chances here. Perhaps White has to pay attention to the sharp 8 e4 if he wants to keep this gambit line alive. 58

The 7 ttJa3 line i s a very eccentric option which leads to sharp positions that can fea­ ture odd pawn structures. For example after 7".iLxa3 8 bxa3 White's queenside is de­ stroyed. As compensation White has the bishop pair, a strong pawn centre and poten­ tial play on the semi-open b- and c-ftles. Nevertheless Black has different ways to neutralize White's initiative. The continuation 7 ttJe5 looks quite natural. White attacks the c4-pawn and increases the pressure on the long diagonal. For a long time this variation was considered to be the most effective weapon against the Sc. Then GM Kuzmin discovered an eccentric queen ma­ noeuvre that begins with 7...'it'd6. Now after 8 ttJxc4 'ii'a6 the queen both vacates the d-file for the rook and protects b7. However, White can reach an advantage in numerous ways. The old-fashioned 7".c5 has practically disappeared from modern practice. After 8 dxc5 White easily gets a slight but enduring edge. In fact in this line White can even gen­ erate a kingside offensive, e.g. 8".iLxcs 9 ttJc3 'ifc7 1 0 ttJxc4 iLxf2 1 1 ktxf2 'it'xc4 1 2 ktxf6! gxf6 1 3 iLh6 (Game 30) . In the game Gheorgiu-Robatsch, Helsinki 1 972 the Austrian GM played the unbeliev­ able 7".ttJc6!? - Black returns the pawn and sees his queenside ruined in return for

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 � g 2 � e 7 6 0 - 0 0 - 0

slightly better development and pressure down the b-file. Here White has a major choice between 8 jLxc6 bxc6 9 4Jxc6 �e8 10 4Jxe7 'iVxe7 1 1 'iVa4 (Game 31), 8 4Jxc6 bxc6 9 e3 (Game 32) and the more popular 8 ttJxc6 bxc6 9 4Ja3 (Games 33 & 34). The main line these days is 7 'tic2, with the obvious aim of restoring material equity. If Black tries to protect his extra pawn with ... b7-b5 he can quickly lose material after 8 a4 c6? 9 axb5 cxb5 1 0 4Jg5. After 7 'iVc2 c5 White has a steady edge similar to that after 7 4Je5 c5. The most common reply to 7 'iVc2 is 7 ... a6. Now White has to decide whether he wants to take on c4 immediately or first prevent ... b7-b5 with 8 a4. The 8 a4 line has, over time, become very popular. The defences 8 ... b6 and 8 ... 4Jbd7 have disappeared from recent practice, and even 8 ... 4Jc6 is seldom seen. After 9 'iVxc4 'iVd5 10 4Jbd2 l:td8 1 1 e3 Black rehabilitated 1 1 ...i.d7 in Lahner-Michenka with tactics, replying to 12 4Je5?! 4Jxe5 13 'iVc7 with 13 ... 4Jc6! etc. However, after the quiet 12 b3 Black's position is ive. In the case of 8 ... c5 (Game 35) 9 4Jbd2 cxd4 1 0 4Jxc4 4Jc6 1 1 l:tdl 4Jd5 it is interest­ ing to play 1 2 4Jd4, but it seems that this is not enough for an advantage. Instead White obtains the better prospects after 9 dxc5 i.xc5 10 tLlbd2 4Jc6 1 1 4Jxc4 'ii'e7 12 4Jfe5. Today, after 8 a4, most players prefer 8 ... jLd7 to try to place the bishop on the long diagonal. White gets very little from exchang­ ing this bishop with 9 4Je5 i.c6 1 0 4Jxc6 tLlxc6 because Black is well developed. After 8 ... i.d7 White should choose between 9 l:tdl i.c6 10 4Jc3 (Games 36 & 37) and 9 'it'c4 (Games 38-40) . The immediate 8 'iVxc4 permits Black's bishop to occupy the long diagonal after 8 ... b5 9 �c2 jLb7. The drawback of this continuation from Black's point of view is the weakening of the queenside pawns. The main plan for Black here is the advance ... c7c5. For example after 10 i.g5 Black often

plays ... c7-c5 without any trouble. Such posi­ tions tend to be equal. This is why most players now prefer to develop the dark­ squared bishop on f4 (Games 41 & 42) or d2 (Games 43-45) . After 1 0 jLf4 i.d6 or 1 0 i.d2 jLd6 the further 1 1 i.g5 makes more sense as it is difficult for Black to achieve ... c7-c5 (see the notes to Game 43). After 1 0 i.f4 w e consider 1 O ...4Jc6 i n Game 41 and 1 O ... 4Jd5 in Game 42. We discuss various possibilities for Black after 1 0 jLd2 in Game 43, and various 1 1 th moves for Black after 1 0 i.d2 i.e4 1 1 �c1 in Game 44. The main answer, the subtle manoeuvre 1 1 ...i.b7, is considered in Game 45.

Game 27 Wojtkiewicz-Dzhandzhgava

Hastings 1989/90 1 d4 d5 2 c4 e6 3 liJf3 liJf6 4 g3 dxc4 5 �g2 �e7 6 0-0 Another move is 6 4Jc3, one sample game going 6...0-0 7 4Je5 c5 8 dxc5 'tic7 9 4Jxc4 'ii'x c5 1 0 �3 4Jc6 1 1 jLe3 4Jd4 1 2 i.xd4 'ii'xd4 1 3 0-0 l:tb8, when White had some pressure on the queenside in Gelfand­ Speelman, linares 1 99 1 , although Black managed to free himself in the event of 1 4 l:tfd1 �c5 1 5 l;Iac1 jLd7 1 6 4Je4 'iib5 1 7 'ii'd 3 l:tfd8 1 8 4Jed6 'iVa6 1 9 4Je3 b 5 20 'iVa3 'ii'b 6, with an equal middlegame position. 6 . . . 0-0

59

Th e C a t a la n

7 tDc3 Also possible is 7 ttJbd2 bs S a4 c6 9 ttJes ttJds 10 e4 ttJf6 1 1 ds with a typical break in the centre in Kostic-Cvetkovic, Yugoslavia 1 993. There followed 1 1 ...iVc7! 1 2 ttJxc6 ttJxc6 1 3 dxc6 iVxc6 1 4 axbs iVxbs 1 5 es ttJds 1 6 iVg4 i.a6 1 7 ttJe4 'it>hS 1 S ttJc3, when Black sacrificed the exchange and had a good position after 1 S ... ttJxc3!? 19 i.xaS l:txaS 20 bxc3 i.b7. The advance 7 a4 ttJc6 S as looks hann­ less and Black shouldn't experience any problems. S ....l:!.bS 9 'iVc2 ttJxd4 (9 ... bs!? is interesting, , 1 0 axb6 axb6 1 1 iVxc4 i.b7 1 2 ttJc3 ttJas evaluated as equal b y Kozul) 1 0 ttJxd4 iVxd4 1 1 i.e3 'iVd6! 1 2 i.xa7 .t!.aS 1 3 i.e3 ttJds i s pleasant for Black according to Kotronias. Nor does 1 2 'iVxc4 promise White more than equality after 12 ...ttJds 1 3 i.xa7 :taS 1 4 i.d4 c s 1 5 i.c3 bs!? 1 6 iVxbs i.a6, when the situation is unclear. 7 . . . tDc6 7 ... cs S dxcs ttJc6 9 'iVa4 'iVas 1 0 'iWxc4 es!? (10 ...iVxcs 1 1 'iVxcs i.xcs 1 2 i.gs slightly favours White) 1 1 i.e3 i.e6 1 2 iVa4 'iVxa4 1 3 ttJxa4 ttJds 1 4 i.d2 f5 was Scheeren-O.Rodriquez, Eindhoven 1 9S6. It looks as if Black has compensation for the pawn, but with active play White managed to drum up an initiative with 1 5 e4! fxe4 1 6 ttJgs i.xgs 1 7 i.xgs h 6 1 S i.e3 ttJxe3 1 9 fxe3 .u.xf1+ 20 ':xfl i.xa2 2 1 ttJc3 i.c4 22 l:td1 1IdS 23 IIxdS+ ttJxdS 24 ttJxe4, with the superior ending (his knight is very good). 7 ... ttJbd7 transposes to 4 ... i.e7 5 i.g2 0-0 6 0-0 ttJbd7 7 ttJc3 dxc4. 8 e3 Also possible is S e4 :tbS 9 i.e3 (9 es ttJds 10 ttJe4 gives White a space advantage and control of the important cs-square, invit­ ing an assessment of unclear) 9 ... bs 1 0 a3 ttJas 1 1 'iVc2 i.b7 1 2 l:!.ad 1 , when White is well developed and will soon be ready for action in the centre. Filippov-Kiriakov, Mos­ cow 1 995 continued 1 2 ... a6 1 3 ds exds 1 4 exds ttJb3 1 5 ttJes l:!.eS 1 6 ttJc6 i.xc6 1 7 60

dxc6 and the bishop pair compensated for Black's extra pawn. 8 . . . tDd7 ? ! Black has other options. After S ... l:tbS 9 'iVa4 ttJb4?! 10 'iVxa7 i.d7 1 1 ttJes White won back the pawn with a superior position in Gleizerov-Berzinsh, Berlin 1 995. There followed 1 1 ...ttJfds 1 2 'iVxbS! iVxbS 1 3 ttJxd7 'iVa7 1 4 ttJxffi i.xffi 1 5 ttJxds exds 1 6 i.d2 and the rooks were stronger than the queen - 1 6 ... c6 1 7 a3 ttJd3 1 S i.c3 bs 1 9 �fd1 'iVd7 20 b3 hs 21 bxc4 dxc4 22 h4 i.d6 23 e4 etc. Sorokin-Zarnicki, Buenos Aires 1 995 went instead 9 ... i.d7 1 0 'iVxc4 ttJas 1 1 'iVe2 cs 1 2 dxcs i.xcs 1 3 ttJes i.eS 14 .l:!.d1 iVc7 1 5 ttJd3 i.d6 1 6 e4 and White was in control: 1 6 ... ttJc4 1 7 i.gs i.c6 1 S .l:.ac1 'WaS? (1 S ... ttJd7 1 9 b3 'iVas 20 i.f4 improves, with a plus for White) 1 9 h4? (after 1 9 i.xf6 gxf6 20 ttJe 1 ! bs 21 b3 i.b4 22 'iVg4+ 'it>hS 23 'Wf3 White obtains a winning position) 1 9 ...ttJes 20 i.d2 'iVa6 21 ttJxe5 with a draw. S ... i.d6 9 ttJd2 es 1 0 ttJxc4 exd4 1 1 exd4 i.g4 1 2 1ib3 (Kotronias gives 1 2 'iVa4 ttJxd4 1 3 ttJxd6 'Wxd6 1 4 i.f4 'iVcs!? 1 5 i.e3 l::t adS 1 6 i.xb7 1ib6, when Black is doing well thanks to his well placed knight) 1 2 ... ttJxd4 1 3 'iVxb7 i.f3! and now Lautier-Karpov, Dos Hermanas 1 995 went 1 4 'Wa6? i.b4 1 5 i.e3 i.xg2 1 6 %:tfd 1 cs 1 7 i.xd4 cxd4 1 8 'it>xg2 Ircs with a n advantage t o Black in view of the bishop and the somewhat ex­ posed position of White's king. Additionally, the d-pawn is strong rather than weak in this position. Karpov proposes 1 4 i.xf3! l':.bS 1 5 'iVxa7 ttJxf3+ 1 6 'it>g2 'Wd7! and Black forces the draw - 1 7 ttJxd6! ttJh4+ 1 8 gxh4 iVg4+ etc. 9 'iVa4 es 10 ds ttJb4 1 1 ttJxes leads to complications after 1 1 ...a6 12 f4 ttJbxd5 13 ttJxc4 ttJxc3 14 bxc3 i.cs (Korchnoi) with an unclear position. Not good here is 1 2 ttJf3? i.f5 1 3 ttJd4 i.d3 1 4 : e 1 i.cs 1 5 e4 ttJbxds! and Black won a pawn in Lautier­ Korchnoi, Moscow 1 994, when 1 6 ttJf5

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 i. g 2 i. e 7 6 0 - 0 0 - 0

tDxc3 1 7 bxc3 ttJg4 1 8 ttJe3 ttJxf2! was terri­ ble for White. 9 'i'e2 4Jb6 1 0 %:td 1 i.d7 1 1 e4 %:teS

1 2 i.e3 12 �f4!? is worth a try. 1 2 . . . i.fS 1 3 h4 White uses his space advantage to engi­ neer an attack. 1 3 . . . 'i'eS 1 4 i.f4 a6 1 S hS h6 1 6 a3 f6 Black has no time to push the c-pawn immediately as 1 6 ...ttJa5 1 7 ttJe5! c5 1 8 dxc5 �xc5 1 9 �g4! (Wojtkiewicz) is very promis­ ing for White. 1 7 eS fS 1 S 4Jh4 4Je7 1 9 g4! 4JbdS 20 i.d2 i.e6 After 20 ... ttJxc3 21 bxc3 fxg4 22 'i!Vxg4 White still has a dangerous attack, although this looks better than what happens in the game. 21 gxfS exfS 22 'ii'x e4 White has won back the pawn and Black still hasn't completed development. 22 . . . %:tdS 23 4JxdS! i.xdS 23 ...ttJxd5?! 24 i.h3 loses .the pawn on fS. 24 i.xdS + 4JxdS 2S 'i'd3 f4 26 'iWfS! In the ending White keeps the advantage. 26 . . . 'ii'x fS 27 4JxfS
29 4Jg6 i.gS 30 fS 32 J::!e 1 ! J::!d 7 33 l:te4 and f4 falls. 32
Game 28 Poluljahov-Nikolaev

Belorechensk 1988 1 d4 dS 2 e4 e6 3 4Jf3 4Jf6 4 g3 dxe4 S i.g2 i.e7 6 0-0 0-0 7 4Ja3

7 . . . i.xa3 Also possible is 7 ... c3 8 bxc3 c5, although this seems to help White consolide the cen­ tre. Grabuzov-Vavra, Pardubice 1 993 con­ tinued 9 ttJe5 ttJbd7 (9 ... ttJd5 10 'iib 3 cxd4 1 1 cxd4 ttJc6 12 ttJxc6 bxc6 1 3 e4 ttJb6 1 4 �e3 i s a shade preferable for White) 1 0 ttJac4 ttJxe5 1 1 ttJxe5 'fIc7 1 2 'iVb3 l:td8 1 3 61

Th e C a t a la n

.i.f4 lLlhS 1 4 .i.e3 lLlf6 1 S :fd 1 and White's central control was enough for an edge. Black also has 7 ... .i.d7 8 lLlxc4 .i.c6 9 .i.f4 lLlbd7 1 0 11c1 .i.dS 1 1 lLlfeS iLxg2 1 2 'it>xg2 lLlxeS 1 3 iLxeS cS (Krasenkow-Gagarin, USSR 1 998) 1 4 iLxf6 gxf6 1 S dxcS iLxcs 1 6 e 3 with a minimal advantage to White in view of Black's damaged structure. 7 ... cS and now 8 lLlxc4 is a transposition from 7 lLlbd2 cS 8 lLlc4, Shtyrenkov­ Mikulchik, Zlin 1 995 continuing 8 ...lLlc6 9 dxcS .i.xcs 1 0 .i.f4 lLldS 1 1 l::. c 1 !? lLlxf4 1 2 gxf4 'iVe7 1 3 lLlceS lLlb4 1 4 a 3 lLldS 1 S e3 iLb6 16 'itb3 .i.d7 17 lLlxd7 'iVxd7 1 8 lLleS when, due to the Catalan pawn cluster with £2, e3 and f4, White had a firm grip on the centre. In Kozul-Ivanovic, Yugoslavia 1 989 White took the other c-pawn: 8 dxcS iLxcs 9 lLlxc4 lLlc6 1 0 a3 as 1 1 iLgS h6 1 2 'iVxd8 lIxd8 1 3 iLxf6 gxf6 14 .l:tfd1 .i.d7 1 S lLlfd2! and (again) Black's weakened pawns tipped the balance in White's favour, while White's more active forces are also significant. 8 bxa3 b5 Another possibility is 8 ... lLldS 9 'iYc2 bS 10 i:tb 1 a6 1 1 i:td 1 h6 12 lLleS iLd7 13 a4 lLlc6 14 .i.a3 with compensation for the pawn in Kamsky-Gogoladze, Ivano­ Frankovsk 1 988. The earlier encounter Sveshnikov-Ubilava, USSR 1 98 1 went 8 ... lLlbd7 9 a4 as 1 0 .i.a3 Ue8 1 1 .l:lc1 .l:ta6!? (1 1 ...lLlb6 12 lLleS iLd7 13 iLxb7 iLxa4 14 'it'd2 and Black keeps the pawn but at the cost of damaging his own pawns, and White also has the bishop pair) 12 .l:txc4 lLlb6 1 3 .l:tcs iLd7 1 4 lLleS lLlxa4 1 S iLxb7 lLlxcs 1 6 iLxcs iLc8 1 7 lLlc6 'iYd7 1 8 lLlb8 'iVbs 1 9 iLxa6 .i.xa6 2 0 lLlxa6 "iVxa6 2 1 'iVc2 and the bishop and superior formation left White on top. 9 a4 An interesting plan is 9 lLlgS!? c6 1 0 e4 h6 1 1 eS lLldS 12 lLle4, White's knight exploiting the weakened dark squares in Black's camp. Kinsman-Stefansson, Stockholm 1 993 saw White emerge from 1 2 ... f5 1 3 exf6 lLlxf6 1 4 62

lLlxf6+ 'iVxf6 1 S .i.f4 'ii'e 7 1 6 a4 iLb7 1 7 .l:te1 with definite compensation. 9 a6 1 0 �a3 �e8 1 1 ttJe5 ttJd5 1 2 e4 . . .

1 2 ttJb6 In the event of 12 . ..ttJ f6 White can sacri­ fice a piece with 1 3 lLlxf7!? 'it>xf7 1 4 eS lLldS 1 S 'ii'hS+ 'it>g8 1 6 .i.e4 g6, when Krasenkow­ Arbakov, Moscow 1 989 witnessed 1 7 iLxg6! hxg6 1 8 'iWxg6+ 'it>h8 1 9 i:tfe 1 lLlc3 (White's threat was l:te4) 20 .l:te3 .i.b 7 (20 ... 'it'd7 21 "iVhS+ 'it>g7 22 :0 lLldS 23 g4! - Krasenkow - wins for White) 21 I!.xc3 lLld7?! 22 'iYh6+ 'it>g8 23 g4 lie7 24 l:th3 with a dangerous attack. Black can improve with 2 1 ...lLlc6!, forcing the draw: 22 'iVh6+ 'it>g8 23 "iVg6+ 'it>h8 etc. (Mitenkov) . Note that Black must accept the second sacrifice, e.g. 1 7 ...l:te 7?! 1 8 iLxe7 'ii'x e7 1 9 .i.e4 (also strong i s 1 9 iLxh7+!? "iVxh7 2 0 'iiVe8+ 'it>g7 21 'iVxc8 and Black's forces lack co-ordination) 1 9 ...lLlc6 20 f4 'iif7 , Kuzmin-Anand, Frunze 1 988. Then Anand gives 21 .i.xh7+! 'iYxh7 22 'iVe8+ 'it>g7 23 f5! iLb7 24 f6+ 'it>h6 2S f7 with a near decisive lead. If this is not your style, then there is the quieter 1 3 l:[b1 iLb7 1 4 :e1 , after which Black must play precisely to maintain the balance. Gleizerov-Zhukhovitsky, USSR 1 986 continued 14 ... lLlc6 1 S lLlxc4 'iVxd4 1 6 'it'c2 "iVxc4 (16...'iWa7 1 7 lLld2 i s interesting) 1 7 'it'xc4 bxc4 1 8 .l:txb7 and now instead of 1 8 ... lLleS?! 1 9 .l:tc1 .l:tec8 20 f4 lLlc6 21 .l:tbb1 as 22 l:txc4 lLlb4 23 iLxb4 axb4 24 l:tbxb4, . . .

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 SL g 2 SL e 7 6 0 - 0 0 - 0

Black should have played 1 8 ...ttJd4 1 9 e5 lDd5 with chances for both sides. White has the bishop pair to compensate the two knights and the ed pawn on the c-file. 1 3 a5 An interesting option is 13 'it'h5!? f6 14 lDg4 c6 15 as ttJa4. 1 3 . . . 4J6d7 1 4 f4 But now the same manoeuvre is not to be recommended for White: 14 'it'h5?! ttJxe5 1 5 dxe5 �d3 1 6 Jte7 (1 6 jLc5 and the queen sacrifice 1 6 ...ttJd7 1 7 lladl ttJxc5!? is good) 16...l:txe7 1 7 llad 1 c5 1 8 l:txd3 cxd3 and Black has the better chances. The pawn pha­ lanx is very strong. 14 . . . SLb7 1 5 'ifh5

1 5 . . . g6 15 ...ttJxe5 16 fxe5 'iVxd4+ 17 'it>hl wins a pawn for Black but the f-file is opened for attack. 1 6 'ifh6? A poor move. White finds himself with the initiative after 1 6 'it'h3 ttJxe5 1 7 dxe5 'i'd4+ 1 8 'it>hl jLxe4 1 9 .l:tadl 'iVe3 20 l:tfel 'i'xa3 21 jLxe4. 1 6 . . . 4Jf6? 16 ... ttJxe5 1 7 dxe5 'iVd4+ 1 8 'itth l jLxe4 19 lladl jLxg2+ 20 'itt xg2 'iVe4+ 21 'it>h3 ltJc6 wins for Black. 1 7 d5 c3 The capture 17 ... exd5 18 exd5 Jtxd5 1 9 l:tadl c 6 20 Jtb2 simply creates attacking chances for White.

1 8 SLb4 exd5 1 9 exd5 SLxd5 20 SLxc3 SLxg2 21 4Jg4! 21 'it'xg2 'iVd5+ 22 llf3 is a more quiet approach, the dark-squared bishop promising White good chances. Nevertheless, White's attack in the game looks formidable. 21 . . . 4Jh5 22 l:tad 1 ! SLd5 23 f5! l:te2 24 4Je3 c6 25 fxg6 fxg6 26 4Jxd5 cxd5 27 l:txd5

27 . . . 4Jd7?? The oruy route is 27 ...'iVe8 28 l:txh5 'iVe3+ 29 ..wxe3 l:txe3, when White keeps an advan­ tage in the endgame due to his more active forces after 30 l:tc5 ttJd7 31 l:I.c7 ltJe5 32 jLxe5 l:txe5 33 l:tm l:tc8 34 l:txh7 l:txc7 35 l:txc7 b4 36 l:tc6 l:txa5 37 l:txg6+ 'it>f7 38 l:tg4 l:txa2 39 l:txb4 as. 28 l:txh5! 'ife7 29 'iVxg6 + ! 1 -0

Game 29 Khalifman-Portisch

Rqkjavik 199 1 1 d 4 d 5 2 c4 e 6 3 4Jf3 4Jf6 4 g 3 dxc4 5 SLg2 SLe7 6 0-0 0-0 7 4Ja3 SLxa3 8 bxa3 SLd7 After 8 ... ttJc6 9 jLb2 Y.Mikhalevsky­ Pigusov, Ubeda 1 997 went 9 ... ttJd5 10 l:I.cl ttJb6 1 1 e4 i.. d7 1 2 l:te l , when White's cen­ tral control served as sufficient compensation for the pawn. There followed 12 ...ttJe7 1 3 ttJe5 i..b 5 1 4 'iVh5 'iVe8 1 5 Jth3 f5 1 6 'iVg5 h6 1 7 'iNh4 l:tf6 (17 ... ttJg6 1 8 ttJxg6 'iVxg6 1 9 63

Th e Ca t a l a n

dS is well balanced) 1 8 �g2 c 6 1 9 d S ! ? cxdS 20 exdS ttJexdS 21 ttJxc4 �xc4 22 �xf6 ttJxf6 23 l::txc4 ttJxc4 24 'ii'x c4 .u.c8 25 1ib3 cJ;; f7 26 'ii'xb7+ 'ii'd 7, the smoke clearing to reveal a level ending. Another path is 9 ...!tb8 1 0 'ii'c 2 bS 1 1 llad1 when Korchnoi-o.Rodriquez, Barce­ lona 1 992 continued 1 1 ...ttJe7 (1 1 ...�b7?! 1 2 ttJgS! h6 1 3 dS! ttJb4 1 4 axb4 hxgS 1 5 dxe6 'ii'e 7 1 6 exf7+ 1hf7 1 7 �xb7 l:hb7 1 8 'ii'g6! left Black with weak pawns .in Glek-Klovans, Tashkent 1 987) 1 2 e4 �b7 1 3 l:tfe 1 ttJg6 1 4 h4. White's latest thrust i s quite typical. Hav­ ing completed development and with control of the centre, White begins a kingside offen­ sive: 1 4 ... .:te8 1 5 ttJh2 'ii'e 7 1 6 'ii'e2 eS 1 7 hS ttJf8 1 8 dxeS ttJ6d7 1 9 ttJg4 h6 20 f4 with a menacing initiative.

9 lLle5 Zaichik-Soffer, Rishon Ie Zion 1 998 went 9 'ii'c 2 �c6, Black surrendering the c4-pawn to complete development. This policy proved successfill after 10 1fxc4 ttJbd7 1 1 �gS �dS 1 2 'ii'b4 b6 1 3 !tfd1 h6 1 4 �xf6 'ii'xf6 1 5 !iac 1 llfc8 1 6 'ii'b 1 �b7 1 7 e4 'ilie7 1 8 'iiid3 ttJf8 1 9 dS l:td8 20 'ii'b 3 .l:tac8, when Black had a solid position.

9 . . . .ltc6 1 0 lLlxc6 lLlxc6 1 1 .ltb2

1 7 �xeS for White, whose bishops dominate in this wide open position. Another try is 1 1 .. .'ilid6!? 12 ltc1 bS 13 e4 eS! 14 f4 l:tad8! 1 5 dS (1 5 fxeS ttJxeS 16 'ii'c2 'ilVb6 favours Black, whose knight is heading for d3) 1 S ... exf4 1 6 gxf4 ltfe8. Here White's centre looks dominating, but Wnite must keep his eyes peeled, e.g. 1 7 cJ;; h 1 ttJxe4! 1 8 �xe4 l:txe4 1 9 dxc6 'ilixc6 2 0 'ii'f3 l:td3 21 'iit'g2 'iit'g6 with the superior ending for Black. White was more circumspect in Alburt­ Geller, New York 1 990, albeit only for a few moves: 1 7 'ii' c2 ttJe 7 1 8 'ii'f2 0g4 1 9 'iiig3 f5 20 llcd1 'ii'b6+ and now 21 �d4? met with 2 1 . . .'ii'g6! 22 �f3 hS! and Black assumed control because White was unable to hold on to the centre and his king was also suscepti­ ble to attack. Geller offers 21 cJ;; h 1 as an improvement, when 2 1 . . .'ii'e 3!? 22 �f3 c3 23 �c1 'ilics 24 �xg4 fxg4 25 'ii'xg4 ttJg6 is unclear.

1 2 l:1.b 1 Glek-Klovans, Frunze 1 988 went 1 2 l::tc 1 ttJb6 13 e3 'ilid7 14 1fc2 ltab8 1 S l:tfd1 l:iJe7 16 � f1 'ii'a4 17 'iit'b 1 , White uninterested in the exchange of queens as this would relieve some of the pressure. After 1 7 .. .l::tfd 8 1 8 e4 lld6 1 9 'ii'a 1 f6 20 J:te1 ttJc6 21 �h3 .l:!.f8 22 eS!? fxeS 23 dxeS %:td2 24 �xe6+ 'It>h8 25 �f1 'ii'b s 26 �c3 l:td3 the situation remained unclear.

1 2 . . . lLlb6

1 1 . . .lLld5 1 1 . ..'ilid7?! looks wrong. KozulKir.Georgiev, Sarajevo 2002 went 1 2 'ilVa4 ttJxd4 1 3 'ilixc4 l:.ad8, when Georgiev pro­ poses 14 l:tfd 1 eS 1 5 e3 'iiie6 1 6 'ii' xe6 ttJxe6 64

1 2 ... bS?! 1 3 e4 ttJb6, as in Krasenkow­ Timoscenko, Tashkent 1 987, appears dubi­ ous. Krasenkow analyses the following: 1 4 d S ! ttJe7 1 5 'ii'g4 ttJg6 1 6 dxe6 with the more pleasant game for White, or 1 4 ... exdS 1 5 exdS ttJe7 1 6 d 6 'ii'xd6 1 7 'ilixd6 cxd6 1 8 �xa8 ttJxa8 1 9 lHd 1 when Black has two pawns for the exchange but White's activity and Black's vulnerable pawns combine to favour the first player.

1 3 e3 'iid 6 1 4 'iic 2 .l:!.fd8 1 5 J:!.fd 1 J:!.ab8 1 6 e4 e5! 1 7 d5 This is better than 1 7 dxeS?! 'it'cs 1 8 e6 fxe6 1 9 l:txd8+ l::tx d8 20 eS ttJd4 21 i.xd4

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 iL g 2 iL e 7 6 0 - 0 0 - 0

.l:!.xd4 with an aggressive stance for Black.

17 . . . ttJd4! 1 8 iLxd4 exd4 1 9 J:txd4 c6 The greedy looking 19 ......xa3 20 e5 'iVc5 is an option.

20 a4 'ii'c 5 20... cxd5 2 1 exd5 g6 22 as ifc5 23 'iWc3 tZlxd5 24 ifxc4 ifxc4 25 l:txc4 was seen in Krasenkow-Flear, Paris 1 990, the Catalan bishop affording White the lead.

retained excellent winning chances: 33 Wg4! 'iVg6+ 34 Wf3 �f6+ 35 We4 g6 36 .ltd3 iff5+ 37 Wd4 iff6+ 38 Wc5 "'e5+ 39 <>t>c6 "'f6+ 40 Wc7 ife7+ 41 Wb8 "'d8+ 42 <>t>xa7 "'xa5+ 43 "'a6 ifc5+ 44 Was etc.

White plays 7 ttJe5

Came 30 Psa khis-Stefansson

21 'ii'c 3 cxd5 2 1 . . .ttJxa4 22 'iVxc4 'iVxc4 23 l:Ixc4 ttJb6 24 dxc6 ttJxc4 25 c7 .l:!.dc8 26 cxb8'ii l:txb8 (Korpics-Gyimesi, Balatonbereny 1 992) leads to a similar minor piece endgame, with the better prospects for White.

Winnipeg 1997 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 iLg2 iLe7 6 0-0 0-0 7 ttJe5

22 exd5 J:tbc8 23 J:tb5

7 . . . c5 23 . . . 'ii'e 7? ! 23. . .ttJxa4 2 4 l:Ixc5 ttJxc3 2 5 lldxc4 nxc5 26 l:txc5 ttJxa2 and a draw was agreed in Nesis-Andonov, Correspondence 1 997.

24 a5 ttJa4 25 'ii'c 2 'ii'e 1 + 26 iLf 1 ttJc3 27 J:txb7 ttJxd5 28 J:txc4 With 28 l:Id1 ! 'ili'e5 (28 .. .'iVxa5? loses to 29 'i'f5!) 29 .ltxc4 White keeps an advantage.

2B . . . ttJe3 ! ? Black has to try something. The straight­ forward 28 . . .l:Ixc4 29 �xc4 'ili'xa5 30 �a6 leads to a very difficult ending.

29 fxe3 J:txc4 30 'ii'xc4 'ii'x e3 + 3 1 'itg2 J:td2+ 32 'ith3 'ii'h 6+ 33 'ii'h 4? % - % Inviting Black to force perpetual check. However, with the following amazing king march White could have kept the piece and

Another option is the rather unorthodox 7 .. .'iWd6 8 ttJxc4 �a6. Tukmakov-G.Kuzmin, Tallinn 1 998 continued 9 �3 �d8 10 ttJc3 ttJc6 (1 O .. .l::txd4? 1 1 ttJb5 l:Id7 12 ttJe5 and Black loses the exchange) 1 1 e3 ttJa5 1 2 ttJxa5 'iix a5 1 3 .ltd2 'iWb6 1 4 'iVc2 .ltd7 1 5 b4 .ltc6 1 6 ttJa4 .ltxa4 1 7 �xa4 e 5 1 8 .ltc3 e4 19 J:tfc 1 �e6 20 �c2 .ltd6 21 b5 l:te8 22 .ltd2 h5 23 h3 a6 24 a4 and White had the initiative.

8 dxc5 iLxc5 It appears that Black's queen is awk-wardly placed after 8 . . .�c7 9 ttJxc4 ifxc5, as was demonstrated in Pinter-Kallai, Hungary 1 988: 10 b3 l:td8 1 1 ttJbd2 ttJc6 12 .ltb2 l:Ib8 1 3 l:tc1 with a definite lead for White. This in­ creased after 13 . . .ifb5 14 a3 .ltd7 1 5 b4 .lte8 16 e4 h5?! 17 h3 ifg5 1 8 e5 ttJh7 19 'iVe2 h4 65

Th e C a t a la n

20 liJe4 � 6 2 1 liJcd6 f5 22 liJc5, when White's forces occupied influential posts.

9 lLle3 'fIe7 Vaulin-Feoktistov, Moscow 1 998 took a more sober course after 9 ... 'iVd4 to 'iWxd4 .i.xd4 1 1 liJxc4 liJc6 1 2 liJb5 e5 1 3 e3 i.c5 14 b3 liJe8 (14 ... i.e6 allows White to put a knight on d6: 1 5 liJbd6 l:tab8 1 6 i.b2 etc.) 1 5 .i.a3 .i.xa3 1 6 liJbxa3 f6 17 .i.d5+ �h8 1 8 .i.xc6 bxc6 and White 'sacrificed' the Catalan bishop in order to inflict upon his opponent a weakened queenside. This pro­ vided a target in the corning ending after 1 9 liJa5 i.d7 2 0 l:tfdl l:t f7 2 1 l:tac 1 l:.c8 22 �g2 etc. 9 ... 'iVe7 allows the pin on the knight with to i.g5, when Ribli-Gopal, Calcutta 1 992 continued 10 ... h6 1 1 .ixf6 'iVxf6 12 liJxc4 liJc6 1 3 .ixc6! bxc6 1 4 liJe4 'iie 7?! 1 5 liJe5 'iic7 1 6 liJxc5 'iVxe5 1 7 'iVc2 l:tb8 1 8 b3 'iVh5 1 9 l:tfdl Itb5 20 .l:i.ac 1 as 21 ':'d6 and Black's weaknesses on the queenside were more important features in the position than the weakened light squares around White's king. Black can restrict his opponent to a slight edge with 1 4 ...'iVf5 1 5 liJxc5 'iWxc5 1 6 'iWa4.

(we are familiar with this manoeuvre by now; White is slightly better) 14 ... .id4 1 5 liJe4 .i.a6 1 6 e3 .ixc4 17 'iVxc4 .ixb2 1 8 nab 1 'iVa3, which was seen in Vaulin-Stanojevic, Nis 1 996. There followed 1 9 .ixf6 gxf6 (1 9 ... .i.xf6 20 liJxf6+ gxf6 21 'iVg4+ �fS 22 'ir'h5 �g7 23 .l:i.b7 .l:i.fS 24 J::td l gives White a dangerous initiative) 20 'iWxc6 f5?! (20 .. J:tac8 21 �7 l:tb8 22 'iVc7 and White shifts the queen to the kingside) 21 liJg5 .if6 (2 1 . ..l:tac8 22 'i¥b7 with double threats on b2 and f7) 22 liJxe6! J::ta c8 23 liJxd8 l:txc6 24 liJxc6 with a very nice ending for White, his two rooks being more than a match for the queen and Black's pawns being weak.

1 1 J:!.xf2 'ifxe4 1 2 J:txf6 ! gxf6 1 3 �h6 lLld7 ! The natural looking 1 3. . .liJc6? i s in fact poor in view of 1 4 e3 .l:!.d8 1 5 'it'h5 e5 1 6 liJe4 'iVe6 1 7 'iVh4 �h8 1 8 liJxf6 liJe7 (Thin­ ius-Comp 'Mephisto', Berlin 1 992) 1 9 .i.fS!, when White wins at least a piece.

1 4 lLle4

1 0 lLlxe4

1 0 . . . �xf2 + Black should avoid following to . . . l:td8 1 1 .if4 with l 1 . . .e5? 1 2 .i.xe5! .i.xf2+ 1 3 l:txf2 .uxd1 + 14 .l:!.xdl 'iVe7 1 5 .ixf6 gxf6 16 liJd5 with a decisive lead for White. This leaves 1 1 . ..�e7 1 2 'iVc2 liJc6 1 3 .i.xc6 bxc6 14 i.g5 66

Maintaining the tension poses the most problems for Black. The straightforward 14 .i.xfS 'iitx fS 15 'it'd6+ �e8 16 .l:i.d 1 'iWc5+ 1 7 'ii'x c5 liJxc5 1 8 liJb5 'iit e 7 1 9 l:tc1 liJd7 is unclear as White is active but Black can hold on to his pawn for a while .

1 4 . . . 'fIb4! 1 4. . .l:td8? runs into 1 5 liJxf6+ 'iit h8 1 6 e3 and Black loses the f6-pawn for nothing: 1 6 . . . e5 1 7 �h5! 'iie6 ( 1 7 ... liJxf6? 1 8 'i'g5

Th e S e m i- Op e n Ca t a la n : 4 . . . dx c 4 5 � g 2 � e 7 6 0 - 0 0 - 0

ttJe8 1 9 'ii'x d8 wins for White) 1 8 ttJxh7! 'it'g6 1 9 'ii'h4 l:f.e8 20 l:If1 ! 'ii'x h7 2 1 �e4 (psakhis) and White has a menacing attack.

'It>d8 28 'iVfS+ 'litc7 29 'iid6+ �xc8 30 'ti'd7 mate!

26 'iVdS + 'it>g7 27 'iVg5 +

1 5 �c 1 f5 Black defends well. Again 1 5 . . . .l:!.d8? is punished after 16 ttJxf6+ 'It>h8 17 'it'd3! 'iVb6+ 1 8 �e3 ttJxf6 19 'ti'c3! 'i&'d6 20 'iix f6+ 'it>g8 21 l:Ifl ! l:Id7 22 l:tf4 'It>fS 23 J:f.g4 and White wins.

1 6 ltJd6 White should force matters with 1 6 .lrLxfS!, e.g. 16 ... ttJxfS (1 6 ... 'litxfS?! 17 'iVd6+ 'iVxd6 1 8 ttJxd6 ttJb6 1 9 J::t c 7 .lrLd7 20 lhb7 clearly favours White, while 1 6 . . .'iVxfS 1 7 ttJd6 'ikd8 18 e4 is also nice) 17 ttJf6+ 'litg7 (1 7 . . . 'lith8 1 8 'iVd8 is worse) 1 8 'iVd8 'iib6+ 1 9 'iixb6 axb6 20 ttJe8+ 'it>g6 21 ttJd6 1i.d7 22 a3 and White wins back the pawn with interest in this end­ game thanks to the queenside majority and active pieces.

Yo - Yo

Game 3 1 G elfand-Aseev

Kla;peda 1988 1 d4 d5 2 c4 e6 3 ltJf3 ltJf6 4 g3 dxc4 5 �g2 �e7 6 0-0 0-0 7 ltJe5 ltJc6 S �xc6 bxc6

16 . . . ltJe5! 1 7 �xfS 't'ib6 + 1 S e3 ! ? O r 1 8 'lit h 1 ttJg4 1 9 ttJxc8 ttJf2+ 2 0 'litg 1 ttJh3+ and Black can force the draw.

1 S . . . 't'ixe3+ 1 9 'it>h 1 ltJg4 20 �c2 'it>xfS 9 ltJxc6 9 ttJc3?! c5 10 dxc5 .lrLxc5 1 1 �a4 ttJd5! 12 ttJe4 ttJb6 13 'iic 2 was the dubious (for White) course of Korchnoi-Petrosian, Can­ didates '/dinal Match (game 1), Ciocco 1 977. Now 1 3. . .'ilVd5!? 1 4 ttJg5 f5 15 1l.f4 .lrLd6! (Kotronias) allows Black to keep both his pawn and the light-squared bishop. His chances are better.

9 . . . �eS 1 0 ltJxe 7 +

Black has left the worst of his problems behind him.

21 ltJxcS ltJf2 + 22 �xf2 �xf2 23 �xb7 �bS White has two minor pieces for the rook but the exposed position of his king forces him to take a perpetual.

24 �dS + 'it>g7 25 'iVg5 + 'it>hS Not 25 ... 'litfS?? 26 'i&'h6+ 'lite8 27 1l.c6+

1 0 'iVa4?! has been tried. Pigusov­ Naumkin, Belgrade 1 988 continued 1O . . . 1l.d6 1 1 'iix c4 a5 1 2 ttJe5 c5 1 3 ttJf3 1i.a6 14 'iVc2 .l:!.c8 and Black had the initiative: 1 5 ttJc3 cxd4 16 ttJxd4 1l.e5 17 .l:!.d 1 and now 17 . . . ttJd5 (pigusov) leaves Black on top.

1 0 . . . 't'ixe7 1 1 't'ia4 Also possible is 1 1 b3 cxb3 1 2 1i.a3 b2 1 3 �xb2 l:Id8 1 4 'ilVc2 l:tb8 with an equal posi­ tion. White has the better pawn structure but Black is active.

1 1 . . . c5 67

Th e C a t a l a n

Black has several other options. The queen manoeuvre 1 1 ...'iVd6 12 lidl 1Va6 isn't quite satisfactory in this position, either. White had the better prospects after 1 3 'iVxa6 .Jtxa6 1 4 ltJc3 .:tab8 1 5 e4 h6 1 6 f3 :fd8 1 7 :b 1 ltJd7 1 8 .Jte3 ltJb6 1 9 .:tbel in Malakhov-Dervishi, Montecatini Terme 1 998. Black's pawns on the c-me are weak but b2 is also an object of attack, while White enjoys more space. Stohl-Polak, Czechoslovakia 1 990 saw 1 1 ...a5 12 1Vxc4 .Jta6 13 'it'c2 .ufd8 14 ltdl c5! (this freeing pawn break secures Black equal chances) 1 5 dxc5 .uxd 1+ 1 6 'it'xdl 'iVxc5 17 ltJc3 ltJg4 18 'it'fl and now Kotronias gives 1 8 ...!lc8!? 1 9 .Jtd2 'iVh5 20 h3 ltJe5 21 1Vg2 ltJc4 22 .Jtel 'iVfS 23 e4 'ii'f6, when Black's activity fully compensates the missing pawn. 1 1 .. .e5 1 2 dxe5 1Vxe5 is also possible, and leads to complex play. Korchnoi-Kotronias, Haifa 1 989 went 1 3 ltJc3 ltJe4 14 'iVxc4 ltJxc3 1 5 'iVxc3 1Vxe2 1 6 nel 'iVb5 1 7 J:te5 'iVd7 1 8 .Jtf4 .Jtb7 1 9 llc5 !lfd8 20 f3 'iVd3 21 .Jtxc7 l:td7 when the opposite coloured bishops offered Black adequate counterplay. The other option is 1 3 'iVxc4, when 1 3 ... .Jte6 leads to a further branch. Black has compensation after 14 1Vc2 .JtfS 1 5 'iVd2 :fe8 1 6 ltJc3 ltJe4 17 ltJxe4 1Vxe4 due (again) to the presence of opposite coloured bishops. This leaves 1 4 1Vd3 !lad8 1 5 'iVe3 'iih 5 1 6 f3, e.g. 1 6 ... .Jtc4! 1 7 ltJc3 ltfe8 1 8 'iVg5 'iih 3 1 9 It £1 (1 9 .Jte3 .Jtxe2!? 20 ltJxe2 l:td5 21 'iVf4 ttde5 22 .Jtxa7 Ihe2 23 .Jt£1 l:hb2 24 a4 and this pawn looks like a killer but Black generates counterplay on the op­ posite flank with 24 ... h5 - Kotronias) 1 9 ... .Jtxe2!? and now 20 .Jte3?! was Kachiani­ Gersinska-Bojkovic, Belgrade 1 990, when Black could have secured the advantage with 20 ... .Jtxf3! 21 ltxf3 ltJg4, forcing White to part with material due to the threat against h2. However, 20 .Jtf4 .Jtd l !? 2 1 .Jte5 h6 22 'iVf4 ltJd5 23 'it'c4 is unclear. 1 6 ... 1tfe8 has also been played, although the lines seem to 68

favour White. 1 7 'iVg5 i s too slow but went unpunished in Gelfand-Timoscenko, Sverd­ lovsk 1 987, when after 1 7 ...1Vh3?! 1 8 'iVc5! .Jtd5 19 e4 the only way to break the domi­ nation of the light squares was to sacrifice, but this did not prove correct after 1 9 ... .uxe4?! 20 fxe4 ltJg4 21 1:t£1 ltJxh2 22 'iVxc7 .uc8 23 .l:txh2 .l:txc7 24 ltxh3 11xc1+ 25 'it>£1 .Jtxe4 26 g4, when White was winning. Black should follow the example of Pigusov­ Raetsky, Voronezh 1 988, 1 7 ... .Jtc4 1 8 'iVxh5 ltJxh5 1 9 lte 1 .uxe2 20 ltxe2 .Jtxe2 21 'it>£1 .Jtc4 22 ltJc3 ltJf6 23 .Jte3 leading to a nice ending for White in view of the better pawn structure and the presence of knights on the board.

1 2 'ii'x c4 1 2 'it'a3 .Jtb7 1 3 ltJc3 ltfc8 is equal.

1 2 . . . cxd4 1 3 'ii'x d4 e5 After 1 3 ... ,Ud8!? 1 4 'iVh4 J::tb 8 1 5 b3 .Jta6 1 6 ltJc3 'iVc5 1 7 ltJe4 ltJxe4 1 8 'iixe4 the exchange of knights should guarantee Black good counterplay due to the presence of opposite coloured bishops. The subsequent 1 8.J:tb4 1 9 1Ve3 1Vc2 20 .Jta3 l:te4 21 ltfel ii'xe2 resulted in a draw in Krasenkow-Ribli, 1 998.

1 4 'ii'h 4

1 4 . . . 'ii'e 6 Hiibner-Handke, 2001 contin­ ued 14 ...ltb8 1 5 b3 ltd8 16 .Jtb2 h6 17 ltJa3 .Jtg4 1 8 ltJc4 .Jtxe2 1 9 1tfe 1 .Jtxc4 20 'it'xc4 ltd5 21 J:tad l and White's forces had finally

Th e S e m i- Op e n Ca t a la n : 4 . . . dx c 4 5 iL g 2 iL e 7 6 0 - 0 0 - 0

ed the battle. After 2 1 . . .Ita5 22 .ic3 l:k5 23 'iVd3 White's bishop and queenside ma­ jority afforded him the better prospects. White can try to repair the light squares with 15 ttJc3 .l:tb4 16 e4 .ib 7 1 7 f3 (1 7 .ig5 ':xb2 18 i.xf6 'iVxf6 1 9 'iVxf6 gxf6 20 l:tab 1 leads to an equal ending) 1 7 ...'iVc5+ 1 8 �g2 l:td8 19 l:t£1 lld3, when Black is active but White seems solid. Shipov gives 20 .ig5 .l:txc3!? 2 1 bxc3 ttJxe4!, breaking the light-squared domination and forcing White to find 22 .te3, when 22 ... 'ii'd 5 23 'iie 7 h5 24 .l:td 1 !? 'iVxd1 25 'iixb4 ttJx£1 26 'iVxb7 ttJd3 27 'iVc8+ forces the draw.

Itxg3 ttJe6, again with compensation.

1 5 ttJe3 iLb7 1 6 e4 l:tfe8

Necessary is 37 e5 'iVd3! 38 'iVxf4 l:t£1 39 ttJe4 lhf3 40 ttJf6+ �g7 41 ttJh5+ �h7 with equality.

Worth considering is the plan to transfer the knight to d3 with 1 6 ...ttJd7!? 1 7 g4 ttJc5 18 f3 ttJd3 (Aseev) with compensation.

28 ':'xg2 ttJg6 29 'iiVg 4 'iiVa 6 30 'iiVg 3 iLe8 3 1 nf2 31 .tg5 allows the nice trick 3 1 ....ih3!? 32 'iVxh3 ::'xc3! etc.

31 . . . 'iiVd 6 32 iLg5?! Preferable is 32 l:tf1 .ia6.

32 . . . ttJf4 33 iLxf4 exf4 34 'ilfh4 iLe6 35 l:tf 1 l:td2! 36 l:tg 1 g6? Black misses a great opportunity in 36 . . J::tx b2 37 'ii f6? .ig4!! 38 'iix d6 .ixf3+ 39 l:tg2 l:!.xg2 40 'iVd8+ �h7 41 'iVh4+ �g6 and wins.

37 'iiVf 6??

37 . . . .i.g4! 0-1

1 7 f3 'iiVb 6 + 1 8 l:tf2 The continuation 1 8 'it>g2 h6 19 g4 .ta6 20 l:td1 l:td8 21 'iV£1 gives White a small advantage.

1 8 . . . h6 1 9 'it'g2 l:td8 20 g4 l:td3 2 1 g 5 hxg5 22 iLxg5 ttJ h 7 23 l:tg 1

See the lines In the note to Black's 36th move.

Game 32 Pi ket-Adams

23 . . . l:te8

Wyk aan Zee 2000

23 ... ttJxg5 24 'iVxg5 lld6 25 �h 1 1lg6 26 'iVd2 (Gelfand, Kapengut) is also a little bet­ ter for White.

1 d4 d5 2 e4 e6 3 ttJf3 ttJf6 4 g3 dxe4 5 iLg2 iLe7 6 0-0 0-0 7 ttJe5 ttJe6 8 ttJxe6 bxe6 9 e3

24 'it'h 1 l:te6 25 iLe1 ttJf8 26 l:tfg2 l:tg6

9 'iia4 'iixd4 10 ttJc3 .id7 1 1 i.e3 'iVg4 1 2 .if3 'iVh3 1 3 'irVxc4 e5 is quite different, Khurtsidze-Moiseenko, Ukraine 1 999 con­ tinuing 14 l:tfd1 ttJg4 1 5 .ixg4 'iVxg4 1 6 'ii'xg4 .txg4 1 7 f3 .te6 1 8 b3 as, when the

Black enjoys some activity and therefore has compensation for the pawn.

27 'iiVh 5 l:txg2 Less forcing is 27 ... 'iVf6 28 l:tg3 .l:txg3 29

69

Th e C a t a l a n

bishop pair and possibilities o n the queenside left Black in charge. 9 SLxc6 l:tb8 10 tiJc3 SLb7 1 1 SLxb7 l:txb7 12 l:tb l 'iWd7! 1 3 e4 l:tfb8 is equal according to Gulko.

9 . . . ttJd5 9 ...SLa6 1 0 SLxc6 c3 (1 O ...l:tb8 1 1 'ii'a4 .l:tb6 1 2 Itdl favours White) 1 1 SLxa8 cxb2 12 SLxb2 SLxfl 1 3 SLf3 and White's superior pawn structure is the main difference.

1 0 'ilVa4

1 1 . . .J:f.b8 After 1 1 ...'iVd7 12 tiJd2 i.a6 1 3 .l:td1 l:tfd8 1 4 b3 Black should prefer the lesser evil 1 4... 11ab8 1 5 tiJxc4 tiJxc4 1 6 bxc4 c5 1 7 SLa3 cxd4 1 8 :xd4 'iWe8 (Ribli) 1 9 SLxe7 'iNxe7 20 'i¥a4, when White has something to bite on on the queenside, to 14 ... cxb3 1 5 'iVxc6 'ii'xc6 16 SLxc6 I1ab8 17 axb3 SL b 7 1 8 SLxb7 l:txb 7 1 9 SLa3 SLxa3 20 l:txa3 e5!?, when a draw was agreed in Cs.Horvath-Yu Shaoteng, Gyula 2000, but where White could have had the slightly better prospects with 21 tiJf3 exd4 22 lhd4 l:txd4 23 tiJxd4 (due to the weaknesses on a7 and c7) .

1 2 J:f.d 1

1 0 . . . ttJb6 10 ... a5 1 1 l:td l 'ii'd 6 12 'it'xc4 SLa6 1 3 'iNc2 tiJb4 1 4 'it'd2 e 5 1 5 a3 tiJd3 1 6 SLe4 tiJxc1 1 7 l:txc 1 exd4 1 8 exd4 SLf6 1 9 l:txc6 'iNxd4 20 'it'xd4 SLxd4 21 tiJc3 and here a draw was agreed in Ippolito-Gormally, Lon­ don 1 999. The ending is equal; Black has the bishop pair while White has the better pawn structure.

1 1 'ilVc2 1 1 �xc6 l:tb8 1 2 'iib 5 SLb7 1 3 SLxb7 i:txb 7 with play for both sides. Meins­ Chandler, 2000 saw the alternative 1 1 'ifa5 SLd7 1 2 tiJa3 SLd6 1 3 SLd2 'ife7 1 4 .u.fc1 l:[fb8 1 5 l:tab 1 e 5 1 6 tiJxc4 exd4 1 7 exd4 tiJd5!? 1 8 tiJe3 (1 8 SLxd5 cxd5 1 9 'iWxd5 SLe6 20 'iVc6 SLf5 21 nel 'ii'f6 is unclear) 1 8 ... l:tb5 1 9 'iWa4 h6 20 'ii'c 2?! tiJxe3 21 SLxe3 SLf5 22 'iVxc6 l:!.ab8 23 l:tal l:!.xb2 with ad­ vantage to Black. Also possible is 1 1 ...i:tb8 1 2 tiJa3 SLd7 1 3 SLd2 i.d6 1 4 l:tfc 1 'iVe7 1 5 �ab 1 c5 1 6 dxc5 SLxc5 1 7 b4 SLd6 with equality in Wells-Arlandi, Escaldes 1 998. 70

1 2 SLxc6 SLb7 1 3 i.xb7 l:txb7 is equal.

1 2 . . :iVe8 Markowski-J aracz, Warsaw 2001 went 1 2 ... a5?! 1 3 tiJd2 SLa6 1 4 b3 cxb3 1 5 tiJxb3 SLb5 1 6 a4! SLxa4 1 7 �xa4 tiJxa4 1 8 tiJxa5 tiJb6 1 9 tiJxc6 'ii'd7 20 tiJxb8 l:txb8 21 i.d2 SLd6 22 l:[al h5 23 h4 'iNe7 24 'iVc6, White's bishop pair and pawn structure contributing to his lead .

1 3 ttJd2 c5 Kotronias gives 13 ...SLa6 14 b3 c5 1 5 i.b2 cxd4 1 6 SLxd4 f6!?, planning ... e6-e5.

14 dxc5 14 tiJxc4 leads only to a level game after 1 4 ... cxd4 1 5 exd4 (1 5 llxd4 SLf6 is also equal) 1 5 ... tiJxc4 1 6 'ifxc4 i.d6.

1 4 . . . iLxc5 1 5 ttJxc4 ttJxc4 1 6 'ii'xc4 'i'b5 1 7 'ilVxb5 J:f.xb5 1 8 b3 iLb7 1 9 iLxb7

Th e S e m i - Op e n Ca t a la n : 4 . . . dx c 4 5 ii g 2 ii e 7 6 0 - 0 0 - 0

l:.xb 7 20 iib2 iid6 2 1 l:.ac 1 21

lbf6 1 9 e3 as 20 .i.f3 h6 21 a4 etc.

White has the superior pawns.

1 0 bxa3 iia6

. . . .l::. b 6 22 e4 f6

An interesting line is 10 ... l:tb8 1 1 'iia4 'iVxd4 1 2 lLe3 'iie 5 1 3 'iixa7!?, e.g. 1 3. .. l:tb7 (13. .. lLd7 1 4 lLd4 'it'xe2 1 5 l:tae l 'iVxa2 1 6 lLxf6 gxf6 1 7 'iVd4 .i.c8 1 8 'iixf6 l:tb5 1 9 l:te4 h 5 20 g4 saw White step up the pace in Istrate-Livner, Rimavska Sobota 1 996) 1 4 'it'd4 'iih 5 1 5 lLf3 e5 1 6 'iNxc4 lbg4 1 7 .i.xg4 'it'xg4 1 8 l:tac1 with chances for both sides. In Raetsky-Namyslo, Biel 1 994 White ig­ nored the a7-pawn, settling for equality after 1 3 li'xc4 lbd5 1 4 lLd4 .i.a6! 1 5 .i.xe5 lLxc4 1 6 l:tfel f6 17 lLd4 l:tfd8, when White had the bishop pair but Black's forces were well placed.

22 ... e5 23 l:td5 f6 24 l:tc4 slightly favours White.

23 iid4 23 f4!? is an interesting try. 23 .. J;:ta6 24 l:.c2 .l:!.b8 25 Wf 1 .l:!.b5 26 'it'e2 Wf7 27 .l:!.d3 We8 28 a4 .l:!.b7 29 .l:!.c4 iie7 29 ...'it>d7 30 e5 fxe5 31 lLxe5 g6 32 .l:.h4 h5 33 .l:.f4 We7 34 .i.f6+ 'it>e8 35 l:te4 helps White a little as it results in the further weak­ ening of Black's pawns.

30 e5

%-%

White is still slightly better, but the players agreed to call it a day.

Game 33 Ribli-Bonsch

Thessaloniki 1988 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 iig2 iie7 6 0-0 0-0 7 ttJe5 ttJc6 8 ttJxc6 bxc6 9 ttJa3

11

iixc6

Black is better after 1 1 'iia 4?! lLb5 1 2 'iia 5 'iVxd4! 1 3 lLe3 'iid 6 1 4 a4 'iNa3, but a more viable, albeit rather tame option is 1 1 lLg5 h6 1 2 lLxf6 'iVxf6 1 3 'iVa4 (1 3 lLxc6 l:tab8 14 'iVa4 l'1b6 1 5 11fdl lId8 16 .i.f3 c6! led to a plus for Black after the further moves 1 7 Wg2 li'e7 1 8 e3 l:tc8 1 9 h4 .i.b5 20 'ifb4 c5 in Antunes-Karpov, Tilburg 1 994, Black with a ed pawn and White's dou­ bled a-pawns of no use) 1 3 . . . lLb5 14 'iia 5 c3 with equality according to Karpov. 11

. . . .l:!.b8 1 2 'iWa4 .l:!.b6

9 . iixa3 . .

After this exchange the pawn structures of both sides are totally ruined. Another move is 9 ... lbd5?!, when 1 0 lbxc4 c5 1 1 dxc5 lLxc5 12 b3 should favour White. Ths was borne out in Ribli-Wolf, 1 989 after 12 .. :iVe7 1 3 lLb2 l:tb8 1 4 'iVc2 l:td8 1 5 l:tfd l .i.a6 1 6 a 3 .i.b6 1 7 I1ac 1 .i.xc4 1 8 'it'xc4

1 3 iig2 13 lLf3 is a Raetsky speciality. Then 13 ... lbd5 14 'iVa5 brings us to a crossroads: 14 . . .'�·f6?! . 1 5 a4 .i.b7 (1 5 ... 'it'xd4 16 .i.a3 71

Th e C a t a l a n

l:.a8 1 7 � c 5 and White wins the exchange) 1 6 �a3 l:tb8 1 7 �c5 nb2 1 8 nab 1 c3 1 9 Ihb2 cxb2 20 l:tb1 'figS 2 1 �a3! �c6 22 'iVc5 �xa4 23 e4 was excellent for White in Raetsky-Sammalvuo, HafnarfJordur 1 998. White has control of the centre and Black cannot hold on to b2 forever. 14 ... f5 fights for e4. Raetsky-Schoenthier, Mannheim 1 998 was unclear after 1 5 �d2 'iVd7 16 lIab 1 l:tfb8 17 lIxb6 .l:txb6 1 8 a4 'iVd6 19 l::t c 1 , when White's bishop pair and piece play compensated for the ed pawn. 14 ... c3! is the best move. Black sacrifices the pawn in order to activate his the pieces. In Raetsky-AI.Schneider, Cappelle la Grande 1 999 there followed 1 5 l:te 1 f5 1 6 �xd5 'fixd5 1 7 'fixc3 l:tc6 1 8 'iib 2 11b6 1 9 'iVc3 and a draw was agreed.

.l:!.d 1 h 6 2 6 'ii'b 3 ttJxc5 Another good option is 26 ... ttJb6.

27 dxc5 'i'e7 28 .l:!.d6 'ii'f 6 Black has to play precisely to avoid an in­ ferior position. 28 ...l:ta4!? 29 l:tbd 1 l:tc4 30 'iVe3 'fif6 looks well balanced.

29 .l:!.bd 1 .l:!.a4 30 �xc6 ! ?

1 3 . . :�c8 1 4 �d2 .l:!.d8 1 5 �b4 1 5 1Ifc1 lIxd4 1 6 �e3 c5 1 7 �xd4 cxd4 is given by Ribli. Black has clear compensation for the exchange due to the strong centre pawns.

1 5 . . . c3!

A temporary piece sacrifice that should lead to equality with accurate defence from Black.

30 . . . �xc6 3 1 .l:!.xc6 .l:!.xc6 32 'i'xa4 .l:!.xc5 33 'ii'x a7 White has the easier game thanks to the a­ pawn.

33 . . J:tc2 34 a4 'i'f3 34 . . . .:.xa2?? loses to 35 'iVb8+ 'it>h7 36 'iib 1+.

35 .l:!.f 1 g6 36 'ii'b8 + 'iii> h 7 37 'i'b3 :c3 Black manages to avoid 37 ... 'iVxb3? 38 axb3 l1b2 39 Ita1 11xb3, when 40 a5 l:tb7 41 a6 lla7 42 'it>f1 'it>g7 43 'it>e2 'it>f6 44 'it>d3 'it>e 7 45 'it>c4 'it>d6 46 'it>b5 is close to win­ ning for White.

38 'i'b2 e5 16 .l:!.fc 1 �xe2 1 7 �c5 .l:!.a6 1 8 'i'c2 �b5 The redeployment of the bishop with 1 8 ... �c4 1 9 'iVxc3 �d5, with equality, is a good idea.

1 9 'ii'x c3 ttJd5 20 'i'd2 20 'iVe 1 !? is interesting.

20 . . . 'ii'd 7 2 1 .l:!.ab 1 c6 22 h4 White seizes space on the kingside.

22 . . . ttJb6 23 'ii'e 1 .l:!.c8 24 'ii'b4 ttJa4 25 72

Missing the best defence in 38 . . .l::t a3 39 lIc1 llxa4 40 l:tc8 'iVd1+ 41 'it>h2 'fid4 with equality.

39 .l:!.e 1 'ii'c 6? ! A clear path t o equality i s 3 9 . . . e4 40 'iVe2 (40 l:he4?? 'fid1+ 41 'it>g2 l:tc1 and Black has a decisive attack) 40 ... 'ilfxe2 41 l:txe2 11c4, when the rook ending is completely drawn.

40 'ii'b 5! 'i'c7?

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 ii.. g 2 ii.. e 7 6 0 - 0 0 - 0

After the inaccuracy o n the 39th move Black could still play 40 .. .'ii'c 5 41 'it>g2 l:tc2 42 'iix c5 l:txc5 43 .l:!.b1 and, despite the fact that White is better here, Black has good chances of saving the game. After the text the struggle is probably over.

the opposite coloured bishops ensured that Black was okay. Browne gives 12 'iVxc4 .ia6 1 3 'iVc2, when White has the upper hand he has the bishop pair and the black pawns are very weak.

41 l:txe5 l:tc1 + 42 'It>g2 .l:I.d 1 ?

12 ... .ia6 13 'it'c5 c6 14 .ig5 f6 15 .id2 and White is better.

42 ...l:tc2 43 h5 is a slight improvement for Black.

43 l:td5 l:te 1

44

l:td7 �c 1 45 �d5 1 -0

Game 34 Manor-Anand

London 1987 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 .1g2 ii.. e 7 6 0-0 0-0 7 ttJe5 ttJc6 8 ttJxc6 bxc6 9 ttJa3 ii.. x a3 1 0 bxa3 ttJd5 1 1 �a4

1 2 �xc6 l:tb8

1 3 �c5 ii.. b 7 Posting the bishop more actively with 1 3 . . . .ia6 seems to work out well for Black here, e.g. 14 e4!? c3!? (14 .. .'it'd7?! 1 5 .if4 l:tfc8 1 6 :fd1 4Ja4 1 7 'iVg5 l:tb2 1 8 d5 h6 1 9 'iVg4 h 5 20 'iVxh5 4Jc3 2 1 .ie5 4Jxd1 22 .uxd 1 was difficult for Black in Sorg-Oettel, Schwabisch Gmuend 1 996, but 17 ... 4Jb2 1 8 l:td2 c 3 1 9 .l:tc2 .l:tb5 20 d5!? i s interesting) 1 5 .l:!.e1 4Ja4 1 6 'WaS .ib5 1 7 l:tb1 c 6 and Black is very solid. Also possible is 14 .if4 4Jd5 1 5 .ixd5 'iVxd5 1 6 'Wxa7 l:tb6 1 7 .ixc7 .ib7 1 8 f3 �a6 (1 8 ... l:tb2? 1 9 :tab 1 lIa8 20 e4 'iVb5 21 'iVb6 'iVxb6 22 .ixb6 l:!.xb1 23 l:txb1 f5 24 .ic5 and White had good chances thanks to his two extra pawns in Palatnik-Razuvaev, Thilisi 1 973) 1 9 'iVc5 1:[c8 20 'iVxd5 .ixd5 21 .if4 when a draw was agreed in Ftacnik­ Kurajica, Dortmund 1 98 1 . However, the following line is slightly better for White: 21 .. .c3! 22 e4 .ic4 23 l:tf2 1:txa3 24 .id6 :ta6 25 .ic5 e5 26 l:tc1 exd4 27 .ixd4 .l:txa2 etc.

1 4 e4 1 1 . . . ttJb6 Preferable to 1 1 ...4Jc3?! 1 2 'iVc2 (1 2 'iVxc4? 'iVxd4! and Black wins material) 12 ... 4Jb5 1 3 e3 4Jd6 1 4 a4 as 1 5 .ia3 'iVd7 16 .ixd6 cxd6 1 7 'iVxc4 .ib7 1 8 l:1fc1 lUc8 19 :tab 1 l:ta7 20 l:tb3, when Black's extra pawn was less important than his develop­ ment problems in Vera-Moran, Alcobendas 1994. In fact he was under considerable pres­ sure after 20 ... c5 21 'iVb5 .ixg2 22 'it>xg2 g6 23 dxc5 dxc5 24 Ihc5 etc. 1 1 ...a5 12 'iVxc6 .ia6 13 'iVc5 was played in Alburt-Browne, Philadelphia 1 989. After 1 3. .. 'iVd6 1 4 'iVxa5 c3 1 5 .ixd5 'ii'x d5 1 6 'i'xc3 .ixe2 chances were even - once again

1 4 . . . �d6 Another option is 14 ... f5 15 f3 fxe4 1 6 73

Th e C a t a l a n

fxe4 .l:txf1+ 1 7 'it>xfl tDa4 (17...'it'f6+?! 1 8 'it>g 1 l:t fS 1 9 i.f4 was seen in Raetsky­ AI. Schneider, Megeve 1 994, when White would have been better after 1 9 ...'it'g6 20 :te l) 1 8 'iVe5 'it'd6 and White is slightly bet­ ter due to the bishop pair. An improvement is 1 4".'iVd7, an unclear situation resulting after 1 5 i.f4 l:tbc8 1 6 l:tfd 1 h6 1 7 l:td2 l:tfd8 1 8 l:tc1 i.a6 1 9 d5 exd5 20 exd5 i.b7 in G.Agzamov-Sveshnikov, Tashkent 1 980.

White plays 7 'ilfc2 a 6 8 a4

Game 35 Filippov-Sulskis

Poland 1999 1 d4 d 5 2 c4 e6 3 ltJf3 ltJf6 4 g3 dxc4 5 i.g2 i.e7 6 0-0 0-0 7 'ii'c 2

1 5 i.f4 'ii'x c5 1 6 dxc5 ltJa4 1 7 .l:!.fc 1 i.a6 1 8 i.f 1 ltJxc5 1 9 i.xc4 .l:!.b1? Much better is 1 9" .i.xc4 20 l:txc4 tDa6 2 1 l:tac 1 1:tfd8 22 l:ta4 l:tb6 2 3 i.xc 7 tDxc7 24 l:txc7 a6 25 l:tb4 l:tbb8!, as in Raetsky­ Ulfarsson, Reykjavik 1 996. White is better here, but Black can fight for the draw. 1 9 "J1fc8 has also been played. Raetsky­ Naumkin, Budapest 1 99 1 continued 20 i.xa6 tDxa6 21 l:tc6 l:tb6 22 l:tac 1 'it>fS 23 i.e3 l:txc6 24 .l:txc6 tDb8 25 l:tc4. Again White's rook and bishop are stronger and more ac­ tive than Black's forces. There followed 25".a6 26 i.f4 'it>e7 27 l:txc7+ l:txc7 28 i.xc7 tDc6 29 f3 �d7 30 i.b6 tDe5 31 'it>f2 tDc4 32 i.c5 when the endgame with an extra a­ pawn was clearly better for White, thanks in no small part to the bishop. Hjartarson­ Schussler, Gausdal 1 985 saw instead 26".c6! 27 i.xb8 l:txb8 28 l:txc6 l:td8! 29 l:txa6 lId4 30 l:ta8+ 'it>e7 31 e5 g5!, this time with an edge for White, but fighting chances for Black. Returning to 1 9 " .l:tfc8, another option is 20 i.e3 i.xc4 21 .l:txc4 tDd7 22 l:tac 1 , e.g. 22".c5!? 23 i.xc5 tDxc5 24 l:txc5 l:txc5 25 l:txc5 when White is in the lead but will have a hard time trying to win the race, or 22" .l:tb7, when 23 .l:tb4! l:txb4 24 axb4 tDb6 25 l:tc5 'it>fS 26 i.f4 c6 27 .l:ta5 l:ta8 28 l:ta6 was very pleasant for White in Chekhov­ Baikov, Moscow 1 979.

Kochiev-V.Orlov, St. Petersburg 1 999 went 7".b5 8 a4 i.a6 (8".c6? 9 axb5 cxb5? 1 0 tDg5 and White i s winning due t o the threats against a8 and h7) 9 axb5 i.xb5 1 0 tDe5 tDd5 1 1 tDa3 i.xa3 1 2 bxa3 c3 1 3 a4 i.a6 1 4 i.a3 l:te8 1 5 l:tfc1 f6 1 6 tDd3 tDc6 1 7 e 3 with an edge for White as Black will find it diffi­ cult to hold on to the c3-pawn. After 7" .c5 8 'iVxc4 cxd4 9 tDxd4 e5 1 0 tDb3 tDc6 1 1 i.e3 i.e6 1 2 'iib 5 White has some pressure on the queenside but Black is well developed. White still needs to get the rest of this forces into play. White was more successful in Kengis-Meijers, Riga 1 984: 8 dxc5 i.xc5 9 'iVxc4 �e7 1 0 .i.g5 tDbd7 1 1 tDbd2 h6 1 2 i.h4!? g5 1 3 tDxg5 hxg5 1 4 .i.xg5 'iWd6 1 5 'iWh4! 'iVd4 1 6 i.f4 .i.e7 1 7 'iVg5+ �h8 1 8 l:tac 1 ! and White had created a strong attack for the sacrificed piece.

20 i.e3!

8 a4

White wins the exchange.

20 . . . i.xc4 21 i.xc5 i.d3 22 i.xf8 'it'xf8 23 f3 'it'e7 24 .l:!.c3 i.a6 25 .l:!.ac 1 'it'd7 26 a4 'it'd6 1 -0 74

7 . . . a6

Also possible is the complex 8 i.g5 b5 9 i.xf6 i.xf6 1 0 tDg5 i.xg5 1 1 i.xa8 'iVxd4 1 2 i.g2 tDd7 1 3 tDc3 f5! 1 4 b 3 ! cxb3 1 5 axb3 'YtVc5 1 6 l:ta2 i.e 7 1 7 e3 .i.d6 as in Korchnoi-

Th e S e m i- Op e n Ca t a la n : 4 . . . dx c 4 5 � g 2 � e 7 6 0 - 0 0 - 0

Vaganian, Montpellier 1 985. For the ex­ change Black has two pawns and decent pieces, inviting an evaluation of unclear. Shabalov-Ivanchuk, Novosibirsk 1 986 went 8 ltJbd2 b5 9 ltJg5 l:ta7 10 b3 cxb3 1 1 ltJxb3 ..tb7 1 2 ..txb7 l:txb7 1 3 e4 ltJfd7! 1 4 ltJ f3 c 5 1 5 dxc5 ltJxc5 1 6 ltJxc5, when Sha­ balov gives 1 6 ...l:!.c7 1 7 ltJxe6 l:!.xc2 1 8 ltJxd8 .l:!.xd8 as equal.

S c5 . . .

Quite different is 8 ...ltJc6 9 'it'xc4 'ii'd 5 10 tLlbd2 (1 0 'it'd3 l:!.d8 1 1 ltJc3 'iih 5 12 'iVc4 ..td7 1 3 i.f4 i.d6 1 4 i.g5! is an attractive alternative for White, who has more room for manoeuvre) 1 O ....:td8 1 1 e3, e.g. 1 1 ...i.d7 (1 1 ...'ii'h 5 1 2 e4 ..td7 1 3 b3 b5 14 'it'c3 i.e8 15 axb5 axb5 1 6 1ha8 l:!.xa8 17 i.b2 l:!.a2 1 8 'it'c 1 ! 'ii'h 6 1 9 h3 g6 20 l:tel 'ii' rn 2 1 'fibl tDb4 22 ltJe5 ltJd7 23 i.c3 favoured White in Kramnik-Piket, Dortmund 1 995) 1 2 ltJe5 (12 b3 is also good) 12 ... ltJxe5, when White should prefer 1 3 'ii't 3 ltJf3+ 14 i.xf3 'it'fS 1 5 'iVxc7, with insufficient compensation for Black, to the continuation in Lahner­ Michenka, Karvina 2001 , where 1 3 'it'xc7? tLlc6! 14 ..txd5 ltJxd5 1 5 'iWxb7 ltJa5 1 6 'iVxa8 Iha8 left Black better because the rook and two pawns were no match for the bishop and knight. After 9 'ii'xc4 Black has also tried 9 ...ltJb4. For example Filippov­ A.Rodriguez, unares 1 997 continued 1 0 tLlc3 b 5 1 1 'ii'b 3 bxa4 1 2 ltJxa4 l:tb8 1 3 l:td 1 ..tb7 1 4 ltJc3 'it'c8 1 5 ..tf4 ltJd3 1 6 l:txd3 ..txf3, when Rodriguez gives 17 'it'a2!? i.xg2 18 'it>xg2 'ii'b7+ 1 9 �gl 'it'xb2 20 'it'xa6 as somewhat better for White in view of the superior structure.

9 ltJbd2 After 9 dxc5 ..txc5 10 ltJbd2 ltJc6 1 1 tDxc4 'iJle7 it seems that 1 2 ..tg5 is good enough only for equality, e.g. 1 2 ... h6 1 3 ..txf6 'i'xf6 1 4 l:!.fdl e5 1 5 e3 ..tfS 1 6 iVb3 e4 1 7 tDfd2 'ii'e 7 1 8 ltJb6 l:!.ad8 1 9 ltJd5 'ife6 20 tLlf4 'iVxb3 21 ltJxb3 and the game is level, or 18 'ifc2?! ltJb4 1 9 'iVc3 l:!.ac8, when Black enjoyed more space in Anastasian-Asrian,

Yerevan 2000. In fact 20 'ife5? 'it'xe5 2 1 ltJxe5 ltJc2 2 2 l:!.abl ltJxe3! 2 3 fxe3 ..txe3+ 24 'it>f1 ..txd2 25 l:!.xd2 e3 saw Black win material. Another option is 1 2 b3 ltJd5 1 3 ..tb2 l:!.d8 14 1:tac 1 ltJdb4 1 5 lic3 ltJd4 16 ltJxd4 ..txd4 1 7 'it'xd4!? .uxd4 1 8 i.xd4 and White had a rook and bishop for the queen in Aro­ nian-Bick, Ubeda 2000. This seems okay here as Black is weak on the dark squares and lags behind on the queenside. The situation re­ mained unclear after 1 8 ...ltJc6 1 9 i.e3 e5 20 ltJb6 l:!.b8 21 i.c5 'ii'f6 22 ltJd5 etc. The natural 12 ltJfe5 is White's most prorrusmg try for advantage. Gulko­ I vanchuk, unares 1 990 was a shade better for White after 1 2 ... ltJxe5 1 3 ltJxe5 ltJd5 1 4 ltJd3 ..td6 1 5 ..td2 ltJf6 1 6 :tac 1 , while 1 2 ... ltJd4 1 3 'it'd l l:!.d8 1 4 e3 ltJfS 1 5 'it'c2 receives a similar assessment.

9

. . .

cxd4 1 0 ltJxc4 ltJc6 1 1 l:!.d 1 ltJd5

1 2 ltJxd4 ! ? ltJdb4 1 3 ltJxc6 ltJxc2 1 4 ltJxdS ltJxa 1 1 5 ltJxb 7 �xb 7 Interesting is 1 5 ... ltJb3 1 6 ltJba5 ltJxa5 1 7 ltJxa5 l:!.b8 1 8 ltJc6 l:!. b7 1 9 b 4 and White has some compensation for the exchange. Also good enough for Black is 1 5 . . ..l:!.b8 16 ..tf4 ..txb7 1 7 i.xb8 l:!.xb8 1 8 ..txb7 l:!.xb7 1 9 11xal 11b4 20 l:!.c1 l:!.xa4 21 ltJb6 l:!.d4 when the ending is even.

1 6 �xb7 tradS 1 7 l:!.f 1 Or 1 7 .l:.xd8 Ihd8 1 8 i.f4 with compen­ sation. The a6-pawn is difficult to protect. 75

Th e Ca t a l a n

17 ... ttJb3 18 iLxa6 J:ta8 1 9 iLb 5 ttJd4 20 iLe3

3 1 . . .J:tb 7 32 J:td3 'iit h 6 33 f4 f6 34 J:te3 iLe5 35 J:txe6 J:txb3 36 J:te6 J:tb5 37 J:te 7 iLb4 3 8 e4 J:te5! N o t 38 ... l:txhS? 3 9 ttJe3 �f8 4 0 l:t f7 , and White wins the g7 -pawn and with it the game.

39 J:txe5 iLxe5 40 'iitf 3 'iit x h5 White has only a modest advantage.

41 ttJe3 'iit g 6 42 ttJf5 'iitf 7 43 'iit g4 iLg 1 44 h3 iLe5 45 ttJh4 iLf2 46 ttJf3 iLe5 47 ttJe 1 iLf2 48 ttJd3 iLe3 49 f5 From here White tries hard to win the game but his efforts are met with some im­ pressive defence. . .

20 . . . ttJxb5 20 ... ttJxe2+?! takes the wrong pawn. Black's majority on the kingside will not be very dangerous compared to White's a- and b-pawns. After 21 'iitg2 l:!.ad8 22 .Jtb6 White has the better chances.

21 axb5 J:tfb8 22 b6 J:tb7 23 J:tb1 iLb4 24 iLf4 h5 25 iLe7 h4? ! Too ambitious. After 2S ...g6 the situation is about equal; Black returns the exchange at the right moment.

26 gxh4 J:te8 27 l:I.d 1 'iit h 7 28 'iit g 2 iLe7 29 h5?! Simple and good is 29 l:td7.

29 . . . J:texe 7 30 bxe 7 J:txe 7 31 b3

49 ... iLb6 50 ttJf4 iLe7 5 1 'iitf 3 iLd6 'iite 3 iLe5 53 ttJg6 iLe7 54 'iitd 4 iLb8 'iit e 5 iLg3 56 'iitd 5 iLh2 57 'iit e 6 'iite 8 ttJh4 'iite 7 59 ttJf3 iLf4 60 ttJd4 iLe3 ttJe6 g6 62 ttJe7 'iitf 7 63 'iitd 7 gxf5 exf5 iLd4 65 ttJb5 iLb2 66 ttJd6 + 'iitf8 'iit e 6 iLd4 68 ttJb5 iLb2 69 ttJe7 iLe5 ttJd5 'iit g 7 71 ttJe3 iLe3 72 ttJg4 iLd4 h4 iLe3 74 h5 iLb2 75 ttJxf6 'iit h 6 ttJg4+ 'iit x h5 77 ttJe5 'iit h 6 78 f6 'iit h 7 f7 'iit g 7 Yo - Yo

52 55 58 61 64 67 70 73 76 79

Game 36 Ftacnik -Dutreeuw

Batumi 1999 1 d4 d5 2 e4 e6 3 ttJf3 ttJf6 4 g3 dxe4 5 iLg2 iLe7 6 0-0 0-0 7 'ilVe2 a6 8 a4 iLd7 9 J:td 1

White is tw o pawns up and has genuine winning chances, although the win will not come easily. 76

In Urban-Rozentalis, Lubniewice 1 998 White unleashed the bishop with 9 ttJeS, when there followed 9 ... �c6 10 ttJxc6 ttJxc6. Then 1 1 �xc6 bxc6 12 l:!.d1 'iVdS 13 ttJa3 iVhS! is nice for Black, so Urban chose 1 1 e3 ttJaS 1 2 ttJd2 cS 1 3 dxcS 'iVc7 1 4 ttJxc4 'iVxcs 1 5 b3 .l:'tac8 1 6 iVb 1 ttJxc4 1 7 bxc4 ':c7 1 8 �b2 ttJd7 1 9 a s l:!.fc8 2 0 �a3 'iYgs 21 h4 'iYf6 22 �b2 'iYh6 and the game was still well balanced. The b 7 -pawn is a bit weak, but so is c4 and, while Black's knight will find a good outpost on cS, White has the bishop pair. 1 3 ...l:tc8 has also been tried here. After

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 iL g 2 iL e 7 6 0 - 0 0 - 0

14 b4 cxb3 I S lbxb3 lbxb3 1 6 'iWxb3 lbd7 17 c6 bxc6 I S l:tdl White emerged with compensation in the form of the bishop pair and activity in Anastasian-Lutz, Batumi 1 999, the subsequent I S ... 'iVc7 19 .i.d2 lHdS 20 as liJcs 21 'ilVc2 .t:!.bS 22 i.c3 adding to the mix. 9 . . . iLe6

1 2 d5 exd5 1 3 e5 ttJe4 1 4 ttJxd5 iLxd5 1 5 J:txd5 ttJee5 1 6 axb5 'ilfeS 1 7 J:td 1 ttJd3 1 S e6! White crashes through. 1 S . . . ttJb6 Dubious is IS ... fxe6?! 19 'iVxc4 liJ3cS 20 bxa6 and White has a winning position. 1 9 exf7 + J:txf7 20 iLe3 axb5 21 J:txaS ttJxaS 22 ttJd4 ttJb6 23 ttJxb5 Black's pawns are simply too weak. 23 . . . 'ilfd7 24 ttJe3 e5 25 b3 'ii'e 6 26 bxe4 ttJb4 27 �e4! �xe4 The lesser evil is 27 ... 'iVxe4 2S liJxe4 liJxc4 29 i.xcs, although this is anyway poor for Black. 2S iLh3! �xe3 29 J:tdS + ! 2 9 . . . iLfS 3 0 iLe6 'ilfa 1 + 3 1 �g2 1 -0 White has too many threats.

10 ttJe3 10 'iWxc4 i.dS 1 1 'iVc2 lbc6 12 lbbd2 lbb4 13 'itOl cS 14 e4 i.c6 I S eS lbd7! 1 6 lbe4 h6 17 dxcS .i.xe4 IS 'ii'x e4 lbxcs 19 l:txdS liJxe4 20 l:txaS l:txaS is level according to Ribli. White's Catalan bishop is good but Black's knights are well posted and White's queenside might prove vulnerable. However, Black would prefer to see 17 b3?! .i.dS I S liJc3 .i.xf3 1 9 .i.xf3 cxd4 20 l1xd4 'iWc7 a s in Donguines-Villamayor, Calcutta 2001 , when the weakness on eS tipped the scales in Black's favour. Indeed after 21 l::t c 4 'ilVaS! 22 i.xb 7 l:tabS 23 .i.g2 lbxeS 24 .i.f4 liJxc4 2S i.xbS liJd2! Black collected a pawn. 1 0 . . . ttJbd7 1O ... .i.xf3 is dealt with in the next main game. Another possibility is 1O ...liJdS 1 1 e4 liJxc3 1 2 bxc3 bS 1 3 lbeS i.b7 14 irb l - a typical manoeuvre for White. After 1 4 ... lbc6 IS liJxc4 eS 16 .i.e3 exd4 17 cxd4 bxc4 1 8 'iWxb7 lbaS 1 9 'iYb2 ltbS 20 'iVc2 .l:tb3 21 i.f1 i.b4 22 ltac1 ltc3 Black does have some compensation for his weaknesses, but perhaps not enough. 1 1 e4 b5

Dizdar-Sadler

Game 37 Pula 1997 1 d4 d5 2 e4 e6 3 ttJf3 ttJf6 4 g3 dxe4 5 iLg2 iLe7 6 0-0 0-0 7 'ii'e 2 a6 S a4 iLd7 9 J:td 1 iLe6 1 0 ttJe3

1 0 . . . iLxf3 1 1 iLxf3 ttJe6 1 2 iLxe6 bxe6 1 3 iLg5 13 as 'iVbs 14 'iVa4 (14 lta4 'iVb3! I S 'iVxb3 cxb3 1 6 ltc4 c S 1 7 dxcS ltfdS 1 8 ltxdS+ ltxdS saw Black generate sufficient activity in N.Pert-Asrian, Yerevan 1 999, the ending being equal after 19 i.f4 liJdS 20 77

Th e Ca t a l a n

tLJxd5 :xd5 21 l:tc3 e5 22 .i.e3 �f8) 1 4 ... c5 15 �xc4 cxd4 1 6 l:txd4 c5 1 7 lld3 'iib4 1 8 :a4 'iWb7 1 9 l:!.a3 "iib4 20 �a2 l::ta b8 was seen in Beliavsky-Stohl, Portoroz 1 999, the bacl...-ward b-pawn affording Black equal chances. Black has also used the b8-square for his rook: 1 3. .. l::tb 8 14 'iVa4 :b4! (it is important to keep the c4-pawn rather than its less important partner on c6) 1 5 'iYxc6 �d6 1 6 �f3 and now Romanishin-Brunner, Altensteig 1 992 went 16 ... :fb8?! 1 7 e4 'it'c6 1 8 d5 with an excellent game for White, whose fortunes quickly improved after the faulty 1 8 ... exd5?, when the punishment came in the form of 1 9 e5 lbe8 20 lbxd5 'it'f8 21 'iVh5 l:t4b5 22 'ilVxh7 f6 23 �g8+! - a nice little trick that wins a piece. However, Kotronias gives 1 6 ...l:'td8 1 7 e4 'ilVc6 1 8 'iVe2 �7 1 9 l::ta2 c5!? 20 d5 exd5 21 exd5 h6! with an acceptable position for Black. 1 3 . . . l:tb8 1 3 . .. tLJd5 14 .i.xe7 'ilVxe7 1 5 tLJe4 l::t fb 8 1 6 'ilVxc4 l::txb2 1 7 lIab 1 'iib4 1 8 it'xc6 l::txb 1 1 9 'iYxa8+ 'iWb8 2 0 'iixb8+ l::tx b8 21 tLJc5 tLJc3 22 l::td2 as led to a level ending in Ti­ moshenko-Asrian, Ubeda 1 998. 14 e3

White's b-pawn can also b e attacked, while Black has an extra pawn. At some point he might play ... c6-c5. 1 4 . . . 4:Jd7 ? ! Black had a slight edge i n Ftacnik­ P.Nielsen, 1 999 after the interest­ ing 1 4 ... h6!? 1 5 .i.xf6 .i.xf6 1 6 lbe4 .llb4 1 7 lbcs 'iVc8 1 8 .l:!.ac1 .i.e7 1 9 lbe4 'ilVb7. Here c4 is no weaker than b2, and Black has a good bishop and decent piece play. 1 5 �xe 7 'it'xe 7 1 6 4:Je4 l:tb4 1 7 4:Jd2 e5 1 8 4:Jxe4 l:tfb8 1 9 l:tae 1 h6 20 dxe5 20 d5 exd5 21 J:hd5 is playable, when 21 ...lbb6 22 lbxb6 cxb6 23 l::tc dl 'iVe4 is roughly level but 2 1 ...11xc4? 22 'ilVxc4 tLJb6 23 iVb3! c4 24 l::tx c4 l:ta8 25 l::tc 6 lbxd5 26 'ii'x d5 simply leads to the loss of a pawn. 20 . . . 4:Jf6 20 ... lbxc5? 21 lbaS and lbc6 is coming. 21 4:Je5 l:txb2 22 'ife4 22 lbc6? runs into 22 ... �e8 23 �xb2 l::txb2 24 l::td 8 'ilVxd8 25 lbxd8 lbe4 and White loses a pawn. Both players should be careful in this seemingly quiet position. 22 . . . 4:Jd5 23 e4? A mistake. After 23 lbd3!? White enjoys a smal1 but solid edge. 23 . . . 4:Je3!

Illescas-Beliavsky, Madrid 1 998 continued 14 .i.xf6 .i.xf6 1 5 lbe4 l::tb4 1 6 e3 'iVd5! and after the further 1 7 lbxf6+ (1 7 as is interest­ ing) 1 7...t,1Xf6 1 8 e4 'ilVhS!? the position is unclear. Black has vulnerable pawns but

Now Black is in charge! 24 fxe3 'ifg5 ! N o t 2 4... 'ilVf6? 25 lbd3 'ik f3 2 6 lbxb2 l:!.xb2 27 'tin 'it'xe3+ 28 �hl 'iixe4+ and White escapes with a draw.

78

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 ii.. g 2 ii.. e 7 6 0 - 0 0 - 0

2 5 'iWc3 'iWh5 2 6 nd8 + Wh7 2 7 nd2 nxd2 28 'iWxd2 'iWxe5 White's king is exposed and his centre pawns are weak. Black has a decisive advan­ tage. 29 'iWd4 'ilVg5!

No exchange of queens! 30 nf 1 e5 31 'iWc3 f6 32 Wg2 'iWg4 33 'i'c4 a5 34 h3 'iWd7 35 'iWc2 'ilVc6 36 nd 1 . J:!.b4 37 nd8 nxa4 38 nc8 nb4! 0- 1

Game 38 Lputian-J . Polgar

WiJk aan Zee 2000 1 d4 d5 2 c4 e6 3 liJf3 liJf6 4 g3 dxc4 5 ii..g 2 ii.. e 7 6 0-0 0-0 7 'iWc2 a6 8 a4 ii.. d 7 9 'ilVxc4 ii.. c 6

10 ii.. g 5 In the event of 1 0 lZ'lc3?! b 5 ! 1 1 'iVd3 b4

1 2 lZ'lbl .i.e4 1 3 'lidl c5 1 4 lZ'lbd2 .i.d5 1 5 dxc5 lZ'lbd7 1 6 lZ'lb3 lZ'lxc5 1 7 lZ'lxc5 .i.xc5 Black had a fine position in Ara.Minasian­ Yegiazarian, Yerevan 200 1 . There followed 1 8 lZ'le5 .i.xg2 1 9 'iVxd8 .l:.fxd8 20 �xg2 l:tac8 21 lZ'ld3 .i.rn 22 .i.g5 .l:tc2 23 l:tac1 l:tdc8 24 l:txc2 l:txc2 25 .l:tc 1 with an equal ending, but 20 ...l:td5!? 21 lZ'ld3 .i.d6 22 l:tdl l:td8 offers Black chances of finding more. Also possible is 10 .i.f4 as 1 1 lZ'lc3 lZ'la6, when Hjartarsson gives 12 l:tad .i.d5 1 3 lZ'lxd5 exd5 1 4 � 5 �c8 with a balanced game. Delchev-Flear, Creon 2001 continued instead 1 2 .:tac1 lZ'lb4 1 3 .l:tfe l lZ'lfd5 1 4 lZ'lxd5 'iVxd5 1 5 1Wxd5 (1 5 .i.xc7 �xc4 1 6 l:txc4 .i.xa4 1 7 lZ'le5 f6 with approximate equality, White's control of the centre being weighed against Black's queenside majority) 1 5 ... lZ'lxd5 1 6 lZ'le5 lZ'lxf4 1 7 lZ'lxc6 (better than 1 7 .i.xc6 .i.b4! 1 8 gxf4 .i.xel 1 9 .i.xb7 l:ta7 20 l:txc7 l:tb8 21 l:tc8+ l:txc8 22 .i.xc8 l:tc7 with an edge for Black as White's pawns are not quite enough for the exchange) 1 7 ...lZ'lxe2+ ( 1 7 ... bxc6 1 8 gxf4 l:tab8 1 9 l:txc6 l::t xb2 20 l::t x c7 .i.b4 looks drawn) 1 8 l:txe2 bxc6 19 .i.xc6 l:tad8 20 d5 .i.f6 21 dxe6 fxe6 22 .i.b5 and White had a minimal advantage in the shape of Black's pawn weaknesses. Black opted for the Catalan trade of knight for dark-squared bishop in Hulak­ Psakhis, Zagreb 1 993, emerging from 1 O ... lZ'ld5 1 1 lZ'lc3 lZ'lxf4 1 2 gxf4 as!? with control of b4. There followed 1 3 l:!.fdl .i.d5 14 �5 lZ'la6 1 5 e4 .i.c6 16 "iVc4 lZ'lb4 1 7 lZ'le5 .i.e8 with more space for White but a solid set-up for Black, the try 1 8 d5 exd5 1 9 lZ'lxd5 lZ'lxd5 20 Itxd5 .i.d6 21 lZ'l f3 .i.c6 22 l:td3 'iVc8 23 e5 .i.e7 producing unclear play. 1 0 . . . a5 The more modest 1 0 ... lZ'lbd7 has also been played, e.g. 1 1 lZ'lc3 l:tc8 12 'ii'd3 lZ'ld5 13 .i.xe7 'ii'xe7 14 e4 lZ'lxc3 1 5 bxc3 l:!.fd8 1 6 l:tfe l lZ'lb6 1 7 'iVc2 .i.e8 1 8 l:teb l l:!.b8 1 9 c4 e5 20 d5 lZ'ld7 21 'iVc3 f6 22 lZ'ld2 b6 23 as with an easier game for White (space) In Wojtkiewicz-Watanabe, Merida 2000. 79

Th e C a t a l a n

11 ... h6 is the other means of dealing with the bishop. In Filippov-Kiriakov, St. Peters­ burg 2000 White dropped back to f4, and after 1 2 .1f4 .1d6 1 3 .l:!.fdl as 1 4 ttJe5 .1xg2 1 5 �xg2 ttJb6 1 6 'ifb3 ttJfd5 1 7 ttJe4 .1b4 1 8 ttJc5 ttJxf4+ 1 9 gxf4 'it'd5+ 20 'it'xd5 ttJxd5 21 �f3 .1xc5 22 dxc5 l:[fd8 23 l:i.d2 f5 chances remained even as both players had a knight in the centre. White can also use the tempo to turn his attentions to the queenside with 12 i.xf6 ttJxf6 1 3 b4 .1d5 14 ttJxd5 exd5 1 5 'ifb3 c6 1 6 e3 ttJe4 1 7 It.fc1 'it'd6 1 8 nabl (Beliavsky-Karpov, Yugoslavia 1 996), when Beliavsky gives the continuation 1 8 ... .1f6 1 9 ':c2 'it'e6 20 b5 cxb5 21 axb5 as as unclear, or 1 3 as 'it'd6 14 .l:!.fc1 .l:!.fd8 1 5 e3 I:tac8 16 'ifb3 llb8 17 'it'c2 .1xf3 1 8 .1xf3 c5 19 ttJe2 cxd4 20 ttJxd4 ttJd5 21 'iib 3 ttJb4 22 1:[c4 b6 23 axb6 'it'xb6 24 :xb4 .1xb4 25 ttJc6 as 26 ttJxb8 'it'xb8, which was equal in Romanishin-Ivanovic, Belgrade 2000. 1 1 ltJc3 ltJa6 1 2 ii.xf6 Of course White is not obliged to make this capture immediately. 12 l:tac1 .1d5 (1 2 ... 'it'd6 13 ttJe5 .1xg2 14 �xg2 c6 1 5 .1xf6 gxf6 1 6 ttJ f3 gave White a small edge in Kasparov-Karpov, World Championship (game 22), Moscow 1 984) 1 3 'it'd3 ttJb4 1 4 'ifbl c 6 1 5 e 4 .1c4 1 6 l:tfdl i.b3 1 7 :el .1c4!? is a line given by Chernin. Black is solid, although the position is unclear. The alternative rook development is 1 2 �fd 1 . Then 1 2 .. :iVd6 1 3 e 3 h 6 1 4 .1xf6 .1xf6 1 5 ttJd2 .1xg2 1 6 �xg2 'ii'b4 1 7 'ii'e2 c6 1 8 ttJc4 is Tukmakov-Kir.Georgiev, Szirak 1 995, when 1 8 ... .1d8 1 9 e4 .1c7 20 d5 exd5 21 exd5 Ilfe8 22 'i¥f1 nad8 could have earned Black equality according to Kir.Georgiev. This leaves 1 2 ... .1d5 1 3 'i¥b5 ttJb4 1 4 .1xf6 .1xf6 1 5 e4 .1c6 1 6 'it'c4 g6 1 7 .l:Iac1 I:te8 1 8 d5 exd5 1 9 exd5 .1d7 20 ttJd4 .1e5 21 b3 'it'e7 22 liel 'iVf6 23 ttJe4 'it'g7 24 ttJc5 with an edge for White in Ftac­ nik-Beliavsky, 2000. White's pieces stand well but Black has his chances, too. 1 2 . . . ii.xf6 1 3 e4 ltJb4 1 4 l:tfd 1 g6 80

I n Delchev-Lputian, Ohrid 2001 the con­ tinuation 14 ... b6 1 5 ttJel 'it'd7!? 1 6 ttJb5 :fc8 1 7 ttJc2 ttJxc2 1 8 'it'xc2 .1e7 1 9 .1h3 .1xb5 20 axb5 c6 21 'iVe2 .l:IfS 22 d5 cxd5 23 exd5 .1c5 24 dxe6 'it'e8 left Black with compensa­ tion for the pawn due to the presence of opposite coloured bishops, while there is also pressure against £1. With 1 5 .l:!.d2 .1b7 1 6 'it'e2 g6 17 'iVe3 'iVe7 1 8 h4 White starts an attack on the kingside, but Black can defend. Romanishin­ Kir.Georgiev, Ohrid 2001 went 1 8 ...l:tad8 1 9 lladl c 5 20 e 5 .1g7 2 1 dxc5 .1xf3 22 nxd8 .l:!.xd8 23 llxd8+ 'it'xd8 24 .1xf3 bxc5 25 'it'xc5 ttJd3 26 'iVd6 .1f6!, the latest clever move maintaining the balance. 1 5 h4 15 .:tac1 .l:!.e8 16 d5 exd5 17 exd5 .1d7 1 8 ttJd4 was seen i n Gelfand-Anand, Shenyang 2000, when 1 8 ... .1e5! 1 9 ttJcb5 .l:!.c8 20 ttJc6 bxc6 21 dxc6 i.e6 22 �hd8 Ilcxd8 23 'iVe2 .1b3 would have given Black compensation for the material. Although there is only a rook and bishop for queen and pawn at the moment, Black's forces look menacing . 1 5 . . . l:tc8 1 6 nd2 ii.g7 1 7 .l::ta d 1 White is slightly better due to his presence in the centre. 1 7 . . . 'ii'e 7 1 8 ltJg5 h6 1 9 ltJf3 b6 20 d5 ii.d7 2 1 ltJd4 exd5 22 exd5 ii.g4

23 ltJc6! With this move White gets rid of Black's light-squared bishop and thus strengthens his

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 !il.. g 2 !il.. e 7 6 0 - 0 0 - 0

control o f these squares. 23 .. )t:Jxe6 24 'ii'x g4 tLle5 25 'ii'e 2 h5 26 tLlb5 'i'b4 27 lId4 'i'e5 28 lIe4 lIee8 29 'ii'e 3 'i'xe3 30 lIxe3 tLle4! Better than the ive 30 ... l::re7 31 l::r e l . 3 1 lIxe8 lIxe8 32 tLlxe7 lIe8 33 d6 1Id8 33 ... llJxd6 34 nxd6 l:txc7 35 �d5 and, if anyone, White has only a minimal advantage in this endgame. 34 tLlb5 tLlxb2 35 lIe 1 tLlxa4 36 lIe7 !il..f 8 37 !il.. d 5 !il.. x d6 38 lIxf7 'it>h8 White is a pawn down but his initiative is strong. The knight on a4 is far from the bat­ tlefield on the kingside. 39 !il.. e4 !il.. e 5 Another option is 39 ....ic5 40 �xg6 llJb2 with unclear play. Black has a dangerous a­ pawn. 40 !il.. x g6 1Id5 41 tLla7 Or 41 .l:th7+ 'It>g8 42 .if7+ 'It>xh7 43 .ixd5 llJc3 and the resulting opposite col­ oured bishop ending is drawn. 41 . . . lIe5 42 lIb7 !il.. d 4 43 !il.. e 8 lIe 1 + 44 'it>g2 tLle3 45 tLle6 !il.. e 5 46 tLle5 a4 47 94! lIe 1 48 tLlg6 + 'it>g8

49 !il..f 7+ 49 gxh5!? Ihe8 50 h6 .ie7 (forced) 5 1 tL\xe7+ l::r x e7 5 2 l:!.xe7 a 3 and Black's ad­ vanced pawns give her adequate counterplay. After 53 l:ta7 a2 54 'It>f3 b5 55 'It>f4 llJa4 56 .l:tg7+ Wh8 57 .l:tgl b4 58 'i£tg5 b3 59 nel b2 White has to force the draw: 60 l:te8+ 'It>h7 61 l::re 7+ 'It>h8 etc.

49 . . . 'it>g7 50 gxh5 tLle4? The situation is unclear after 50 ... 'lt>h6. 51 !il.. e 8+ 'it>h6 Or 5 1 ...'lt>f6 52 l:!.f7+ 'It>e6 53 llJf4+ 'It>d6 54 lld7+ 'It>e5 55 llJd3+ and White is win­ rung. 52 tLle7 'it>h7 53 !il.. g 6+ 'it>g7 54 tLld5 + 'it>f8 55 h6 tLl d 6 56 lIb8 + tLle8 57 tLlf6 !il.. d 4 58 tLlxe8 'it>e7 59 tLlg7 'it>f6 60 h7 1 -0

Game 39 Karpov-Milos

Buenos Aires 2000 1 d4 d5 2 e4 e6 3 tLlf3 tLlf6 4 g3 dxe4 5 !il.. g 2 !il.. e 7 6 0-0 0-0 7 'i'e2 a6 8 a4 !il.. d 7 9 'i'xe4 !il.. e 6 10 !il.. g 5

1 0 . . . !il.. d 5 1 1 'i'd3 e5 1 1 ...�e4 1 2 'iVe3 llJbd7 13 llJc3 �c6 1 4 'iVd3 .ib4 1 5 .l:tfe1 �xf3!? 1 6 �xf3 c 6 1 7 .l:tedl 'iiVa 5 1 8 h 4 h6 1 9 �d2 e 5 2 0 �el .l:tfe8 21 e3 .l:tad8 was equal in Beliavsky-Z.Almasi, Ubeda 1 997, Black being ultra-solid. This worked out better for Black than 1 4 ....l:tb8 1 5 'it'c2 b 5 1 6 axb5 axb5, which was the course of Akopian-Ghaem Maghami, Yerevan 200 1 . After 1 7 llJe5 �xg2 1 8 'It>xg2 llJxe5 1 9 dxe5 llJd5 20 �xe7 't!Vxe7 21 llJe4 b4 22 l::r fel b3 23 'iic 5 iYd7 24 'iic 6 'ilie7 25 h4 h6 26 .l:ta7 White's activity gave him the better chances. 1 2 dxe5 12 llJc3 .ic6 13 l::r fd 1 cxd4 14 llJxd4 81

Th e C a t a la n

i.xg2 1 5 'it;>xg2 'ii'a s 1 6 i.xf6 i.xf6 1 7 lLle4 .ixd4 1 8 'iVxd4 lLlc6 1 9 'ii'c 5 was level in Tukrnakov-Beliavsky, Portoroz 1 996, while Black also has the immediate 1 2 ... cxd4. Then Karpov-Beliavsky, linares 1 994 continued 1 3 lLlxd5 'it'xd5?! 14 h4! lLlbd7?! (1 4 ... lLlc6 1 5 i.xf6 i.xf6 1 6 lLlg5 'ti'fS 1 7 i.e4 'iVe5 1 8 i.xc6 i.xg5 1 9 i.xb7 l:.a7 20 .ixa6 i.e7 21 l:.fdl is a shade favourable for White thanks to the ed queenside pawns) 1 5 lLlxd4 'ii'd 6 1 6 l:.fdl lLlc5 1 7 iVc4 I:.fd8 1 8 b4! with a clear lead for White, who won after 1 8 ... lLlxa4? 1 9 'iib 3! 'ii'b 6 20 e3 etc. Black is only slightly worse after the correct 1 3. .. lLlxd5 14 i.xe7 'ii'x e7 1 5 lLlxd4, when the Catalan bishop comes into play. 1 2 . . . tDbd7 1 2 ... i.xc5 1 3 lLlc3 i.c6 is sensible. In On­ ischuk-Zvjaginzev, Poikovsky 2002 a draw was agreed after 1 4 'iVc4 lLlbd7 1 5 b4 .ie7 16 b5 i.xf3 1 7 i.xf3 lLle5 1 8 'ii'f4 lLlxf3+ 1 9 'iix f3 axb5 2 0 axb5 'iVb6 21 i.e3 'iVc7 22 i.f4 'iVb6 23 i.e3. White can try 1 4 lLle5 �xd3 1 5 lLlxd3 i.d4 16 .ixf6 .ixf6 1 7 lLle4 but 1 7 ... i.e7 is acceptable for Black, who was able to neutralise White's attempted initiative in Papaioannou-Beliavsky, Istanbul 2000 after 1 8 l:.fc 1 lLld7 1 9 lLlec5 lLlxc5 20 lLlxc5 i.xg2 21 'it;>xg2 l:tfd8 22 lLld3 'it>f8 23 l:.c7 l:.ab8 24 l:.ac1 �e8 25 I:t7c4 f6 26 f4 :d7 27 l:.c8+ l:.xc8 28 l:!xc8+ l:.d8 29 :c7 l:!.d7 etc. 1 3 tDc3 tDxc5 1 4 �e3

1 4 . . . �a5 82

Another possibility is 1 4. . ..::.c 8 1 5 l:.fd 1 :e8 1 6 lLle5 'iWb6 1 7 i.h3! l1cd8 1 8 as 'ii'b 3 19 i.xf6 i.xf6 20 lLlxf7!, White's brilliant sacrifice securing an advantage in Smyslov­ Nogueiras, Graz 1 984 after 20 ... 'it>xf7 21 'iVxc5 .uc8 22 'it'e3 i.xc3 23 bxc3 'ifxc3 24 'ii'f4+. 1 5 .ixfS 1 5 :fdl .ic6 1 6 'iWf4 lLlb3 17 :abl h6 1 8 i.xf6 i.xf6 1 9 lLle4 i.xe4 20 'ii'x e4 lLlc5 2 1 'ii'e 3 'ii'b4 22 lLle5 :ad8 2 3 lLld3 lLlxd3 24 l:txd3 'ii'xa4 25 l:.xd8 I:.xd8 26 i.xb 7 was agreed drawn in Adianto-Short, Beijing 2000. However, an interesting suggestion is Pigu­ ' SOy S 1 5 lLlxd5!? lLlxd5 1 6 �a3 i.xg5 1 7 lLlxg5 h 6 1 8 :fc1 lLld7 1 9 lLle4 b 5 20 axb5 'iVxb5 21 lLlc3 lLlxc3 22 bxc3, when the combination of the unbalanced pawn struc­ ture and bishop versus knight looks nice for White. 1 5 . . . .ixfS 1 S tDxd5 exd5 1 7 �a3 �bS 1 7 ... lLle4!? is worth a try. 1 8 l::tadl l:.fd8 1 9 e3 l:.ac8 2 0 lLld4 11c4 2 1 b 3 l:tcc8 2 2 b 4 'ii'b6 23 as 'ii'd 6 24 :c1 g6 25 'iVb2 'ii'd 7 26 b5 i.xd4 27 'ifxd4 'it'xb5 28 i.xe4 dxe4 and a draw was agreed in Karpov-Gelfand, Monaco (rapid) 200 1 . This is a bit premature as there is still a lot to play for in this posi­ tion. 1 8 .l:tab 1 tDb3 1 9 .l:tfd 1 l:!.fe8 20 e3 .l:tac8

21 tDd2! After the knight exchange it might seem difficult for White to generate an initiative

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 iL g 2 iL e 7 6 0 - 0 0 - 0

with opposite coloured bishops on the board, but this is not the case. 21 l:lxdS?! offers White nothing after 21 .. . .ie7 22 as 'i'g6 23 'iWa2 lbc1 24 'iVa1 lbb3 2S lbeS 'iWc2 26 'iVa2 lbc1 . 2 1 . . . tt'lxd2 2 2 l1xd2 Black's light-squared pawns are easier to attack. 22 . . . d4 23 exd4 iLxd4 24 a5 �g6 25 l1bd 1 iLe5 26 b4 iLf8 27 iLxb7 White has a big advantage. 27 . . . l1e2 28 l1d4!

As long as White protects the cS-square f2 be quite safe. 28 . . . l1ee2 29 'iWf3 :e 1 + 30 'it>g2 l1xd 1 31 'iWxd 1 l1b8 32 'iWf3 iLe7 33 b5! axb5 34 a6 iLe5 3 5 l1d5 'iWe2 Or 3S ....ia7 36 lld7 'iVe6 37 1:[xf7! and White is winning. 36 :d7 'it>h8 37 iLe4 'iWa2 38 'iWf5 1 -0

will

Game 40 Kramnik -Gelfand

Astana 200 1 1 d4 d5 2 e4 e6 3 tt'lf3 tt'lf6 4 g3 dxe4 5 .ig2 iLe7 6 0-0 0-0 7 'iWe2 a6 8 a4 iLd7 9 �xe4 iLe6 1 0 iLg5 .id5 1 1 �e2 .ie4 1 2 �d 1 The alternative is 1 2 'iYc1 with a wide range of possibilities. 12 ... lbbd7 13 lbc3 .ic6 14 'iVc2 .i.b4 1 5 �fe1 .ixf3 1 6 .ixf3 c6 and Black had a solid

Slav-like position in Tkachiev-Z.Almasi, Se­ nec 1 998. Indeed after 1 7 1:[ed1 'i'aS 1 8 lba2 .id6 1 9 .i.d2 'i'c7 20 .i.g2 as 21 1:[ac1 l:tfd8 22 h3 .i.f8 23 e4 eS 24 .i.e3 the point was shared. Black also had nothing to fear after 12 ... cS 13 dxcs lbbd7 14 .ixf6 .ixf6 1 s lbbd2 .idS 16 e4 .ic6 17 'iNc2 'iWaS 18 1:[fc 1 l:tac8 1 9 lbb3 ifb4 20 'iWc4 'iWxc4 2 1 l:txc4 .ixb2 22 l:tb1 .ia3 23 .l:ta1 .ib2 in Kir.Georgiev­ Bruzon, Moscow 200 1 . The proceedings during the series 1 2 ... lbc6 1 3 lbbd2 lbxd4 1 4 lbxd4 .ixg2 1 5 lbxe6! fxe6 1 6 'it>xg2 'iWdS+ 1 7 lb f3 c S 1 8 as h6 19 .ixf6 .ixf6 20 'iWc2 c4 21 1:[fd1 'iWbs 22 lbd4 .ixd4 23 .l:txd4 1:[ac8 24 'ij'd2 l:lcs 25 e4 'i'c6 proved interesting but still led to equality in Kir.Georgiev-Beliavsky, 2001 , while 1 7 ... .id6 1 8 'it'c2 'it'e4 1 9 l:tfc1 'it'xc2 20 .l:txc2 lbdS 21 .id2 eS! left Black with an isolated pawn in the centre but com­ pensation for this weakness in the form of active pieces in Kir.Georgiev-Brenjo, Yugo­ slavia 200 1 . After the further 22 lbgs lbf6 23 lbe6 1:[f7 24 .:tac1 c6 25 1:[d1 l:.e7 26 lbgs e4 27 .iaS lbdS 28 l:.c4 1:[ae8 the situation re­ mained unclear. Also possible is 1 3. .. .ig6 14 lbb3 h6 1 5 .ixf6 .ixf6 1 6 1:[d1 lbb4 1 7 lbe1 c6 1 8 as SLhS 19 'it'd2 lbdS 20 lbd3 when White had a firm grip on the dark squares in Ivanisevic­ Beliavsky, Belgrade 2000. 12 ... h6 is sufficient for equality, e.g. 1 3 .ixf6 SLxf6 1 4 lbc3 SLxf3 1 5 .ixf3 c 6 with the same type of solid position for Black in Beliavsky-Lutz, Leon 200 1 , which continued 1 6 e3 as 1 7 l:.d1 lba6 1 8 'i'c2 lbb4 1 9 'iVe4 'i'e7 20 1:[d2 1:[ad8 21 l:tad1 1:[d7 22 "ii'g4 1:[fd8 23 lbe4 b6 24 .ie2 g6 25 .ic4 .ig7 26 'iVe2 lbdS 27 lbc3 'iWb4 with a good game for Black. Dropping back led to equality in Khalifman-Bologan, Panormo 2001 : 1 3 .i.f4 lbdS 1 4 lbc3 lbxf4 1 5 'iYxf4 .ixf3 1 6 .ixf3 c6 17 .l:tfd 1 as! 18 h4 lbd7 19 'it'e4 etc. 1 2 . . . e5 12 ... h6 13 .i.xf6 .ixf6 14 lbc3 .ixf3 1 5 83

Th e C a t a l a n

i. x f3 c 6 1 6 'iVb3 l:ta7 1 7 a s ! and White can hope for an edge. Then 1 7 ... i.xd4 1 8 l:.fd 1 'i'f6 1 9 l:ta4 e5 20 lZ'le4 is good for White, so Azmaiparashvili-Gelfand, Moscow 2001 went 1 7 ... lZ'ld7 1 8 l:tfdl 'i'c7 1 9 lZ'le4 i.e7 20 lZ'ld2 .l:.aa8 21 lZ'lc4 :ad8 22 :ac1 lZ'lf6 23 e4, although White's space advantage meant something. 1 3 dxc5 i.xc5 Again Black is not without alternatives. 1 3 ...'i'xd l ?! is not the best option. Pigu­ sov-Kruppa, Panormo 2001 continued 1 4 l:txd 1 i.xc5 1 5 lZ'lc3 .i.c6 1 6 lZ'le5 .i.xg2 1 7 'it>xg2 b 6 1 8 i.xf6 gxf6, and now White could have obtained the chances after 1 9 lZ'lc4 lla7 2 0 lZ'le4 f5 21 lZ'lxc5 bxc5 22 l:td6 with considerable play and the superior struc­ rure. Black was okay in Kozul-Ribli, Solin/Split 2001 after 1 3 ... lZ'lbd7 1 4 lZ'lc3 i.c6 1 5 b4 as 16 b5 i.xf3 17 i.xf3 lZ'lxc5 1 8 .i.e3 l:Ic8 1 9 'iVbl i.d6 2 0 lZ'le4 lZ'lcxe4 21 i.xe4 'i'e7 22 i.g2 i.c5 23 .i.d2 b6 24 'iWb3 .l:tfd8 etc. 1 3 ... h6 14 i.e3 lZ'ld5 1 5 'i'c1 'i'a5 1 6 lZ'lbd2 lZ'lxe3 1 7 lZ'lb3 'iVb4 1 8 'i'xe3 lZ'ld7 sees Black develop and thus earn a level game. However, in Romanishin-Beliavsky, Belgrade 2000 White's less precise play saw Black emerge from 1 6 lZ'la3?! lZ'ld7 1 7 lZ'lc4 'i'b4 1 8 lZ'le 1 lZ'lxe3 1 9 lZ'lxe3 lZ'lxc5 20 lZ'l1 c2 'iVb6 21 as 'ii'c7 22 i.xe4 lZ'lxe4 23 l:ta4 lZ'ld6 24 lZ'la3 l:tac8 25 'i'xc7 l:!xc7 with a slight lead, the knight rather misplaced on a3. 14 'ili'xdS J:txdS 1 5 4Jbd2 i.c6 1 6 4Jb3 4Jbd7 16 ... i.b4!? 17 nfc 1 lZ'lbd7 1 8 i.d2 i.xd2 19 lZ'lfxd2 i.xg2 20 'it>xg2 l:tdc8 21 lZ'lc4 l:tab8 22 lZ'lca5 g5 23 'it>f1 'it>f8 24 'it>e 1 led to a draw in Savchenko-Pigusov, Ohrid 200 1 . 1 7 J:tfc 1 i.b6 1 S 4Jfd2 i.xg2 1 9 'it'xg2 J:tdcS 20 i.xf6 4Jxf6 21 4Jc4 i.c 7 22 4Jc5 J:tabS?! 22...b6? is poor because after 23 lZ'lxa6! l:txa6 24 lZ'ld6 i.xd6 25 Ihc8+ .i.f8 26 :d 1 h5 27 l:tdd8 lZ'lh7 28 l:ta8 the rook and ed a-pawn are clearly superior to the two 84

minor pieces o n the other flank. However, an improvement on the game is 22 ... b5! 23 lZ'ld2 i.d6 24 lZ'ldb3 'it>f8 with equality. 23 b4 'it'fS 24 J:tab 1 'it'e 7 In reply to 24 ... b5 White sacrifices once again: 25 lZ'lxa6!? bxc4 26 lZ'lxb8 .i.xb8 27 b5 and his pawns are dangerous. 25 b5 axb5 26 J:txb5 b6 27 4Jd3 i.dS 2S 4Jde5

2S . . . 'it'fS? 28 ... lZ'le4 29 l:tdl f6 is better, when 30 �d7+ 'it>e8 31 l:.d4 lZ'lc3 32 lZ'ld6+ 'it>e 7! 33 lZ'lxc8+ l:txc8 34 l:tc4 .l:l.xc4 35 lZ'lxc4 lZ'lxb5 36 axb5 slightly favours White as the b6pawn needs protection. The bishop is ive and it is not easy to drive the knight away from c4. Nevertheless, Black has good chances of saving the game. 29 4Jxb6 ! J:txc 1 Or 29 ... .i.xb6 30 l:txb6 .l:txb6 31 l:txc8+ 'it>e7 32 l:!a8 when, with accurate play, White should win. 30 4Jbd 7 + 4Jxd7 3 1 4Jxd 7 + 'it'eS 32 4JxbS J:tcS 33 4Ja6 l:!.c2 34 e3 J:ta2 35 4Jc5 i.c7 36 J:!.b7 'it'dS 37 l:!.b4 'it'e7 37 ... .i.d6 is another option. After 38 lZ'lb7+ 'it>c7 39 lZ'lxd6 'it>xd6 40 l:tb7 l:ha4 41 l:txf7 J:tg4 42 f4 White has a decisive advan­ tage. 3S 4Je4 f5 39 J:tb7 J:tc2 39 ... fxe4 40 l:hc7+ 'it>f6 41 l:tc4 'it>f5 42 h3 and Black's pawns are very weak. 40 4Jg5 h6 41 4Jf3 'it'f6 42 4Jd4 J:tc4 43

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 ii.. g 2 ii.. e 7 6 0 - 0 0 - 0

ttJb5 ii..e 5 44 f4 ii.. c 3 4 5 �f7 + ! The minor piece ending is winning for White. 45 . . .'it'xf7 46 ttJd6 + We7 47 ttJxc4 Wd7 48 Wf3 Wc6 49 e4 Wc5 50 ttJe5 fxe4+ 51 Wxe4 Wb4 52 ttJc6 + Wxa4 53 ttJd4 Wb4 54 ttJxe6 Wc4 55 g4 ii.. f 6 56 h3 ii.. b 2 57 h4 ii.. c 3 58 f5 ii.. b 2 59 ttJxg7 ii.. x g7 60 g5 1 -0 White plays 7 'i'c2 a6 8 'i'xc4

Game 4 1 Jo . H orvath-Welis

Odorheiu 5ecuiesc 1993 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 ii..g 2 ii.. e 7 6 0-0 0-0 7 'i'c2 a6 8 'i'xc4 b5 9 'i'c2 ii.. b 7

10 ii..f4 1 0 i.g5 ttJbd7 1 1 ttJbd2 l:tc8 (1 1 ...c5!? 1 2 i.xf6 gxf6! i s interesting) and White must address Black's desire to push the c-pawn. 1 2 ttJb3!? i.e4 1 3 'it'c1 c 5 1 4 ttJxc5! ttJxc5 1 5 dxc5 l:!.xc5 1 6 'iWe3 'it'a8 1 7 i.xf6 gxf6 was seen in Larsen-Ribli, Amsterdam 1 980, when 1 8 .l:tad1 should secure a slight edge. White can also throw in 1 2 i.xf6 ttJxf6 in order to reduce his opponent's influence on c5. Kas­ parov-Karpov, World Championship (game 20), Leningrad 1 986 went 1 3 ttJb3 c5 14 dxc5 i.d5 (1 4 ... i.e4 1 5 'iic 3 i.d5 16 lIac1 i.xb3 17 'iix b3 i.xc5 1 8 e3 'iib 6 19 l:tc2 .l:tc7 20

ttJe5 l:tfc8 21 l:td1 h6 22 .l:tcd2 was drawn in Schandorff-Nielsen, Esbjerg 2001) 1 5 :tfd1 (1 5 ttJe1 i.xg2! 1 6 ttJxg2 ttJd7 1 7 .l:!.fd1 'it'c7 1 8 c6 ttJb8 1 9 .l:tac1 'it'xc6 20 'iix c6 Itxc6 21 ttJf4 led to a draw in Khalifman-Yusupov, Moscow 1 988) 1 5 ... i.xb3 1 6 'iix b3 'iic 7 1 7 a4 'it'xc5 1 8 axb5 axb5 1 9 ttJd4 b 4 20 e3 .l:tfd8 21 l::t d2 'iib 6 and the position was symmetrical, equal and, consequently, drawn. Also possible is 1 3. .. i.e4 1 4 'it'c3 (14 'it'c 1 c5 1 5 dxc5 as! 1 6 a4 'iid 5! 1 7 'tie3! is un­ clear) 1 4 ... ttJd5 1 5 'it'd2 i.b4! 1 6 'it'd1 c5, the thematic break equalising for Black in Spiri­ donov-Kotronias, Corfu 1 989 after 1 7 a3 c4! 1 8 axb4 cxb3 19 ii'xb3 'it'd6 20 ttJe5 i.xg2 21 �xg2, when 21 ...'iix b4 is even slightly better for Black according to Kotronias. 1 0 . . . ttJc6 10 ... i.d6 1 1 i.g5 transposes to 10 i.d2 ..id6 1 1 i.g5. 1 1 �d 1 In the event of 1 1 ttJbd2, worth consider­ ing is 1 1 ...ttJxd4 1 2 ttJxd4 i.xg2 1 3 ttJxe6 fxe6 14 �xg2 c5 1 5 ttJf3 with unclear play. Black's queenside majority compensates for the slight weakness on e6. In Piket-Lautier, Wijk aan Zee 1 997 Black opted for 1 1 ...l::t c 8 12 l:!.ad1 ttJb4 13 �1 ttJbd5 14 i.e5 ttJd7 15 ttJb3 'it'e8! 16 e4 ttJ5b6 17 .l:tfe 1 , when 1 7 ...ttJc4 would have guaranteed equality according to Lautier. Kotronias gives 1 2 ttJb3 ttJb4 1 3 'it'c1 i.d5 1 4 ttJfd2 i.xg2 1 5 �xg2 c5! 1 6 dxc5 as! 1 7 a4 'iVd5+ 1 8 f3 i.xc5! 1 9 e4 'it'h5 20 g4 ttJxg4 21 fxg4 'it'xg4+ 22 ..ig3 i.b6 with an initiative for Black thanks to White's exposed king. Also possible is 1 1 ttJc3 ttJb4 (1 1 ...ttJxd4 1 2 ttJxd4 i.xg2 1 3 ttJxe6 fxe6 14 'i.t>xg2 c5 1 5 .l:tad1 'it'e8 and the queen i s heading for c6, giving Black a good game) 12 �1 c5 1 3 dxc5 i.xc5, when the most narural course is 14 ttJg5 ..ixg2 1 5 �xg2 ttJbd5 etc. Instead after 1 4 i.e5?! ttJg4 1 5 ttJe4 i.xe4! 1 6 'iix e4 f5 Kirov-Vera, Timisoara 1 987 went 17 'iib 1 , White being punished for his ivity after 1 7 ... 'iib 6 1 8 e3 ..ixe3! 1 9 fxe3 ttJxe5 20 85

Th e C a t a l a n

ttJxe5 'iVxd+ 21 l::t f2 �xe5 2 2 �xa8 l:txa8, when the knight and two pawns fully com­ pensated the rook, the secure outpost on d5 and the exposed white king adding to the mix to leave Black better. Kotronias gives 1 7 'it' f4 ttJc2! 1 8 l::t ab 1 (1 8 l::t ac1 ? ttJxf2! 1 9 .l:!.xf2 �e3 and Black wins) 1 8 ... 'iVb6 and Black enjoys the initiative. 1 1 . . . 4:Jb4

1 2 ff'c 1 1 2 �xc7? �xc7 1 3 �xc7 ttJc2 1 4 ttJe1 ttJxa1 15 �xb7 l::t a7 and Black emerges the exchange up. 1 2 . . .l1c8 Also possible is 12 ... 'it'c8 13 �g5 c5 1 4 �xf6 �xf6 1 5 dxc5 a s with play for the pawn in Sploshnov-Berzins, Trinec 1 998. There followed 16 ttJa3 1l.c6 17 ttJc2 l:!.a7! 1 8 ttJcd4 l::t c7 1 9 a3 ttJa6 20 'ife3 �xf3 2 1 i.xf3 l::t x c5 when Black won back the pawn with equal chances after 22 ttJb3 l::t c2 23 ttJxa5 'ii'c 7. 1 3 4:Jc3 4:JbdS 1 4 .lteS ! ? After 1 4 ttJxd5 i.xd5 1 5 �e3 c 6 1 6 ttJ e 1 'i¥b6! 1 7 ttJd3 �xg2 1 8 �xg2 ttJd5 1 9 f3 c5 the thrust helped free Black in Khalifman­ Lutz, Wijk aan Zee 1 995, 20 dxc5 ttJxe3+ 21 'it'xe3 �xc5 22 ttJxc5 'ii'x c5 23 'iVxc5 .l:!.xc5 24 l::t a c1 l:tcc8 25 b4 g6 steering the game to an equal ending. This is certainly preferable (for Black) to 1 4 ... ttJxd5 1 5 e4 ttJf6 1 6 e5 ttJd5 1 7 �g5. Black should keep an eye on e4. 86

1 4 . . . c S 1 S dxcS .ltxcs 1 6 .ltxf6 ifxf6 1 7 4:Je4 'fie7 1 8 4:JxcS l:txcS 1 9 ifd2 White is better due to the slight weakness of Black's queenside. These pawns are more advanced and therefore more exposed. 1 9 . . . tDf6 19 ... .l:!.fc8 20 ttJe1 h6 21 l::t a c1 l::t x c1 22 l::tx c1 l::t x c1 23 'iVxc1 ttJf6 24 ttJd3 is Anders­ son-Kir.Georgiev, Sarajevo 1 985, again with the better prospects for White. 20 it'd6 ! 2 0 l::t ac 1 led only t o equality after 2 0... h6 21 l::tx c5 'iVxc5 22 l::t c 1 'iVb6 23 ttJe5 �xg2 24 'it>xg2 in Andersson-Beliavsky, Debreceen 1 992. 20 .. :ii'x d6 21 l:txd6 l:tfc8 22 4:Je 1 .ltxg2 23 Wxg2 as 24 l:ta6 a4 2S b3 axb3 26 axb3 gS 27 4:Jd3 l:tc2 28 l:ta8 �xa8 29 l:txa8 + Wg7 30 Wf 1 4:Je4 3 1 b4 4:Jd2+ 32 We 1 4:Jb3 33 .l:ta7 Wf6 If Black tries to get rid of the well placed knight on d3 with 33 ... ttJc1 ? he is punished after 34 �d1 l::t a2 35 l::txa2 ttJxa2 36 'it>c2, when the knight is caught. 34 l:td7 l:ta2?! Dubious. 34...g4 is more appropriate, when 35 e3 l::t d2 36 ttJc5 l::t xd7 37 ttJxd7+ �e7! 38 ttJe5 f6! 39 ttJxg4 ttJc1 40 �d2 ttJa2 sees Black regain the pawn with immediate equality. 3S f3 hS 36 Wf2 g4? ! After the more stubborn 36 ...l::t c2 37 f4 gxf4 38 gxf4 ttJc1 39 ttJe5 'itfS 40 �f3 l::t c3+ 41 e3 f6 42 l::t c 7! Black anyway finds his king in trouble. 37 f4 l:td2 38 .l:!.b7 4:Jd4? 38 ... l::tc2 39 ttJe5 is also good for White, although the text loses. 39 We3 4:Jb3 40 4:JeS 1 -0 White wins at least one pawn.

Game 42 Kramnik -Svidler

Linares 1998 1 d4 dS 2 c4 e6 3 tDf3 tDf6 4 g3 dxc4 S

Th e S e m i - Op e n C a t a la n : 4 . . . dx c 4 5 JL g 2 JL e 7 6 0 - 0 0 - 0

JLg2 JLe7 6 0-0 0-0 7 'W!Vc2 a 6 8 'ii'x c4 b5 9 'ii'c 2 JLb7 10 JLf4 ttJd5 1 1 ttJc3 Another option for White is 1 1 ttJbd2 ttJd7 1 2 ttJb3 c5 1 3 dxc5 lIcs 1 4 l:tadl ttJxc5 15 e4 ttJd7 1 6 'iWd3 ttJ5f6 (1 6 ...ttJxf4 1 7 it'xd7 ttJe2+ I S 'lithl �xe4 1 9 'iVa7 gives White too much play) 1 7 e5 ttJd5, when Black equalized in Beliavsky-I.Ivanov, Minsk 1 976. After I S l:tfel ttJxf4 19 'iVxd7 ttJxg2 20 'iYxb7 ttJxel 21 l'lxdS ttJxf3+ 22 'litg2 l:tfxdS 23 'litxf3 .ifS 24 'iVxa6 1:[c2 White had a pawn and prospects of a ed pawn on the queenside, while Black had more active pieces. 1 1 . . .ttJxf4 1 2 gxf4

1 2 . . . ttJd7 Stangl-Ruf, Kecskemet 1 990 went 12 ...ttJc6 1 3 l:tfdl ttJb4 14 'iWc1 'ifbs 1 5 a3 ttJd5 1 6 ttJe5 ttJxc3 1 7 'i'xc3 �xg2 I S 'litxg2 and White had a finn grip on the dark squares which soon translated to an advan­ tage after I S ... �d6 1 9 l:tac1 'i'b7+ 20 ttJc6 �hS (20 ... .ixf4? 21 'i!Vf3) 21 'i!Vf3. The knight on c6 is very annoying for Black. 1 3 l:!.fd 1 1 3 ttJg5?! is best avoided, as demonstrated in Gutman-Kochiev, Ashkhabad 1 975: 13 ... �xg5 14 .ixb7 .i.xf4! 15 �xaS 'iVh4 1 6 l:tfd 1 l:txaS and Black had a pawn and an attack brewing. The further 1 7 'i!Ve4 'i!Vxh2+ I S 'litft l:teS 1 9 it'f3 .i.d6 20 e3 f5 left both sides with something to bite on. Stangl­ Blauert, Dortmund 1 992 continued 1 3 ttJe4

l:tcS 1 4 ttJc5?! (too ambitious - 1 4 l:tfdl c5 is equal) 1 4 ... ttJxc5 1 5 dxc5 �xf3 1 6 �xf3 'iVd4 17 e3 'i!Vxc5 I S 'iVxc5 .ixc5 19 l:tfc1 .i.d6 20 lIc6 lIaS 21 l:tc2 l:tabS 22 :c6 and Black was slightly better due to his extra pawn, but the game was drawn. 1 3 . . . JLxf3? ! 1 3. .. 'i'cS 1 4 ttJe4 c5 1 5 dxc5 ttJxc5 1 6 ttJxc5 'i'xc5 1 7 'iVxc5 �xc5 I S l:tac1 l::t fcS 1 9 ttJe5 �xg2 20 'litxg2 f6 21 ttJf3 .ifS 22 e3 was the course of Ribli-Karpov, Amsterdam, 1 9S0. White has a small advantage here and he eventually won the game. In Sosonko-Dutreeuw, Brussels 1 993 White threw in 14 a4 b4 before 15 ttJe4. Af­ ter the dubious 1 5 ... .id5?! 1 6 l:Iac1 l::t a7 1 7 ttJe5 ttJxe5 I S fxe5 'iVaS 1 9 'iVd3! f5 20 exf6 gxf6 21 .if3 l:tdS 22 'iVe3 'lithS 23 'lithl l:tgS 24 l:tgl c6 25 ttJd2 Black had numerous weaknesses in this semi-closed position. Sos­ onko offers 1 5 ... c5 16 dxc5 ttJxc5 17 ttJd6 .ixd6 IS .l!txd6. 1 4 JLxf3 l:!.b8 1 5 e3 1 5 ttJe4 .id6 16 e3 'iVh4 17 'lith 1 ttJf6 I S ttJxf6+ 'iVxf6 1 9 l:tgl 'iVe7 20 .ic6 with the better prospects for White in Smyslov­ Barczay, Kapfenberg 1 970. The pawn pha­ lanx dominates the enemy bishop, while White is able to attack Black's queenside pawns. 1 5 . . . ttJf6 1 5 ... 'i'cS leaves White in charge after 1 6 l:tac1 c 5 1 7 d5 exd5 I S ttJxd5 .id6 1 9 .ig4 'i!VdS (19 ... 'i'c6?! 20 'iVe4 'lithS? 21 ttJf6! and White wins a piece) 20 e4 (Sosonko). Black has problems on the d-ftle. 1 6 l:!.ac 1 'ii'd 6 1 7 ttJe2 l:!.fc8 1 8 e4 'ii'd 7 1 9 d5! White breaks through in the centre and obtains an advantage. 1 9 . . . exd5 Or 19 ... .ifS 20 ttJd4! exd5 21 ttJc6 l:tb6 22 e5 ttJeS 23 l:txd5 'iVh3 24 'i'e4. Notice how well White's pieces are posted compared to Black's uncoordinated forces. 20 e5 ttJe8 21 l:!.xd5 'ii'h 3 22 JLg2 'ii'h 4 87

Th e Ca t a la n

23 lLld4! White wins material. 23 . . :ii'xf4 24 lLlc6

1 0 . . . lLlc6 1O ... i.d6 1 1 i.g5 tbbd7 is a different set­ up. Ricardi-Soppe, San Fernando 1 993 went 1 2 tbbd2 h6 1 3 i.xf6 tbxf6 1 4 l:Hdl l:tc8 1 5 tbb3 i.e4 1 6 'iVc3 'ii'e 7 1 7 J::tac1 and White had a finn grip on c5, securing an advantage after 1 7 ... i.d5 1 8 tbe5 i.xg2 1 9 'it'xg2 i.xe5 20 dxe5 tbd5 21 'iic 5 'iVg5 22 'iVd4 tbb6 23 tbc5. Nor did Black have any joy in Hlibner­ Eng, Gennany 1 985 after 1 2 ... c5 1 3 i.xf6 'ii'xf6?! 1 4 tbe5! i.xg2 1 5 tbxd7 'ii'd 8 1 6 tbxfB i.xfl 1 7 'iix h7+ 'it>xfB 1 8 l:!.xfl cxd4 1 9 'iib 8+ 'it'e7 20 'ii'xg7, when White won a pawn, 20 ... 1i'h8 21 'iig5+ 'ii'f6 22 'ii'xf6+ 'it'xf6 23 l:!.c1 adding the only open file to his collection. Black has an improvement in 24 . . . .lth4 1 3 ... gxf6 1 4 tbe4, although White is still bet­ After 24 ... i.g5 25 IIcd 1 �b6 26 l:t5d4 the ter. queen is in trouble - 26 ... J::tx c6 27 'ii'x c6 'irfS After 10 ... tbbd7 White sends his bishop 28 J::td 8! and White wins (Marciano) . elsewhere - 1 1 i.a5. Chetverik-Ortmann, Gyula 2000 continued l 1 ...tbb6 1 2 tbbd2 25 l:tcd 1 l:tb6 26 l:t5d4 l:txc6 Both 26 ... 'ii'h 6 27 'ii' f5 and 26 ...'iVg5 27 l::t c 8 1 3 i.xb6 cxb6 14 'ii'd 3 'ii'c 7 1 5 l:tfdl 'it'e4 win for White. .l:f.fd8, White's majority in the centre counting 27 .ltxc6 �xe5 28 .ltd7 l:td8 29 l:txh4 for more than Black's on the queenside. With 1 -0 1 6 a3 'ii'c2 1 7 'iVxc2 k[xc2 1 8 tbel ttc7 1 9 i.xb7 l:!.xb7 2 0 e 3 White's chances would Game 43 have been better. Kobalija-Kiriakov 1 1 ...l:tc8 12 tbbd2 l2Jb8 13 a3 l2Jc6 14 i.c3 b4?! 15 axb4 l2Jxb4 1 6 'iib 3 i.d5 1 7 Dubai 2002 '--------------... 'ii'a4 i.c6 1 8 'iVa5 i:tb8 1 9 l2Je5 i.xg2 20 'it>xg2 l:tb5 21 'iVa4 c 5 2 2 dxc5 i.xc5 23 1 d4 d5 2 c4 e6 3 lLlf3 lLlf6 4 g3 dxc4 5 l2Jdf3 saw Black's attempt to steal the advan­ .ltg2 .lte7 6 0-0 0-0 7 �c2 a6 8 �xc4 b5 tage leave him only with a weak a-pawn in 9 �c2 .ltb 7 1 0 .ltd2 I.Almasi-Lauber, Gyula 1 997. I.Almasi gives 1 4 ... l2Jd5 1 5 b4 l2Jxc3 1 6 'iVxc3 l:ta8 1 7 l2Jb3 when White is better, again due to control of c5 and the c-ftle. 1 1 ...l:!.a7 looks artificial, e.g. 1 2 l2Jbd2 'ii'a8 1 3 b4 l2Jb8 1 4 a3 l2Jc6 1 5 l2Jb3 l2Jxa5 1 6 l2Jxa5, Chetverik-B.Funnan, Karvina 1 998. I f Black cannot move the c-pawn then the rook looks a bit silly on a7. There followed 1 6 ... i.e4 1 7 l2Jel ! (White goes for an advan­ tage in the ending) 1 7 ... i.xc2 1 8 i.xa8 l:taxa8 1 9 l2Jxc2 i.d6 20 f4 with total control of the dark squares. After 1 3 ... e5 1 4 dxe5 l2Jxe5 1 5 88

Th e S e m i - Op e n C a t a la n : 4 . . . dx c 4 5 � g 2 � e 7 6 0 - 0 0 - 0

a3 ttJc6 1 6 i.xc7 ttJd4 1 7 'iWd3 i.xf3 1 8 ttJxf3 ttJxe2+ 1 9 'iWxe2 !:txc 7 20 ttJd4 Black is under pressure on the queenside. 1 1 e3 Preferable to 1 1 l:[d I ?! ttJxd4 12 ttJxd4 .txg2 1 3 i.a5 (1 3 ttJxe6?! i.e4! 1 4 'iWxe4 lLlxe4 1 5 ttJxd8 l:1fxd8 and the pin on the d­ ftle is annoying for White) 1 3 ... i.e4 14 'iVxc7 'i'xc7 1 5 i.xc7 l:tac8, when Black was hap­ pier in Plaskett-Adams, Great Britain 1 997. White's queenside needs help and Black en­ joys a pull . 1 1 . . :i'b8 The mysterious 1 1 ....l::!.a7 has been played here. Salov-A.Rotstein, 1 993 contin­ ued 12 i::t c 1 'iWa8 1 3 ttJel ttJb8 14 i.xb7 'i'xb7 1 5 i.a5! c5 (it is difficult to suggest an improvement for Black) 1 6 dxc5 l:tc8 (1 6 ... ttJc6 17 i.b6 l:ha8 1 8 a4 ttJd7 19 as!? l:tac8 20 ttJd2 favours White) 17 ttJd3 ttJc6 18 i.b6 .l:!.aa8 19 a4 bxa4 20 ttJc3 ttJd7 2 1 ttJxa4 and White kept the extra pawn. Also possible is 1 1 ...ttJb4 1 2 i.xb4 i.xb4 13 a3 i.d6 1 4 ttJbd2, when Beliavsky­ Karpov, Brussels 1 988 went 1 4...l:tc8 1 5 b4 as! 1 6 e4 i.e7 1 7 !:tab 1 axb4 1 8 axb4 .l:!.a8 1 9 l:!.fel l:ta4 2 0 'iWc3 'ilVa8 2 1 ttJe5 l:td8 22 'i'xc7 '>t>f8, White winning the weak c7-pawn but Black having counterplay against the pawns on b4, d4 and e4. After 23 d5 Be­ liavsky gives 23 ... exd5 24 exd5 .txd5 25 ttJg6+ fxg6 26 'i'xe7+ '>t>g8 as equal. Pigusov­ Rozentalis, Sevastopol 1 986 saw more action on the other flank: 1 4...'ii'e 7 1 5 e4 e5 1 6 ttJh4! g6 1 7 f4 exd4 1 8 e 5 d3 1 9 'ii'x d3 i.c5+ 20 '>t>hl nfd8 21 'ii'c 2 ttJd5 22 l:tael and White's initiative in the centre and on the kingside looked dangerous. 1 2 a3 J:tc8 1 3 lLlc3 Timman-Hjartarson, Amsterdam 1 989 went 1 3 l:tel !? as 14 ttJc3 ttJd8 1 5 e4 b4 1 6 ttJa4 ttJd7 1 7 i.f4 i.d6 1 8 i.e3 i.c6 1 9 d5 b3 20 'ilVc4 'iWb5 21 'ilVxb5 i.xb5 22 ttJc3 .td3 23 ttJd4 with an advantage to White thanks to his better placed pieces. 1 3 . . . lLld8

1 4 b4 14 e4!? invites the liberating 14 ... c5 1 5 i.f4 'i'a7 1 6 dxc5 i.xc5 1 7 e5 ttJd7 1 8 l:tadl (Short), with only an edge for White. 1 4 . . . a5 1 5 bxa 5 ! ? 1 5 .l:!.abl axb4 1 6 axb4 ttJc6 1 7 ttJa2 lIa6 1 8 ttJc1 �a8 1 9 'i'dl .u.a3 20 ttJel e5! saw Black free himself and equalize in Karpov­ Short, Tilburg 1 988, but 1 5 'iWb2!? axb4 1 6 axb4 ttJe4 1 7 ttJe5! i s interesting. Black can then sacrifice the queen with 17 ...ttJxd2 1 8 ttJd7 ttJxfl 1 9 ttJxb8 i.xg2 20 '>t>xg2 ttJxe3+ 21 fxe3 .l:taxb8 22 .l:ta6 but White emerges in front as the rook and bishop are no match for the queen in this position, where Black is quite ive. In Chetverik-M.Ponomariov, Briansk 1 995 Black chose 1 7 ... ttJxc3 1 8 'i'xc3 i.xg2 19 '>t>xg2 i.d6 20 ttJd3 but after 20 ...'it'b7+ 21 f3 lIxal 22 l:hal :ta8 23 e4 .l:!.xal 24 'it'xal 'it'c6 25 '>t>f2 f5 26 '>t>e3 White again stood better. 1 5 . . . J:txa5 15 ... c5 16 dxc5 l:!.xc5 17 'iWbl favours White, Kobalija offering the subsequent 1 7 ... l:!.xa5? 1 8 ttJd5! etc. 1 6 J:tfb 1 ! 1 6 ttJd5? runs into 1 6 ... i.xd5 1 7 .ltxa5 'it'a8 1 8 e4 i.xe4 19 'i'c3 ttJd5 20 'iWd2 ttJc6 21 i.b4 ttJcxb4 22 axb4 'iWb7 when White loses the b-pawn. Black has obvious com­ pensation for the exchange and the semi­ closed nature of the position is to his benefit. 1 6 . . :i'a8 1 7 'it'd 1 c6 89

Th e Ca t a l a n

Or 1 7 ....tc6 18 d5! exd5 1 9 ttJd4 lIxa3 20 l:!.xa3 .txa3 21 ttJxc6 'iVxc6 22 ttJxb5 when White's activity outweighs the pawn. 1 8 a4 b4 1 9 tDa2 J:!.xa4 20 tDxb4 l:ba 1 2 1 J:!.xa 1 'iWb8 22 tDaS i.xaS 23 J:!.xaS c5

24 'iWa 1 ! 24 ttJe5 offers no advantage for White af­ ter 24 ... cxd4 25 exd4 'iVb5 26 l:!.a7 .td6 27 'fWal ttJd5 and the position is equal. 24 . . . cxd4 Better is 24 ... ttJc6 25 dxc5 .txc5 26 .tc3 ttJb4 27 .l:!.a4. 25 tDxd4 ' f8?! A strange decision. The lesser evil is 25 ...fNc7 26 .tf3 with the more pleasant po­ sition for White. 2S i.f3 26 .ta5! ttJd5 27 .txd8 .txd8 28 .txd5 exd5 29 ttJf5 secures a definite edge. 2s . . :i'c7 27 ' g2 'ikc5 28 i.a5 'ikc4 29 i.xd8! i.xd8 30 .!:!.cS .l:!.xcS 31 tDxcS ' e8? 32 'i'a8 'ikd3 33 tDd4 'ikc4 Or 33 ... fNc3 34 .tc6+ We7 35 .ta4 .tb6 36 ttJc6+ and White wins. 34 i.cS + 1 -0

Game 44 Gelfand-Lutz

Dortmund 2002 1 d4 d5 2 c4 eS 3 tDf3 tDfS 4 g3 dxc4 5 i.g2 i.e7 S 0-0 0-0 7 'ikc2 as 8 'ikxc4 b5 9 'ikc2 i.b7 10 i.d2 i.e4 90

1 1 'ii'c 1 tDbd7 Black has a range of options. 1 1 ....l:!.a7 (again!) was played in Nielsen­ Krivonosov, Panormo 200 1 . After 1 2 iLe3 ttJd5 1 3 ttJc3 ttJxc3 1 4 'i*'xc3 .l:!.b 7 1 5 .l:!.fd 1 c6 1 6 ttJe5 iLxg2 1 7 c,t>xg2 �d5+ 1 8 f3 b4 1 9 'iib 3 ttJd7 2 0 ttJd3 'ii¥b 5 21 :ac 1 Black was a little worse. l 1 ...c6 12 ttJc3!? .txf3 13 .txf3 'ii'xd4 14 .te3 'iVd8 15 a4 and Black had won a pawn but White had sufficient play in Nielsen­ Rozentalis, Esbjerg 200 1 . There followed 1 5 ... b4 1 6 ttJe4 ttJd5 1 7 .td4 'iVc7 1 8 ttJc5 .l:!.d8 19 ir'c4 ttJd7 20 ttJxd7 lIxd7 21 .tc5 as 22 .l:!.ac1 .txc5 23 'i*'xc5 .l:!.d6 24 .l:!.c2 .l:!.ad8 25 .l:!.fc1 and White had compensation for the pawn. Chetverik-Namyslo, Budapest 2002 con­ tinued 1 1 ...b4 1 2 .tg5 ttJbd7 1 3 ttJbd2 .tb7?! 14 ttJb3 c5? 1 5 .txf6 gxf6 1 6 dxc5 .l:!.c8 17 .l:!.dl fNc7 18 'i*'h6! and the weakness of Black's kingside was cause for concern, the further 1 8 ... ttJxc5 1 9 ttJxc5 'i*'xc5? 20 lId4! increasing White's lead to decisive propor­ tions: 20 ... iLxf3 21 lIh4 .th5 22 l:.xh5 'fWc2 23 e4 etc. Black is advised to follow the ex­ ample set in Sigurjonsson-Gruenfeld, Rand­ ers 1 982 after 1 3 ... .td5, although 1 4 'iVc2 c5 1 5 e4 .tb7 16 e5 ttJd5 17 iLxe7 �xe7 1 8 ttJe4 cxd4 1 9 .l:!.ad 1 .l:!.fc8 2 0 'iVb 1 proved good for �'hite thanks to the great outpost on d6. Finally, Black has tried 1 1 ...ttJc6, e.g. 1 2

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 iL g 2 iL e 7 6 0 - 0 0 - 0

.ie3 lbb4 1 3 lbbd2 .i.b7 1 4 1LgS l:!.c8 1 5 a3 lLlbdS 1 6 b4!? (White clamps down on cS and thus maintains a pull) 1 6 ... aS 1 7 e4 lbb6 18 bxaS lba4 and now Sosonko-Zuckerman, New York 1 984 continued 1 9 eS lbd7?! 20 .ixe7 'iix e7 21 iVbl .i.a6 22 lbe4 h6 23 '1'b4 cS 24 dxcs lbdxcS 2S l:!.fdl with serious problems for Black, who should have gone with 19 ...lbdS 20 lbe4 (although this still favours White) . After 1 2 ...l:!.c8 1 3 lbbd2 .i.dS Van der Wiel gives 14 .i.gS!? lbd7?! 1 5 e4 .ixgS 1 6 exdS .i.xd2 1 7 'ii'x c6, when White wins material. Mishuchkov-Raetsky, Gorki 1988 went 1 4 lbb3 lbb4 1 5 �dl .i.e4 1 6 lLlel c 6 1 7 .i.d2 1Lxg2 1 8 lbxg2 � 6 1 9 .ixb4 .i.xb4 2 0 lbf4 .i.d6 2 1 lbd3 lbd7 22 Itc1 l:!.fd8 23 'ii'c 2 eS 24 dxeS lbxeS 2S lbxeS .ixeS 26 lbcs with the more pleasant pros­ pects for White. 1 2 iLa5 l:lcS 1 3 ttJbd2 In Beliavsky-Speelman, Amsterdam 1 989 13 lbc3 1Lb7 14 a4 b4 l S lba2 'iWe8 16 .i.xb4 c5 17 dxcS lbxcs 1 8 'iVd 1 l:!.d8 19 'iiic2 1Le4 20 'iic 4 .i.dS 21 'ilif4 .i.d6 22 'ii'e 3 lbb3 gave Black compensation for the missing pawn. 16 ... 1Lxb4!? 1 7 lbxb4 cS is given by Kotromas. 13 . . . iLaS 1 4 'iiVc 2 Also possible is 14 l:tel �e8 15 b4 lbe4 16 a4 f5 17 axbS axbS 1 8 e3 .i.d6 19 l:!.a2 .idS 20 l:!.c2 c6 21 lbxe4 .i.xe4 22 l:!.b2 'ii'e 7 with equality in Kasimdzhanov-Lutz, Essen 2002. In Chetverik-Harasta, Slovakia 1 997 White emerged with advantage after 14 .I:tdl 'l'e8 15 b4 lbb8 (l s ...lbdS 16 a3 f5 1 7 lbb3 'i'hs 1 8 lbcs is good for White) 1 6 lbb3 lLlc6 17 a3 lbxaS 1 8 lbxaS 1LdS 19 lbeS .ixg2 20 'it'xg2 lbdS 21 e4 lbb6 22 'i'c6. 14 . . . 'iiVe S 1 5 b4 ttJd5 ls ... lbb8 16 a3 lbc6 17 lbb3 .i.d6 1 8 l:i.ac1 e S gave Black some freedom in Harik­ rishna-Beliavsky, Moscow 200 1 , although 1 9 e3 lbxaS 20 lbxaS e 4 21 lbd2 .i.dS 22 lbbl c6 23 .i.h3 l:!.c7 24 lbc3 'ilie7 25 lbxdS cxdS 26 'i'a2 hS 27 l:!.c2 was a shade better for White.

1 6 a3 f5 1 7 l:lae 1 ! ? 1 7 l:!.ac1 'ii'h S 1 8 lUe 1 lbSf6 1 9 lbeS lbxeS 20 dxeS lbg4 21 h3 .i.xg2 22 'li;>xg2 lbxeS 23 'it'a2 'ii' f7 was equal in Andersson­ Lutz, Pamplona 1 998. Black has chances on the kingside to make up for the weaknesses on the other flank. 1 7 . . :ii'g 6 Another possibility is 17 ...'i'hS 18 e4 fxe4 19 l:!.xe4 'iVf5 20 'iVb3 and White is slightly better. 1 S e4 fxe4 1 9 'iiVx e4 l:lf6 20 'iiVx g6 l:lxg6 21 ttJe4 This queenless middlegame favours White. Black has many weaknesses . 2 1 . . . l:lfS Black can put a knight on c4 with 2 1 ...lbSb6 22 lbcs lbxcs 23 dxcS lbc4 24 lbeS lbxeS 25 l:txeS .i.xg2 26 'it'xg2 but this ending also favours White. 22 h4 iLd6? ! 22 ...lbsb6 23 lbfd2 doesn't allow ... lbc4. 23 h 5 ! l:lh6 24 ttJc5 ttJ5b6 25 ttJg5 Now e6 is doomed. 25 . . . iLxg2 26 'iit x g2 l:lxh5 27 ttJgxe6 l:lcS 2S ttJxd7 ttJxd7 29 l:lc 1 c6 30 l:lfe 1 l:ld5 31 ttJdS!

Highlighting the weakness of c6. 31 . . . l:lxd4? Or 3 1 ...lbb8 32 l:e6! lhd4? 33 l:!.e8+ .i.f8 34 l:!.xf8+ and White wins a piece. 32 J:ted 1 Black has problems on the d-ftJ.e. 91

Th e C a t a la n

3 2 . . . i.e5 3 3 1:txd4 i.xd4 3 4 1:td 1 e5 3 5 bxe5 tUxe5 36 1:txd4 tUb3

Perhaps Black relied on this fork, but... 37 1:te4! 1 -0 White keeps the extra piece due to the mate threat on e8.

Game 45 Pigusov-Aseev

Sevastopol 1986 1 d4 d5 2 e4 e6 3 tUf3 tUf6 4 g3 dxe4 5 i.g2 i.e7 6 0-0 0-0 7 'We2 a6 8 'Wxe4 b5 9 'We2 i.b7 1 0 i.d2 i.e4 1 1 'We 1 i.b7

1 2 1:td 1 1 2 �f4 �d6 1 3 ttJbd2 ttJbd7 1 4 ttJb3 and now Kasparov-Kramnik, World Champion­ ship (game 1 5), London 2000 went 14 ... �d5 1 5 i:tdl 'ii'e 7 (1 5 ... 'ito8 1 6 ttJe5 �xg2 1 7 xg2 �xe5 1 8 dxe5 'ito7+ 1 9 f3 ttJd5 20 e4 92

favours White) 1 6 ttJe5 �xg2 1 7 �xg2 ttJd5 18 ttJc6 ttJxf4+ 19 'it'xf4 'iNe8 20 'iNf3 with a pull for White, whose knight on c6 will prove annoying. The later game Nielsen-Goldin, Moscow 200 1 , continued instead 14 ...:c8 1 5 i:td1 �d5, Black preparing to push the c­ pawn. There followed 1 6 ttJc5 ttJxc5 17 dxc5 �xf4 1 8 'ii'x f4 'ii'e 7 19 ttJe5! (a strong pawn sacrifice) 1 9 ... 'ihc5 20 l:tac1 'ii'e 7 21 ttJc6 �xc6 22 J:txc6 l:tfd8 23 I:!.dc1 h6 24 a3 with an advantage to White, who should be in no hurry to win back the pawn. 1 3 ttJc3 ttJbd7 1 4 :dl 'iNb8 1 5 ttJe5 �xg2 1 6 ttJxd7 ttJxd7 1 7 �xg2 c5 1 8 dxc5 'irb7+ 1 9 f3 �xc5 20 ttJe4 �e7 21 �c7 �xc7 22 �xc7 ttJf6 23 �d6 �xd6 24 ttJxd6, Tukma­ kov-Lalic, Palma de Mallorca 1 989 resulted in an ending that was favourable for White. Black has another option in 1 2 ... ttJd5 1 3 ttJc3 ttJxf4 1 4 'iixf4, e.g. 1 4. . .'iid 6 1 5 lUd l ttJd7 1 6 l:tac1 i:tac8 1 7 d 5 exd5 1 8 'it'xd6 �xd6 1 9 ttJxd5 �xd5 20 l:txd5 ttJb6 21 i:tdd 1 b4 22 i:tc6 ttJa4 23 .l:td2 as 24 l:ta6 !la8 which led to an edge for White in Fominyh­ Pihlajasalo, Elista 1 998. Black's pawn struc­ ture on the queenside is a bit exposed but, on the other hand, he has active pieces. In Piket­ Karpov, Dortmund 1 995 Black took the opportunity to push with 14 ... c5 1 5 dxc5 �xc5 1 6 %:tac1 'ii'e 7 1 7 l::tfd l .l:ta7, when Piket proposes 1 8 a4! b4 1 9 ttJe4 �xe4 20 'iVxe4 as 21 ttJe5 with advantage to White. His co-ordination is better and (again) Black's queenside pawns are the more vul­ nerable. 1 2 . . . 'We8 1 3 i.a5 13 a4!? is interesting. 1 3. .. c5 14 dxc5 �xc5 1 5 axb5 axb5 16 %:txa8 �xa8 17 �e3 �e7 1 8 'it'xc8 l::tx c8 1 9 ttJd4 �xg2 20 �xg2 b4 21 :c1 l:hc1 22 �xc1 gave White something in Bareev-Aseev, Irkutsk 1 986, an important factor being the king's quick access to the centre. In Pigusov-Fokin, Kemerovo 1 985 Black pushed with 13 ... b4, but after 14 �g5 ttJbd7 1 5 ttJbd2 c5 1 6 ttJb3 �d5 1 7 ttJxc5 i.xc5?! 1 8 �xf6 gxf6 19 dxc5 ttJxc5 20 'iVh6

Th e S e m i- Op e n C a t a la n : 4 . . . dx c 4 5 i. g 2 i. e 7 6 0 - 0 0 - 0

(Black should always be aware o f this ma­ noeuvre) White had generated a strong of­ fensive: 20 ... ttJb3 21 .l:!.xd5! exd5 22 �h3 'i'c2 23 ttJh4 ttJd4 24 e3 'ii'd2 25 �f5 and White won. An improvement is 1 7 ... ttJxc5 1 8 dxc5 'ii'x c5 1 9 'it'xc5 �xc5 20 �xf6 gxf6 21 lLld2!, although after the exchange of light­ squared bishops Black will have some prob­ lems on this colour complex on the queen­ side. 1 3 . . . i.d6 13 ... c5 1 4 dxc5 'it'xc5 1 5 ttJbd2 ttJbd7 1 6 lLlb3 'it'xc1 1 7 .l:!.axc1 .l:!.fc8 1 8 �c7 'it' f8 was Chernin-M.Gurevich, Vilnius 1 985. Black's king is approaching the centre so this end­ game is equal. 14 ltJbd2 ltJc6 1 5 ltJb3 ltJxa5 1 6 ltJxa5 .id5 1 7 ltJc6 a5 1 8 ltJfe5 na6 18 ... �xg2 19 'it'xg2 lla6 20 'it'c2 with a slight advantage for White. 19 i.xd5 exd5 1 9 .. .'�Jxd5, to exchange one knight, is a dubious idea as Black loses control of e4: 20 e4 ttJb4 21 a3 ttJxc6 22 ttJxc6 �7 23 d5 exd5 24 exd5 is given by Aseev. White is better. 20 'ii'c 2 20 'it'g2!? is worth a try after 20 ...l!e8 the position is unclear. 20 . . . 'ilh3

Not 21 llac 1 ? in view of 21 ....l:!.xc6! 22 'iVxc6 �xe5 23 dxe5? ttJg4 and Black wins. With 21 f3 White protects g4. After 2 1 ...ttJh5!? 22 'iib 3?! �xe5 23 ttJxe5 ttJxg3 24 hxg3? 'i*'xg3+ 25 'it'f1 .ll h 6! Black is win­ ning, so 22 e4 is better, with an unclear situa­ tion. 21 . . . a4 22 'ilf3 ltJe4 23 nac 1 a3 24 b3 b4 25 ltJd3 ltJc3 26 nxc3! White has to be careful here: 26 ttJdxb4?! �xb4 27 ttJxb4 .l:!.h6 28 .l:!.xc3 (28 .l:!.e 1 ? ttJxe2+!) 28...'it'xh2+ 2 9 'it'f1 'it'h 1 + 3 0 'it'xh 1 l:txh1+ 31 'it'g2 .l:!.xd1 when Black is the ex­ change up but many pawns are being ex­ changed, promising White hopes of a draw. 26 . . . bxc3 27 'ilxd5 'ile6 28 'ilf3 ne8 29 e3 c2 30 nc 1 'ilg6 This is better than 30 ... �xg3?! 31 ttJdb4 llxc6 32 ttJxc6 �d6 33 'it'f1 , which nets White the c2-pawn and with it the slightly superior ending. 3 1 ltJde5 i.xe5 32 ltJxe5 nxe5 33 dxe5 'ii'd 3 34 'ilb7 g5

-

35 nxc2 White forces an equal queen ending. More ambitious is 35 'it'g2 l:te6 36 'it'xc7 'it'd5+ 37 e4! 'i*'xe4+ 38 'it'gl l:tc6 39 'ii'd8+ cJ;;g7 40 'it'xg5+ 'it'f8, although the c2-pawn remains dangerous. 35 . . . 'ilxc2 36 'ilc8 + ! % - % 21 'ilb3

93

Th e C a t a l a n

Summary One of the main positions in the 7 ttJa3 line arises after the natural moves 7 ... �xa3 8 bxa3 bS 9 a4 a6 10 �a3 l:te8 1 1 ttJeS ttJdS 12 e4. White has the very interesting sacrifice 13 ttJxt7!? after 12 ... ttJf6, and to us it seems that this is enough only for a draw. The jump to the other wing, 12 ... ttJb6 (Game 28) leads to very complex play too, and requires more practical tests. Another popular system in this line is 8 ... �d7, when after 9 ttJeS �c6 10 ttJxc6 ttJxc6 1 1 iLb2 White prepares e2-e4 and has genuine compensation for the pawn (Game 29) . Practice has shown that against 7 ttJeS the best continuation is Robatsch's surgical decision 7 ... ttJc6!? In the case of 8 �xc6 bxc6 9 ttJxc6 'iVe8 10 ttJxe7+ 'iVxe7 1 1 'iVa4 Black can achieve a good game in more than one fashion. For instance Black can choose to exchange queens with 1 1 ...'iVd6 12 lIdl 'iVa6 or prepare the development of the bishop to a6 with l 1 ...aS. Black also has good prospects if he chal­ lenges the d4-pawn with 1 1 ... cS or 1 1 ... eS, thereby facilitating development. In the variation l 1 ...cS 1 2 'iVxc4 cxd4 1 3 'ifxd4 Black has two reliable possibilities, 1 3 ... :d8 and 1 3 ... eS 1 4 'iVh4 'iVe6 I S ttJc3 �b7 (Game 3 1 ) . In the complicated variation l 1 ...eS 1 2 dxeS 'iVxeS 1 3 'iVxc4 �e6 1 4 'ir'd3 l:[ad8 I S 'iVe3 'iVhS 1 6 f3 �c4 (Kotronias) looks quite convincing. After 8 ttJxc6 bxc6 9 ttJa3!? (Kuzmin) 9 ... �xa3 1 0 bxa3 White accepts the doubled a­ pawns, the resulting situation appearing rather exotic. Then 1O .. JIb8 1 1 'iVa4 'iVxd4 1 2 �e3 'iVeS 1 3 'iVxa7 deserves more practice, while in the case of 1 0 ... �a6!? 1 1 �xc6 nb8 1 2 'iVa4 lIb6 there is not much difference between 1 3 �g2 and 1 3 �f3. In both cases Black tends to equalize by playing ... c4-c3 at the right moment. In order to assess the variations with 1 O ...ttJdS it is important to evaluate the endgame after 1 1 'iVa4 ttJb6 12 'iVxc6 Itb8 1 3 'iVcs �b7 14 e4 'iVd6 1 S �f4 'iYxcs 1 6 dxcS ttJa4 1 7 .l:tfc 1 �a6 1 8 �f1 ttJxcS 1 9 �xc4 (Game 34) . Black must strive for a rook ending similar to Hjartarsson-Schussler. It does not seem likely that anything can be made of the extra, doubled pawn. As Black very often reaches a good position in the 9 ttJa3 line, White has tried the solid 9 e3 during recent years (Game 32). White has a normal queenside structure but also an obvious problem with the development of the dark-squared bishop. Black has counter-chances after something like 9 ...ttJdS 1 0 'iVa4 ttJb6 1 1 'iVc2 .l:Ib8 1 2 l:tdl 'iVe8 followed by... c6-cS. In the modern line with 7 'iVc2 a6 8 a4 iLd7 Black solves his opening problems quite suc­ cessfully. After 9 lIdl iLc6 10 ttJc3 ttJbd7 1 1 e4 bS 12 dS exdS 1 3 eS White has a dangerous initiative, although in the event of 1 O ... �xf3 1 1 �xf3 ttJc6 1 2 �xc6 bxc6 the b-ftle means that Black's chances are not worse. After 9 'iYxc4 �c6 the natural 1 0 ttJc3 allows Black to play 1 O ... bS with a good game, while in reply to 1 0 iLf4 Black attains equal prospects with 1 0 ... aS 1 1 ttJc3 ttJa6. Black should not develop in the same way against 1 0 �gS, where White can grab the centre with 1 1 �xf6 �xf6 1 2 e4. Instead of 1 0 ... aS it is better to continue 1 O ... ttJbd7 1 1 ttJc3 h6 or to follow the main line with 1 O ... �dS. After 1 1 'iVd3 Black can continue to chase the enemy queen with 1 1 ...�e4 or immediately play l 1 ...cS with equality. The advance ... c7-cS is also strong in the case of 1 1 'it'c2 �e4 12 'iVdl and deserves further testing after 1 1 'iVc2 iLe4 12 'iVc1 . Also quite acceptable for Black here are 1 2 ... ttJbd7 and 12 ... h6, and a later ... �xf3 and ... c7-c6 with a solid set-up. Black obtains the bishop pair and destroys White's pawn chain after 8 'iVxc4 bS 9 'iVc2 �b7 10 �f4 with 1 0 ... ttJdS 1 1 ttJc3 ttJxf4 12 gxf4. However, the pawn formation is quite typical for the Catalan. The pawns on d4, e3, f4 and f2 provide White with something to bite on because they offer good control over key squares in the centre. But we believe that White does not have an advantage after 10 ... ttJc6 1 1 tIdl ttJb4 1 2 94

Th e S e m i - Op e n C a t a la n : 4 . . . dx c 4 5 il. g 2 il. e 7 6 0 - 0 0 - 0

'it'c 1 'ii'c 8!? 1 3 �gS cS!? etc. Note, however, that . . .c7-cS does not promise full equality in the case of 1 2 . . . l::t c 8 1 3 ttJc3 ttJbdS 14 �eS. The deeper idea behind the quite ive look­ ing 10 i.. d2 is to pin the c7-pawn with i.. d 2-aS, and this comes into play after 1 0 ... ttJbd7. Another central idea in this line is to address the queenside with b2-b4. It is probably better for Black to continue 10 ... ttJc6 1 1 e3 ttJb4 12 �xb4 i.. xb4 1 3 a3 i..d 6 14 ttJbd2 l::t c 8 1 5 b4 as! with good prospects of equality, as in the other lines White has a mini­ mal edge, for example 1 0 ... �e4 1 1 'ii'c 1 b4 1 2 i..gS!? followed by ttJbd2. The problem with the subtle manoeuvre 1 O ... �e4 1 1 �c1 �b7 is the possibility of 1 2 'ii'c2 with a repetition, which might occasionally be an attractive option for White. Moreover White has something after 1 2 .tf4 ttJdS 1 3 ttJc3 ttJxf4 1 4 'it'xf4 o r after 1 2 l::t d l 'ikc8 1 3 a4!?

1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 il.g2 i.e7 6 0-0 0-0 7 'it'c2 7 ttJc3 7 ...ttJbd7 Game 3 (Chapter 1); 7 ...ttJc6 - Game 27 7 ttJa3 �xa3 8 bxa3: 8 ... bS Game 28; 8 ... �d7 Game 29 7 ttJeS 7 ... cS Game 30 7 ... ttJc6 8 �xc6 Game 3 1 8 ttJxc6 bxc6 (D) 9 e3 - Game 32 9 ttJa3 i.. x a3 1 0 bxa3: 1 O ... .ta6 - Game 33; 1 O ... ttJdS - Game 34 7 a6 8 ii'xc4 8 a4 8 ... cS Game 35 8 ... �d7 (D) 9 .l:tdl �c6 1 0 ttJc3: 9 ... ttJbd7 Game 36; 1 O ... �xf3 Game 37 9 'ikxc4 �c6 10 �gS 1 0 ... aS Game 38 10 ... i..d S: 1 1 'iVd3 - Game 39; 1 1 'ii'c2 Game 40 8 b5 9 ii'c2 i.b7 10 i.. f4: 1O ... ttJc6 Game 4 1; 10 ... ttJdS Game 42 10 il.d2 i.e4 1 0 ...ttJc6 Game 43 1 1 'it'c 1 (D): l 1 ...ttJbd7 - Game 44; l 1 ...i.b 7 Game 45 -

-

-

-

-

. . .

-

-

-

-

-

. . .

-

-

-

-

8 . bxc6 . .

8

. . .

i.d7

1 1 ii'c 1 95

CHA PTER FIVE

I

4 . . . dxc4 5 jLg2 a 6

1 d 4 d 5 2 c4 e 6 3 4Jf3 4Jf6 4 g 3 dxc4 5 i.g2 a6 S ... a6 is one of the most popular tries to hold on to the pawn in what can rightly be called the Catalan Gambit. Black wants to defend the c4-pawn with the help of ... b7-bS, with White's compensation coming in the shape of a formidable looking centre and a lead in development. Not surprisingly the result is complex play in which tactics and aggression are major factors. After S ... a6 there are move order issues and lines which could belong under other systems such as the Slav Defence. Fortu­ nately this is not too confusing, and we have endeavoured to indicate the most essential transpositions. This has been done by focus­ ing on plans rather than concrete lines, which seems to be the logical approach to positions that can arise via numerous routes. White can choose to prevent ...b7-bS by spending a tempo on 6 a4 (Game 46), but this cannot be recommended because after 6 .. .tbc6 the weaknesses of both b3 and b4 will prove significant. The difference be­ tween this system and S ... cS can be appreci­ ated in the following variation: 6 ... cS 7 0-0 cxd4!? 8 'ii'xd4 'ii'xd4 9 ttJxd4 eS when the inclusion of a2-a4 and ... a7 -a6 benefits Black as bS is unavailable to the knight and a4 to 96

the queen. More attractive is 6 ttJeS, opening the h 1 a 8 diagonal and simultaneously attacking the c4-pawn. Then only vigorous measures allow Black to keep his material advantage. 6 ... �b4+ (Game 47) is related to the varia­ tion beginning S ... c6 6 ttJeS �b4, but there are also some differences. White does not have to enter the line with 7 �d2?! 'ii'xd4 8 �xb4 'ii'x eS because, compared with S ... c6, the d6-square is not weak here. But after 7 ttJc3 ttJdS 8 �d2 bS White generates an en­ during and satisfying initiative for the sacri­ ficed pawn. The logical reaction to 6 ttJeS is 6 ... cS (Games 48-S0), since White has surrendered some control of d4. Defending with 7 e3 closes in the d-bishop and, after the strong­ est reply, 7 ... .J:!.a7! (followed by ... b7-bS), it is White who has difficult issues to address. Practical experience also offers little value to 7 .Jl.e3 (Game 48) as after 7 ...ttJdS! Black will find the most appropriate moment to ex­ change on e3, thus damaging White's struc­ ture. 7 ttJa3 is the most promising continuation. White sacrifices the d4-pawn in return for rapid development, in the case of 7 .. .'it'xd4 responding with the strong 8 'ii'a 4+ !? (Game 49), when Black's 'centralized' queen will not

4 . . . dx c 4 5 i. g 2 a 6

fmd a safe haven. After 7 ... cxd4 8 ttJaxc4 i.c5 the reader should pay attention to the sacrifice of a second pawn with 9 0-0 0-0 1 0 b4!? with a considerable lead i n development, the critical position arising after 8 .. J::ta 7!? 9 i.d2 b6 1 0 'iVb3 .tb7 1 1 .txb7 l:[xb7 1 2 ttJa5. Black should b e safe after both 12 ... l::ta7 (quiet) and 1 2 ... l:te7!? (complex). The most popular answer to 5 ... a6 is 6 0-0, White continuing development. Now Black has two different paths down which he can walk - advancing the b-pawn ( ... b7-b5) at once or keeping his options open with the sensible (developing) 6 ...ttJc6. The line 5 ... a6 6 0-0 ttJc6 is very similar to the system with 5 ... ttJc6 (Chapter 6) and there are often transpositions. little attention has been given to 7 ttJa3 .txa3 8 bxa3 - es­ pecially when compared to 5 ... .te7 6 0-0 0-0 7 ttJa3 .txa3 8 bxa3. Black has gained a tempo with ... a7-a6, but this seems to be of little value - the notes to Gleizerov-S.Ivanov (Game 5 1 ) illustrate that it is not dangerous for Black. In that game we have also consid­ ered 7 a4, when after 7 .. J!b8 8 as it is better not to open the game with 8 ... b5, preference being for the completion of development with 8 ... .th4. After 5 ... a6 6 0-0 ttJc6 the requirements of the position are best addressed by concen­ trating on the centre with e2-e4. White can choose to prepare this expansion with the help of 7 e3, 'iVe2 and .l:!.fdl or try the quicker but riskier 7 ttJc3. After 7 e3 (Games 52-53) Black is unable to hit d4 and should turn his attention to ... b7-b5. This can be achieved with 7 .. .J::t b 8, but this is probably not good enough for equality in view of the manoeuvre ttJf3-d2-c4 (rather common In the Catalan). The main continuation is 7 ... i.d7 8 'iVe2 b5 9 l:[d l (Game 53) . After 7 ttJc3 l:tb8 8 e 4 (Games 54-56) ex­ perience has shown that e4-e5 is not the most appropriate option. White needs to break in the centre at the best possible mo­ ment with d4-d5. Practice has given us the

following possibilities: a) 8 ... .te7 9 d5 exd5 10 exd5 ttJb4 1 1 ttJe5 .tfS 12 a3 ttJd3 13 ttJxc4 ttJxc1 14 l:txc 1 . This has been played often. Some commen­ tators evaluate it as even, while others believe White's greater share of territory is enough for an edge. The latter assessment seems more accurate. b) 8 ... .te7 9 'iVe2 b5 (9 ... ttJxd4 is not so clear and obviously need more tests) 10 :dl and then d4-d5 with boundless complica­ tions (Game 54) . c) 8 ... b5 9 'iVe2 ttJxd4 1 0 ttJxd4 'iixd4 (Game 55), which is closely related to 8 ... .te7 9 'iVe2 ttJxd4. d) 8 ... b5 9 d5 (Game 56) . The system with 5 ... a6 6 0-0 b5 (Games 57 -60) resembles 5 ... b5, occasionally with possible transpositions between the two lines. However, 5 ... a6 is different in that Black has a wider variety of possibilities. In reply to 7 ttJe5 Black has to choose between 7 ... c6 and 7 ...ttJd5. After 7 ... c6 (Game 57) play used to continue 8 ttJxc6 'iVb6 9 ttJe5 .tb7, after which the exchange of bishops and a strong central pawn formation af­ forded White a modest but enduring advan­ tage - all the way to the ending. The confi­ dence in 7 ... c6 was challenged by Razuvaev, who came up with 8 b3!, the point being that after 8 ... cxb3 the knight receives an alternate route in ttJc6-a5-b3. Here the pawn forma­ tion in the centre clearly favours White, thus leaving 7 ... c6 out in the cold. After 7 ttJe5 ttJd5 White should play 8 a4 as 8 ttJc3 (Game 58) is probably less accu­ rate. First there is 8 ... c6 9 ttJxd5 exd5 10 e4 .te6 1 1 a4, which was considered dangerous for Black since the game Sosonko-Hiibner. In light of 1 1 ...11a7!? it seems that White's initiative has been overestimated. Addition­ ally, Black also has 8 ... .tb7 9 ttJxd5 exd5. Here White can win the exchange with 10 e4 dxe4 1 1 'iVh5 g6 12 ttJxg6 fxg6 13 'iVe5 'iie 7 14 'iix h8 but Black should receive more than sufficient counterplay. 97

Th e Ca t a l a n

Therefore after 5 ...a 6 6 0-0 b5 the main line is 7 ltJe5 ltJd5 8 a4 .ib7, when White has two quite different paths. One idea is to play for positional compensation against Black's pawns and weakened squares with 9 b3 (Game 59) . The second possibility is to play in the centre with 9 e4 ltJf6 10 axb5 axb5 1 1 l:!.xa8 .ixa8 1 2 ltJc3 (Game 60) .

Game 46 Rashkovsky-K . G rigorian

Kishinev 1975 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 i.. g 2 a6 6 a4

6 . . . c5 6 ... ltJc6 7 0-0 ltJa5 (7 .. J::tb 8 transposes to 6 0-0 ltJc6 7 a4 l:!.b8) 8 ltJbd2 c5 9 dxc5 .ixc5 10 ltJe5 c3! was Gulko-Bronstein, Vil­ nus 1 975, Black inducing a weakness in the enemy structure and in so doing obtaining equal chances. After 1 1 bxc3 0-0 1 2 ltJec4 .ie7 1 3 ltJxa5 'iVxa5 14 '�3 e5 1 5 ltJc4 'iVc7 1 6 ltJb6 l:!.b8 1 7 c4 .ig4 Black solved his queenside problems. 7 0-0 ttJc6 7 ... cxd4!? 8 'iVxd4 (White can try 8 ltJxd4 e5 9 ltJc2 'iVxd l 10 l:!.xd l but should expect no more than equality after 1 0 ... ltJc6) 8 ... 'iVxd4 9 ltJxd4 e5 10 ltJc2 ltJc6 1 1 .1i.xc6+ bxc6 12 ltJe3 .ie6 13 ltJd2 and now in Rashkovsky-Sveshnikov, Moscow 1 976 Black employed the same idea as in Gulko98

Bronstein, above: 1 3 ... c3! 1 4 bxc3 ltJd5 1 5 ltJe4 h 6 1 6 .ia3 .ixa3 1 7 ':xa3 ltJxe3 1 8 fxe3 � e7 1 9 l:!.aa 1 l:!.ab8 2 0 ltJ c 5 .1i.c4 21 l:!.fd 1 l:!.hd8 with an edge. White can improve with 1 0 ltJf3 ltJc6 1 1 ltJbd2 .ie6 1 2 ltJg5 .ib4 1 3 ltJxe6 fxe6 1 4 .ih3 when Black's pawns are weak and his extra material lacks relevance, leaving White on top. 8 dxc5 8 ltJe5 ltJa5 9 dxc5 �xdl 10 l:[xdl trans­ poses to 8 dxc5 'ii'x dl 9 l:!.xdl ltJa5 1 0 ltJe5, while 8 ... .td7 9 ltJa3 cxd4 10 ltJaxc4 is better for White because of the weakness on b6. This leaves 8 ... ltJxd4 9 e3 ltJb3 1 0 �xd8+ '.t>xd8 1 1 l:!.a3! (1 1 ltJxt7+ �e8 1 2 ltJxh8 ltJxa 1 1 3 ltJa3 l:!.a7 1 4 ltJxc4 b5 1 5 ltJe5 .ib7 favours Black) 1 1 ...'.t>e8 1 2 ltJxc4 ltJxc1 13 Ihc1 ltJd7 1 4 l:!.b3 J:tb8 15 as, as in Gavrilov-Zhukhovitsky, Dubna 1 97 1 . Black has a pawn but problems with queenside development. The position is unclear. 8 . . . 'i!Vxd 1 9 �xd 1 ttJa5 Karpov-Sveshnikov, Moscow 1 976 con­ tinued 9 ... .txc5 10 ltJbd2 ltJg4 1 1 e3 ltJge5 1 2 ltJxe5 ltJxe5 1 3 ltJe4 .ib4 1 4 .id2 .ixd2 1 5 l:!.xd2 �e7 1 6 ltJd6 l:[b8 1 7 f4 ltJg4 18 ltJxc4 .1i.d7 and White won the pawn back with better prospects. 1 0 ttJe5

1 0 . . . ttJd5 Black can also play 1O ... ltJd7 1 1 ltJxd7 .ixd7 1 2 .id2 ltJb3 1 3 .ixb 7 .ixc5! (1 3 ... l:!.d8?! 14 c6!) 14 .txa8 ltJxa 1 1 5 ltJc3

4 . . . dx c 4 5 iL g 2 a 6

tLlb3 1 6 �b 7 when �'hite keeps a small edge due to the vulnerable black pawns. 1O .. .'�jb3!? 1 1 c6 ttJd5 12 l:ta2 bxc6 is worth a try. 1 1 e4 �te can also try the exchange 1 1 i.xd5 exd5 12 ttJc3, but 12 ... i.e6 13 ttJxd5 l:tc8 equalizes. Note that continuing development with 12 ... �xc5 1 3 ttJxd5 O-O? is unsound in view of 14 �e3 �d6 (1 4 ... .l1.xe3? 1 5 ttJe7+ Wh8 16 ttJxt7+! and Black is mated!) 1 5 tLlxt7! 'it>xt7 1 6 ttJb6 etc. 1 1 . . . ttJb4 1 2 ttJd2 iLxc5 1 2 ...ttJb3 13 ttJxb3 cxb3 invites 14 l:!.a3 i.xc5 15 l:txb3 f6 16 ttJc4 with an edge for White. Again Black is a little uncomfortable on the queenside. 1 3 ttJdxc4 ttJc2? 1 3. .. ttJxc4 14 ttJxc4 �d7 1 5 ttJe5 i.c6 1 6 i.f4 and �te i s slightly better, a s i s the case after 13 ... ttJb3 1 4 l:tbl 0-0 1 5 i.e3 .l1.xe3 1 6 tLlxe3. 14 l:tb1 ttJxc4 1 5 ttJxc4 iLd7 15 ... b5 is premature due to 16 e5 l:ta7 1 7 tLld6+ �xd6 1 8 l:txd6, when the bishop pair and Black's vulnerability on the dark squares combine to favour �te. 16 b3 iLc6 1 7 iLb2 0-0 1 S l:tbc 1 ttJb4

1 9 ttJa5 1 9 ttJe5!? is an interesting option. 1 9 . . . iLb6 19 ... i.e7 20 ttJxb7! helps �te. 20 ttJxc6 ttJxc6 21 l:td7 l:tabS 22 e5 l:tfeS

The lesser evil is 22 ...l:tfdS. 23 iLe4?! Not the best. 23 i.xc6 bxc6 24 l:txc6 i.a5 25 l:txa6 .uxb3 26 l:txa5 l:txb2 27 l:!.aa7 gives �'hite a winning rook endgame. He is active on the seventh rank and has a dangerous a­ pawn. 23 . . . g6 24 Wg2 .l:!.edS 25 iLxc6 bxc6 26 l:txdS+ l:txdS? Now Black loses, but 26 ... �xd8 27 ILxc6 l:txb3 28 �c1 'it>g7 29 l:txa6 is scarcely an improvement. 27 l:txc6 iLxf2 2S iLa3 iLd4 29 J:!.d6! litxd6 30 exd6 iLb6 31 d7 iLdS 32 b4 f6 33 a5! Wf7 34 b5 1 -0

White plays 6 ttJe5

Game 47 Fominyh -Sveshnikov

Elista 1995 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 iLg2 a6 6 ttJe5

6 . . . iLb4+ Also possible is 6 ...l:ta7 7 0-0 b6 (7 ... b5?! is dubious in view of 8 a4) 8 ttJc3 i.b 7 (Be­ liavsky gives 8 ... b5 9 a4 b4 10 ttJa2 c5 1 1 ttJxc4 cxd4 1 2 as with compensation for the pawn) 9 'iVa4+ ttJfd7 and now Beliavsky­ Portish, Hungary 1 999 went 1 0 i.xb 7?! l:txb7 1 1 ttJc6 'iVc8! 12 ttJxb8 b5! with good counterplay for Black. Instead after 1 0 l:tdl 99

Th e C a t a la n

iLxg2 1 1 'it>xg2 c 5 1 2 iLe3 White i s better, suggesting that Black's litde queenside adven­ ture is dubious. 7 ltJc3 ltJd5 8 ..td2 b5 Protecting the c4-pawn with the knight has also been tried, 8 .. .'!iJb6 9 e3 lb8d7 1 0 lbxd7 �xd7 I 1 lbe4 iLxd2+ 1 2 lbxd2 0-0 1 3 'it'c2 the course o f Bandza-Kaidanov, Torcy 1 99 1 . After 1 3 ... e5!? 1 4 dxe5 .l:[d8 1 5 0-0-0 'it'e8 1 6 lbxc4 iLf5 1 7 �xf5 lbxc4 1 8 'it'c2 lbxe5 1 9 'it>b 1 c5 20 .l:td5 .l:.xd5 21 iLxd5 g6 22 'it'c3 the game was agreed drawn, al­ though White stands better here, primarily because the bishop is superior to the knight. 9 0-0 White should avoid 9 a4 iLb7 10 e4?! lbf6 1 1 iLg5 h6 1 2 iLxf6 gxf6 1 3 lbg4 f5 when Black has a huge advantage. Neistadt gives 1 0 lbxd5 iLxd2+ 1 1 'it'xd2 iLxd5 1 2 e 4 iLb7 1 3 axb5 axb5 1 4 l:txa8 iLxa8 1 5 'iVa5 'it'xd4! 1 6 'it'xa8 0-0, Black's pawns and the activity of his queen affording him good compensation for the piece. In the event of 9 ... iLxc3 10 bxc3 f6 1 1 e4 lbe7 1 2 lbg4 iLb7 1 3 'iib l c6 14 lbe3 �c8 1 5 0-0 0-0 Black runs the risk of coming under fire on the dark squares, as was dem­ onstrated in Tukmakov-Lputian, Rostov-on­ Don 1 993: 1 6 iLc1 :e8 17 iLa3 lbd7 1 8 iLh3 lb ffi 1 9 iLc5 �c7 20 'iib4 Uad8 2 1 l:tfd 1 and these weak dark squares provided White with good play for the pawn. 9 . . . ..txc3 1 0 bxc3

1 00

1 0 . . . f6 Black has also resisted the urge to hit the knight. 1 0 ... 0-0 1 1 a4 c6 1 2 e4 lbe7 1 3 iLc1 lbd7 14 lbxd7 (14 f4!? is worth a try) 1 4 ... iLxd7 1 5 iLa3 l:1e8 1 6 l:tel lbc8 1 7 'ili'h5 lbb6 1 8 as lba4 1 9 l:te3 favoured White in Razuvaev-Sveshnikov, Tbilisi 1 978. Black's queenside pawn majority lacks mobility. White, on the other hand, has a good dark­ squared bishop and possibilities of a kingside attack. In Sulava-Sveshnikov, Pula 2000 Black chose 1 2 ... lbb6 1 3 as lb6d7. Then 1 4 lbxd7 lbxd7 1 5 e5 (with compensation) is playable, while the game continued 1 4 lbg4 f5 1 5 lbe5 lbf6 1 6 iLg5 h6 1 7 i.xf6 'iixf6 1 8 ':e 1 1:ta7 1 9 l:ta2 with compensation for the pawn due to the presence in the centre. However, White can win back the pawn with 1 6 exfS exfS 1 7 iLxc6 and emerge with an advantage. 1 1 e4 ltJe7 l 1 ...fxe5 12 exd5 0-0 13 dxe5 favours White. Black shouldn't open up the centre too early. 1 2 ltJg4 c6 Another possibility is 1 2 ... e5 1 3 dxe5! iLxg4 14 'iVxg4 �xd2 1 5 exf6 gxf6 16 e5, when Black wins a piece but White's attack looks rather menacing. 1 2 ... iLb7 1 3 a4 0-0 14 'iWe2 is sensible as Black's bishop is not ob­ structed. 1 3 a4 0-0 Also interesting is 1 3 ... e5!? with variations similar to those in the previous note. After 14 dxe5 iLxg4 15 �xg4 �xd2 16 exf6 gxf6 17 axb5 �xc3 18 e5! there is compensation for the piece. 1 4 axb5 cxb5 1 5 e5 ltJd5 1 6 exf6 gxf6 1 7 ltJe3 ltJc6 17 ... i.b7 18 �g4+ 'it>f7 19 i.e4 leaves White with a promising attacking stance, while 1 7 ... lbxe3 1 8 iLxe3 .l:i.a7 1 9 .l:i.e 1 also looks dangerous. 1 8 ltJxd5 exd5 1 9 .l:l.e 1 .l:l.f7 Another try is 1 9 ... i.fS 20 �f3 i.e4, but White sacrifices and gets the better chances

4 . . . dx c 4 5 iL. g 2 8 6

after 21 1he4! dxe4 22 'iYxe4. 20 'i'hS fS? 20...ltJe7 doesn't save Black: 21 l:he7! 'i'xe7 22 .l:!.e1 'ii'd 8 23 i.xd5 l:taa7 24 i.h6. 21 iL.xdS! 'i'xdS 22 .l:teS + .l:tfS 23 iL.hS! iL.b7 24 .l:txfS + Or simply 24 'iYg5+ <j;f7 25 J:.ae1 and wins. 24 . . . .l:txfS 2S 'i'gS + �f7 2S 'i'g7 + �eS 27 .l:te 1 + �dS 2S 'i'xb7 It is interesting to note that material is level now! Black quickly collapses. 2S . . . .l:tgS 29 iL.f4! "i'd7 30 'i'bS + �cS 3 1 d5 1 -0

Game 48 Ivanchuk - K uporosov

Tallinn 1986 1 d4 dS 2 c4 eS 3 ttJf3 ttJfS 4 g3 dxc4 S iL.g2 as S ttJeS cS

This logical central thrust addresses the slight, voluntary weakening of d4 caused by 6 ttJe5. 7 iL.e3 One of the two possible ways to guard the d-pawn, and clearly better than 7 e3, if not enough for an advantage. After 7 e3 Black can go for immediate equaliry with 7 ... cxd4 8 exd4 i.e7 9 0-0 0-0 1 0 ltJxc4 ltJc6 or choose the more interesting 7 ... .l:!.a7!, which had an outing in Konopka-Kucera, Decin 1 995 when 8 ltJxc4 b5 9 ltJe5 i.b7 10 i.xb 7 .l:!.xb 7

1 1 dxc5 �xd 1+ 1 2 <j;xd1 i.xc5 1 3 ltJc3 ltJbd7 14 ltJd3 i.d6 1 5 <j;e2 <j;e7 16 i.d2 l:tc8 17 Ilhc1 l::t b c7 perhaps left Black with something. 7 ltJa3 is considered in the next two games. 7 . . . ttJdS! This is the downside to \X'hite's previous move. After 7 ... cxd4?! 8 i.xd4 i.c5 9 i.xc5 'iYa5+ 10 ltJd2 'iix c5 1 1 ltJdxc4 \X'hite had a useful lead in development in Stean­ Rodriguez, Las Palmas 1 978. S dxcS This is the serious move here. 8 0-0 ltJd7 9 ltJxc4 b5 10 ltJcd2 ltJxe3 1 1 fxe3 cannot be right. In Hort-Portisch, Nice 1 974, 1 1 ....l:tb8 12 ltJe4 f5 13 ltJf2 i.e 7 14 e4 0-0 1 5 e3 i.g5 16 "iVd3 ltJb6 favoured Black. 8 ltJc3 ltJxe3 9 fxe3 ltJd7 also looks harmless, 1 0 ltJxc4 i.e7 1 1 0-0 0-0 12 ltJe4 .l:!.b8 13 a4 'iic 7 being equal in Verat- Rotstein, Cannes 1 992. S . . . ttJd7 ! Experience has shown that this is the most natural move, after which Black is genuinely threatening to take on e3. The immediate 8 ... ltJxe3?! is very risky, e.g. 9 'iixd8+ <j;xd8 1 0 fxe3 <j;e8 1 1 c6 bxc6 1 2 ltJxc4 l:ta7 1 3 ltJc3 l:td7 1 4 0-0, Birnboim­ Rapoport, Jerusalem 1 996, when Black lagged behind in development and \X'hite's knights were productively posted. However, 8 ... f6 is playable. After 9 ltJxc4 ltJxe3 10 'iixd8+ <j;xd8 1 1 fxe3 .ltxc5 1 2 ltJc3 l:!. a7 Black i s still playing catch-up but it looks less serious here, as was demonstrated in Beliavsky-Csom, Vilnius 1 978, when 1 3 ltJe4 .ltb4+ 1 4 'it> f2 b 5 1 5 a 3 bxc4 1 6 axb4 ltJc6 1 7 b5 axb5 1 8 ltJd6 <j;e 7 1 9 ltJxb5 .l:!.xa 1 20 lIxa1 ltJe5 21 .l:!.a7+ i.d7 was even. 9 iL.d4 Practically forced. In reply to 9 ltJxd7 Black has 9 ... ltJxe3! 10 fxe3 .ltxd7, when 1 1 i.xb7 is too riskY in view of 1 1 ....ltxc5 1 2 i.xa8 �xa8 1 3 .l:!.g1 i.xe3 1 4 1:1. f1 �g2 1 5 �d6 "iVb7 and Black has a strong attack for the exchange. Dokhoian-Ftacnik, 1 993 continued instead 1 1 �d4 "iVa5+ 1 2 ..

101

Th e C a t a l a n

lbc3 'iix c5 1 3 i.xb7 (1 3 O-O!? is a possible improvement, but Black should be fine) 1 3. .. �xd4 1 4 exd4 .l:!.b8 1 5 i.xa6 .l:!.xb2 with an edge for Black. 9 . . . ttJxe5 1 0 �xe5

lbe4 �c7 1 8 b4! in Ivanchuk-Novikov, Tal­ lin 1 986, White securing the c5-square for the knight. After 1 8 ... 11d8 1 9 'ifb3 'it'f7 20 lbc5 .Jl.c8 21 l:tfdl White was doing well. An al­ ternative is 1 6 .. J:td8 1 7 bxc4 tt:ie3!?, when play might develop 1 8 cxb5! 'iVb6 1 9 fxe3 'iVxe3+ 20 .l:!.f2 with a mess. 1 7 fxe3 'ii'x e3 + 1 8 �f2 �xg2 19 'it>xg2 �d8 20 �c2 c3 21 �f3 �xd2 22 �xe3 �xd 1 23 �exc3 'it>e7 ! 24 �c6 J:!.d6 25 �c7 + �d7 26 �7c6 �d6 26 . . . .l:.a8 27 .l:.b6 and White's activity as­ sures him of the draw. 27 �c7 + Y> - Y>

Game 49 Raetsky-Naiditsch

1 0 . . . f6 10 ... i.xc5!? is a sharp but apparently per­ fectly sound opportunity. White will get no­ where unless he accepts the challenge: 1 1 i.xg7 .l:Ig8: 1 2 i.d4?? lbf4! 1 3 'iHa4+ i..d 7 1 4 �xc4 lbxg2+ 0- 1 is the miniature Verat-Antunes, Royan 1 989, while 12 i.. e 5 "iVg5 13 f4 'iid 8 14 i.xd5 "iVxd5 15 'iit'x d5 exd5 was also very short but agreed drawn In Tratar­ A.Petrosian, Ljubljana 1 995, where White could have tried 1 4 lbc3!?, e.g. 1 4 ... lbe3 1 5 'iixd8+ 'it'xd8 1 6 i.. f6+ 'lite8 1 7 i.e4 with a murky situation. Salov-Portish, Brussels 1 988 continued 1 2 i.c3 b 5 1 3 i.d4 'iVb6 1 4 i.. x c5 'iVxc5 1 5 lbc3 i..b 7 1 6 lbxd5 i.xd5 1 7 .Jl.xd5 l:.d8 1 8 0-0 .l:!.xd5 1 9 �c2 h 5 with sufficient counter­ play. 1 1 �d4 �xc5 A common trick in these positions. 1 2 �xc5 'ii'a 5 + 1 3 ttJd2 13 lbc3 lbxc3 14 "iVd6 lbd5+ 1 5 'it'f1 'iVd8 is unclear according to Ruban. 1 3 . . . 'ii'x c5 1 4 �c 1 b5 1 5 b3 �b7 1 6 0-0 ttJe3 ! An important improvement. Earlier in the same tournament Black played 1 6 ... c3?! 1 7 1 02

Dortmund 2000 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 �g2 a6 6 ttJe5 c5 7 ttJa3

7 . . . 'ii'x d4? ! Black has problems with his queen after this capture. More natural is 7 ... cxd4, as in the next main game. After 7 ... .l:.a7 8 dxc5 'iixd 1+ 9 'it>xd 1 .Jl.xc5 10 tt:iaxc4 b5 1 1 lbd3! White has the advantage. In Rashkovsky­ Beliavsky, Baku 1 977 the subsequent 1 1 ....l:!.d7 1 2 lbce5 .l:!.d8 1 3 i.d2 i.b6 1 4 a4 left Black's queenside under terrible pressure. 8 'ii'a4+ ! ? Although this check i s not necessarily the strongest move, in practical it is the

4 . . . dx c 4 5 i.. g 2 a 6

most unpleasant to face. After 8 'iYxd4 cxd4 9 lbaxc4 lbbd7 10 lbxd7 lbxd7 1 1 0-0 .i.c5 12 e3 0-0 13 exd4 .i.xd4 1 4 l:!.d 1 e5 1 5 .i.e3 .ixe3 1 6 fxe3 White was a little better in Agzamov-Azmaiparashvili, Yerevan 1 98 1 . There is no doubt that he will regain his pawn, and without losing his lead in devel­ opment. 8 . . . lLlfd7?! Not best. The outcome of 8 ...lbbd7 9 tt'laxc4 b5 has also proved unsuccessful. Note that the following line is not forced: 1 0 'iVa5 lbxe5 1 1 lbxe5 lbd5 1 2 lbc6 'iYf6 1 3 a4! b4 (1 3. . .bxa4! i s better, but White has the initiative) 14 e4 .i.b7? (14 ... .i.d7! is far from ideal but not decisive) 1 5 exd5 exd5 1 6 .i.g5!! 'i'xc6 (1 6 .. .'iVxg5? 1 7 'iVc7 .i.c8 1 8 lbe5 'iVf6 19 .i.xd5 and White wins) 1 7 0-0-0 f6 1 8 .!:the 1 + with a strong attack, e.g. 1 8 ...� f7 1 9 .ixd5+ 'iVxd5 20 l:!.xd5 .i.xd5 2 1 'ii'c7+ �g6 22 'it'd7! .i.g2 23 .i.e3 and White has a fierce offensive against the enemy king, as well as a nice material situation. Meanwhile, there is also to �c2!? lbxe5 1 1 lbxe5 lbd5 1 2 lbc6 "i'f6 1 3 e4 lbb6 14 e5 'it'g6 1 5 .i.e4 �5 1 6 .ie3 with a potent initiative. 8 ... .i.d7 is probably safest. After 9 lbxd7 lbbxd7! (9 ... 'iVxd7 to 'iVxd7+ lbbxd7 1 1 .ixb7 'ua7 1 2 .i.g2 is somewhat better for White) to .i.xb7 l:!.b8 1 1 .i.xa6 c3 1 2 b3 'iVxa4 1 3 bxa4 .i.d6 it is not clear that White is better. Black has good development. 9 lLlaxc4 b5 1 0 'iYa5

1 0 . . . bxc4 Black has a difficult choice, one pitfall be­ ing 1O ... .l:la7? 1 1 e3! and the queen is trapped. After to ... lbxe5 1 1 lbxe5 .ll a7 1 2 'iVb6 lbd7 1 3 'i'xa7 'iVxe5 1 4 .i.f4 'iVxb2 1 5 t[dl Black will not survive for long, but perhaps 1 2 ... .i.d6!? is worth a look. 1 1 lLlxd7 'iYxd7 1 2 i.. x a8 lLlc6 1 3 'iYb6 lLld4 1 4 i.. e 3! White exploits his lead in development. 14 0-0 'it'd6 15 �a5 lbxe2+ 16 �hl is only slighdy favourable. 1 4 . . . 'iYb5 Black gains litde from 1 4 ... lbc2+ 1 5 'it'f1 lbxe3+ 1 6 fxe3 iVd6 1 7 .i.c6+ �e7 1 8 'iVa7+ �f6 1 9 .i.f3, when his chances for survival are slim. 1 5 'iYc7 'iYd7 After 1 5 ... .i.d7 White wins with 16 0-0 lbxe2+ 17 'it'g2 lbd4 1 8 a4 iVb4 19 l:!.fd 1 , leaving Black's kind with inadequate defence. 1 6 'iYxd 7 + i.. x d7 1 7 J:tc 1 i.. d 6 1 8 i.. b 7?! 18 .i.xd4 cxd4 1 9 l:!.xc4 e5 20 .i.c6 is sim­ pler because Black has no counterplay. White WIns . 1 8 . . . <;t>e7 1 9 i.. x a6 l:tb8 20 i.. xc4 l:txb2 21 i.. x d4 cxd4 22 0-0

White is still close to winning but now there are additional technical challenges. 22 . . . i.. a4 23 l:tb 1 l:td2 24 l:tb7 + <;t>f6 25 l:ta7 i.. d 1 26 a4 i.. x e2 27 i.. x e2 l:txe2 28 J:td7 i.. c 5 29 J:tc 1 i.. a 3 After 29 ... .i.b6 30 l:!.bl .i.c5 3 1 as d3 32 1 03

Th e C a t a l a n

l::t xd3 l:txf2 33 l:tfl White is easily winning the ending. 30 �cc7 We5 30 ... e5 3 1 l:tc6+ 'l¥tf5 32 as and the a-pawn is a runner! 3 1 a5 �a2 32 a6 .i.b2 33 a7 d3 34 �xd3 .i.d4 35 �xd4 Wxd4 36 .l:txf7 g5 37 .l:!.xh7 e5 38 �d7 + We4 39 Wg2 g4 40 h3 gxh3+ 4 1 Wxh3 Wf5 42 f3 �a4 43 g4+ Wg5 44 Wg3 .l:ta3 45 Wf2 Wf6 46 We2 We6 47 �b7 Wf6 48 Wd2 1 -0

Came 50 G irnza-Linder

Cerma'!Y 1998 1 d4 d5 2 c4 e6 3 tLlf3 tLlf6 4 g3 dxc4 5

ii g2 a6 6 tLle5 c5 7 tLla3 cxd4 8 tLlaxc4

8 . . . �a7 lbis developing move has been consid­ ered the most natural for some time, Black in danger of running into trouble if he is not sufficiently acquainted with certain factors. For example the ostensibly natural 8 ... �e7 is typical of what might happen if Black be­ lieves the position is harmless. I.Almasi­ Vadasz, Budapest 1 997 went 9 �d2 as 1 0 'iVb3 tLlbd7 1 1 �xb7 a4! (1 1 ...l:tb8? 1 2 �xa5 �xb7 1 3 �xd8 �xhl 14 �b6 and Black doesn't have enough for the queen - the a­ pawn is strong and f2-f3 is coming) 1 2 'it'f3 �xb7 13 'it'xb7 tLlxe5 14 tLlxe5 'it'd5 1 5 'iVc6+ 'it'xc6 1 6 tLlxc6 with a more pleasant 1 04

endgame for White. In reply to 8 ... �c5 9 �d2 is less effective now as Black has control over the b6-square. Consequently in Alburt-I.Ivanov, New York 1 983 the subsequent 9 ...tLld5 1 0 tLld3 i.a7 1 1 �a5 'it'd7 (1 1 ...'it'e7? 1 2 �xd5! exd5 1 3 �b4 and Black is i n trouble) 1 2 tLlde5 'iVe7 1 3 'iVb3 tLlc6 1 4 0-0 0-0 1 5 l:tac1 I:tb8 1 6 tLlxc6 bxc6 1 7 'iVc2 c5 1 8 l::t fd 1 left the onus on White to demonstrate compensation. lbis leaves 9 0-0 0-0 1 0 b4!? (White needs to act energetically) 1 O ... �xb4 1 1 lIbl �c5 1 2 tLld3 �e7 1 3 tLlb6 l::t a7 1 4 fib3 tLlc6 1 5 �b2, Russo-Haro, Guarapuava 1 992. An accurate evaluation of the position is difficult as only one game has been played, but White has compensation here. After 1 5 ... e5 he found 1 6 tLlxc8 'it'xc8 1 7 tLlxe5! tLlxe5 1 8 �xd4 �c5 19 i.xe5 with the advantage of the two bish­ ops in an open position. 9 .i.d2 9 a4 b6 1 0 'iib 3 �b 7! should be okay for Black. Berezin-Novikov, Donetsk 1 998 went 1 1 0-0 �xg2 1 2 'l¥txg2 'iVd5+ 1 3 f3 �c5 1 4 e 4 'iib 7 1 5 tLld3 0-0 1 6 a s with play for both sides. 9 . . . b6 10 iVb3 .i.b7 1 1 .i.xb7 �xb7 1 2 tLla5 �e7 ! ? A provocative move, and also one that looks very strange when we consider the plight of the ffi-bishop. Av.Bykhovsky­ Kaidanov, Irkutsk 1 983 continued 12 ...l:tc7 1 3 'it'xb6 �c5 (1 3 ... Itc 1+?? 14 l:[xc1 'it'xb6 1 5 l:tc8+ 'l¥te7 1 6 l:hb8! and the threat of �b4+ decides the game) 1 4 'iVb3 0-0 1 5 0-0 'it'd5 1 6 'iVxd5 tLlxd5 1 7 �fc1 lIfc8 1 8 tLld3 �b6 1 9 l::t xc7 l:txc7 20 l:tc1 with the stan­ dard Catalan ending advantage. 1 2 ... 11a7 1 3 0-0 (1 3 tLlac6?! tLlxc6 1 4 tLlxc6 'iVa8 1 5 'it'xb6? l:lb 7 1 6 'it'xd4 .l:!.c7 cannot be recommended) 1 3 ... tLle4 1 4 'iVa4+ b5 1 5 'iVc2 tLlxd2 1 6 'iixd2 'iid 5 1 7 tLlf3 �c5 1 8 tLlb3 0-0 1 9 tLlxc5 'iix c5 20 l:tfc1 'iVd5 21 'it'xd4 'iixd4 22 tLlxd4 l:td7 23 tLlb3 l:tfd8 with a draw to come in Dizdar-Novikov, Lucerne 1 997.

4 . . . dx c 4 5 iL g 2 a 6

1 3 0-0 White achieves nothing from 1 3 ttJac6. Kopylov-Yakovich, Hamburg 1 999 contin­ ued 1 3. .. ttJxc6 1 4 ttJxc6 'Was! 1 5 Itel �c7 1 6 'i'xb6 ttJd5 1 7 'iib8+ 'Wxb8 1 8 ttJxb8 l:!.xc1+ 19 i.xel i.d6 20 ttJc6 e5 21 ttJa5 'i;e 7 and Black was better thanks to his development lead. 1 3 . . . tLle4 13 ... bxa5? does not work in view of 1 4 'i'a4+ ttJbd7 1 5 i.xa5 'iib 8 1 6 l:tae l , when White's attack is decisive. 14 'i'a4+ b5 1 5 'i'c2 tLlxd2 1 6 'i'xd2

ttJxd4 ':'xd4 27 l::t c8+ 'i;g7 28 tIxb8 tId 1+ 29 'i;g2 l:!.d2 Black wins a pawn and maybe the game. 24 . . . tLlxc6 25 J::tx c6 'i'b4 26 J::t x a6 it'xb2 27 J::ta 7 b4 28 'itg2 Perhaps Black is a shade better here, but White has sufficient play to hold. However, this is a road down which it might be dan­ gerous to travel. Who knows where Black had improvements? 28 . . . b3 % - % Now he settles for a draw.

Black meets 6 0-0 with 6 . . . tLlc6

Game 5 1 G leizerov-S . Ivanov

Poland 1992 1 d4 d5 2 c4 e6 3 tLlf3 tLlf6 4 g3 dxc4 5 iLg2 a6 6 0-0 This simple move is more dangerous for Black. 6 . . . tLlc6

1 6 . . J:tc7 16 ... 'Wd5?! 17 'iYel ! exposes Black's vul­ nerable back rank. 1 7 J::ta c 1 i.c5 1 8 'i'f4 Perhaps White should put the brakes on here. 1 8 .l:f.xc5 leads to equality after 1 8 . ..l:txc5 1 9 ttJb7 'ifc7 20 ttJxc5 �xe5 21 ttJb3 0-0 22 ttJxd4 tId8 23 tIel , when a draw is on the cards. 1 8 . . . 0-0 Also possible is 18 ... f6 19 ttJec6 ttJxc6 20 l:txc5 ttJe5 21 l:txc7 �xc7 22 ttJb3 and Black IS no worse. 19 llxc5 J::t x c5 20 tLlb7 'i'c7 21 tLlxc5 it'xc5 22 J::t c 1 'i'd6 23 'i'e4? ! White could also try 23 ttJc6 �xf4 24 ttJe7+ 'i;h8 25 gxf4 g6 26 'i;g2 with play for the pawn, although Black still looks better. 23 . . . g6 24 tLlc6 After 24 ttJf3?! tIdR 25 "it'xd4 'iYxd4 26

7 a4 Relatively harmless. 7 e3 and 7 ttJc3 are considered below. Also lacking punch is 7 ttJa3 .i.xa3 8 bxa3, e.g. R ... O-O 9 i.b2 tIbR 1 0 �c2 b 5 1 1 tIad l , when Shipov-S.Ivanov, St. Petersburg 1 996 went l 1 ...tIb6!? 12 e4 .i.b7 13 ttJg5, the main idea behind this question­ able knight manoeuvre perhaps being 13 ... a6 1 4 d5!? Instead there followed 1 3 ... ttJd7 1 4 1 05

Th e C a t a l a n

h 4 h 6 1 5 tt:Jh3 l:e8 1 6 tt:Jf4 e S ! (the downside to the long knight journey) 1 7 dxeS tt:JcxeS and Black was slightly better. There is some compensation for the pawn, but not really enough to recommend trying this position again. In Khalifman-Akopian, Manila 1 992 the situation was a complete mess after 1 1 ...tt:Je7 1 2 e4 .i.b7 1 3 lIfel tt:Jg6 1 4 .i.c3 'iWd7 1 5 h4 �fd8 16 dS. 7 . . . l:tb8 Also interesting is 7 ... .i.e 7!? 8 tt:Ja3 tt:JaS. 8 a5 White is trying to prevent ... b7-bS. Black also has a good game after the alternatives. 8 .i.gS .i.e7 9 tt:Jc3 bS to axbS axbS 1 1 tt:JxbS!? l:xbS 12 'iVa4 l:b4 13 �xc6+ .i.d7 14 �a6 l:xb2 1 5 'it'xc4 0-0, with chances for both sides, needs practical testing, while 8 tt:Ja3 .i.xa3 9 .l:!.xa3 bS to axbS axbS 1 1 tt:JeS tt:JxeS 12 dxeS tt:JdS 13 'iVd4 0-0 14 'iVcs �e7 1 5 .i.e3 �xcS 1 6 .i.xcs i:td8 1 7 .l:[dl left White with good play for the exchange but hardly an advantage in Prokopchuk-Daschian, Mos­ cow 1 999. 8 tt:Jc3 .i.b4 9 l:el tt:JdS 10 �c2 .i.e7 1 1 l:dl 0-0 12 e4 tt:Jcb4 13 �1 tt:Jxc3 14 bxc3 tt:Jd3 1 5 tt:Jd2 tt:Jxc1 16 l:xc1 eS saw Black generate excellent counterplay in Bu Xiangzhi-Drei, Reykjavik 2000. 8 . . . �b4 The most logical move now the a-pawn is exposed. After 8 ... bS 9 axb6 cxb6 White has a chance to create an initiative, e.g. 10 .i.f4 .i.d6 1 1 tt:JeS tt:Je 7 1 2 e4 .i.b 7 1 3 tt:Jc3 0-0 1 4 .i.gS! b S 1 5 .i.xf6 gxf6 1 6 tt:Jg4 tt:Jg6 1 7 h4 with a good attack for the pawn in Peturs­ son-Oil, Groningen 1 997. 12 tt:Jxc4 .i.xf4 1 3 gxf4 0-0 was more peaceful in Timoshenko­ Novikov, Nova Gorica 1 997, when 14 e3 tt:JfdS 1 5 tt:JeS .i.b 7 16 tt:Jd2 tt:Jg6 17 tt:Jxg6 hxg6 1 8 tt:Jf3 �e7 1 9 tt:JeS tt:Jb4 was no worse for Black. 9 'iWc2 There is no route to an advantage here. After 9 'it'a4 0-0 to tt:JeS .i.d7! 1 1 tt:Jxc4 'iYe7 12 'iVdl l:fd8 Black is fIne according to S.Ivanov and Yuneev. Also harmless is 9 1 06

.i.d2 .i.xd2 1 0 'iWxd2 0-0 1 1 'iYc3 bS 1 2 axb6 cxb6 1 3 'iVxc4 .i.b7 14 e3 as 1 5 tt:Jc3 tt:Jb4 1 6 .t!.ad1 .i.a6 1 7 'iVb3 l:tc8 with equality In Maiwald-S.Ivanov, Katowice 1 993. 9 . . 4Jxa5 .

The safest option. Black can also consider taking with the bishop: 9 ... .i.xaS!? to tt:JeS (it is diffIcult to evaluate the compensation after to 'iVxc4 .i.b6 1 1 l:td1 - enough for a pawn, but does it result in an advantage?) to .. tt:Jxd4!? 1 1 'iVa4+ bS 1 2 �xaS 'iVd6! 1 3 tt:Jf3 tt:Jxe2+ 1 4 �hl tt:Jxc1 1 S l:txc1 0-0 with an unclear game (S.Ivanov and Yuneev) . 1 0 'ii'a 4+ 4Jc6 1 1 4Je5 4Jd5 After 1 1 ...0-0 White gets an advantage with 1 2 .i.xc6! bxc6 1 3 tt:Jxc6 .i.d7 1 4 tt:Jxd8 .i.xa4 1 5 tt:Jxe6 fxe6 1 6 .l:txa4, when his supe­ rior pawn structure will help in the ending. 1 2 4Jxc6 bxc6 1 3 'ii'x c6 + �d7 1 4 'ii'xc4 0-0 .

4 . . . dx c 4 5 ii. g 2 a 6

1 S l:!.d 1 Accepting the pawn with 1 5 �xd5 exd5 16 'i'xd5 leads to equality after 1 6...�b5 1 7 'i'xds IibxdS I S lbc3. 1S . . . cS!? 1 6 e4 Or 16 dxc5 i.b5 17 �c2 �e7 and Black is doing fine - Gleizerov. 16 . . . lLlb6 1 7 'ii'b 3 'Wie7 The pawn sacrifice 17 ... a5 I S dxc5 i.xc5 19 .l:Ixa5 lbd5! is interesting, 20 'ifa2 lbb4 21 'i'a3 producing a messy position. 1S ii.f4 J:!.aS 1 9 dS

2S 'litg2? .l:Ixfl ! 29 c;t>xfl l:!.a 1 30 'it>g2 .l:lxb 1 and Black makes a draw. 2S . . .l::t a a 1 29 'iit g 2 J:!.d 1 30 lLlxb3 White wins easily with 30 'iVdS .l:Iabl 31 �d4 e5 32 �e3 .l:Ixb2 33 lbxb3 etc. 30 . . . J:!.xf 1 31 lLlxa 1 J:!.xa 1 32 b4 J:!.b 1 33 bS f6 34 'Wic6 'iitf 7 3 S 'Wic7 + 'iite S 36 it'c2! l:!.a 1 37 b6 lLld7 3S b7 'iite 7 39 it'b2 nd 1 40 bait' lLlxbS 41 'WixbS J:!.dS 42 'Wic7 + nd7 43 'WicS + 'iitf 7 44 it'hS+ 1 -0

Game 52 Wells-Barsov

York 2000 1 d4 dS 2 c4 e6 3 lLlf3 lLlf6 4 g3 dxc4 S ii.g2 a6 6 0-0 lLlc6 7 e3

19 . . . c4? A horrible blunder, after which White wins by force. 19 ... exd5 is necessary, when 20 exd5 �g4 21 d6 (21 Iiel !?) 21 ...�xdl 22 'i'xdl .l:IadS 23 Iixa6 .l:IfeS is balanced. 20 d6! Perhaps Black forgot about this move. 20 . . . cxb3 20 ... �xd6 21 �xd6 'iVxd6 22 .l:Ixd6 cxb3 does not work in view of 23 .l:Ixb6, when White emerges with an extra piece. 21 dxe7 l:rfcS 21 ...i.xe7 22 �e3. 22 ii.e3 ii.cs 23 ii.xcs l:!.xcS White is less well developed but has the significant advantage of a pawn on the 7th rank, so now he decides the game. 24 J:!.xa6! J:!.xa6 24 ....l:IbS 25 .l:Ixb6! is no improvement. 2S J:!.xd7 J:!.c 1 + 26 ii.f 1 lLlxd7 27 eS'Wi + lLlfS 2S lLld2

7 . . . ii.d7 The logical, calm approach. Black has no problem parting with the exchange on as as can be seen in the notes to White's 9th move in the next main game. 7 ... l:!.bS is less accurate, White achieving a slight advantage after S lbfd2!, e.g. S ... e5 9 �xc6+ bxc6 1 0 dxe5 lbg4 1 1 lbxc4 i.e6 1 2 �c2 h5! 1 3 .l:Idl �c8 14 �e4! �d5 15 .l:Ixd5 cxd5 16 'it'xd5 iLe7, Tukmakov-V.Mikhalevski, Biel 1 998. Now after 17 e6! �xe6 IS �xe6 fxe6 19 h3 White is a little better according to Tukma­ kov. White's knights got busy in Lputian­ Y.Milov, Istanbul 2000: S ... �d7 9 lbxc4 b5 10 lbcd2 �b7 1 1 lbc3 lbdS 1 2 lbf3 c5 1 3 e4 107

Th e Ca t a l a n

cxd4 1 4 ttJxd4 ttJc6 15 �f4 l:tc8 1 6 ttJb3 �b4 1 7 e5 'iVxd1 1 8 .l:i.fxd1 ttJd7 19 ttJe4 ttJdxe5 20 ttJec5 �xc5 21 �xe5 �b6 22 �xg7 J:tg8 23 �f6 and Black still had some problems to solve. 8 ttJc3 8 'iVe2 is considered in the next game. 8 . . . ttJd5 One of a number of options open to Black: 8 ... b5 9 ttJe5 ttJd5 10 ttJxd5 exd5 1 1 �xd5 ttJxe5 12 dxe5 c6 1 3 i.e4 and White has the better prospects. 8 ....l:!.b8 looks dangerous for Black. Cs.Horvath-Marciano, Austria 1 998 continued 9 ttJe5 ttJa5 10 e4 b5 1 1 g4! h6?! (risky) 12 f4 c5 13 d5 b4 (1 3 ... exd5 14 ttJxd7 ttJxd7 1 5 ttJxd5 gives White a promising initiative) 14 dxe6 �xe6 1 5 'iVa4+ ttJd7? (the decisive mistake, but after 1 5 ... �d7 1 6 ttJxd7 'iVxd7 1 7 'iVxa5 bxc3 1 8 bxc3 White never­ theless has a terrifying attack) 1 6 f5! 'i;; e 7 1 7 l:t d 1 ! bxc3 1 8 fxe6 'i;;x e6 1 9 ttJxd7 cxb2 20 ttJxffi+ lIxffi 21 :xd8 bxa1'iV 22 iVd7+ 1 -0. Black has to play 1 1 ...b4 12 g5 bxc3 1 3 bxc3! l:tb5! (13. .. ttJg8? is dangerous - after 14 'iVf3 "iie 7? 1 5 �a3 White wins) 14 gxf6 gxf6 1 5 ttJxd7 'iVxd7 1 6 'iVf3 J:tg8 1 7 'i;; h 1 when a draw was agreed in Sosonko-Piket, Rotter­ dam 1 997. However, White would be quite justified in trying for more in this fmal posi­ tion. 8 ... �d6 is a reasonable alternative. White might need to improve on 9 ttJe5 i.xe5 1 0 dxe5 ttJxe5 1 1 �xb 7 l:tb8, when Black is fine. Then 12 'ikd4 .l:!.xb7 13 'ikxe5 0-0 1 4 l:td1 "iie 7 1 5 e 4 �c6 1 6 'iVd4 'iVb4 1 7 �e3 ttJd7! proved unreliable and left White worse in Karpov-Anand, Leon 1 999. The later game Franco-Korneev, Madrid 2000 went instead 1 2 �g2 'iVe7 1 3 'iVd4 'iVd6 14 l:td1 'iVxd4 1 5 l:Ixd4 l:tb6 1 6 ttJe4 �c6 1 7 f4 �xe4 1 8 �xe4 ttJxe4 1 9 ':xe4 ttJf3+ 20 'i;; f2 ttJxh2 21 l:txc4 0-0 and, this time, chances were even. 9 ttJd2 9 "iVe2 ttJxc3 10 bxc3 �d6 1 1 .l:!.d1 b5 1 2 1 08

ttJe5! was the course of Khalifman­ A.Petrosian, Moscow 1 987. Then 1 2 ...ttJxe5 1 3 dxe5 �xe5 14 �a3! l:tb8 1 5 f4 �d6 1 6 �xd6 cxd6 1 7 l:txd6 gives White good play for the pawn, so Black chose 1 2 ... �xe5 1 3 dxe5 ttJxe5, when 1 4 ifh5! ttJd3 1 5 �a3 .l:!.b8 1 6 �e4 favoured White. 9 ttJxc3 1 0 bxc3 b5 1 1 a4 . . .

1 1 . . . l:!.b8 1 2 e4 White can also try 1 2 iVg4 g6 1 3 axb5 axb5 14 'iVe2 ttJe7 1 5 e4 �g7 1 6 e5 ttJd5 1 7 ttJe4 with good play for the pawn. The weak­ ness of the dark squares in Black's camp is significant. 1 2 . . . e5 The open position after 1 2 ... b4 1 3 ttJxc4 bxc3 14 d5 is not in Black's interest. 1 3 axb5 axb5 1 4 d5 ttJe 7 1 5 �a3 ttJg6 1 6 �xf8 ttJxf8

1 7 f4

4 . . . dx c 4 5 iL g 2 a 6

White begins his kingside attack, exploit­ ing the development lead. Also strong is 1 7 'i'h5 tL'lg6 1 8 na7 0-0 1 9 l:tfal 'iVc8 20 h3, when Black is under pressure. 1 7 . . . tUgS 1 8 f5 tUe 7 1 9 fS gxfS 20 J:taS lbg8 This seems to be very ive, and that in itself is a risk. 20 ... c6!? is less clear. 21 �h5 iLe8 22 J:teS iLd7 23 tUf3 ! �e7? Black cracks under pressure, missing 23 ... .i.xc6 24 tL'lxe5! iLxd5! 25 exd5 'iWe7 26 tL'lc6 'iVe3+ 27 �hl .l:Ib6, which is still very difficult but at least forces White to work. 24 J:txe7 �dS 25 tUg5 ! White comes crashing through. 25 . . . tUhS 25 ... fxg5 26 'iVxf7+ �d8 27 l:txd7+ 'iVxd7 28 'iWf8+ �c7 29 'iVc5+ �b7 30 .l:tb l ! and Black cannot defend himself. 2S J:txd 7 ! �bS + There is no defence, e.g. 26 ... 'iVxd7 27 l:!.xf6 'iVa7+ 28 �hl and White's attack de­ cides. 27 'ith 1 fxg5 28 J:tfxf7 �gS 29 J:tfe7 + 'itf8 30 'ti'f3 + 'itg8 3 1 J:teS 1 -0

Game 53 H a ba-Gorin

Pardubice 1999 1 d4 d5 2 e4 eS 3 tUf3 tUfS 4 g3 dxe4 5 iLg2 as S 0-0 tUeS 7 e3 iLd7 8 �e2 b5

9 J:td 1

Another natural move for White is 9 tL'lc3, when 9 ... iLd6 invites 10 e4 e5 1 1 dxe5 tL'lxe5 12 tL'lxe5 .i.xe5 13 f4 with a fluent initiative for White, who generated an offensive in Tkachiev-Beliavsky, Enghien 1 999 after the subsequent 1 3 ... .i.xc3 1 4 bxc3 c6 1 5 f5! 0-0 1 6 iLg5 h6 1 7 .i.h4 'iib6+ 1 8 'iit h l tL'lh7 1 9 l:tadl 'iVc7 2 0 e 5 etc. Beliavsky-Goldin, Nik­ sic 1 997 went 9 ... .i.e 7 10 tL'le5 tL'lxe5 1 1 dxe5 tL'ld5 12 tL'lxd5 exd5 1 3 .i.xd5 c6 14 iLe4 0-0 1 5 b3 iLe6 1 6 1lVc2 h6 with an unclear game. 9 tL'le5 wins the exchange but is not the strongest opportunity available. 9 ...tL'ld5 is the weak response, Tratar-A.Petrosian, Ptuj 1 998 turning out advantageous for White atter 1 0 tL'lxd7 'i'xd7 1 1 b3 cxb3 1 2 axb3 .i.e7 13 iLb2 0-0 14 ttdl tL'lcb4 1 5 e4 tL'lb6 1 6 tL'lc3 l:tad8 17 tL'la2 as 1 8 tL'lc3. The cor­ rect approach is 9 ... tL'lxe5! 1 0 .i.xa8 'iVxa8 1 1 dxe5 tL'le4 1 2 a4 (12 b3! is better) 1 2... 'iVb7 1 3 axb5 axb5 1 4 f3 tL'lc5 1 5 e 4 tL'lb3 1 6 l:ta2 tL'lxc1 1 7 .l:txc1 .i.c5+ 1 8 �g2 iLd4 1 9 'ifd2 c5 and the two bishops and queenside major­ ity gave Black an excellent game in Janssen­ Piket, Rotterdam 1 999. 9 . . . iLe7 Also possible is 9 ... .i.d6 10 e4 e5 1 1 dxe5 .i.xe5 (not 1 1 ...tL'lxe5? 12 tL'lxe5 .i.xe5 13 f4 .i.d6 1 4 e5 .i.c5+ 1 5 �f1 tL'lg4 1 6 e6 and White wins) 1 2 tL'lc3 (1 2 tL'lxe5 tL'lxe5 1 3 h3 c6 1 4 iLe3 is unclear) 1 2 ... 0-0 1 3 .i.g5 iLxc3 14 bxc3 'ife8 1 5 eS .i.g4 16 h3 iLxf3 1 7 iLxf3 'iVxe5 1 8 'iVxe5 tL'lxe5 1 9 .i.xa8 l:txa8 with enough compensation for the exchange according to Kaidanov. 1 0 tUe3 An interesting line is 10 e4 0-0 1 1 tL'lc3 l:re8 12 tL'le5 l:tb8 1 3 iLe3 h6, when White came up with the complicated 1 4 tL'lxf7!? �xf7 15 e5 in Cvitan-Dautov, Dresden 1 998. Then 15 ... tL'lg8 16 'ii' f3+ �g6 17 'iVg4+ �f7 18 .i.xc6! .i.xc6 19 d5 gives White a winning attack, the main point being 1 9 ... iLxd5 20 tL'lxd5 exd5 21 e6+ �f8 22 'ifg6, when mate cannot be avoided. Dautov played 1 5 ... tL'lb4 1 6 exf6 iLxf6 1 7 tL'le4 'iVe7 1 8 tL'lc5 and both 1 09

Th e Ca t a l a n

sides had decent chances. Notice that in many of these positions a pawn is only a positional factor alongside many others. 1 0 . . . ttJb4 A very standard manoeuvre to trade off White's bishop, but this is also rather time­ consuming. Beliavsky gives 1O .. .'�Jd5 1 1 e4 tiJxc3 1 2 bxc3 0-0 1 3 h4 with compensation for White. Who this exchange benefits is not clear. 1 1 e4 ttJd3 1 2 ttJe5 ttJxc 1 1 2 ... b4 1 3 tiJxd7 'iNxd7 1 4 e5 tiJd5 1 5 tiJe4 0-0 1 6 'iNc2 seems to offer White good chances of creating an initiative. Without his bishop Black is unable to maintain sufficient influence on the light squares. 1 3 �axc 1 0-0 14 g4 White has addressed development and needs to set in motion some kind of aggres­ sive action in order to justify the pawn sacri­ fice. 1 4 . . .'i!t'c8 The potency of W'hite's attack becomes evident in the following line: 14 ... c6 1 5 g5 tiJe8 16 d5! �d6 17 tiJxd7 �xd7 1 8 e5 cxd5! (the bishop is going, and this way is best) 1 9 exd6 tiJxd6 20 'iVe5 and White's prospects are slightly preferable. Note that 20 ... b4? 21 tiJxd5! wins for White. 1 5 g5 ttJe8 1 6 ttJxd7 'llVx d7

1 7 e5? An unsound pawn sacrifice which Black should exploit. 17 h4 f6 1 8 e5 is complex. 1 10

1 7 . . . c6? ! Black should accept the invitation and play 1 7 ... .ll xg5! ! 1 8 .ll x a8 .ll x c1 1 9 kIxc1 c6!, trapping the bishop. Then 20 tiJe4 (20 'iVe4 �d8! and ... �g5+) 20 ... �c8 21 .ll x c6?! (21 b3!? and White is not so much worse) 21 ...'iNxc6 22 b3 �d5! favours Black. 1 8 ttJe4 f5 1 8 ... tiJc7 19 tiJf6+ .ll xf6 20 gxf6 looks very dangerous. 1 9 exf6 gxf6 20 ttJc5 �xc5 21 dxc5 'ilie7 ? ! The queen is i n the way o n this rank, and the c-pawn is hanging. After the superior 21 ...�c8 22 gxf6 kIxf6 23 kIc3 kIa7! the bat­ tle rolls on. 22 �xc6 �c8 23 �d7 'llVx c5 24 'llVx e6 + 'it>h8 25 �d5 ttJg7

26 g6 ! ! A fantastic move that opens up the king­ side. 26 . . . �fd8? ! 26 ... tiJxe6 is better. After 27 kIxh7+ �g8 28 .ll xe6+ kIf7 29 kIc3! White wins. One line goes 29 ... .l:i.cc7 30 kIxf7 kIxf7 31 gxf7+ 'it>fB 32 kIg3 and it is all over. 26 ... .l:Ic7 was the best try. Now after 27 �h3 h5 28 kIxc7 �xc7 29 'iNe3 tiJf5 30 �f3 tiJg7 31 kIel White has a fantastic position, probably win­ ning, but there is still some work to do. 27 'llVh 3 1 -0 Black resii-,'11ed because 27 ... h5 is met by 28 'iNxh5+ tiJxh5 29 kIh7 mate.

4 . . . dx c 4 5 iL g 2 a 6

Game 54 Raetsky-Ekstrom

Ziirich 1998 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 iLg2 a6 6 0-0 ttJc6 7 ttJc3

7 J:!.bS Experience has made this the main line. 7 ... fi.e7 allows S 'ilVa4 0-0 9 'iYxc4 when Black has no natural plan of development. 7 ... fi.b4 also looks too optimistic. S 'ikc2 0-0 9 J::!.d l 'ii'e 7 10 i.g5 h6 1 1 i.xf6 'it'xf6 12 e4 nbS 1 3 e5 'it'e7 14 d5 tZJa5 1 5 tZJd4 i.d7 1 6 f4 J::!. fdS 1 7 'it>h 1 i.eS I S rs saw White gener­ ate a promising kingside attack in Shipov­ Bashkov, Cheliabinsk 1 9 9 1 . Note that the two bishops are of little importance here as the position is still somewhat closed. Another strong option is S i.g5 0-0 9 J::!.c 1 i.e7 10 e3, when Bronstein-Kholmov, Moscow 1 957 continued 1O ... tZJd5 1 1 i.xe7 tZJcxe7 12 tZJe4 b5 13 b3 cxb3 14 �xb3 as 15 tZJe5 b4 1 6 tZJc5 and White had good play o n the c-file and in the centre. S e4 iLe7 S ... b5 is considered in the next game. S ... i.b4 can be met with 9 d5!? i.xc3 1 0 bxc3 tZJe7 (1 0. . . exd5 1 1 exd5 tZJxd5 1 2 tZJg5 i.e6 looks risky but is not necessarily un­ sound) 1 1 l:te1 0-0 12 i.a3, when Petursson­ Lechtynsky, Smederevska Palanka 1 9S4 was rather unpleasant for Black after 12 ... exd5?

1 3 exd5 J::!. e S 1 4 d6 tZJrs. Now White wins with 1 5 dxc7 'it'xc7 1 6 g4! J::!. d 8 1 7 'it'e2 tZJd6 1 8 i..x d6 �xd6 19 J::!. ad1 'iVf8 20 l:!.xd8 'iVxdS 21 g5 tZJd7 22 'it'e8+ and Black loses material. An improvement is 12 ... l:teS 1 3 d6 cxd6 14 i.xd6 J::!. a S, when White has clear compensation for the pawn but Black is not necessarily so much worse. 9 'ilVe2 ! ? White needs t o develop his initiative pa­ tiently. After 9 i.g5 0-0 1 0 e5 tZJd5 1 1 h4 tZJxc3 12 bxc3 b5 there is not enough for the pawn, 1 3 tZJd2 i.b7 14 tZJe4 i.xg5 1 5 hxg5 tZJe 7 1 6 'ii'h 5 tZJf5 leaving Black slightly bet­ ter in Franco-Korneev, San Sebastian 2000. A famous main line is 9 d5, which steers play by force to a position where White has a little extra space and Black has the two bish­ ops: 9 ... exd5 10 exd5 tZJb4 1 1 tZJe5 i.rs 1 2 a 3 tZJd3 1 3 tZJxc4 tZJxc1 1 4 J::!. xc 1 0-0

. . .

This position has appeared many times in practical play. Some commentators believe the situation to be equal while others prefer �'hite. Perhaps an assessment somewhere between the two is fair, which means that Black should be able to equalise in the next ten or so moves with accurate play, but will suffer more from inaccuracies. Kaidanov­ Shabalov, Seattle 2000 went 1 5 �d4 tZJeS 1 6 tZJe3 i.d7 1 7 tZJe4 tZJd6 I S tZJc5 i.b5 1 9 l:tfdl i.f6 2 0 �d2 b 6 21 tZJb3 �d7 22 tZJd4 ItbeS 23 b3 as 24 tZJxb5 and a draw was agreed. Another line goes 1 5 b4 l:te8 16 'iVd4 1 1 1

Th e C a t a la n

liJd7 1 7 d 6 cxd6 1 8 liJxd6 .txd6 1 9 'fixd6 liJe5 with complete equality according to Haba. 9 . . . b5 It is dangerous to grab the pawn with 9 ... liJxd4 1 0 liJxd4 'fixd4, e.g. 1 1 l::.d 1 'fic5 1 2 .te3!? (this packs more punch than 1 2 e5 liJd7 1 3 liJe4 'fixeS 1 4 .tf4 'fia5 1 5 'ii'x c4 e5 1 6 .td2 'iVb6 1 7 .te3 c5 1 8 liJc3 liJf6 which was unclear in Klinger-Bonsch, Lugano 1 989) 1 2 .. :iVa5 1 3 'ii'x c4 0-0 1 4 a3 and White has good compensation for the pawn as Black's queen is in need of a safe haven. 1 0 �d 1 0-0 10 ... liJb4 1 1 liJe5 liJd7! 12 b3?! (1 2 'ii'g4 g6 1 3 .th6 liJxe5 1 4 dxe5 liJd3 gives White a slight edge) 1 2 ... liJxe5 1 3 dxe5 liJd3 1 4 .te3 0-0 1 5 f4 ..tb4 16 'fic2 'fie7 17 liJe2 as! 1 8 h4 ..ta6 and Black was clearly better in Lalic­ Chandler, Hastings 1 999/2000. 1 1 d5 exd5 1 2 e5!

1 2 . . . d4! ? Black has to give up the knight one way or the other. 1 2 ...liJd7?! 1 3 e6?! liJb6! benefits only Black, so White should play 13 liJxd5. Then 1 3. ..liJb4 14 e6 liJxd5 15 exd7 ..tb7 1 6 liJe5 c 6 1 7 liJxc6?! (1 7 .tf4, with a mess, is better) 17 ...i.xc6 18 ..txd5 .txd7 19 ..tc6 ..txc6 20 l:txd8 tIbxd8 gives Black all the chances, which leaves 14 liJd4! liJxd5 (14 ... i.b 7?! 1 5 liJxb4! ..txg2 1 6 �xg2 .txb4 17 liJc6 and White wins a piece) 1 5 liJc6 'it'e8 16 .ll. x d5 liJb6 1 7 i.e4! tIa8 18 'iVh5 g6 19 1 12

'ii'h 6 .tg4 20 .l:.e1 with a clear advantage to White according to Haba. An interesting alternative is 12 ... liJb4!? 1 3 exf6 .txf6 1 4 liJe5 when Black has three possibilities. 1 4 ... .l:te8?! 15 liJxd5 liJxd5 (1 5 .. J::tx e5? 16 liJxf6+ 'ifxf6 1 7 'ii'x e5! etc.) 16 l:Ixd5 'fie7 1 7 f4 and White i s well ahead. 1 4. . ..t b7?! 1 5 a3! tIe8 1 6 axb4 J:Ixe5 1 7 'iVc2 c6 1 8 liJe4! was clearly better for White in Raetsky-Astrom, HafnarfJordur 1 997, Black without counter­ play on the dark squares. The sound option is 14 ... d4!, e.g. 1 5 a3 l:te8 1 6 axb4 d3 1 7 .tf4 dxe2 1 8 .l:txd8 tIxd8 1 9 liJxe2 with chances for both sides, although White's task is probably the easier. 1 3 exf6 i.xf6 1 4 it'e4 After 1 4 liJe4 .te7 1 5 a4 d3 1 6 'iVd2 liJb4 Black has good counterplay for the piece. 1 4 . . . i.b 7 1 5 ttJd5 d3 1 5 ... liJb4 16 liJxf6+ 'ii'xf6 17 'ii'f4 c5 1 8 'fixf6 gxf6 1 9 .tf4 and Black needs to be careful. 1 6 'li'f5? ! White chooses the wrong strategy i n try­ ing to deliver mate. Preferable is 1 6 liJxf6+ 'ii'xf6 1 7 'ii' f4! when White has some advan­ tage in the endgame due to his control over the dark squares. 1 6 . . . i.d4 1 7 ttJg5 g6 1 8 it'h3 h 5 1 9 it'h4 i.g7

20 ttJf4?! White continues down the wrong path.

4 . . . dx c 4 5 it.. g 2 8 6

After 20 i.f4 tLJe5 21 tLJc3 i.xg2 22 �xg2 J:1eS he is worse but can at least fight. 20 . . . t"t::le S? A big mistake that justifies White's ap­ proach. Much stronger is 20 ... tLJd4! 21 tLJe4 (21 i.xb7 .l:!.xb7 22 tLJxh5 f6! 23 tLJxg7 fxg5 and Black wins, as is the case after 21 tLJxh5 ttJe2+ 22 �h1 i.xg2+ 23 �xg2 'ii'd5+ 24 f3 gxh5 25 'iVxh5 'iVf5 etc.) 21 ..."ifxh4 22 gxh4 !:tfeS and White will not last long. 21 it.. x b7? No doubt overlooking Black's 22nd move. 21 tLJxh5! is correct, when Black needs to play 21 ...f6! (21 ...gxh5? 22 i.xb7 J::.x b7 23 'i'e4! and White wins) 22 tLJxg7 fxg5 23 i.xg5 'it'd7 24 i.f6 with a complex position. A possible continuation is 24 ... i.xg2 25 i.xe5 i.b 7 26 tLJe6!! 'iVxe6 27 'iYhS+ �f7 2S 'i'h7+ �eS 29 'iYxc7 'it'd5 30 "iixbS+ �f7 3 1 'ifc7+ �e6 3 2 f3 l:txf3 33 'iVb6+ i.c6 34 .l:!.d2 �xe5 and Black is a little better. Both sides have alternatives along the way, of course, but this variation illu strates the nature of these positions. 21 .. ..l::!.x b 7 22 t"t::lx hS?

22 tLJfh3 c5 23 i.e3 .l:!.c7 with a clear ad­ vantage to Black. White no longer has an attack. 22 . . . 'ii'x gS ! ! This simplification completely destroys White's offensive. 23 �xgS t"t::lf 3+ 24 Wg2 t"t::l x gS 2S t"t::l x g7 t"t::le4 26 .l:!.e 1 t"t::lf 6 27 it.. h 6 .l:!.d8 28 .l:!.e7

b4 29 h3 c3 30 bxc3 bxc3 3 1 t"t::le 6 fxe6 32 .l:!.g7 + Wh8 33 .l:!.e7 t"t::l g 8 0 - 1

Game 55 K halifman-Ivanchuk

Minsk 1986 1 d4 dS 2 c4 e6 3 t"t::lf 3 t"t::lf 6 4 g3 dxc4 S it.. g 2 a6 6 0-0 t"t::l c 6 7 t"t::l c 3 .l:!.b8 8 e4 bS!?

Black wants to keep his pawn and emerge with the better chances. 9 'ii'e 2 9 e5?! is unnatural. After 9 ...tLJd5 1 0 tLJg5 i.e7 1 1 h4 h6 1 2 tLJh3 tLJcb4 Black is already on top. 9 . . . t"t::l x d4 1 0 t"t::l x d4 'ii'x d4 1 1 it.. g S White achieves nothing from 11 i.e3?! 'it'd3! 12 'it'el (1 2 'it'xd3 cxd3 13 e5 tLJd7 1 4 i.a7 .l:!.b7 1 5 i.xb7 i.xb7 1 6 f4 c 5 1 7 b 3 g5! and White is in trouble according to Milov) 1 2 ... 'iVdS! 1 3 .l:!.d1 tLJd7 with a better game for Black (Khalifman) . A reasonable alterna­ tive is 1 1 .l:!.d 1 'iVb6 1 2 e5 tLJd7 1 3 tLJe4 tLJxe5 ( 1 3 ... i.b7!? 14 i.e3 'ii'a 5 is unclear) 14 i.e3 c5 1 5 .i.f4 tLJd3 (1 5 ... tLJd7!? 16 i.xbS "iVxbS - unclear - is also sound) 1 6 i.xbS 'it'xbS 1 7 b 3 i.e7 1 S bxc4 bxc4 1 9 'i'c2 0-0 20 'ifxc4 tLJe5 21 'ilVc3 'iYc7 22 l1ac1 and White had a pull in Miton-J akovenko, Yerevan 2000. 1 1 . . . 'ifb6 The messy 1 1 ...i.b7!? 12 .l:!.fd1 'ii'e 5 1 3 i.e3 b4 1 4 f4 "ifa5 1 5 e 5 needs investigating. 1 2 eS 1 13

Th e C a t a l a n

1 2 .txf6?! gxf6 1 3 e5 runs into 1 3. .. fS! and White does not get to the e4-square, while 1 2 l:tad 1 ?! meets with 1 2. . .e5! 1 3 .txf6 'iixf6 1 4 ttJd5 'ii'd 6, nullifying White's development lead. 1 2 . . . ttJd5 Khalifman gives the following line as win­ ning for White: 1 2 ...ttJd7!? 1 3 l:tadl .tc5 1 4 ttJe4 0-0 1 5 l:txd7!? .txd7 1 6 ttJf6+ gxf6 1 7 .txf6. However, Black can play 1 7 ... l:tfd8 1 8 'iih 5 WfS 1 9 'iVxh7 We8 2 0 'iVg8+ .tfS 21 .tg7 'iVc5 and we see no win - after 22 .tf3 .tc8 23 .tc6+ .td7 24 .tf3 a draw is all there

Forced. 1 7 ... l:tc8 1 8 ttJxc 7+! and Black is mated on d8. 1 8 'i'xg7 �d7 1 9 ttJf6 + �c6 20 ttJe4! White has a strong attack, against which defence is very aw1.-ward. 20 . . . �c8? ! 2 0... l:thg8 21 ii'f6 l:tg6 2 2 'ir f3 i s neces­ sary, although Black's chances of survival are slim . 2 1 'i'd4 �d7 22 ttJf6 ! �c8 23 i.e3 b4 24 a4 24 'iVe4+ 'lii>b 5 25 l:td5+! was simpler. 24 . . . bxa3 25 bxa3 J::r b 3 26 �e4+ �b5

IS.

1 3 i.xd5! exd5 14 ttJxd5 �b7 14 .. :�c6 15 l:tadl .te6 (1 5 ... .tb7 1 6 ttJf6+! gxf6 1 7 exf6+ �e6 1 8 'ii'd2 .td6 1 9 l:tfel ii.e4 2 0 ii'd4 puts Black under pres­ sure) 16 'i'f3! .txd5 17 lhd5 'i'e6 1 8 lHdl .te7 1 9 .txe7 iYxe7 20 l:td7 'i'e6 21 l:txc7 and White is better. 1 5 J::rad 1 i.h3? ! Black must play 1 5 ... h6!?, when White can try 1 6 ttJxc 7+ iYxc7 1 7 .l:.d8+ 'i!Vxd8 1 8 .txd8 Wxd8 19 �f3 We8 and the position is rather unclear. 1 6 e6!

27 J::rd 5 + ! i.xd5 28 �xd5 + c5 29 a4+ �b6 30 �xd6+ �b7 31 'i'd5 + �c6 32 �xc4 1 -0

Game 56 Tukmakov-Hulak

Croatia 1999

Prising open the enemy defences with de­ cisive effect. 1 6 . . . i.xe6 16 ... ii.xfl loses to 17 'ii'e 5! .td6 1 8 �xg7 l:lfS 1 9 e7! etc. 1 7 �e5! i.d6 1 14

1 d4 d 5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 i.g2 a6 6 0-0 ttJc6 7 ttJc3 J::r b8 8 e4 b5 9 d5!? White does not waste a second, immedi­ ately pushing forward. The reasoning behind this thrust is that Black has spent time with ... b7-b5. 9 . . . ttJb4 Black does not benefit from opening the centre: 9 ... exd5 10 exd5 ttJb4 1 1 b3 cxb3 (l 1 .. .ttJfxd5 1 2 ttJxd5 'ii'x d5 1 3 ttJd4 places Black under heavy attack) 12 "i:Vxb3 ttJd3 1 3

4 . . . dx c 4 5 i. g 2 8 6

.JigS iLe7 1 4 ttJd4 iLd7 1 5 l:!.ad1 ttJc5 1 6 'i'c2 and Black was under pressure in Gore­ lov-Abramov, USSR 1 99 1 . 1 0 b3! ? White wants to blow the position open as he has completed kings ide development. There is also 10 ttJe5 iLd6 1 1 f4 exd5 12 a3 ttJd3 1 3 ttJxd3 cxd3, when Miton-Jakovenko, Oropesa 1 999 went 14 exd5?! iLc5+ 1 5 Wh 1 .Jif5 1 6 1't'e1+ 'iie 7 1 7 1't'e5 g6 1 8 h3 h5 1 9 l:!.e1 Wd7 20 'ii'xe7+ iLxe7 with advantage to Black. Neistadt suggests 1 4 ttJxd5 ttJxd5 1 5 exd5 0-0 1 6 'ii'x d3 c6 with equality. 1 0 . . . cxb3

1 1 'iWxb3 White also has a very interesting option in the sharp 1 1 ttJd4!?, e.g. 1 1 ...c5 (1 1 ...e5 1 2 ttJxb3 c 5 1 3 a 3 ttJd3 1 4 1't'xd3 c4 1 5 1't'd1 cxb3 1 6 'iVxb3 and White has a structural advantage) 1 2 dxc6 e5 1 3 ttJd5. In Cvitan­ Luther, 1 998 White emerged from 13 ... iLd6 14 c7 iLxc7 1 5 ttJxb4 'iVxd4 1 6 'i'xd4 exd4 1 7 axb3 in front. 1 3. . .ttJbxd5!? is an interesting alternative, e.g. 14 exd5 exd4 1 5 l:!.e1+ iLe7 16 d6! (1 6 iLf4? ttJxd5! 1 7 i.xd5 1't'xd5 1 8 iLxb8 b2 1 9 .l:1b 1 'iVxc6 and Black has healthy counterplay) 1 6 .. .'iVxd6 1 7 iLf4 'iVb4! 1 8 a 3 'i'a4 1 9 iLxb8 iLe6 2 0 iLe5 and White stands better, albeit in a tricky position. 1 1 c5 Greed is dangerous - 1 1 ...exd5 12 exd5 ttJbxd5 13 l:!.e1+ iLe7 14 ttJd4 iLb7 1 5 tt'lrs . . .

and White has a very dangerous attack ac­ cording to S.Ivanov. Black is forced to play 1 5 ... 'it>fB, when White has 1 6 iLg5 tt'lxc3 1 7 iLxb 7 l:!.xb 7 1 8 'i¥xc3 iLd6 1 9 iLxf6 gxf6 20 'iVf3 l:!.b6 21 1't'g4 .J:tg8 22 'iWh4 with a deci­ sive attack. The problem for Black in this line is where improvements might be found. 1 2 dxc6 tDxc6 1 3 nd 1

1 3 . . . 'iWb6 Black is in a difficult situation here. For example 1 3 ... 'ii'c 7 14 iLf4 e5 1 5 tt'ld5 tt'lxd5 1 6 exd5 exf4 1 7 'i'c3! is excellent for White according to Tukmakov. However, 1 3 ...tt'ld7!? might be an improvement. After 1 4 i.f4 l:!.b 7 1 5 l:!.ac1 tt'la5 1 6 'iVc2 iLa3 White went for it in Raetsky-Barsov, Abu Dhabi 2001 with 1 7 tt'lxb5!? axb5 1 8 'ii'x c8 iLxc1 19 "it'xc 1 0-0 20 iLd6 lle8 21 e5 tt'lb6 22 tt'lg5 (22 tt'ld4!? tt'ld5 23 iLxd5 exd5 24 tt'lf5 looks very dangerous for Black) 22 ... tt'ld5 23 l:!.xd5!? exd5 24 iLxd5 h6 25 tt'le4 with compensation for the two ex­ changes. If Black survives he is doing well, but it is not clear that he will. 1 4 i.f4 nb7 1 5 e5 tDd7 After 15 ... tt'lg4 16 tt'le4 White has a very promising position, a sample line being 1 6 ...tt'la5 1 7 'iYc2 .:!.c7 1 8 'iYe2 tt'lc4 (1 8 ... i.b7 19 tt'lfg5! h5 20 h3 iLxe4 21 tt'lxe4 tt'lh6 22 tt'ld6+ and White wins) 1 9 h3 tt'lh6 20 iLxh6 gxh6 21 tt'lf6+ We7 22 'iVe4 and there is no adequate defence against 23 'iih 4. 1 6 tDe4 tDc5 1 7 'iWe3 ! tDxe4 1 8 'iWxe4 1 15

Th e Ca t a l a n

l:tc7 1 8 ... i.c5 19 :ac 1 !? offers White good compensation. 1 9 l:tac 1 ..tb7 ? ! 1 9 ....Jl. c 5 2 0 .Jl.e3! .Jl.xe3 21 'iWxe3 'it'xe3 22 fxe3 :d7 (directed against lbd4) 23 lld6 lbe7 24 lbd4 and White seems to have more than enough compensation for the missing exchange. 20 ..te3 if'aS 21 ltJd4 ltJdS 21 ...lbe7 walks into 22 'iVxb7!! lhb7 23 .Jl.xb7 when White is winning, e.g. 23 ...'tia4 24 lId2 f6 25 lbxe6 rt;; f7 26 lbd8+ rt;;g8 27 .Jl.c5 and Black's king will never be safe. 22 if'd3 l:txc 1 22 ... .Jl.xg2 23 lbb3 and Black must resign. 23 ..txc 1 ..tdS 24 ..tgS!

24 . . . h6 In reply to 24 ... i.e7 White wins with 25 i.xe7 rt;;x e7 26 i.xd5 exd5 27 lbf5+ and Black is about to be destroyed. 2S ltJxe6! fxe6 26 if'g6 + ltJf7 27 ..txdS 27 'iVxe6+! i.xe6 28 i.c6+ .Jl.d7 29 i.xd7 mate. 27 . . . ..te7 2S ..tc6 + 'ilo>fS 29 ..txe 7 + 1 -0 Black meets 6 0-0 with 6 . . . bS

Game 57 llincic-Djuric

Vmjacka Ba'!Ja 1999 1 d4 dS 2 c4 e6 3 ltJf3 ltJf6 4 g3 dxc4 S 1 16

..tg2 a6 6 0-0 bS 7 ltJeS c6

S ltJxc6 Razuvaev's brilliant idea was 8 b3! cxb3 9 lbxc6 'it'b6 l O lba5 l:ta7 (lO ... bxa2 1 1 'ii'c2 axbl 'ili' 1 2 'ili'xc8+ 'iVd8 1 3 'iVxd8+ rt;;x d8 1 4 .l:!.xb 1 lbd5 1 5 1i.d2 with compensation for the pawn) 1 1 lbxb3, when Razuvaev­ M.Gurevich, Riga 1 985 continued 1 1 ....i.e 7 1 2 e4 0-0 1 3 i.e3 J::.d7 1 4 lb l d2 'iVd8 1 5 a4! bxa4 1 6 ':xa4 1i.b7 1 7 lba5! .Jl.c6 1 8 lbxc6 lbxc6 19 lbb3 �6 20 l:tal ! �5 21 lIbl with an advantage to White thanks to the bishop pair and the centre. Another possibil­ ity is l 1 ...lId7 1 2 e4! i.b7 1 3 llel ! i.e7 1 4 e5 tbd5 1 5 'ii'g4 rt;; f8 16 .Jl.g5 h5 17 'iWh4 when the plight of Black's king promises White the more pleasant prospects, although this is an improvement for Black on the dubious 1 5 ...g6?!, when Krasenkow-Kohlweyer, Os­ tend 1 990 continued 16 i.h6 lbb4 17 i.xb7 'iVxb7 1 8 'iVe2 lb8c6 1 9 lbc3 and, once again, Black's poor king was a significant factor. S . . . if'b6 9 ltJeS 9 lbxb8 lIxb8 lO b3 cxb3 1 1 axb3 .Jl.b7 1 2 i.xb7 �xb7 1 3 i.f4 l:tc8 with equality. 9 . . . ..tb7 1 0 ..txb7 White can also try lO e4?! lbxe4 1 1 "iVh5 lbd6 1 2 d5 g6 (the immediate capture 1 2 ... .Jl.xd5 1 3 .Jl.xd5 exd5 1 4 lbc3 'iib7 1 5 lIdl i.e7 1 6 lbxd5 i s nice for White) 1 3 �e2 i.g7 and now 1 4 i.e3?! was tried in Aseev­ Novikov, Lvov 1 984, which went 14 .. .'ikc7

4 . . . dx c 4 5 � g 2 a 6

1 5 ii.f4 0-0 1 6 ttJc3 .l:!.e8 and Black was on top. There is no reason to lose a tempo with the bishop, which brings us to 14 dxe6 O-O! 15 �e3 �c7 16 ii.xb7 'iVxb7 17 �c5 'iVd5 18 �xd6 1Vxd6 19 ttJxf7 'it'e7, when White keeps the extra pawn but the situation re­ mains unclear due to Black's strong bishop. Returning to 1 1 'iib 5, we should consider 1 1 ...g6!?, when 12 ttJxg6?! fxg6 1 3 'ii'e 5 J:!.g8 14 .Jixe4 .Jig7 1 5 1V f4 .tIf8 1 6 't't'g4 .Jixd4 favours Black. 1 0 . . . 'iWxb7 1 1 a4 ttJc6 Also possible is l 1 ...ttJbd7 12 axb5 axb5 1 3 l:'txa8+ 'ii'x a8 14 ttJc3 b4 1 5 'iVa4 1Vxa4 1 6 ttJxa4 ttJxe5 1 7 dxe5 ttJd7 1 8 .Jif4 �e7 1 9 .tIel , given by Cebalo. White picks up a pawn on the queenside and has excellent winning chances. 1 2 axb5 axb5 1 3 J:ba8 + 'iWxa8 14 ttJc3

1 4 . . .�b7 Another try is 14 ... b4 15 ttJxc6 'tWxc6 1 6 �a4 1Vxa4 1 7 ttJxa4 .Jid6 1 8 .Jie3 <j;e7 1 9 .l:!.el c 3 2 0 bxc3 b 3 21 ttbl .l:!.b8 2 2 �g5 h6 23 �xf6+ <j;xf6 24 <j;f1 and White is slightly better in the ending. His pawns are on dark squares and thus dominate the bishop, while the far advanced b3-pawn might prove vul­ nerable. 1 5 ttJxc6 �xc6 1 6 e4! b4 Or 1 6 ... ttJxe4 1 7 'it'D f5 1 8 .l:!.dl ! .Jie7 1 9 d 5 exd5 2 0 ttJxd5 and White's activity i s suf­ ficient to keep the fire burning. 1 7 d5 'iWa6

Also possible is 1 7...�d7 1 8 ttJa4 �e7 1 9 ttJb6 � 5 20 ttJc8 exd5 21 ii.g5 dxe4 22 �xf6 gxf6 23 ttJd6+ .Jixd6 24 'it'xd6 when, despite being two pawns down, White has compensation due to his opponent's poor co-ordination. 1 8 dxe6 fxe6 1 8 ... bxc3 gives White a strong initiative af­ ter 1 9 e5! c2 20 'it'd4 �xe6 21 exf6 etc. 1 9 e5! ttJd7 20 ttJe4 ttJxe5 21 'iWh 5+ ttJg6?! Correct is 21 ...ttJf7, when after 22 ttJg5 g6 23 it'D ttJxg5 24 �xg5 �e7 25 �xe7 'i;;x e7 26 1Ve4 White still has compensation due to Black's insecure king. 22 l:te 1 �e7 23 �g5 �a5 23 ... 0-0 does not help Black in view of 24 .Jixe7 ttJxe7 25 ttJg5 h6 26 ttJxe6 .l:!.f7 27 ttJd8 and White wins material. 24 �g4 'iWd5 24 ... iff5? 25 ttJd6+ 24...iVb6 25 ii.xe7 <j;xe7 26 ttJc5! wins for White. 25 �xe7 ct;xe7 26 ttJg5 e5 27 h4!

27 . . . h6 No better is 27 ... h5 28 it'f5 .l:!.h6 29 .l:!.al . 28 ttJf3 �e6 29 �e4 l:td8 30 h5 ttJf8 3 1 ttJxe5 l:td5 3 2 'it'h4+ 1 -0

Game 58 Vajnerman-Novikov

Lvov 1984 1 d4 d 5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 1 1 7

Th e Ca t a l a n

�g2 a6 6 0-0 b5 7 lLle5 lLld5

8 lLlc3 �b7 Black can also play 8 ... c6 but, after 9 lLlxd5 should avoid 9 ... cxd5 1 0 e4, when disaster can soon strike, e.g. 1O ... i.e7? 1 1 exd5 exd5 1 2 lLlxf7! 'it>xf7 1 3 'iVh5+ g6 1 4 i.xd5+ etc. Best i s 9 . . .exd5 1 0 e 4 iLe6 1 1 a4, when an interesting option is 1 1 ...1:.a7!? 1 2 axb5 cxb5 1 3 exd5 i.xd5 with unclear play. Sosonko-Hiibner, Tilburg 1 979 went l 1 ...b4?! 12 exd5 i.xd5?! 1 3 'iVg4! h5 (or 1 3 ... iLxg2 1 4 l:tel !) 14 iLxd5! cxd5 1 5 'iVrs Ita7 16 l:tel Ite7 17 i.g5 g6 1 8 iLxe7! and White won. 9 lLlxd5 After 9 e4 lLlf6 10 d5 lLlbd7 the sacrifice 1 1 lLlxf7 'it>xf7 12 dxe6+ 'it>xe6 1 3 e5 is not good enough in view of 1 3 ... i.xg2 14 exf6 lLle5 with an advantage to Black. Speelman­ Van Der Sterren, Baku 1 983 continued 1 1 lLlc6 iLxc6 1 2 dxc6 lLle5 1 3 iLf4, when Black could have secured an edge after 1 3 ... lLld3 1 4 e 5 lLld5 according to Speelman. 9 . . . exd5 Not 9 ... iLxd5? 10 e4 i.b7 1 1 �h5! g6 1 2 lLlxg6! fxg6 1 3 1Ve5 lLld7 1 4 'iVxh8 'ii'e7 1 5 h4! 0-0-0 1 6 iLg5 'iVf7 1 7 d5! 1:.e8 1 8 dxe6 Itxe6 19 Itadl i.c6 20 iLh6 'it'b7 21 Itxd7 and Black resigned in Sosonko-Scheider, Buenos Aires 1 978. 1 0 b3 10 e4 dxe4 1 1 'iVh5 g6 12 lLlxg6 fxg6 1 3 �e5+ 'iVe7 1 4 'ii'x h8 lLld7 needs investigat1 18

ing. Beliavsky-Sveshnikov, Frunze 1 98 1 went 1 5 iLh3 lLlf6 1 6 iLg5 'it'f7 1 7 iLxf6 'iVxf6 1 8 �xf6+ 'it>xf6 1 9 a4 W e7 20 axb5 axb5 21 1:.xa8 i.xa8 with compensation in the form of the active king: 22 1:.al iLc6 23 1:.a6 Wd6. Moiseev-Brilla Banfalvi, Correspondence 1 984 took the following course: 1 5 h4 0-0-0 1 6 iLg5 �f7 1 7 iLh3 (1 7 iLxd8 i.g7 1 8 'iVxh7 lLl fS 1 9 'ii'xg7 'ii'xg7 2 0 iLg5 lLl h7 21 i.e3 g5 is unclear) 17 ....l:!.e8 18 �f6 'iVd5 19 i.e3 'it'b8 20 'ii'g5 lLlb6 and, in this semi­ closed position, Black had enough for the exchange, the players agreeing a draw after 21 iLg4 'iVg8 22 h5 iLe7 23 'iVh6 iLfS. Black's approach in H.Olafsson-Van Beek, Antwerp 1 998 is worth repeating: 1 O ... iLe7!? 1 1 exd5 iLxd5 1 2 lLlxf7 iLxf7 1 3 iLxa8 c 6 1 4 iLf4 iLd5 1 5 nel 0-0 1 6 i.xb8 and White was left to contemplate the fate of his trapped bishop. Chances were even after 1 6 ... i.b4 1 7 'YWg4 iLxel 1 8 .l:!.xel 1:.f7 1 9 Ite5 'ii'x b8 20 1:.xd5 �xa8 21 Ite5 �d8. 1 0 . . . c3 In the event of 10 .. .f6 1 1 e4! Huzman and Schneider give l 1 ...fxe5 1 2 'i!Vh5+ 'it>d7 1 3 exd5 'it>c8 1 4 dxe5 lLld7 1 5 bxc4 bxc4 1 6 i.e3 lLlb6 1 7 iLxb6 cxb6 1 8 d 6 iLxg2 1 9 \t>xg2 �d7 2 0 .l::t.fc1 b5 21 a 4 with a n unclear position. White has good play in return for the material deficit, and Black still needs to complete his kingside development. Instead of the capture on e5, 1 1 . ..g6 leads to the following position:

4 . . . dx c 4 5 Ji.. g 2 a 6

Gulko-I.Novikov, Volgodonsk 1 983 con­ tinued 12 exd5!! fxe5 13 dxe5 iLg7 14 iLa3!, White sacrificing another pawn to keep Black's king in the centre of the board. After 14 ... c3 1 5 'iVg4 'iVc8 1 6 e6 c5 there followed 17 d6 0-0 1 8 d7 4Jxd7 1 9 exd7 'iVc7 20 i.xb7 'it'xb7 21 'iVe6+ .l:!.f7 22 l':.ad l .l:!.d8 23 i.xc5 iLf6 with a slight advantage to White, but even better is 1 7 'iWf4! .l:!.f8 1 8 'iih 4 when Black is deprived of the right to castle and White's attack continues. 1 1 e4 dxe4 1 2 'iWh5 g6 1 3 ltJxg6 fxg6 1 4 'iWe5+ flie7 1 5 flixh8 ltJd7 1 6 d5! flig7 ! Black shouldn't give u p the c3-pawn. 16 ... 0-0-0?! 17 �xc3 iLg7 1 8 iLg5! favours White. 1 7 flixg7 Ji.. x g7 1 8 Ji.. xe4 1 8 l':.e 1 c2 1 9 l':.xe4+ �f7 20 iLg5 iLxa 1 21 l':.e7+ '.tg8 22 l':.xd7 iLf6! 23 iLd2 iLc3 and the strong c-pawn secures Black equal play. 1 8 . . . ltJc5 1 9 �e 1 ? ! 1 9 iLg2 c2 20 iLa3 iLxa 1 21 l':.xa 1 4Jd3 22 i.e4 b4 23 iLxd3 bxa3 24 iLxc2 0-0-0 25 l:td 1 l':.xd5 26 l:txd5 iLxd5 results in a level endgame, while 1 9 f3!? is interesting. 1 9 . . .'.t>f7

20 Ji.. g 5?! Again White has a better move at his dis­ posal in 20 iLa3, although 20 ... 4Jxe4 21 .l:!.xe4 i.f6 still leaves Black holding his own thanks to the far advanced c-pawn. 20 . . . h6 21 Ji.. f4 ltJxe4 22 �xe4 Ji.. x d5 23

�ee 1 c5 Black's dangerous queenside pawns tip tlle scales in his favour. 24 Ji.. e 5 Ji.. x e5 25 �xe5 �d8 26 f4 b4! 27 <.t>f2 After 27 l':.d 1 c2 28 l':.c1 Black has 28 ... iLxb3!! (very nice!) 29 axb3 .l:!.d1+ 30 lIel ':xe1+ 31 .l:!.xel c4 and wins. 27 ... c4 28 bxc4 Ji.. x c4 29 �b 1 �d2 + 30 <.t>e3 30 �el l':.b2 31 l':.dl iLb5 doesn't help White. 30 . . . �b2 31 �e 1 Ji.. x a2 32 <.t>d4 �d2 + 33 <.t>c5 c2! 34 �e7 + <.t>f6 0 - 1

Game 59 Rogers-Chandler

Wellington 1986 1 d4 d5 2 c4 e6 3 ltJf3 ltJf6 4 g3 dxc4 5 Ji.. g 2 a6 6 0-0 b5 7 ltJe5 ltJd5 8 a4

8 . . . Ji.. b 7 Another form of is offered by 8 ... c6. Then White can try 9 axb5 cxb5 1 0 4Jc3 iLb7 (l 0. . .4Jxc3?! 1 1 bxc3 l':.a7 1 2 iLc6+ and 'W'hite picks up the b5-pawn with a supe­ riour position) 1 1 4Jxd5 exd5 12 e4, when 12 ... dxe4 1 3 'iVh5 'iVf6 14 iLg5 'iVe6 remains complicated. Black fared worse in Heck­ Zude, Hessen 1 994, which continued 12 ... iLd6?! 1 3 exd5 0-0 14 i.d2 f6 1 5 4Jc6 �c7 1 6 �f3 'iVf7 17 l':.ae 1 . White's doubled pawns in the centre are vulnerable but this 7 79

Th e Ca t a l a n

factor i s outweighed b y his aggressive stance. The immediate 9 e4 puts the question to Black's knight, e.g. 9 ... ttJb6 1 0 d5!? 'i*'c7 1 1 Sl.. f4, when White is doing well. This brings us to 9 ... ttJf6, with a further branch. Zaichik­ A.Ivanov, Beltsy 1 979 continued 10 d5!? cxd5 1 1 exd5 exd5 12 axb5 Sl.. e7 13 ttJc3 Sl..b 7 14 bxa6, when Black could have limited his opponent to a slight edge with 1 4 ... ttJxa6 1 5 ttJxc4 ttJb4 16 lha8 Sl.. x a8 1 7 Sl..e 3. White delayed d4-d5 for a move in Sosonko­ Bouwrnester, Netherlands 1 973, 1 0 ttJc3 Sl..b 7 1 1 d5!? cxd5 12 exd5 exd5 1 3 axb5 axb5 1 4 l:txa8 Sl.. x a8 1 5 ttJxb5 'i*'a5 1 6 ttJd4 Sl..b4 17 Sl..g5 seeing the active blockading strategy providing good play for the pawn. 9 b3 c3 9 ... cxb3 10 axb5 axb5 1 1 l:txa8 Sl.. x a8 1 2 'it'xb3 c 6 1 3 ttJc3 has also been tried. After 1 3 ... ttJd7 1 4 ttJxd5 exd5 1 5 'i*'e3 Sl.. e 7 1 6 Sl.. a3 White's pressure outweighs the pawn deficit, while Romanishin-Marjanovic, Yere­ van 1 989 went 1 3 ... Sl.. e 7 1 4 ttJxd5 exd5 1 5 'iVa2! Sl..b 7 1 6 'i*'a7 'i*'c7 1 7 Sl.. f4 Sl.. d 6 1 8 :c1 'ii'e7 19 .l:.a1 with a big edge for White in view of Black's problems on the queenside. Meanwhile, White also has 10 'i*'xb3, when 10 ... c6 transposes to 5 ... b5 6 a4 c6 7 0-0 Sl.. b 7 8 ttJe5 ttJd5 9 b3 cxb3 1 0 'it'xb3 a6 (Game 85 in Chapter 8), and 1 0 ...ttJc6 1 1 ttJxc6 Sl.. x c6 1 2 e4 ttJb6 1 3 'i*'c3 Sl.. b 7 1 4 axb5 axb5 1 5 llxa8 Sl..x a8 1 6 Sl.. a3 provides enough for the pawn. 1 0 axb5 10 e4 b4!? is an interesting piece sacrifice, trapping the b l -knight. Oil-Hoelzl, Pula 1 997 continued 1 1 exd5 Sl.. x d5 12 'tWh5 g6 1 3 'it'h3 Sl..g7?! (Sosonko suggests 1 3 ... c6!?) 14 ttJxc3! bxc3 1 5 Sl.. a3 .l:ta7 16 l:!.fel l:!.b7 1 7 Sl..x d5 'iVxd5 1 8 'iVh4 c6 19 'i*'f4 Sl.. x e5 20 dxe5 c2 21 Sl..d 6 c5 22 J:tac1 with the better game for White due to his influence on the dark squares. 1 0 " .axb5 1 1 l:txa8 iLxa8 1 2 e4 b4! ? Lputian gives 1 2. . .ttJf6 1 3 ttJxc3 b 4 1 4 ttJb5!? c 6 (not 1 4. . .Sl.. xe4? 1 5 Sl.. x e4 ttJxe4 1 6 1 20

'it'f3 ttJf6 1 7 ttJc6 with a clear advantage to White according to E.Vladimirov) 1 5 d5! cxb5 16 dxe6 (E. Vladimirov-Torhallsson, Gausdal 1 991) 1 6 .. .'ii'x dl 1 7 exf7+ We7 1 8 l:txd1 when White has sacrificed a piece for two pawns but the superior development and Black's poor co-ordination provide adequate compensation. 1 3 exd5 iLxd5 1 4 iLe3 Novikov suggests 14 Sl.. x d5 'it'xd5 1 5 'i*'g4 (more active than 1 5 Sl.. e 3 Sl..d 6 1 6 'i*'g4 0-0 1 7 l:!.c1 Sl.. x e5 1 8 dxe5 ttJc6!, after which Black seized the initiative in G.Kuzmin­ Novikov, Kharkov 1 985) 1 5 ... ttJc6 1 6 ttJxc6 'iix c6 as unclear. The knight is caught again and Black has strong pawns on b4 and c3, but he also has some development problems. 1 4 . . . iLe7 14 ... Sl.. xg2 15 Wxg2 'i*'d5+ 16 Wgl with play along the lines of 1 4 Sl.. xd5. 1 5 iLxd5 'ii'x d5

1 6 'ii'c 2 16 'i*'d3 f6 17 ttJc4 0-0 favours Black. White's knight should retreat to d3. 1 6 " .iLf6 1 7 4Jxc3 17 ttJd3 ttJc6 18 ttJf4 'i*'d7! is not to be recommended for White (Rogers believes Black is clearly better) . The knight is not better on f4. 1 7 . . . bxc3 1 8 'ii'x c3 iLxe5 Or 1 8 ... c5 19 J:tc1 0-0 20 'iix c5 'iixb3 with equality. 1 9 dxe5 4Jc6 20 f4 0-0 21 l:tc 1 4Je 7

4 . . . dx c 4 5 ii.. g 2 a 6

22 ii.. c 5 White should be careful here. 22 'it'xc7 'it'xb3 23 i.c5 liJd5 24 'it'a7 and, due to his exposed king, White cannot take on £8. Black keeps his knight on the board and White's king is more insecure - Black is better. 22 . . . .l:.eS 23 ii.. x e7 .l:!.xe7 24 'ii'c 5 'ii'd S 25 .l:!.a 1 Black gets no opportunity to exploit the enemy king position. 25 . . . h5 2S .l:!.aS 'ii'x aS 27 'ii'x e7 'ii'c s 2S 'ii'd S+ 'iit h 7 29 'ii'd 3+ gS 30 h4 % - %

Game 60 Romero-Antu nes

Havana 199 1 1 d4 d 5 2 c4 eS 3 4Jf3 4JfS 4 g 3 dxc4 5 ii..g 2 as S 0-0 b5 7 4Je5 4Jd5 S a4 ii.. b 7 9 e4

9 . . . 4JfS Khalifman-Ruban, USSR 1 985 went 9 ... liJb6? 10 axb5 axb5 1 1 Iha8 i.xa8 1 2 'iVh5 g6 1 3 liJxg6! fxg6 1 4 'it'e5 with a deci­ sive lead for White. 1 0 axb5 axb5 1 1 .l:!.xaS ii.. x aS 1 2 4Jc3 cS White is well on top after 12 ... b4 13 'it'a4+ liJbd7 1 4 liJb5!, e.g. 1 4 ... i.xe4? 1 5 i.xe4 liJxe4 1 6 liJc6 'it'c8 17 iVa8! etc. 1 3 d5 Also possible is 1 3 i.g5 i.b7. Then a pa­ tient continuation is 14 d5 i.e7 1 5 dxe6 fxe6 16 iVe2 0-0 17 h4 'it'e8 1 8 h5 liJbd7 1 9 liJxd7 liJxd7 2 0 i.xe7 'it'xe7 21 e5 liJc5 a s in Smejkal-Chandler, 1 985, when 22 f4 would have provided compensation for the pawn. More adventurous is 1 4 liJx£7!? 'it'x£7 15 e5 h6, with a choice for the bishop. The faulty 1 6 i.h4?! was seen in Khalifman­ Novikov, Lvov 1 985, when Black should have continued 1 6 ...g5! 1 7 exf6 'it'xf6 1 8 liJe4 iVg6 19 iVaI liJa6. Later, in Nesis-Block, Correspondence 1 987, White sacrificed a piece for an initiative after 1 6 .ixf6 gxf6 1 7 'iVh5+ 'it'g7 1 8 l:tal , the subsequent 1 8 .. .f5! 1 9 liJe2 l:tg8 20 liJf4 'it'd7 21 oUa7 'it'h8 22 liJxe6 'it'xe6 23 l:txb7 maintaining the dy­ namic balance. Black can play the immediate 13 ... i.e7, when after 14 'iVaI i.b7 15 iVa7 'iVc8 Flear gives 1 6 liJf3 h6 1 7 i.xf6 i.xf6 1 8 e5 i.e 7 1 9 liJd2 0-0 20 liJxb5 cxb5 21 iVxb 7 i.b4! as equal. Stajcic-Hoelzl, Austria 1 997 went 1 6 d 5 0-0 1 7 dxe6 fxe6 1 8 i.h3 liJa6 1 9 i.e3 i.d6 20 liJf3 .l:.e8 21 e5 i.b8 22 'iVd4 liJd7 23 l:1dl liJ£8 with chances for both sides Black has a pawn, but White is active. 1 3 . . . ii.. d S Another option is 13 ...cxd5 14 exd5 i.xd5 15 liJxd5 exd5 16 liJg4 i.e 7 17 liJxf6+ i.xf6 1 8 l:1e 1+ 'it'£8 1 9 i.xd5 h5 20 'iVf3 g6 21 b3 'it'g7 22 bxc4 bxc4 23 i.xc4 'iVc7 24 i.d5 with the more harmonious force for White in Cvitan-Ekstroem, Dresden 1 998. 1 3 ... i.e7 1 4 dxe6 fxe6 1 5 'iVe2 0-0 1 6 i.h3 'iVc8 1 7 liJf3 liJa6 18 liJg5 liJc7 19 i.f4 was the course of 12 1

Th e Ca t a la n

Zilberstein-Novikov, Blagoveschensk 1 988, 19 ... h6 20 .1Lxc7 hxg5 21 1l.e5 offering White sufficient play, with Black's damaged pawn structure to aim at. 1 4 ttJg4!? 1 4 .1Lf4 �c7! 15 �al is unclear, rather than 1 4 ... exd5?! 1 5 exd5 cxd5 1 6 iVaI ! ..txe5 17 .1Lxe5 ttJc6 18 .i.c7! �xc7 19 'it'xa8+ with a big advantage to White in Ulibin-Antunes, Bayamo 1 99 1 . 1 4 . . . b4 Black can avoid spoiling the queenside pawns and instead play 14 ...ttJxg4 1 5 'it'xg4 iVf6 1 6 dxe6 fxe6 1 7 l:tdl (1 7 .i.h3!? is an interesting option) 1 7 ... 0-0 1 8 .i.e3 with an unclear position. White has the usual com­ pensation in his more threatening forces. 1 5 ttJxf6 + gxf6 15 ...'it'xf6 16 dxc6 .i.c7 17 ttJb5 favours White. 1 6 'if'a4 1 6 ttJb 1 cxd5 1 7 exd5 '!ie 7 1 8 dxe6 .1Lxg2 1 9 '!ixg2 fxe6 20 �g4 sees White both win back the pawn and maintain the initiative. 1 6 . . . i.b7 ? ! Preferable is 1 6 ... bxc3 1 7 �xa8 cxb2 1 8 .i.xb2 cxd5 1 9 exd5 when White can claim compensation for the pawn. In the game he is just better. 1 7 e5!

1 7 . . . bxc3 Or 17 ... .i.xe5 18 dxc6 0-0 19 cxb 7 bxc3 20 .i.h6 and the bishop pair and ed pawn 1 22

on b 7 give White a clear advantage. 1 8 exd6 1 8 dxc6!? looks pronusmg. After 1 8 ... .i.xe5 1 9 c7+! 'iVd7 20 cxb8'iV+ ..txb8 21 'it'xd7+ '!ixd7 22 .i.xb7 cxb2 23 .i.xb2 1l.e5 24 .i.xe5 fxe5 25 ..ta6 White is close to win­ ning the ending. 1 8 . . . 'iVxd6 1 9 dxe6 fxe6 20 bxc3
No exchange of rooks! Black's king is ex­ posed so White makes sure to keep as many pieces as possible on the board. 23 . . . ttJd7 24 i.e3 ttJe5? The lesser evil here is 24 ... 'it-g7 25 �4 �xb4 26 cxb4 with a problematic ending for Black. 25 'iVh4
29 l:ta5 i.xf 1 30 l:ta8 + l:td8 31 'iVb4 'iVd7 32 'iVf8 + ! 1 -0

4 . . . dx c 4 5 i.. g 2 a 6

Summary Let us compare the two S ... a6 main lines after 6 ttJeS and 6 0-0. If 6 ttJeS practice has demonstrated that after 6 ... cS 7 ttJa3 cxd4 8 ttJaxc4 J:Ia7 White has difficulties achieving an advantage. The variation with 6 0-0 features more variations and promises much more interesting play. After 6 ... ttJc6 White can choose between 7 e3 and 7 ttJc3 and, in the case of the former, with 7 e3 i.d7 8 'iVe2 bS 9 l:!.d1 , there is no reason to write off the older continuation 9 ... i.d6. Cer­ tainly after 10 e4 the departure 1O ... i.e7?! is simply a loss of time, but the break in the centre with 10 ... eS seems to be reasonable. Meanwhile, after 9 ... i.e7 10 ttJc3 a useful manoeuvre was seen in the game Haba-Gorin, where Black found ...ttJb4-d3, exploiting the weakness of d3. The position arising after 7 ttJc3 l:tb8 8 e4 i.e7 9 'iVe2 bS 10 J:f.d1 is critical. The impression today is that after 10 ... 0-0 1 1 dS exdS 12 eS a sharp, open struggle occurs where White's chances are preferable. As for Black, he should be looking out for the aforementioned ma­ noeuvre ...ttJb4-d3. After 8 ... bS 9 'iVe2 ttJxd4 White's initiative and Black's material superiority more or less cancel each other out, while 8 ... bS 9 dS ttJb4 1 0 b3!? seems to give White the bet­ ter chances. Black should be cautious, perhaps avoiding this line altogether. After S ... a6 6 0-0 bS 7 ttJeS ttJdS 8 a4 i.b 7 the challenge with 9 b3 allows Black to try a very interesting piece sacrifice in 9 ... c3!? 1 0 e4 b4 or 10 axbS axbS 1 1 l:txa8 i.xa8 1 2 e4 b4!?, for­ ever leaving the knight entombed on b 1 . Of course White can always return the piece with tiJxc3, but it seems that he can expect little in of an advantage. A kind of tabia occurs after 9 e4 ttJf6 1 0 axbS axbS 1 1 l:ha8 i.xa8 1 2 ttJc3 c6. Previously 13 i.gS was often played here, but it is not clear whether the bishop belongs here. Perhaps this decision should be delayed. After 13 dS i.d6 14 ttJg4!? White has a strong initiative, and al­ though this is not cut and dried, Black should probably resort to 13 ... i.e7, after which a degree of accuracy is required to hold the balance. 1 d4 d5 2 c4 e6 3 lZlf3 lZlf6 4 g3 dxc4 5 i.. g 2 a6 6 0-0 6 a4: 6 ... cS Game 46; 6 ...ttJc6 7 0-0 l:tb8 - 6 0-0 6 ttJeS 6 ... i.b4+ - Game 47 6 ... cS (D) 7 i.e3 Game 48 7 ttJa3: 7 ... 'ifxd4 - Game 49; 7 ... cxd4 - Game 50 6 lZlc6 6 ... bS 7 ttJeS 7 ... c6 - Game 57 7 ... ttJdS 8 ttJc3 Game 58 8 a4 i.b7 (D) 9 e4 - Game 60 9 b3: 9 ... c3 Game 59; 9 ... cxb3 1 0 'iVxb3 c6 Game 85 (Chapter 8) 7 lZlc3 7 a4 - Game 5 1 7 e 3 i.d7: 8 ttJc3 Game 52; 8 'iVc2 Game 53 7 ..l:lbS S e4 8 ... i.e 7 - Game 54 S b5 (D) 9 'iVe2 Game 55; 9 dS - Game 56 -

-

. . .

-

-

-

. .

. . .

-

-

-

CHA PTER SIX

I

4 . . . dxc4 5 i.,g2 CDc6

1 d 4 d5 2 c 4 e 6 3 ttJf3 ttJf6 4 g3 dxc4 5 �g2 ttJc6 5 ... lbc6 is a modern treatment of the Cata­ lan Opening, becoming popular in the 1 980s. There is considerable reasoning behind this development. The Catalan arises mainly from the Orthodox Defence of the Queen's Gam­ bit Declined and from the neo-classical Nimzo Indian Defence. In the Queen's Gambit the move ... lbc6 occurs very rarely at higher levels as it obstructs the c-pawn and therefore hinders the freeing advance ... c7-c5 as well as the often helpful of d5 offered by ... c 7 -c6. (Only in the 1 990s did the Chigorin Defence establish itself as a serious defence for Black, due to the creative input from Grandmaster Alexander Morozevich) . However, in the Catalan ... lbc6 s the thrust ... e6-e5, an alternative to ... c7-c5 in the fight for space in the centre. Additionally the knight attacks the d4-pawn, for which the natural defence is e2-e3, which causes a slight weakening of White's light squares. Finally, the knight is prepared to go to as and defend the c4-pawn and, if the occasion presents itself, begin an offensive with ...lbb4. In reply to 5 ... lbc6 White usually plays ei­ ther 6 0-0 or 6 it'a4. After 6 0-0 l:tb8 with the follow-up ... b7-b5 the lines are very simi­ lar to those of Chapter 8. Moreover, ... a7 -a6 1 24

often transposes to the line 5 ... a6 6 0-0 lbc6 (Chapter 5) and these two chapters are suit­ able for a parallel study in connection with this set-up. Here we shall treat only those cases where Black manages without an early ... a7-a6. White has several continuations at his dis­ posal after 6 0-0 J:tb8, although these options are not of equal value. The aggressive 7 lbe5 is not dangerous here, for after 7 ... lbxe5 8 dxe5 �xd1 9 l:txd1 lbd5! the game is simpli­ fied at White's cost. 7 �g5 enjoys litde popu­ larity as White can wait for this posting. Much more popular is 7 e3 (Game 61). White defends the d4-pawn with the inten­ tion of following up with lbd2xc4, but with the knight going to d2 his control over the centre is reduced and Black will find it easier to execute ... e6-e5, often with the assistance of 7 ... �d6. But here it is perhaps even better to simply protect the c4-pawn with 7 ... b5, which offers sufficient counter-chances. 7 a4 (Game 62), the most natural way to prevent ... b7-b5, weakens both b3 and b4. Then 7 ... a6 is considered in Game 51 in Chapter 5 (after the move order 5 ... a6 6 0-0 lbc6 7 a4 l:tb8), and 7...�b4 and 7 ... lba5 are quite reasonable, but perhaps the most exact is the simple 7 ... b6, when Black appears to have an edge.

4 . . . dx c 4 5 i. g 2 ltJ c 6

White's most popular selection is 7 4Jc3. If Black prefers not to transpose to 5 ... a6 with 7 ... a6 he can choose between 7 ... �b4 (Games 63 & 64) and 7 ... b5 (Games 65 & 66) . 7 ... �b4 prevents White's planned ex­ pansion with e2-e4 and simultaneously pre­ pares kingside castling. Here (unexpectedly) a transposition to the Nimzo-Indian Defence is possible, e.g. 1 d4 4Jf6 2 c4 e6 3 4Jc3 �b4 4 g3 d5 5 �g2 0-0 6 4Jf3 dc4 7 0-0 4Jc6 8 e3 l:tb8. All things considered, the logical follow­ up to playing 6 . ..ll b 8 must be the consistent 7 ... b5. Subsequently, in the case of the natural 8 e4 Black has won a tempo compared to the lines with 5 ... a6, and as a bonus the a6-square remains vacant for the bishop. Not surpris­ ingly the battlefield of the variation starting with 5 ...4Jc6 6 0-0 has not been after 8 e4 (which still earned a place in Game 65) in recent years, but rather the more ambitious 8 lLle5!? (Game 66) . With 6 'i¥a4 White attacks the c4-pawn at once and generates pressure on the a4-e8 diagonal, but Black will also be able to use the early development of the queen to create counterplay. Black has tried many moves in practice - 6 ... �d7 (Game 67), 6 ... 4Jd7 (Games 68 & 69) and 6...�d6 (Game 70) but 6 ... �b4+ (Games 7 1 -75) is seen most often. The position after 6 ... �d7 7 'it'xc4 can also arise after 5 'i¥a4+ �d7 6 'iix c4 4Jc6 7 lLlf3. Black concentrates mainly on queenside development, so 7 ... �d6?! (aimed at ... e6-e5) is insufficient and does not promise equality. More common is 7 ... 4Ja5, which at once attacks the queen and prepares the advance of the c-pawn. Now 8 'iic2 is not the strong­ est because after 8.J::tc 8 and ... c7-c5 the queen feels the pressure down the c-ftle. The main line is 6 ... �d7 7 'ii'xc4 4Ja5 8 'iVd3 c5. The eccentric manoeuvre ... 4Jd7-b6 (as seen in Games 68 & 69) does not impress, the knight being no better on b6 than it was on f6. Nevertheless, 6 ... 4Jd7 7 'iix c4 4Jb6 has

become popular in Grandmaster practice. Black gets rid of the irritating pin and frees his position with the help of ... e6-e5. Let us consider the situation after 8 'it'd3 e5 (8 ...lLlb4, with further selection between ... e6e5 and ... c7-c5, is seen less often). The con­ tinuations 9 dxe5 and 9 0-0 are both harm­ less, but Alexander Raetsky's 9 �e3, consoli­ dating the d4-pawn, introduces interesting challenges. Notice that it is not clear whether White profits from the inclusion of 9 �g5 f6 and then 1 0 �e3. Furthermore, it is reason­ able for Black to play 9 ... �e7. Nevertheless the continuation we see most often is 9 4Jxe5 4Jb4, when White has tried 1 0 'iVd1 , 1 0 iVb3 (so that after 1 O ... �e6 1 1 'iVd1 the b7-pawn has been weakened - an idea meant to discourage 1 1 ...'ii'xd4) and 1 0 'iVc3, which often leads to the exchange of queens after 1 0 ...'ii'xd4 1 1 0-0 'iix c3 1 2 4Jxc3. The main reply to 6 'ii'a4 is 6 ... �b4+. Black prevents 'ii'xc4 and after 7 �d2 4Jd5 (7 ... �xd2?! 8 4Jbxd2 gives White a stable Catalan advantage after winning back the c4pawn) takes the game into wild complica­ tions. Since 1 998 7 .. �d6 has been seen in tournament play (including super-GM events) . Black switches strategies and plays for the realisation of ... e6-e5, providing the Catalan prisoner on c8 with some air. This variation should, of course, be compared to the line with an immediate ... �d6. 6 ... �d6 (Game 64) is seen much less frequently. Here White has the manoeuvre 4Jb 1 -d2xc4, and after 7 0-0 0-0 there is 8 l::!d 1 , which can hinder ... e6-e5. After 6...�b4+ 7 �d2 �d6 White has several possible continuations, the main fea­ ture being whether or not to allow ... e6-e5. I f h e wishes t o take the prophylactic approach, then after 8 �c3 and 8 4Je5 tactics will dominate. But after 8 'iVxc4 e5 and 8 0-0 0-0 9 'iix c4 e5 the game goes down a strategic path, and one that has clear reference points. Although subtle, the bishop manoeuvre has not been able to take its place as the main 1 25

Th e C a t a l a n

continuation, which is still 6 'iVa4 .ib4+ 7 .id2 lbd5. Here, apart from 8 .ixb4, White occasionally essays a strange queen manoeu­ vre with 8 'ifb5 in order to collect the c4pawn. Black can keep the extra pawn by an almost forced sequence: 8 ... .ixd2 9 lbbxd2 c3 1 0 bxc3 lbxc3 1 1 'iVd3 lbd5 1 2 0-0 0-0 when, after having lost a pawn, White relies on his completed development with e2-e4, establishing a strong pawn centre, when Black should try to avoid weaknesses and, with accurate manoeuvring, endeavour to hang on to the pawn by sitting out White's initiative. To play such a position as Black can be rather unpleasant, especially for those players whose style is on the aggressive side. Instead of 8 ... .ixd2 9 lbbxd2 c3 Black often prefers 8 ... .ie7 or 8 ... 0-0 (Game 72). The main position of the 6 'iVa4 complex arises after 6 ... .ib4 7 .id2 lbd5 8 .ixb4 lbxb4. The fascinating complications that develop here have been tested for more than half a century (since Aisenstadt-Taimanov, Leningrad 1 949, in fact) . 9 a3 (Game 73) leads to fireworks after 9 ... b5 1 0 �xb5 lbc2, when Black picks up the rook but loses the c6-knight and, subsequently, the cornered knight. However, the conventional 'advan­ tage' of having two knights against a rook is nullified by the unhappy lot of White's king, which is stripped of the right to castle. The practical results at the end of the last century were so poor for White that the authors of this book produced an article in Chess in Rus­ sia in 1 997 that was dedicated exclusively to 9 a3 and went under the title The Dying Con­ tinuation'. Then in 2001 in Wijk aan Zee Kasparov himself played it! The truth was that he was content with a draw in that game, but of course 9 a3 took a step forward in its development. 9 lbe5 (Game 74), which immediately ex­ ploits the pin on the a4-e8 diagonal, has not received widespread attention. The reason is obvious - Black does not have any problems here. However, simple development of the 126

kingside with 9 0-0 (Game 75) does pose problems for Black. The games played from this position have often continued 9 ... l:tb8 1 0 lbc3 a 6 1 1 lbe5 0-0 1 2 .ixc6 lbxc6 1 3 lbxc6 iLxc6 14 'iVxc4 l:txb2 1 5 l:tabl l:tb6 1 6 �c5, which is an important tabia for the evaluation of the whole variation. For Black, despite the extra pawn, it is necessary to fight for equal­ ity. White has more space and the knight is obviously stronger than the bishop on c8. Black is in the unenviable situation of having to defend with some precision, and it is no wonder that in recent practice Black has avoided this critical position.

White plays 6 0-0

Game 6 1 Rustemov-Sax

Germatry 2000 1 d4 d5 2 c4 e6 3 tZJf3 tZJf6 4 g3 dxc4 5 �g2 tZJc6 6 0-0 l:tb8

7 e3 A major alternative is 7 .ig5 .ie7 8 e3 0-0 9 lbfd2 (9 lbbd2 b5 and Black holds on to the pawn) 9 ... e5! 10 iLxf6 (10 iLxc6 bxc6 1 1 dxe5 lbg4 1 2 .ixe7 'iVxe7 1 3 lbxc4 lbxe5 and White's weak squares leave him worse) 1 0 ... iLxf6 1 1 d5 e4!, which was the course of D.Gurevich-Adams, Biel 1 993. There fol­ lowed 1 2 lbxc4 b5 1 3 lbc3 bxc4 1 4 dxc6 l:lxb2 1 5 'iVc1 l:tb8, when Black had the

4 . . . dx c 4 5 iL. g 2 ljj c 6

bishop pair and White's pieces lacked co­ ordination. Adams offers 12 'it'c1 'ii'x d5 1 3 lLlxe4 lIVeS, when White regains the pawn and mass exchanges will follow, leading to a level rniddlegame. 7 . . . b5 7 ... �d6 8 4Jfd2 e5!? is interesting. Then after 9 �xc6+ bxc6 1 0 dxe5 1I.. x e5 1 1 4Jxc4 1I..d6 1 2 it'c2 the vital light-squared bishop leaves the arena but White has play against Black's damaged queenside. Kozul-Plachetka, Charleville 2000 took a much different course - after 9 d5 4Je 7 10 4Jxc4 0-0 1 1 e4 1I.. d7 12 a4 4Jg6 1 3 4Jxd6 cxd6 14 4Ja3 4Je8 15 �e3 b6 16 lib3 White stood better in this King's Indian structure, with the bishop pair and easy play on the queenside. Note that Black's counterplay on the kingside seems rather slow. 8 b3 8 a4 a6 9 axb5 axb5 10 b3 cxb3 1 1 4Jbd2 1I..e 7 1 2 4Jxb3 0-0 1 3 �a3 �xa3 1 4 Iha3 lIVd6 1 5 'it'a I ! b4 1 6 l::ta2 4Je4! 17 4Jfd2 li'ixd2 1 8 .u.xd2 4Je 7 was instructive in Steckner-Ionov, Dortmund 1 99 1 . Again White is happy to play a pawn down, with pressure on the c-ftle and control of the key c5-square. Black, typically, is not without chances of his own. 8 'ii'e2 �e7 9 lIdl 0-0 1 0 e4 �b7 1 1 4Jc3 b4 12 d5 exd5 1 3 4Jxd5 4Jxd5 1 4 l::tx d5!? 1I..d 6 15 'iVxc4 is interesting, while Kachiani Gersinska-Rabiega, 1 995 continued 14 exd5 4Ja5 1 5 i.e3 i.f6 1 6 1I.. xa7 I.Ie8 1 7 'iVc2 lIa8 1 8 1I.. d4 i.xd4 1 9 4Jxd4 'it'f6 20 li'ic6 4Jxc6 21 dxc6 �a6, Black's pieces tak­ ing up influential posts. 8 . . . cxb3 9 axb3 as 1 0 'ii'e 2 iL.e7 1 1 .l:!.d 1 0-0 1 2 iL.b2 lDb4 1 3 lDbd2 White played the immediate 13 e4 in Tratar-Tukmakov, Bled 1 996, which contin­ ued 1 3 ... �b7 14 4Jc3 �e8 1 5 .:tac1 l:td8 1 6 'Yi'e3 (1 6 e 5 4Jfd5 1 7 4Je4 and White con­ trols the c5-square and therefore has full compensation for the pawn) 1 6 ... h6 1 7 h4 4Jg4 1 8 'ii'f4 fS 19 exfS exf5 (1 9 .. J::tx fS!?

leads to exchanges and a level endgame after 20 'ii'xg4 h5 21 'ii'h 3 i.xf3 22 �xf3 I:txf3 23 'iixe6+ 'it'f7) 20 iVxc7 l::t d7 2-1 'ilib6 l::t f6 22 iVa5 'ii'a 8 with a complex struggle ahead. 1 3 . . . iL.b7 1 4 e4 c5 1 5 e5 lDfd5 Another option is to protect c5 and in­ stead put the bishop on d5, 1 5 ...4Jd7!? 1 6 4Je4 �d5 giving chances for both sides. 1 S dxc5 iL.xc5 1 7 lDe4 iL.e 7 More straightforward is 1 7 ... J:tc8 1 8 4Jxc5 lIxc5 1 9 l:1ac1 ttxc1 20 l::t xc1 with compen­ sation, although it makes a lot of sense to keep the dark-squared bishop. 1 8 h4! ? 'it'e8? Better and more natural is 1 8 ... l::t c 8 1 9 4Je1 'ilib6 20 'it'g4 l::t fd8 21 I.Id2 with com­ pensation. 1 9 lDfg5 1 9 4Jd6 �xd6 20 exd6 'iVd7 21 4Jg5 is another means of attack. 1 9 . . . hS

20 'it'g4! ! f5 20 ... hxg5 21 hxg5 lIVc6 (21 ...g6 22 4Jf6+ 1I.. xf6 23 gxf6 4Jc2 24 iVh4 and Black is about to be mated) 22 l::t ac1 (22 4Jf6+? 4Jxf6 leaves Black a piece up) 22 ... lib6 23 4Jf6+ i.xf6 24 gxf6 g6 25 'it'g5 and (again) the weak dark squares around Black's king will prove decisive. Nor does 20 ...iVc6 21 l::t ac1 'iVb6 22 4Jf6+ �xf6 23 exf6 g6 24 i.d4 help Black - a knight sacrifice on e6 is corning next. 21 exfS lDxfS 22 iL.xfS iL.xfS 23 lDxfS + 127

Th e Ca t a la n

l:txf6 24 �xb 7 l:txb 7 24 ... hxg5 25 hxg5 l:tf8 26 g6 Will S for White. 25 ttJe4 Or 25 'iVe4 hxg5 26 'ii'xb7 gxh4 27 'iVe4 with a huge advantage. 25 . . . l:tg6 26 'i'e2

The party is almost over for Black. 26 . . . l:tc7 27 h5 l:tc2 28 'i'f3 1 -0 The rook is trapped.

Game 62 Sandner-Luther

Bad Zwesten 1999 1 d4 d 5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 �g2 ttJc6 6 0-0 l:tb8 7 a4

7 . . . b6 7 ... a6 transposes to 5 ... a6 6 0-0 ttJc6 7 a4 l:lb8 (Game 5 1 in Chapter 5) . After 7 ... i.b4 1 28

the modest 8 e3 is too slow: 8 ... ttJa5 9 .ltd2 'iVe7 1 0 ttJa3 ttJe4 1 1 i.xb4 'iVxb4 1 2 'iVe1 'iVxe 1 1 3 .l:!.fxe1 ttJd6 and Black holds on to the pawn. Maslov-S.Ivanov, St. Petersburg 1 992 saw the more fluid 8 'iVc2 ttJxd4 9 ttJxd4 'iYxd4 1 0 .lte3 'iNd6 1 1 'iNxc4 0-0 1 2 ttJc3 'iNe7 1 3 .ltxa7 l:ta8 1 4 .ltd4 l:ta5, and now 1 5 .ltxf6 gxf6 1 6 l:tfd1 secures White an edge in view of his superior pawns and de­ velopment. 7 ... ttJa5 looks promising. Then 8 ttJbd2 c5 9 dxc5 .ltxc5 10 'iVc2 �c7 1 1 e4 e5 1 2 ttJxe5? soon led to disaster for White in Gorodilov-Raetsky, Vladimir 1 987 (although Black was already slightly better) after 1 2 ... 'iVxe5 1 3 ttJxc4 ttJxc4 1 4 .ltf4 'iYh5 1 5 .ltxb8 ttJg4 1 6 h 3 ttJge3! etc. Instead the continuation 10 ttJe5 c3 1 1 bxc3 0-0 1 2 'iNc2 b6 steers the game to equality after 1 3 ttJb3 ttJxb3 14 'iVxb3 i.b 7 1 5 ttJc6 .ltxc6 1 6 .ltxc6 'iYc7, when White has the bishop pair but weak pawns on the queenside. Black also has 12 ... 'iVc7 here, Chashev-Raetsky, Makhach­ kala 1 987 producing another equal position after the subsequent 1 3 ttJd3 .ltd6 1 4 c4 e5 1 5 c5 b6 16 .lta3 bxc5 17 ttJxc5 .ltb7 1 8 llfc 1 .ltxg2 1 9 'iitxg2 nfc8. The possession of the c5-square is not so relevant here. 8 ttJa3 Again White can try 8 e3, and again it seems a bit slow. Alburt-I.Ivanov, Long Beach 1 989 is a good example: 8 ... .ltb 7 9 ttJbd2 ttJa5 1 0 'iVc2 ttJd5!? 1 1 ttJe5 (1 1 ttJxc4? ttJb4 1 2 'iVc3 ttJxc4 1 3 'iVxc4 .lta6 and Black is winning) 1 1 ...ttJb4 1 2 'iVd1 ? (White has to b e satisfied with a repetition with 1 2 'iVc3 ttJd5) 1 2 ... .ltxg2 1 3 'iitxg2 'iVd5+ 1 4 'iYf3 f5! (closely monitoring e4) 1 5 e4?! 'iYxd4 1 6 'iVh5+ g6 17 ttJxg6 hxg6 1 8 'iVxg6+ �d7 1 9 ttJf3 'iNxe4 20 l:te1 'iVd5 21 'iNf7+ .lte7 22 .ltg5 nbe8 23 l:le5 'ii'c 6 24 l:tae1 l:lhf8 25 'iNh7 ttJd3 and Black won. White's attack was simply too optimistic. 8 . . . �xa3 9 bxa3 Preferable to 9 l:lxa3 .ltb7 10 .l:!.e1 ?! (10 'iiVc2 ttJxd4 1 1 'iYxc4 offers White some

4 . . . dx c 4 5 iL g 2 tD c 6

compensation for the pawn) 1 O ... ttJe4 1 1 l:te3 f5 12 b3 cxb3 13 .l::!.x b3 0-0 14 'i¥d3 .l:.f7 1 5 J:tbl .l::!.d 7, when Black was in control o f the light squares in the centre in Chetverik­ Siklosi, Kecskemet 1 995. 9 ... iLb7 1 0 iLb2 0-0 1 1 'ilt'c2 tDa5 1 2 ltJe5 iLxg2 1 3 ..t>xg2

1 3 . . . tDe8 The direct approach 13 ... c5!? 14 dxc5 'i'd5+ 1 5 Wgl 'i¥xc5 gives Black more breathing space and better chances due to the extra pawn. 14 e4 tDd6 1 5 iLc3 tDb3 Also possible is the more forcing 15 ... f6 16 �xa5 fxe5 1 7 dxe5 bxa5 1 8 exd6 cxd6 1 9 'i'xc4 with chances for both sides. 16 l:tab 1 'ii'e 8 1 7 tDxc4 tDxc4 1 8 'ii'x b3 ltJd6 1 9 f3 'ilt'd7 20 l:tfd 1 c6

White has the superior minor piece and more space, but Black is solid.

2 1 l:tbc 1 l:tfd8 22 iLb4 tDe8 23 iLd2 l:tbc8 24 as b5 25 iLg5 f6 26 iLe3 tDd6 27 a6 tDc4 28 a4 f5 29 axb5 cxb5 30 exf5 exf5 31 iLf4 Black has secured himself an outpost on c4, while White has a ed pawn in the centre. 3 1 . . . 'ii'd 5 32 l:tb 1 l:te8 33 l:te 1 l:txe 1 34 l:txe 1 h6?! 34 ...ttJb6 35 l:tb 1 b4 36 .1i.d6 'iix b3 37 l:txb3 ttJd5 is equal. 35 l:te5 35 l:te 7!? is interesting. 35 . . . 'iIt'xd4 36 l:txb5 ..t>h7 37 'ilt'b 1 l:tc5 38 l:txc5 'ilt'xc5 39 g4 ..t>g6 40 'ilt'xf5+ 'ilt'xf5 41 gxf5+ ..t>xf5 42 ..t>g3 42 .1i.b8 ttJe3+ 43 �g3 ttJd5 44 .1i.xa7 ttJc7 45 .1i.d4 ttJxa6 46 i.xg7 �g6 presents White with more chances to play for a win, although a draw is the most likely result. 42 . . . tDe5 43 iLe3 % - %

Game 63 Ricardi-Smyslov

Buenos Aires 1990 1 d4 d5 2 c4 e6 3 tDf3 tDf6 4 g3 dxc4 5 iLg2 tDc6 6 0-0 l:tb8 7 tDc3 iLb4

8 e3 Three other important continuations have been tried here. 8 ttJe5 simplifies too much: R ... ttJxe5 9 dxe5 'it'xd l 10 l:!.xd l .1i.xc3 1 1 bxc3 ttJd5 with equal chances. 8 l:tel ttJd5 9 1 29

Th e Ca ta la n

'iic2 .i.e7 1 0 a3 tUa5 1 1 lIbl 0-0 1 2 e4 i s a bit more ambitious, White turning his atten­ tion to the centre after some preparatory work. However, Black is ready to meet this expansion, e.g. 1 2 ... tUxc3 13 'ii'x c3 b6 1 4 .i.f4 .i.d6 1 5 .i.g5, Poluljahov-Raetsky, Sim­ feropol 1 989 with equality after 1 5 ... i.e 7. If Black is not satisfied with a draw, then 14 ... i.b7!? is an interesting alternative here. Moiseenko-C.Horvath, Pula 1 999 contin­ ued 8 .i.g5 0-0 9 lIc1 h6 10 .i.xf6 'it'xf6 1 1 tUe4 'iWrs 1 2 tUed2 'iVa5 1 3 a3 i.xd2 1 4 tUxd2 'iVb6 1 5 i.xc6 �xc6 1 6 l:'txc4 'i!Vd6 1 7 �c2 b6 1 8 lIxc7 .i.b7 1 9 tUf3 .i.xf3 20 exf3 'ii'xd4 21 lIdl 'it'f6 22 'it>g2, when White's activity (rook on the seventh rank) is enough for the better game. Black has a possible improvement in 17 ... i.d7!?, which has the advantage of not comrniting the queenside structure. S . . . 0-0 9 'ife2 9 'ii'c2 is another reasonable move, when 9 ... b5 10 lIdl 'it'e8 1 1 e4 e5 12 dxe5 tUxe5 1 3 tUxeS �xe5 was the course o f Bezold­ Romanishin, Balatonbereny 1 995. After 1 4 .i.f4 'it'e7 1 5 e 5 i.xc3 1 6 �xc3 the bishop pair and activity provide compensation. 9 . . . b5 1 0 a4 a6 1 1 axb5 axb5 1 2 .l:.d 1 'VC'hite controls the a-file. 1 2 . . . ttJa5 As in other lines after a2-a4, the b3-square has become a target. 1 3 ttJe5 i.b7 1 4 i.xb7 �xb7 1 5 e4 1 5 'it' f3 tUd5 1 6 .i.d2 .i.xc3 17 .i.xc3 tUxc3 1 8 l:i.xa5 leads to complicated play where White shouldn't be worse. 1 5 . . .'ifeS 1 6 i.g5 ttJd7 1 7 'i'g4 'it>hS 17 ... rs!? deserves attention. 1 S ttJxd7 'ifxd7 1 9 i.f6! ? Tough play! 1 9 d 5 tUb3 2 0 .l:ta6 'iWc8 21 dxe6 c6 is complex, although Black still has an extra pawn. 1 9 . . . gxf6 20 ttJd5 i.e7 20 ... tUb3? loses after 21 tUxf6 'iVd8 22 �h4 �g7 23 �g5+ �h8 24 �h6 etc. 21 ttJxe7 'i'xe7 1 30

21 ...tUb3!? should be considered, e.g. 22 tUd5! rs (22 ...�d8 sees a forced draw after 23 iVh4! exd5 24 lIa8! 'ii'd 6 25 lIxffi+ �xf8 26 �xf6+) 23 �h4 (23 tUf6?! fxg4 24 tUxd7 tUxa 1 25 tUxffi tUc2 and the extra pawn be­ gins to tell) 23 ... f6 24 tUf4 with a struggle ahead in view of Black's exposed king. 22 .l:.xa5

In this maJor pIece ending White has compensation for the pawn in his active forces and the a-file, while Black also has vulnerable pawns . 22 . . . .l:.dS 23 'ifh5 c6 24 .l:.a6 .l:.c7 25 'it>g2 .l:.ccS 26 'iVh4 .l:.d7 27 .l:.da 1 f5 2S 'i'f4 fxe4 28 ...�f6 29 lIa8 l:'tcd8 30 e5 �e7 31 �h6 demonstrates White's potential. 29 .l:.aS .l:.cdS

30 'i'xe4? White can force a draw with 30 lI l a7!

4 . . . dx c 4 5 $i.. g 2 Ci'J c 6

l:txa7 3 1 l::!. xa7 't!Vxa7 32 "iff6+. 30 . . .'ii'd 6 Bringing the queen into play with advan­ tage. 31 iVf3 iVd5! 32 l:txdS+ �xdS 33 �a7 'it'gS! 34 �xf7 "xf3+ 35 �xf3 �xd4 Converting this rook endgame should be fairly automatic for a virtuoso such as Smys­ lov. 36 �f6 �e4 37 �f 1 �g7 3S l:rf3 b4 0-1

Game 64 G leizerov-Raetsky

Riazan 199 1 1 d4 d5 2 c4 e 6 3 Ci'Jf3 Ci'Jf6 4 g 3 dxc4 5 i.g2 Ci'Jc6 6 0-0 �bS 7 Ci'Jc3 i.b4 S "c2 0-0 9 �d 1

9 . . :�!Ve7 In the event of 9 ... .i.e7 10 e4 b5 1 1 �e2 l:te8 1 2 .i.e3 .i.d7 the course of Summermat­ ter-Sosonko, San Bernardino 1 99 1 is not to be recommended for White because after 1 3 ttJg5?! e 5 1 4 dxe5 ttJxe5 1 5 .i.xa7 l::!. a8 1 6 �d4 ttJd3 White had won back the pawn but the outpost on d3 gave Black the advantage. After 1 3 ttJe5!? b4 1 4 ttJa4 ttJa5 1 5 ttJc5, on the other hand, White is very active. Also possible is 9 ... b5 10 d5!? exd5 1 1 ttJxd5 h6 as in Sorokin-Kacheishvili, New York 1 999, when 1 2 ttJe5!? ttJxe5 1 3 ttJxb4 'iVe7 1 4 .i.f4 would leave White well devel­ oped and ready for action, although Black is

not without chances. 1 0 e4 Another option is lO .i.g5 h6 1 1 .i.xf6 'it'xf6 1 2 ttJe4 �f5 1 3 ttJh4 't!Vb5 1 4 l::!. a c1 with fluid development for White. 1 0 . . . $i.. x c3 Black might consider lO ... e5!? 1 1 dxe5 ttJxe5 12 ttJxe5 �xe5 1 3 .i.f4 "iile7 14 ttJb5 .i.g4 with unclear play. 1 1 bxc3 e5 l 1 ...b5 12 e5 ttJd5 1 3 ttJg5 f5 14 exf6 gxf6 1 5 ttJe4 is quite typical, Black's weakened pawns and the inviting c5-square affording White compensation. 1 2 d5! White wins the exchange. 1 2 . . . Ci'Ja5 1 3 'iVa4 Yusupov gives 1 3 ttJxe5!? �xe5 14 .i.f4 'iVe7 1 5 d6 'iVe6 1 6 dxc7 l::!. a 8 1 7 .i.d6 l::!. e 8 1 8 e 5 ttJg4 with an assessment o f unclear White has only one pawn for the piece but Black lacks co-ordination and is behind In development. 1 3 . . . b6 1 4 $i.. a 3 "eS 1 5 "c2 $i.. g 4 15 ... .i.d7 16 l:te1 c5 17 dxc6 .i.xc6 1 8 ..ItxfB 'it>xfB favours White, for whom the d­ file will be a more significant factor than Black's extra pawn on the queenside (as compensation for the exchange) . 1 5 ... ttJb 7 1 6 .txfB 'it>xfS 1 7 ttJd2 ttJd6 1 8 l::!. ab l and now Yusupov-Agdestein, Belgrade 1 989 saw 1 8 ... .id7?! 1 9 l::!. e 1 .ia4 20 'iVc1 'it'g8 21 .i.fl , when White won the pawn back with interest. The bishop is needed on a6. An improvement is 1 8 .. .'�Jd7!?, e.g. 1 9 .ifl .ta6 20 l::!.b4 tbc5 and White collects the pawn but the closed flavour of the posi­ tion leaves the situation unclear, or 1 9 "iVa4 tbc5 20 �xa7 .ib7 21 "ilVa3 l::!. a8 22 'iVb2 'ilVa4, again returning the pawn but at the cost of sufficient time to permit Black activity (the closed structure and Black's pressure on the a-file offer adequate compensation for the exchange). 1 6 $i.. xfS 'it'xfS 16 ... 'ilVxfB 17 ttJxe5 .i.xd l 1 8 l:!.xdl (White 13 1

Th e C a t a la n

invests hi s exchange to achieve a superior position - the centre, dominating bishop and superior piece placement) 1 6 ... .i.xG!? 1 7 .i.xG 'it'xffi 1 8 .i.e2 i s given by Gleizerov. Black transfers his f6-knight to d6 but, de­ spite the closed position, White's prospects are slightly preferable. 1 7 d6 c5?! More promising is 1 7 ... cxd6 18 l:!xd6 ttJb7 1 9 i::td2 .i.xG (1 9 ... ttJcS 20 ttJe1 .i.d7 21 G is an edge for White) 20 .i.xG ttJcS. Again the d-flie is open, but this time Black is quite active. The subsequent 21 .i.g2 ttJd3 22 .i.f1 ttJxe4 23 .i.xd3 ttJxd2 24 iLe2 :d8 25 i::t d 1 'iVd7 26 'it'xh7 f6 leads to complex play. 1 8 l:td2 l:td8 1 9 4:)h4 g6 20 f4! �d7 2 1 �e 1 �c6 21 ...exf4 22 gxf4 ttJhS 23 l:!.f2 'iiVe 6 24 f5 'it'xd6 25 fxg6 hxg6 26 eS and White has succeeded in opening up the position with a dangerous initiative. 22 4:)f3

27 . . . �xe4 28 fxe5 f5

29 �xf5! White crashes through! 29 . . . gxf5 30 4:)xf5+ �xf5 3 1 l:txf5 4:)c6 No better is 3 1 ...ttJe6 32 i::t f7+! 'iVxf7 33 i::tx f7+ Wxf7 34 'it'xh6 and White is winning. 32 l:txf8 1 -0

Game 65 Yevseev-Goldin

St. Petersburg 1998

22 . . . 4:)d7 22 ... exf4 23 eS ttJdS 24 'iVc1 puts h6 in White's sights and is awk-ward for Black. 23 �h3 f6 24 l:tf2
1 d4 d5 2 c4 e6 3 4:)f3 4:)f6 4 g3 dxc4 5 �g2 4:)c6 6 0-0 l:tb8 7 4:)c3 b5 8 e4 8 ttJgS ttJb4 (8 ... .i.d7 9 dS! exdS 10 iLxdS ttJeS 1 1 .i.f4 is dangerous for Black) 9 a4 a6 1 0 axbS axbS 1 1 e4 h6 1 2 eS ttJfdS 1 3 liJge4 iLe7 1 4 it'g4 g6 was seen in Ulibin­ Krasenkow, Las Palmas 1 993. W'hite's com­ pensation consists of control of key dark squares - such as cS - and better develop­ ment. 8 . . . b4 The quiet 8 ... iLe7 has proved a good al­ ternative. Then 9 dS seemed too early in Yevseev-Brodsky, Nizhnij Novgorod 1998, which continued 9 ... exdS 10 exdS liJb4 11 ttJd4 ii.g4 1 2 it'd2 ii.d7 1 3 b3 0-0 1 4 a4 a6 1 5 axbS axbS 1 6 bxc4 bxc4, when Black had doubled c-pawns but a useful outpost on d3. Nor was White in a healthy state in Ulibin­ Kogan, Cappelle la Grande 1 995: 9 'iNe2 b4 (9 ... ttJxd4 1 0 liJxd4 it'xd4 1 1 iLe3 'it'd3 12

4 . . . dx c 4 5 iL g 2 CO c 6

iVxd3 cxd3 1 3 e5 4Jd5 1 4 4Jxd5 exd5 1 5 .ixd5 and both a7 and d3 are easy to attack) 10 4Ja4 i.a6 1 1 e5 4Jd5 1 2 'iVe4 c3 and Black seized the initiative. 9 tLie2 �b6 9 ... 4Jxe4?! runs into 10 4Jd2 fS 1 1 4Jxe4 fxe4 12 i.xe4 i.b7, when Black has many weaknesses and White steers the game to his advantage after 1 3 i.e3 i.d6 14 'iWc2. 1 0 d5!? The alternative i s 1 0 e5 4Jd5 1 1 i.g5 i. e7 12 i.xe7 'it'xe7 1 3 4Jd2 i.a6. Before White gets his knight to the desired e4-square Black completes development with this active bishop move, threatening a timely ... c4-c3. Black is slightly better. 1 0 . . . exd5 1 1 exd5 'ii'x d5 The recapture with the knight is risky: 1 1 ...4Jxd5 12 4Jf4 (1 2 l:te l I?) 12 ... 4Jxf4 1 3 .ixf4 i.d6 1 4 l:te1+ i.e6 1 5 3Le3 and the initiative is in White's hands. 1 2 tLifd4 'ii'd 7 1 3 tLixc6 �xc6 14 �xd7 + The alternative is to keep the queens on the board, 1 4 i.xc6!? 'iVxc6 1 5 4Jd4 offering White a pull. 1 4 . . .'it>xd7 The king is well placed on c6, making the text preferable to 14 ... i.xd7 15 i.xc6 i.xc6 16 3Lf4 3Lc5 17 l:tac1 i.b5 18 i.e5, which is difficult for Black. 1 5 iLxc6+ Wxc6

Black has clear compensation in the form of his active king, the bishop pair and two

pawns for the exchange. Nevertheless a de­ gree of caution is called for here as White enjoys better development. 1 6 iLe3 iLc5 1 7 b3 The less radical 17 J:tac1 i.xe3 18 fxe3 3La6 19 4Jd4+ <;t>b6 is probably better, al­ though Black should be okay. 1 7 . . . tLid5 1 7 ... i.h3!? 1 8 l::r. fcl i.xe3 1 9 fxe3 c3 20 a3 as 21 axb4 axb4 seems like a good line for Black. 1 8 iLxc5 Wxc5 1 9 bxc4 tLic3 20 tLixc3 bxc3 21 �ac 1 Wxc4 22 �fe 1 �d8 23 �e7 �d7 The more ambitious 23 ... c5!? 24 l:txf7 g5 25 l:txa7 i.fS is worth a try. 24 �e8 iLb7 25 �e3 �d3 26 �e7 White has infiltrated on the seventh rank but the c3-pawn is very dangerous . 26 . . . iLd5 27 �xc7 + Wd4 28 a4 28 l:txa7? c2! and White can do nothing about ... J:Id 1 . 28 . . . a 5 29 Wf 1 iLb3 30 �d7 + Wc4 3 1 �c7 + Wd4 3 2 �d7 + Wc4 3 3 �c7 + % - %

Game 66 Raetsky-Kelecevic

Silvaplana 1997 1 d4 d 5 2 c4 e6 3 tLif3 tLif6 4 g3 dxc4 5 iLg2 tLic6 6 0-0 �b8 7 tLic3 b5 8 tLie5 tLixe5 9 dxe5

9 . . . tLid7 1 33

Th e C a t a l a n

9 ...�xd 1 ?! 1 0 .l:!.xd 1 lbd5 11 a4! sees White open the queenside and in doing so seize the initiative, while after 9 ... lbd5 1 0 lbxd5 exd5 1 1 1Lxd5 1Le6 1 2 1Lc6+ 1Ld7 1 3 �d5 White stands better. Delchev-Genov, Pazardzik 1 99 1 continued 1 3 ... 1i.xc6 1 4 �xc6+ "iNd7 1 5 "it'e4 1i.c5 1 6 .i.g5 0-0 1 7 .l:!.ad1 "iNe6 1 8 .l:!.d5 1i.e7 1 9 1Lxe7 "iNxe7 20 f4 l:tfd8 21 .l:!.fd1 and White was still in the driv­ ing seat. 1 0 'iVd4 ! ? 1 0 1Lc6 has also been played. Kaidanov­ Serper, Gausdal 1 99 1 went 1O ... a6 1 1 'ikd4 .l:!.b6?! 1 2 'ike4 1i.c5 1 3 1Lxd7+! 1Lxd7 1 4 'ilVg4. Thi s i s the usual stuff, Black winning a pawn but White generating sufficient play. Black has problems with g7. Play continued 1 4...'it>f8 1 5 .l:!.d1 h5 1 6 �f4 h4 1 7 lbe4 1Le7 1 8 1i.e3 hxg3 19 fxg3! (the f-file opens and White goes on the offensive) 19 ... .l:!.c6 20 .l:!.f1 1Le8 21 .l:!.ad1 'iVc8 22 'iVf3! 'it>g8? 23 1Lg5 'it>f8 24 1i.xe7+ and White won. Black should instead look to the follow­ ing: 1 1 ...1i.b7 1 2 1i.xb7 l:txb7 1 3 .ll d 1 c5 1 4 'ikg4 'iVc7 1 5 1Lf4 'YWc6 1 6 f3 f5 (Black has to free himself somehow, and this is a creative solution) 1 7 exf6 lbxf6 1 8 'iVg5 1Le7 1 9 "iNxg7 .l:!.g8 2 0 'iVh6 .l:!.g6 2 1 "iNh3 b4!

We are following Gleizerov-Werle, Hoo­ geveen 2000, which witnessed the rather ive 22 lbb 1 ?!, to which Black responded 22 ... .l:!.d7! 23 lbd2 c3 24 lbc4 cxb2 25 lbxb2 .l:!.xd 1+ 26 .ll x d1 c4 with advantage thanks to 1 34

the dangerous queenside pawns. S.Ivanov & Yuneev offer the following: 22 lbe4 lbxe4 23 "it'xh 7 e5! 24 1Lxe5 lbg5 25 "it'h8+ 'it>f7 26 iVh5 and the attack on Black's troubled king compensates for the sacrificed piece, or 25 ... 1i.f8 26 l:td8+ 'it>xd8 27 'ikxf8+ "iNe8 28 1i.f6+, again with pressure against the black king. 1 0 . . . �b7 Also possible is 1O ... c5 1 1 'iVf4 "iNc7 12 .l:!.d1 'iVb6 13 a4 b4 14 lbb5 'ii'd 8 1 5 lbd6+ 1i.xd6 1 6 exd6 0-0 1 7 as when White was in charge in Prakken-Ippolito, Ubeda 2000 . Even worse is 1 2 ...'iVxe5? 1 3 .l:!.xd7!, while 1 2 ... lbxe5 1 3 a4 a6 1 4 axb5 axb5 1 5 1i.c6+! �xc6 1 6 'iVxe5 is very pleasant for White. 1 1 l:td 1 The greedy 1 1 'ii'x a7?! is less convincing as 1 1 ...1Lc5 1 2 'ii'a 5 1Lxg2 1 3 'it>xg2 'iVc8 1 4 f3 iVb7 looks good for Black. White's e5-pawn is weak and Black's queenside majority seems the more useful. 1 1 . . . �xg2 1 2 'iit x g2

1 2 . . . 'iVc8 12 ... c5 is also possible but it is IOglcal to keep this square free for a piece. After the further 1 3 'iVe4 'iVc7 1 4 a4! the opening of the a-file helps only White. 1 3 a4 a6 1 4 axb5 axb5 1 5 f3 �e7 1 5 ... 1i.c5 16 �g4 g6 17 1i.h6 lbxe5 1 8 �e4 lbd7 1 9 'iVc6 and Black i s struggling against the tide . 1 6 l:ta5 b4

4 . . . dx c 4 5 i. g 2 CU c 6

Black seeks counterplay. After 1 6 ... c6 1 7 lLle4 h 6 1 8 lLld6+ i.xd6 1 9 "iVxd6 lIb 7 20 .1i.e3 his dark squares are too weak. 1 7 CUb5 CUb6 1 S 'iWg4 'it>fS Something has gone wrong for Black. 1 9 l:ta7 l:tb7 20 'iWe4 l:txa7 21 CUxa7 'iWa6 22 CUe6 'iWa4 22 ...lLld5 23 lLlxe7 �xe7 24 �g5+ f6 25 exf6+ ttJxf6 26 �xf6+ gxf6 27 'iVg4 leaves Black's king walking on thin ice. 23 l:td4 '/WaS 24 i.e3 f5? The final mistake, although the correct 24 ... h6 25 lId1 'iVa4 26 lIc1 is still good for White. 25 exf6 i.xf6 26 'iWxe6 'iWeS 27 l:tdS! 1 -0 Coming next is i.c5.

White plays 6 '/Wa4

Game 67 Tkachiev-Solozhenkin

2000 1 d4 d5 2 e4 e6 3 CUf3 CUf6 4 g3 dxe4 5 i.g2 CUe6 6 '/Wa4 i.d7 7 'iWxe4

7 . . . CUa5 7 ... i.d6 8 ttJc3 0-0 9 0-0 h6 to .i.f4 .i.xf4 1 1 gxf4 gives rise to White's typical grip on the dark squares in the centre. Consequently Black tried to ... lIb8?! in Tukmakov-Barle, Bled 1 996, but 1 1 lLle5! b5 1 2 �d3 ttJb4 1 3 'ii'd2 ttJbd5 1 4 lIac1 ttJxf4 1 5 iVxf4 nb6 1 6 ttJe4 ttJxe4 1 7 'iVxe4 lIa6 1 8 a3 lIa4 1 9 e3

i.e8 20 'iYb 7 then left White with the supe­ rior pawn structure, while control of the c­ ftle and the c5-square also remained intact. S 'iWd3 The queen is worse on c2 than d3 because of the semi-open c-file, e.g. 8 ... lIc8 9 ttJe5 c5 10 ttJxd7 'ii'xd7 1 1 dxc5 .i.xc5 12 ttJc3 ttJd5! 13 .i.xd5 exd5 1 4 0-0 0-0 1 5 '¥id3 d4 16 ttJe4 .i.e7 1 7 .i.f4 'iVd5 when Black had sufficient play to compensate the isolated d-pawn in Vaganian-Taimanov, USSR 1 974. Tratar­ Plachetka, Charleville 2000 continued 9 0-0 c5 to ttJc3 cxd4 1 1 ttJxd4 i.c5 1 2 lId 1 'iVb6 1 3 e3 0-0 1 4 ttJb3 ttJxb3 1 5 'iVxb3 'iVxb3 1 6 axb3 .i.c6 1 7 .i.d2 .i.xg2 1 8 �xg2 lIfd8 with no problems for Black throughout. S . . . e5 9 0-0 The more aggressive 9 ttJe5 was tried in Shipov-Kacheishvili, Berlin 1 995: 9 ... ttJc6 to 0-0 cxd4 1 1 i.f4 .i.c5 12 ttJd2 0-0 13 lIac1 .i.b6 1 4 ttJdc4 (14 ttJxc6 i.xc6 1 5 �xc6 bxc6 16 lIxc6 '¥idS is only equal) 14 ... ttJd5 1 5 lIfd 1 ttJxf4 16 gxf4 ttJxe5 17 ttJxe5 nb8 is instructive, White exerting some pressure on the queenside, with Black generating suf­ ficient play to maintain the balance. 9 . . . i.e6 1 0 CUe3 Another option is 10 lId1 c4 1 1 'iVc2 .i.e4 (note that this move is made possible by the omission of ttJc3) 12 'iVd2 h6! 1 3 'iVe1 ttJc6 14 'iVf1 .i.d5 1 5 ttJc3 �b4 as in Lautier­ Karpov, Ubeda 1 994, when Black could suc­ cessfully fight for e4 thanks to 1 6 ttJxd5 exd5 17 ttJe5 O-O!, with eyual chances. 1 0 . . . exd4 to ... lIc8 1 1 .l:.d 1 c4 merely gives White a free hand to get busy in the centre, e.g. 1 2 'iVc2 .i.b4 1 3 d5!? 1 3. .. exd5 1 4 e 4 .i.xc3 1 5 exd5 ttJxd5 1 6 bxc3 0-0 1 7 ttJg5 g6 1 8 ttJe4 and (again) the weakness of the dark squares proved a good investment in Ftacnik­ Lechtynsky, Czechoslovakia 1 979. Then 1 8 ... f5!? is interesting, while 1 8 ...'iVe7 1 9 .i.g5 f6 20 l:txd5! was a little combination that returned the pawn, 20 ... i.xd5 21 ttJxf6+ lIxf6 22 i.xd5+ rj;g7 23 i.xf6+ �xf6 24 1 35

Th e Ca t a la n

'it'd2 spelling trouble for Black. 1 1 ttJxd4 ii.c5 1 1 ...i.xg2 12 Wxg2 'it'd7 13 I:tdl .J1i.cs 1 4 ii.gS favours White. 1 2 .l:!.d 1 ii.xg2 1 3 'it'b5 + ! ? ttJd7 1 4 c,t>xg2 a6 14 ... 0-0?? IS lDf3 loses the knight. 1 5 'ifd3 ii.e7

1 6 ii.f4! .l:!.c8 The tempting 16 ... eS? loses material due to 17 lDf5 exf4 18 lDxg7+ WfS 19 'iiVx d7. 1 7 ttJf3 ttJc4 1 8 ttJa4 b5 1 9 b3! ttJcb6 20 ttJxb6 ttJxb6 21 'ifxd8+ ii.xd8?! 21 ...lIxd8 22 l:txd8+ .J1i.xd8 23 .l:!.c1 is the lesser evil. 22 e4 The b6-knight is deprived of the last good square. White has a clear advantage. 22 . . .16 22 ... lDd7 23 l:tac1 l:txc1 24 Ihc1 lDf6 2S eS lDdS 26 :c8 and Black is trailing. After the text the queenside pawns come under attack. 23 .l:!.d6 c,t>e7 24 ttJd4 e5 25 ttJf5 + c,t>f7 26 ii.e3 g6 27 ttJh6 + c,t>g7 28 .l:!.ad 1 Suddenly White is very active. 28 . . . .l:!.b8 29 .l:!.d7 + ! A nice exchange sacrifice. 29 . . . ttJxd7 30 .l:!.xd 7 + c,t>f8 31 ttJf7 .l:!.g8 32 ttJd6! Mate cannot be avoided without heavy loss of material. 32 . . . g5 33 .l:!.f7 mate! 1 36

Game 68 Raetsky-S . Ivanov

Simferopo! 1989 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 ii.g2 ttJc6 6 'it'a4 ttJd7 6 ... lDdS 7 'iiVxc4 lDb6 transposes to 6 ... lDd7 7 'iic4 lDb6. 7 'it'x c4 ttJb6

8 'ifd3 e5 Another idea is 8 ... lDb4, e.g. 9 'iiVb 3 eS 1 0 lDgs "iIIe7 1 1 a 3 lDc6 1 2 .J1i.xc6+ bxc6 1 3 dxeS h 6 1 4 lDf3 .J1i.h3 with unclear play ac­ cording to Shipov, or 9 'iiVc 3 eS 1 0 dxeS i.f5 1 1 0-0 ii.xb 1 1 2 ltxb 1 lDxa2 1 3 'iiVb 3 lDxc1 1 4 l::tb xc 1 , Shipov-Gofstein, Paris 1 995, when White had a lead in development. 9 ii.e3! ? A home-made idea from one of your au­ thors! In Gasimov-Zvjagintsev, Yurmala 1 992 a draw was agreed in the equal position that resulted from 9 dxeS "iIIx d3 1 0 exd3 lDb4 1 1 'it>d2 ii.f5 1 2 a3 lDxd3 1 3 lDh4 lDxc1 1 4 Wxc1 .J1i.xb 1 I S l:txb 1 . Meanwhile, in Poluljakhov-Brodsky, Krasnodar 1 999 the continuation 9 ii.gS ii.e 7 1 0 ii.xe 7 'iVxe7 1 1 lDxeS lDxeS 1 2 dxeS 'iiVx eS 1 3 lDc3 0-0 14 0-0 c6 I S l::tfd l .J1i.e6 16 'iVd4 'iIIc 7 saw Black entering the rniddlegame phase with no problems. 9 . . . ttJb4 9 ... exd4 10 lDxd4 lDb4?! 1 1 "iIIe4+ i.e7 12

4 . . . dx c 4 5 iL g 2 l"iJ c 6

lDc3 ttJc4 1 3 l:td1 !? ttJd6 (13. .. ttJxb2 1 4 ttJc6 lDxd1 1 5 ttJxd8 ttJxc3 16 'iVe5 clearly favours White) 1 4 'iWb1 0-0 1 5 0-0 'iVe8 1 6 .i.f4 .i.d7 17 a3 ttJc6 1 8 ttJdb5 and, after some weird manoeuvring, Black was drifting into trouble in Filippov-Short, Batumi 1 999. In fact after 18 ...l:k8 1 9 ttJxc7! .l::!.x c7 20 ttJd5 .i.f5? 21 lDxc7 White won. A lesser evil is 1 O ...ttJxd4 1 1 .Jtxd4 .Jtb4+ 12 ttJc3 0-0 1 3 a3 .Jte7 1 4 �d1 c 6 1 5 'iVe4!, although Black has prob­ lems with the co-ordination of his pieces. 10 'iWb5 +

1 0 . . . l"iJd7 1O ... .Jtd7!? is a clever move with which Black sacrifices a pawn in order to seek to exploit the insecure position of White's queen. In J.Horvath-Rabiega, Austria 1 996 this did not work out for Black after 1 1 �xe5+ .Jte7 1 2 ttJa3 0-0 1 3 0-0 ttJ6d5 1 4 .i.d2 .u.e8 1 5 .i.xb4 ttJxb4 1 6 'iVf4 .Jte6 1 7 ttJb5! ttJd5 1 8 'iVd2 .Jtb4 1 9 'iVc2 c 6 20 ttJc3. However, there is an improvement in 1 4 ... .Jtf6 1 5 'iVh5 g6 16 'ikh6 .Jtg7 17 'ikh4 'ikxh4 1 8 gxh4 nfe8, when White's kingside structure is compromised and Black has some compensation for the pawn deficit. 1 1 'iWa4 exd4 1 1 ...b5?! 12 'iVb3 .Jtb7 13 ttJxe5! is bad for Black, and after 1 1 ...e4 1 2 ttJg5 ttJd5 1 3 'iVb3 White has numerous threats (e4 is weak and Black is behind in development) . 1 2 l"iJxd4 cS 1 3 l"iJc3 l"iJbS 1 4 'iWb3 c5 Perhaps this is too ambitious as it helps

White's Catalan bishop. Also possible is 14 ...ttJ4d5 15 ttJxd5 ttJxd5 16 .Jtxd5 'iWxd5 17 'iWxd5 cxd5 1 8 ttJb5 with an endgame that is generally slightly better for White and here also features a lead in development. 1 5 l"iJdb5 iLeS 1 5 ... .Jtf5 16 l:td 1 ttJc2+ 17 'it>f1 ttJxe3+ 1 8 fxe3 'ikc8 1 9 ttJd5 and in return for his dam­ aged pawns White has a strong attack. 1 S 'iWd 1 iLf5?! A waste of time, but the alternative 1 6 ... 'ikxd1+ 1 7 .u.xd1 ttJc2+ 1 8 'it>f1 ttJxe3+ 1 9 fxe3 is also good for White. Again the doubled e-pawns are balanced by White's more rapid deployment. 1 7 0-0 White is better, with pressure against Black's queens ide. 1 7 . . . aS 1 8 'iWxd8+ .l:lxd8 1 9 l"iJc7 + �d7 20 l"iJ7d5 l"iJ4xd5 21 l"iJxd5 l"iJc4 22 .l:!.fd 1 l"iJxb2? Losing immediately. After 22 ... .i.d6 23 b3 ttJxe3 24 ttJxe3 Black loses at least a pawn as b 7 is hanging. 23 .l:!.d2 l"iJc4 24 l"iJbS + 1 -0

Game 69 Polovodin -Zviagintsev

St. Petersburg 1994 1 d4 d5 2 c4 eS 3 l"iJf3 l"iJfS 4 g3 dxc4 5 iLg2 l"iJcS S �a4 l"iJd7 7 �xc4 l"iJbS 8 �d3 e5 9 l"iJxe5

137

Th e C a t a l a n

9 . . . ttJb4 After 9 ... ttJxd4 10 e3 ttJe6 1 1 'iVxd8+ ttJxd8 the game is equal, while 1 0 .i.e3 .i.c5 1 1 ttJc3 .i.f5 12 .i.e4 .i.xe4 13 'iWxe4 f5 1 4 'it'd3 'iWd6 i s unclear. White's best i s 1 0 0-0 f6 1 1 ttJf3 ttJxf3+ 1 2 �xf3 .i.d6 1 3 ttJc3 with a slight edge. 1 0 �b3 In the event of 1 0 �dl 'iWxd4 1 1 'it'xd4 ttJc2+ 1 2 'it'd 1 ttJxd4 1 3 e3 ttJe6 Black shouldn't experience any problems. Gelfand­ Ivanchuk, Sochi 1 986 continued 14 ttJc3 f6 1 5 ttJd3 ttJc5 1 6 ttJxc5 .i.xc5 1 7 'it'c2 .i.f5+ 1 8 e4 .i.e6 1 9 f4 0-0-0 20 b3 h5 with equal chances. 10 'iWc3 'it'xd4 needs investigating. After 1 1 'i*'xc7!? .i.e7 1 2 ttJf3 �c4 1 3 'i*'xc4 ttJxc4 14 ttJd4 .i.f6 1 5 a3 .i.xd4!? 1 6 axb4 ttJxb2 1 7 .l::ta2 Black wins the pawn back but White keeps an edge, while Haba-Klovans, Leinfel­ den 2001 saw instead 1 5 ... ttJxa3 1 6 ttJxa3 .i.xd4 17 0-0 0-0 1 8 :d 1 .i.f6 1 9 ttJb5. Again Black managed to collect the invested pawn, but White's pressure on the queenside se­ cured an advantage. The earlier encounter Beliavsky-Adams, Madrid 1 998 went 1 1 0-0 'iVxc3 12 ttJxc3 .i.d6 1 3 ttJf3 c6 14 lIdl .i.e7 1 5 ttJd4 0-0 16 a3 ttJ4d5 17 ttJxd5 ttJxd5 with a level game, but 1 5 e4!? is interesting and seems to give White some advantage after 1 5 ... 0-0 1 6 a3 ttJa6 1 7 .i.f4. 1 0 . . . �e6 1 1 �d 1 �xd4 1 2 �xd4 ttJc2 + 1 3 'it'd 1 ttJxd4 1 4 �xb7 .l:!.d8

1 5 �d2 In Kaidanov-Shariyazdanov, Elista 1 998 White preferred 1 5 ttJd2, and after the fur­ ther 1 5 ... f6 1 6 ttJc6 ttJxc6 1 7 .i.xc6+ �f7 1 8 'it'c2 .i.c5 1 9 ttJb3 .i.xf2 2 0 : f1 .i.d4 21 .tf4 Black returned the material with equality after 21 ....i.xb3+ 22 �xb3 ttJd5. 1 5 . . . f6 1 6 ttJc6 ttJxc6 1 7 �xc6 + 'it'f7 1 8 'it'c2 �c5 Black's activity compensates for the pawn deficit. 1 9 f3 ttJc4 20 �e4 .l:!.d6 21 �d3 .l:!.b8 Now all of Black's pieces are in play. 22 �c 1 22 b3 ttJe5 23 .i.e4 f5 24 .i.c3 fxe4 25 .i.xe5 l:td5 26 f4 e3 is to Black's benefit. White has problems mobilising his forces and his light-squared bishop has a bright future. 22 . . . �d4 23 ttJc3 .l:!.c6 24 �e4 .l:!.c5 25 .l:!.d 1 Another move is 25 e3 but Black has a strong reply - 25 ... ttJa3+! 26 'it'd2 .i.xc3+ 27 bxc3 l::t d8+ 28 'it'e2 ttJb5 and Black both regains the pawn and maintains the offen­ sive. 2S . . ..bs 26 f4 ttJxb2!

A nice little combination that wins back the pawn. 27 �xb2 .l:!.xb2+ 28 'it'xb2 �xc3 + 29 'it'b 1 �c4 Of course Black has no intention of re­ storing the material balance immediately as 1 38

4 . . . dx c 4 5 i. g 2 0, c 6

his bishop is far better than the rook. 30 a4 f5 3 1 l:te 1 ? After the more circumspect 3 1 .i.f3 .i.xal 32 '.t>xal .i.b3 Black wins a pawn but White's forces then enjoy sufficient freedom with which to secure equal chances. 3 1 . . . i.xa 1 32 ':'xe4 l1xe4 33 i.d5 + �f6 34 i.xe4 i.d4 This opposite coloured bishop ending is far from a draw. Had White's king been on the kingside the draw would have been obvi­ ous. Now these pawns make easy targets for the enemy bishop. 35 �e2 i.g 1 36 h3 i.h2 37 �d3 i.xg3 38 e3 h6 39 �e2 g5 40 �f3 i.e 1 41 �e2 i.b4 Black already has one ed pawn but to win the game he has to create a second. 42 �f3 i.d2 43 i.a6?! Better is 43 fxg5+ hxg5 when Black has to work to make progress on the kingside. 43 . . . g4+ !

Forcing an exchange that results in a sec­ ond, decisive ed pawn. 44 hxg4 fxg4+ 45 �xg4 i.xe3 46 �f3 i.e 1 47 �e4 �e6 48 i.e2 a5 49 i.h5 �d6 50 i.d 1 �e5 51 �e5 �b4 52 f5 White also has a ed pawn that needs attention, but the win is uncomplicated. If necessary Black will sacrifice his bishop for the f-pawn so that his two runners cannot be held at bay by White's lone bishop. 52 . . . e5 53 f6 i.b2+ 54 �f5 e4 55 f7

55 'it>g6 i.xf6 56 'it>xf6 'it>c3 57 .lte2 'it>d4 doesn't help White. 55 . . . i.g7 56 �e4 �e3 57 i.g4 �d2 0-1

Game 70 Zsu . Polgar-Benjamin

Dortmund 1985 1 d4 d5 2 e4 e6 3 0,f3 0,f6 4 g3 dxe4 5 i.g2 0,e6 6 'i'a4 i.d6

7 0-0 In the event of 7 ttJbd2 (note that with the moves ... i.b4+ and i.d2 inserted this is not possible) 7 ... 0-0 8 ttJxc4 Black must not allow the exchange of his potentially useful dark­ squared bishop for White's knight. Therefore in this position practice has seen the forcing 8 ... b5!? 9 "iVxb5 �b8 1 0 'iVxc6 i.d7 1 1 'iVa6 i.b5 1 2 "iVa5 .i.xc4 1 3 a3. Then 1 3. .. ttJe4!? is worth a try, with compensation after 14 ttJe5 .i.d5 1 5 0-0 c5. Instead Jo.Horvath-Van Der Werf, Haarlem 1 995 continued 1 3 ... c5 1 4 'iVxd8 �fxd8 1 5 dxc5 .ltxc5 1 6 b 4 .tb6 1 7 .i.b2, when 1 7 . . .ttJe4 1 8 e 3 ttJc5! 1 9 ttJe5 ttJd3+ 20 ttJxd3 .i.xd3 21 �c1 as gives Black slightly less than a pawn's worth of compen­ sation. 7 . . . 0-0 8 l:td 1 I f White is not prepared to forget about the c4-pawn he should play 8 'iixc4 e5 9 ttJc3 exd4 1 0 ttJxd4 ttJe5 1 1 'iVa4 .td7 (1 1 ...c6!? is interesting) 1 2 �b3 c5 1 3 ttJdb5 c4 14 'iVc2 .i.c5 1 5 .i.f4 'iVe7 1 6 �ad l , when 1 39

Th e C a t a l a n

White had a n edge i n Gleizerov-Dzhakaev, Hoogeveen 2000. The ambitious thrust of the c-pawn seemed to have helped White more than Black. 8 . . . tDb4 An interesting approach to the position is 8 ... ir'e8?! 9 'iVxc4 e5 10 d5 li:'lb4, when the queen monitors both the e-file and the a4-e8 diagonal. Filippov-Loeffler, Ubeda 1 998 con­ tinued 1 1 li:'lc3 li:'lc2 12 l:tb1 e4 1 3 li:'ld4 li:'lxd4 14 llxd4 i.f5 1 5 CDxe4 b5 1 6 �d3 i.xe4 1 7 i.xe4 i.c5 1 8 i.f3 i.xd4 1 9 'i*'xd4 and White enjoyed a greater share of territory and decent attacking chances in return for the exchange. 9 tDe5 c6 Another option is 9 ... li:'lfd5!? 10 a3 li:'lb6 1 1 'iVa5 li:'lc2 12 .l:!.a2 f6 and the position remains unclear. 1 0 a3

viding compensation for the missing pawn. Finally, after 1 1 ...�6 1 2 ir'a5 li:'lc2 1 3 l::ta2 li:'ld5 1 4 'i'xb6 axb6, we have another com­ plex battle to look forward to. 1 1 tDc3 "ikc7 Black can try 1 1 ...li:'ld5 1 2 li:'le4 b5 13 ir'c2 but the resulting dark square weaknesses give White something to aim at. You might have noticed that after 1 2 ir'xc4?? li:'lxc3 1 3 bxc3 i.xe5 Black is winning. 1 2 "ikxc4 .i.xe5 1 3 dxe5 "ikxe5 1 4 .i.f4 "ikh5 1 5 .i.d6 The dark squares again give White enough play. 1 5 . . J�e8 1 6 .i.f3 "ikf5 1 7 .i.xc6?! A more quiet approach with 1 7 b4!? should be preferred. 1 7 . . . bxc6 1 8 "ikxc6 �b8 1 8 ... i.d7 1 9 'iVxa6 li:'le4 20 li:'lxe4 'iVxe4 21 'ifd3 slightly favours White. 1 9 .i.xb8 tDxb8 20 "ikb5 "ikh3 After 20 ... 'iVxb5 21 li:'lxb5 a6 22 li:'ld6 l:tfB the exchange of queens helps White's cause, whereas the text at least draws attention to White's king. 21 tDe4 21 'i*'xb8?? li:'lg4 and mate is on the menu. 2 1 . . . tDbd7 22 tDxf6 + tDxf6 23 f3 e5 24 e4 "ike6 24 ... h5 25 l:td6 h4 26 ir'fl sees White ex­ change queens to leave Black with only a small edge. 25 J:tac 1 .i.d7 26 "ikb4 "ika6 27 Wf2 "ike6

1 0 . . . tDa6 Black can also head directly for the end­ game with 1O ... b5!? 1 1 li:'lxc6 bxa4 12 li:'lxd8 li:'lc2 1 3 .l:!.a2 li:'ld5 14 li:'lc6. White's rook looks a bit silly at the moment but he has made progress in the centre and Black's queenside has seen better days. Not surpris­ ingly this is a difficult position to assess. We should not forget that Black has two other possibilities after 1 1 li:'lxc6, the most obvious being 1 1 ...li:'lxc6 1 2 'ii'x b5 i.d7 1 3 i.xc6 ttb8 14 'iVa4 'iVc7 1 5 i.xd7 li:'lxd7 with a lead in development and pressure on the b-ftle pro-

Yo - Yo

1 40

Game 71 Bauer-I ppolito

New York 2000 1 d4 d5 2 c4 e6 3 tDf3 tDf6 4 g3 dxc4 5 .i.g2 tDc6 6 "ika4 .i.b4+ 7 .i.d2 .i.d6 8 tDe5 White has a range of possibilities here. In the case of 8 li:'la3 the most direct response is 8 ... i.xa3, when after 9 "it'xa3 li:'lxd4 10 li:'lxd4 'iVxd4 1 1 lld1 White has sacrificed two

4 . . . dx c 4 5 i. g 2 l"jj c 6

pawns but the lead in development and pres­ sure against Black's queenside make life rather awkward for the defender. An unclear position arises after 8 ... a6 9 ltJxc4 b5 1 0 tLlxd6+ cxd6 1 1 'iVa3 .i.b7, which leaves 8 ... ltJe4 9 ltJxc4 ltJxd2, Black making use of the 'extra' .i.d2 to remove a potentially influ­ ential bishop. After 10 ltJfxd2 .i.d7 Beliavsky gives 1 1 e3!? 0-0 1 2 'itb3 as an interesting choice, while in Beliavsky-Anand, Dortmund 1 998 the continuation 1 1 ltJxd6+ cxd6 1 2 ttJe4 0-0 saw Black sacrifice a pawn for a more fluid deployment of his forces: 1 3 tLlxd6 'it'c7 1 4 ltJe4 l:tfd8 1 5 'it'c4 'it'b6! 1 6 e3 !lac8 17 'iVb3 'ii'a5+ 1 8 ltJc3 e5! 19 d5 ttJd4! 20 exd4 exd4 21 0-0 dxc3 22 bxc3 .l:!.xc3 23 'i'xb7 'it'c5! with a balanced game. Another option is 8 .i.c3 0-0 9 'it'xc4 with a space advantage for White. This was trans­ lated to a general lead in Wojtkiewicz-Goldin, Oak Bridge 2000, which went 9 ... 'ii'e 7 1 0 ttJe5 .i.xe5 1 1 dxe5 ttJd5 1 2 0-0 .l:!.d8 1 3 l:t d 1 i.d7 1 4 ltJd2 ttJxc3 1 5 'fixc3 .i.e8 1 6 ltJb3 .l:!.xd 1+ 17 l:!xd 1 J:td8 1 8 nc 1 . O f course White can capture on c4 imme­ diately. For example D.Gurevich-Beliavsky, Las Vegas 1 999 went 8 'it'xc4 e5 9 dxe5 (9 d5!? is interesting) 9 ...ltJxe5 10 ltJxe5 .i.xe5 1 1 'it'b5+ (the point of the exchange on e5, but Black is still okay) l 1 ...ltJd7 12 .i.c3 O-O! (a temporary piece sacrifice) 1 3 0-0 (1 3 i.xe5 c6!) 1 3 . .. c6 14 'it'c4 l:te8 1 5 a4 ltJb6 16 'it'c5 i.g4! and Black was doing fine. Finally there is 8 0-0 0-0 9 'fixc4, when 9 ... e5 presents White with a decision regard­ ing the d-pawn. 1 0 dxe5 ttJxe5 1 1 ttJxe5 i.xe5 1 2 .i.c3 doesn't seem to pose Black any problems, e.g. 1 2 ... 'it'e7 1 3 .i.xe5 'it'xe5 14 ttJc3 c6 1 5 .l:!.fd 1 .i.e6 1 6 'iVb4 .l:!.ab8 1 7 .l:!.d2 a s 1 8 'iVb6 ltJd7 1 9 'it'd4 'it'xd4 20 l:txd4 J::t fd8 21 l:tadl �f8 with a draw in Karpov-Kramnik, Wijk aan Zee 1 998. In­ stead in Raetsky-Shariyazdanov, Biel 1 999 White pushed with 10 d5, which earns a terri­ torial advantage after 1 O ...ltJe 7 1 1 ttJc3 a6 1 2 i.g5 ttJe8 1 3 e4.

8 . i.xe5 9 i.xc6 + 9 dxe5 ltJd5 1 0 f4 ltJb6 1 1 .i.xc6+ bxc6 1 2 'it'xc6+ 'it'd7 i s unclear because, while Black's pawns look fragile they are not easy to attack and, in the meantime, he has the b-fue. 9 . . bxc6 1 0 dxe5 'ii'd 5 Also possible is 1 O ...ltJd7 1 1 .i.c3 ltJb6 1 2 'fixc6+ 'ii'd 7 1 3 'ii'xd7+ �xd7 1 4 ltJa3 .i.a6 1 5 0-0-0+ �c6 1 6 lld4 (Ilincic) . Compared with the previous note featuring 9 dxe5, White has clearly improved and his forces are much better placed. 1 1 f3 l"jjd 7 1 1 ...'it'xe5? 12 'iWxc6+ �d8 13 'it'xa8 'ii'xb2 1 4 0-0 'iix al 1 5 'it'xa7 nets a pawn for Black but he will soon feel the power of his opponent's pieces. 1 2 l"jjc 3 'ii'c 5 12 ...'it'xe5 13 'it'xc6 l%.b8 1 4 .i.f4 'it'd4 was tested in Ilincic-Marjanovic, Yugoslavia 1 998. After 15 'ii'x c7 0-0 16 l:i.dl 'it'b6 17 'it'xb6 1:txb6 1 8 .i.d6! :e8 1 9 .i.a3 f5 20 l:td2 ltJf6 21 �f2 e5 22 l::t h dl White's control of the d­ fue and more sound structure secured the better prospects. Perhaps Black should look to 14 ... 'iVf5!? 1 5 .i.xc7 .l:txb2 1 6 .i.d6 .i.b7 1 7 'iVxc4 'iVa5! with good counterplay. I n return for leaving his king in the centre Black can generate pressure of his own. 1 3 f4 . .

.

1 3 . . . 0-0 The situation after 1 3 ... ltJb6!? 14 'it'c2 .i.b7 1 5 0-0-0 0-0-0 16 ltJe4 'it'e7 17 .i.e3 14 1

Th e C a t a l a n

.l:Id5 is given a s unclear b y Bauer. 1 4 0-0-0 .l:!.bS 1 5 ttJe4 'iWb5 1 6 'ii'c 2 c3?! After 16 ... .l1.b 7 17 ttJg5 g6 18 .l1.c3 Black keeps the c4-pawn but the question is how much is it worth, as without the pawn Black has more room for manoeuvre - hence the text. 1 7 i.xc3 ttJb6 1 S ttJg5 g6 1 9 h4 ttJd5 20 h5! ! The direct approach. Time is more impor­ tant than material. 20 . . . ttJe3 21 't\Ve4 ttJxd 1 22 .l:!.xd 1 Wg7 Other moves are clearly worse, e.g. 22 ... h6 23 hxg6! hxg5 24 .l:Ih 1 ! with the cruel threat of .l:Ih8+!, or 22 ... c5 23 "iYh1 !, heading for h7. 23 g4 Here 'W'hite misses a great opportunity to continue in gambit style with 23 f5!! gxfS (23 ... exfS 24 e6+ f6 25 'iVh4 wins - h 7 is the problem) 24 h6+ 'it>g8 25 'iVh4 and 'W'hite is winning (it is safe to say that Black is weak on the dark squares!). 23 . . . i.a6 23 ... h6 24 ttJxf7! 'it>xf7 25 'it'xg6+ 'it>e7 26 'iVxh6 and 'W'hite's attack looks dangerous. 24 .l:!.d7 'iWc4 25 h6+ WgS 26 f5!

26 . . . .l:!.bdS Even the exchange of queens doesn't help Black: 26 ... 'ii'x e4 27 ttJxe4 gxf5 28 gxfS �c8 29 .l:Ixc7 .t!.b7 30 .l:Ixc6 with a huge advantage to 'W'hite. 27 fxe6 .l:!.xd7 'W'hite's win is practically forced. 142

2S exd7 .l:!.dS 29 e6! 'ii'xe4 30 exf7 + Wf8 3 1 i.g7 +

3 1 . . . We7 32 fS'iV + .l:!.xfS 33 i.xfS + Wxf8 34 dS'iY + 1 -0 Mate is unavoidable.

Game 72 G rabliauskas-Benjamin

New York 2000 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 i.g2 ttJc6 6 'iYa4 i.b4+ 7 i.d2 ttJd5 8 'iYb5 0-0 8 ... �xd2+ 9 ttJbxd2 c3 10 bxc3 ttJxc3 1 1 'iVd3 ttJd5 nets Black a pawn but 'W'hite has compensation, e.g. 1 2 0-0 0-0 1 3 e4 ttJf6!? 14 .l:Ifd1 'iVe7 15 .l:Iac1 nd8 16 ttJf1 �d7 17 ttJe3 .i.e8 1 8 a3 h6 19 d5 .l:Iac8 20 'iVb 1 ttJa5 21 .l:Ic3 Iljushin-Aleksandrov, Smolensk, 2000. In Van Heste-Barsov, Netherlands 1 998 Black tried 1 3 ... ttJb6 with some patient manoeuvring after 14 ttJb3 'ii'e 7 1 5 .l:Iac1 �d7 1 6 ttJc5 ttJd8 1 7 .l:Ife 1 �e8 1 8 h4 h6 1 9 'iVb3 ttJc8 2 0 ttJe5 c 6 21 ttJc4 b 6 2 2 ttJd3 ttJb7 23 e5, although 'W'hite emerged with a pull. Well worth considering is the retention of dark-squared bishops with 8 ... .l1.e7!?, when Paunovic-Zvjagintsev, Yugoslavia 1 995 went 9 'ii'xc4 ttJb6 1 0 'iVd3 0-0 1 1 0-0 e5 1 2 .l1.c3 exd4 1 3 ttJxd4 ttJxd4 14 �xd4 �fS 1 5 'iVxfS 'iixd4 1 6 ttJc3 c6, Black succeeding in neu­ tralising 'W'hite's aggression. Instead of 12

4 . . . dx c 4 5 iL g 2 l2l c 6

i.c3 White can try 1 2 ttJxe5, e.g. 1 2 ... ttJxe5 (1 2 ... ttJxd4 13 .i.c3 is an edge thanks to White's modest development lead) 1 3 dxe5 'i'xd3 14 exd3 J:td8, when the attack down the d-ftle provides Black with compensation. 9 'WiVxc4 Also possible is 9 .i.c3, when the forcing 9 ... .i.xc3+!? 1 0 ttJxc3 ttJxc3 1 1 bxc3 �d5 is okay for Black. An alternative to the bishop trade is 9 ... e5!?, e.g. 10 dxe5 i.e6 1 1 0-0 a6 12 'ilt'a4 jus 1 3 ttJh4 .i.xb1 (13. .. b5!? might be better, when 1 4 'it'd 1 ..lte6 1 5 a4 l:tb8 is unclear) 1 4 l:taxb1 .i.xc3 1 5 'ii'x c4 ..ltxe5 1 6 i.xd5 with the easier game for White in Ro­ manishin-Brodsky, Nikolaev 1 995. Note that 10 O-O?! is not in White's interest in view of 1O ... a6 1 1 'it'xc4 e4 1 2 .i.xb4 (directed against ...ttJa5) 12 ... exf3 13 .i.xfS fxg2 14 �xg2 ..lte6! and the resulting rook versus rwo minor pieces middlegame clearly favours Black. 9 . . . l2lb6 1 0 'WiVd3 e5 1O ... .i.xd2+ 1 1 ttJbxd2 e5 12 ttJxe5 ttJb4 13 'it'b 1 'it'xd4 14 ttJdf3 'iVd8 1 5 a3 ttJ4d5 1 6 0-0 'it'e7 1 7 �c2 a s was fine for Black in De Boer-S.Polgar, Netherlands 1 996. 1 1 iLxb4 1 1 ttJxe5 i.xd2+ 12 ttJxd2 ttJb4 13 �c3 'iVxd4 1 4 0-0 'l'xc3 1 5 bxc3 ttJ4d5 is pleasant for Black. 1 1 . . . l2lxb4 1 2 'ifd2 a5 1 3 a3 e4! ? 1 3. .. ttJc6 1 4 dxe5 'ilt'e7 1 5 'YWf4 a 4 1 6 0-0 l:ta5 17 ttJc3 ttJxe5 with a draw in Konopka­ Grabliauskas, Litomysl 1 996. The text is more ambitious. 14 axb4 l2lc4! 14 ... exf3? 1 5 ..ltxf3 ttJc4 16 �d3 ttJxb2 1 7 �c3 and the knight i s lost. 1 5 'ifc 1 ? ! 1 5 'iVf4 has more bite. Benjamin gives 1 5 ... exf3 1 6 ..ltxf3 'iVe7 1 7 ttJc3 (1 7 bxa5 'ifb4+ 1 8 ttJc3 'i¥xb2 19 'it'c1 'iYxc1+ 20 l:txc1 c6 leads to equality) 1 7...'iVxb4 1 8 0-0 'iVxb2 19 ttJd5 ttJd2 20 lIfd1 ttJxf3+ 21 'it'xf3 c6 22 ttJe7+ 'i£th8 23 .l:.ab1 'iVc2 24 e3. White wins back the b 7 -pawn with an equal game. 1 5 . . . exf3 1 6 iLxf3 'WiVxd4 1 7 l2ld2 l2le5

1 8 bxa5 iLh3!

Black's is a sensible pawn sacrifice. White's king is kept in the centre for the time being. 1 9 'WiVc3 In the event of the greedy 19 ..ltxb 7 Black generates a formidable attack after 19 .. J::ta e8 20 .i.e4 ttJg4. 1 9 . . . l2lxf3 + 20 l2lxf3 'ife4 21 'ifd4 'ifc6 22 'ifc3 'ifa6 No exchange of queens. By now White really misses his Catalan bishop! 23 l2ld4 l:!.fe8 24 f3?

White tries to repair the light squares, but he cannot afford the time. 24 b4 ':ad8 25 e3 �f6 is necessary. 24 . . . c5! 25 'ifxc5 iLg2 26 'it'f2 The point behind Black's play is seen in the line 26 l:tg1 i.xf3 27 'i£tf2 ':xe2+ 28 'it'xf3 �d3+ 29 'it>g4 ':xh2, when White's 1 43

Th e Ca t a l a n

king is in dire straits. 2S . . . i.xh 1 27 �xh 1 �ec8 28 'ii'g 5 hS 29 'ii'g 4 'ii'x a5 The knight and pawn are no match for the rook in this wide open position. 30 tDf5 'ilfbS + 31 Wg2 'ii'f S 32 Wh3? 32 ttJd4!? puts up more resistance. 32 . . . �c2 33 �d 1 �xe2 34 �dS �eS 35 �d7 b5?! A quicker win results from 35 ... h5 36 'iVxh5 g6 37 ttJh6+ �g7. 3S f4 b4 37 �b7 'ilfxb2 38 �xb4 'ii'fS With or without rooks and queens Black will win the game. 39 �b7 �d8 40 �b5 �d2 41 �b8 + Wh7 42 tDh4 �ee2 43 tDf3 �b2 44 �f8 �f2 45 f5 'ii'e 7 4S �a8 'ii'e 2 0-1

Game 73

Dorfman, Cannes 1 990. The rook looks rather silly on a2. 1 1 . . . tDxa 1 An interesting line is 1 1 ...�d7!? 1 2 'it>xc2 ttJxd4+ 1 3 ttJxd4 �xb5 1 4 ttJxb5 J:!.b8, e.g. 1 5 �c6+ (1 5 ttJ l c3 0-0 1 6 l:tadl 'iVe7!? 17 ttJxa7 'iVc5 18 ttJc6 l:tb6 1 9 ttJa4 'iVf5+ 20 e4 'iNb5 with a highly complex struggle ahead) 1 5 ... �f8 1 6 ttJ l c3 'iNe7 1 7 .l:.hdl 'iNc5 (Chet­ verik-Grabliauskas, Martin 1 996) 1 8 .ltf3 Ihb5 (the less forcing 1 8 ... a6 1 9 ttJd4 'ib6 20 b4 cxb3+ 21 �b2 also merits attention) 19 .l:.d8+ 'i;e7 20 .l:.xh8 .l:tb3. If judged only in material , then White has more than enough for the queen. But Black has an ac­ tive position and White's king could be more secure. 1 2 'it'xcS + i.d7 1 3 'ilfxc4

Ivanch u k - Korchnoi

TilbufJ!, 1989 1 d4 d5 2 c4 eS 3 tDf3 tDfS 4 g3 dxc4 5 i.g2 tDcs S 'ii'a 4 i.b4+ 7 i.d2 tDd5 8 i.xb4 tDxb4 9 a3

9 . . . b5! Simagin. 1 0 'ilfxb5 tDc2+ 1 1 Wd2 This is better than 1 1 �f1 �d7 12 .l:.a2 ttJ2xd4 1 3 ttJxd4 ttJxd4 14 'it'c5 ttJb3 1 5 'iVxc4 l:tb8 1 6 � f3 0-0 1 7 �g2 e 5 1 8 ttJc3 c6 19 .l:.dl 'iNe7 20 �g4 �xg4 21 'iNxg4 nfd8 with an advantage to Black in Bogdanovsky1 44

1 3 . . . c5!? The main alternative is 1 3 ... .l:.b8, e.g. 1 4 b4 c5! 1 5 ttJc3?! cxd4 1 6 ttJe4 .ltb5 1 7 'iVa2 d3 with considerable counterplay for Black in Zilberstein-Raetsky, Voronezh 1 988. After 1 8 l:txa 1 .l:.c8! 1 9 ttJc5 I1xc5! 20 bxc5 'iNa5+ 21 �e3 'iVc3 22 exd3 'iVxd3+ 23 �f4 g5+ 24 ttJxg5 'iVf5+ Black's attack is indeed very dangerous. White should go for the knight immedi­ ately, Kozlov-Chernikov, Vladivostok 1 978 being a good example of model play after 1 5 'iNc3! cxb4 1 6 axb4 as 1 7 'iVxal axb4 1 8 ttJe5 0-0 1 9 .l:.dl ! b3 20 �e1 b2 21 'it'a7 .ltb5 22 ttJc3 'iVd6 23 'iVa5 nfd8 24 �e4 (24 ttJxb5?

4 . . . dx c 4 5 iL g 2 ltJ c 6

'i"xe5! is a nice little trick that illustrates the power of the b2-pawn) 24... �e8. White has two minor pieces for the rook, a fIne centre and his king is about to feel safer on the kingside, but the b2-pawn is dangerous. 1 4 it'a2 'ili'a5 + ! 1 5 b4 1 5 lbc3 is also interesting and has been tried at the highest level recently: 1 5 ... cxd4 1 6 lbxd4 .l:td8 1 7 l:txa 1 e 5 (17...'iib 6!? 1 8 e 3 e5 1 9 lbd5 "iVh6! offers Black counterplay, or 1 9 ltJde2 �e6+ 20 .ltd5 'iVc5 and White cannot hold on to f2) 1 8 b4 iVb6 19 lbc2 .lte6+ (19 .. :ihf2 20 .ltd5 0-0 21 'ii'c4 .lte6 22 lbe3 'iVxh2 23 "iVh4 'iVxh4 24 gxh4 f5 25 'it>c2 'it>f7 26 l:tdl f4 27 lbc4 gave White the smallest of edges in Kasparov-Adams, Wijk aan Zee 2001 , the game eventually ending in a draw) 20 i.d5 "iVxf2 21 J::td l !? (White wants to transfer the king to the gueenside) and now in Kir.Georgiev-Timoschenko, Panormo 200t there followed 21 ...0-0?! 22 'it>c1 i.xd5 23 lbxd5 'it>h8 24 'ii'c4 'iVxh2 25 g4 "iVg2 26 ttJde3, when White was well co-ordinated and had the better prospects. Timoschenko proposes 21 ...'iVxh2! 22 'it>c 1 �xd5 23 ttJxd5 l:tc8 24 'iVb3 'iVxe2 25 'iVa4+ 'it>f8 26 ttJc7 g6 27 'iVd7 rj;;g7 as an improvement (with two pawns and an insecure king to aim at, Black should be satisfIed with his lot) . 1 5 . . . cxb4 1 6 it'xa 1 .l:!.c8 1 7 ltJe5 Not to be recommended is 17 e4 0-0 1 8 'it>e3 bxa3, when the a-pawn looks menacing and White's king might prove awkwardly placed. 1 7 . . . iLb5 1 8 'it>e3? White should address the threatened infil­ tration. 1 8 i.e4?! 0-0 19 'it>e3 f6 20 ttJd3 f5 21 i.f3 �xd3 22 'it>xd3 'iVa4 saw White ex­ perience problems in Raetsky-Pushkov, Po­ dolsk 1 992, when 23 'iib2 bxa3 24 'iVxa3 'iVc4+ 25 'it>e3 e5! 26 'ii'd 6 iVb3+ led to a win for Black. The best course is 1 8 a4!? 0-0 1 9 i.e4 f6 20 lbf3! (Byrne & Mednis) and the struggle continues. 1 8 . . . .l:!.c2 1 9 iLf3 0-0 20 a4 f6 21 ltJd3 21 lbd2 fxe5 22 axb5 J::tc3+ 23 'it>e4 'iVxb5

and with this king position White can't ex­ pect to survive much longer. 2 1 . . . iLc4 22 ltJd2

22 . . . it'g5 + 2 2... .ltxd3! 2 3 exd3 'ii'g5+ wins o n the spot. 23 ltJf4 e5 24 ltJxc4 exf4+ 25 gxf4 'it'f5 26 ltJd6 26 i.e4 .u.c3+ 27 i.d3 (27 'iVxc3 "iVh3+) 27 ... J::te8+ 28 'it>d2 "iVxf4+ 29 'it>dl "iVxd4 and Black has a decisive advantage. 26 . . . it'e6 + 0-1

Game 74 Vakhidov-Ziatdinov

Tashkent 1987 1 d4 d5 2 c4 e6 3 ltJf3 ltJf6 4 g3 dxc4 5 iLg2 ltJc6 6 'ili'a4 iLb4+ 7 iLd2 ltJd5 8 iLxb4 ltJxb4 9 ltJe5

145

Th e Ca t a l a n

9 . . . .ltd7 9 ... 0-0 invites White to contemplate the captures on c6. White can cause considerable structural damage with 10 �xc6?! ttJxc6 1 1 ttJxc6 bxc6 1 2 'iVxc4 but this leaves him open to trouble on the b-ftle, e.g. 1 2 .. J:tb8 1 3 0-0 (1 3 b 3 ..ta6!) 1 3. . .l:txb2 1 4 ttJc3 'iVd6 1 5 l:tabl l:tb6 1 6 l:.fc 1 l:td8 1 7 e 3 e 5 1 8 dxe5 'iWxe5 and the light-squared bishop gave Black a plus in Stajcic-Luther, Kecskemet 1 993. Note that 10 ttJxc6 ttJxc6 1 1 �xc6 is quite different in that it allows the interesting 1 1 ...'it'xd4!?, which leaves 1 1 e3, when Mo­ chalov-Korneev, Minsk 1 995 went 1 1 ...e5 1 2 d 5 ttJe7 1 3 'ii'x c4 ttJ f5 1 4 0-0 ttJd6 1 5 'ii'c 3 l:te8 16 ttJd2 �f5 17 l:tac1 l:tc8 1 8 l:tfd 1 'it'd7 with a very solid set-up for Black. 1 0 12Ja3?! Not the most accurate of the available op­ tions. 10 ttJxd7 'it'xd7 1 1 a3?! b5 12 'it'xb5 ttJc2+ 1 3 ..t>f1 ttJ6xd4 1 4 'it'xc4 l:td8! 1 5 l:ta2 0-0 1 6 ..te4 (1 6 b4 ttJf5 and mate on d 1 is threatened) 1 6 ... f5 1 7 �d3 f4! leaves Black with the initiative. Instead Gulko assesses 1 1 ttJc3 'iVxd4 1 2 a3 ttJd5 1 3 l:tdl 'iVe5 as un­ clear. 10 0-0 appears best, when the following line is given as slightly better for White by V.Mikhaievsky: 1 O ... ttJd5 1 1 'ifxc4 ttJxe5 1 2 dxe5 �c6 1 3 ttJc3 0-0 1 4 l:tfdl ttJb6 1 5 'it'c5 ttJd7 1 6 'it'd4 ..txg2 1 7 'it>xg2 ttJb6 1 8 'ife4 again White enjoys a space advantage. 1 0 ... a5 1 1 ttJxd7 'iYxd7 12 ttJc3 J:td8 1 3 l:tfdl 0-0 1 4 a 3 was the course o f Y.Mikhalevsky­ Beshukov, Saint Vincent 2000, 14 ...ttJd5 1 5 'iix c4 ttJb6 1 6 iYc5 ttJxd4 1 7 l:!.ab l providing White with compensation thanks to Black's loose pawns. Kobylkin-Nadyrhanov, Kras­ nodar 2002 continued 10 ... 0-0 1 1 ttJxc6 ttJxc6 12 'it'xc4 e5 1 3 dxe5 ttJxe5 14 �d5 (more promising than 14 �f4 'ilVe7 1 5 ttJc3 c6 16 .l:.fd 1 f6 17 l:td4 l:tfd8 1 8 l:tad 1 ..te6 with equal chances in O'Cinneide-Adams, Kilkenny 1 999) 14 ... 'iVf6 1 5 �xb7 ..tc6 1 6 ..txc6 ttJxc6 1 7 ttJc3 l:tab8 1 8 'iVxc7 l:tfc8 1 9 'ilVd7 l:td8 20 'ilVg4 l:tb4 2 1 'ilVh5 l:txb2 and 146

Black achieved some chances but was still a pawn down, Nadyrhanov evaluating the po­ sition after 22 l:tfc1 ttJd4 23 nab 1 l:td2 24 l:tb7 g6 25 'iVg4 as slightly favourable for White. 1 0 . . . l2Jd3 + !

1 1 12Jxd3 cxd3 1 2 e 3 d 2 + 1 3 �xd2 e5 1 4 d5 With the king on d2 White has to keep the position closed. 1 4 . . . 12Je 7 1 5 'ii'e4 0-0 Black can opt for an equal endgame with 1 5 ... �c6 1 6 'iVxe5 ..txd5 1 7 ..txd5 �xd5+ 1 8 'ii'x d5 ttJxd5. 1 6 �e2 16 'ii'x e5 runs into 16 ... ttJxd5!, e.g. 17 ..txd5 (1 7 'ilVxd5 �cfj gives Black a superior ending in view of his stronger minor piece) 1 7 ... �e6 1 8 e4 c6 and Black opens the centre and begins an attack against the troubled king. 1 6 . . . c6 1 7 'ii'x e5 17 d6 ttJf5 18 tthdl 'iVf6 and Black has the initiative. The d6-pawn is a bit loose. 1 7 . . . l2Jxd5 1 8 .ltxd5 cxd5 1 9 .l:!.hd 1 .ltc6 20 .l:!.d4?! 20 ttJc2 l:te8 21 �f5 'iYb6 and White has no time to consolidate with ttJd4. Black's initiative is too strong. 20 . . . .l:!.e8 21 'iiVf 5? A lesser evil is 21 'ilVf4 'i!Vb6 22 l:tb4 'it'a6+, although Black retains the upper hand. 21 . . :iWb6 22 12Jc2

4 . . . dx c 4 5 i.. g 2 tLl c 6

22 .l:td2 d4! blows away the defences. 22 . . . g6 23 'ilVg5 'ilVxb2 Black is winning. 24 ..t>d2 .l:!.e4 25 .l:!.c1 .l:!.c8 26 .l:!.xe4 dxe4 27 'ilVa5 i.. b 5 0-1

Game 75 Filippov-Rausis

Dubai 1999 1 d4 d5 2 c4 e6 3 tLlf3 tLlf6 4 g3 dxc4 5 i.. g 2 tLlc6 6 "iVa4 i.. b4+ 7 i.. d 2 tLld5 8 i.. x b4 tLlxb4 9 0-0

l:!.ac1 ttJb3 1 8 .l:tcd 1 ttJb4 1 9 e3 cS with equal chances. 1 0 . . . a6 10 ... 0-0 1 1 a3 ttJdS 12 "it'xc4 ttJaS 13 'ir'd3 favours White, but l O ... .i.d7 leads to compli­ cated play, e.g. 1 1 a3 bS 1 2 ttJxbS ttJdS (1 2 ... a6 1 3 ttJc3 ttJxd4 14 'ilkaS! ttJb3 I S 'fieS f6 1 6 �e4 ttJxal 17 axb4 ttJb3 1 8 �xc4 'i'e7 19 'i¥xb3 and the knights are stronger than the rook) 1 3 ttJc3 (1 3 e4 ttJce7 14 exdS .i.xbS I S 'ilVxa7 c3 and Black has good coun­ terplay) 1 3. .. .l:txb2 14 'iVxc4 ttJaS I S 'i'd3 .l:tb3 1 6 .l:tfc1 cS 1 7 �d2! c4 1 8 ttJeS ttJxc3 19 .l:txc3 and White's centre was the differ­ ence in Romanishin-Aleksandrov, Pula 1 990. 1 1 tLle5 1 1 a3 bS 12 "it'dl ttJdS is nice for Black. 1 1 .l:tfc 1 , on the other hand, is more to the point: 1 1 ...0-0 1 2 'iVdl bS 1 3 b3 ttJaS 14 ttJe4 ttJdS I S ttJcS ttJb6 1 6 ttJeS f6 1 7 b4 fxeS 1 8 bxaS exd4 1 9 axb6 (Nedobora-Sjodahl, La Coruna 1 993) 1 9 ... cxb6!? with an unclear position. 1 1 . 0-0 1 1 ...'�xd4 is thematic, e.g. 12 ttJxc6 ttJxc6 13 .i.xc6+ bxc6 14 'i'xc6+ 'iVd7 IS 'i'xc4 0-0 16 b3 'iVe7 17 ttJa4, Lavrov-Raetsky, Lipetsk 1 993. In this knight versus bishop middle­ brame Black's weaknesses on the queenside are more important than the compromised light squares around the white king. Never­ theless, White's advantage is only small. 1 2 i.. x c6 12 ttJxc6 bxc6 13 a3 ttJdS 14 �xc4 .l:txb2 I S e3 is also possible. White has a good pawn structure and compensation for the pawn. 1 2 . . . tLlxc6 1 3 tLlxc6 bxc6 1 4 "iVxc4 14 �xc6 �xd4 I S .l:tfdl 'iVeS 1 6 .l:td2 .l:tb4 17 .l:tadl .i.b7 1 8 �d7 'i'cs was balanced in Rashkovsky-S.Ivanov, Elista 1 995. 1 4 . . . "iVd6! ? The immediate capture on b 2 needs inves­ tigating: 14 ... .l:txb2 IS .l:tab 1 .l:tb6 (1 5 ... .l:td2!? 16 e3 eS 17 dxeS .i.fS is interesting) 16 'iVcs h6 1 7 .l:tfdl l:!.xbl (1 7 ... .i.d7 1 8 a4 �8 19 as .l:tb2 20 �a3 .l:txbl 21 .l:txb l 'flVa7 22 �cS .

9 .l:!.b8 1 0 tLlc3 10 ttJeS?! is a dubious idea, for after lO ... 'iVxd4 1 1 ttJxc6 ttJxc6 1 2 .i.xc6+ bxc6 1 3 �xc6+ iLd7 1 4 �xc7 0-0 - a s in the previ­ ous game - the light-squared bishop gives Black the better chances. After l O ttJa3 .i.d7 1 1 "itbs White collects the pawn with the usual Catalan space advan­ tage, while Wojtkiewicz-Krasenkow, Manila 1 998 went 1 1 ttJeS 0-0 12 ttJxc6 ttJxc6 1 3 �xc4!, the subsequent 1 3 .. .'iVxd4 1 4 .i.xc6 �xc4 IS ttJxc4 bxc6 16 b3 .l:td8 17 .l:tfdl iLd7 1 8 .l:td2 leading to a pleasant ending for White in view of Black's poor pawns. How­ ever, in Wojtkiewicz-Yuneev, Yerevan 1 996 Black fared better, the simple and sensible 10 ... 0-0 1 1 'iVb5 b6, with the idea of ... .i.a6, being a common theme in the Catalan. There followed 1 2 'flVxc4 .i.a6 1 3 ttJbS 'ir'dS 1 4 'i'xdS ttJxdS I S a4 ttJaS 1 6 ttJeS .l:tbd8 1 7 . . .

.

147

Th e Ca t a la n

'ii'x cs 2 3 dxcS .i.cs favoured White in Cs.Horvath-Luther, Budapest 1 991) I S :xb 1 'iVd6 1 9 ltJe4 'tlidS 20 'tlixdS cxdS 21 ltJcs l:!.eS 22 l:.bS was seen in Khalifman­ S.Ivanov, St. Petersburg 1 996. Despite the pawn deficit White's chances are preferable because Black's bishop is quite poor at the moment and White has a firm grip on the dark squares. 1 5 e3 White can't protect both b2 and d4, so he has to choose. 1 5 b3 I:tb4 16 ltJe4 l:!.xc4 1 7 ltJxd6 llxd4 I S ltJxcs lhcs 1 9 l:[fc 1 a s 20 l:!.xc6 a4 21 .l:tac1 g6 was agreed drawn in Mittelman-Zifroni, Tel Aviv 200 1 , but 1 5 ltJe4!? is interesting, e.g. I S ...'ii'd S 1 6 'ii'c2 'ii'xd4 17 b3 �bS I S l:!.fdl 'ii'e S 1 9 ltJc3 ncS 20 l:!.d3 with an edge for White in Kamenets­ Mohrlock, Correspondence 200 1 . 1 5 . . J:txb2 1 S J:!.ab 1

1 S . . . J:!.bS

148

Also possible is 1 6 .. J::txb 1 1 7 l:!.xb 1 eS I S dxeS 'ii'x eS 1 9 'iVxc6 .i.h3 20 ltJe2 gS 21 'iih 6 'iVe4 22 'iVxgS+ with a draw in Rausis­ Nisipeanu, Cappelle la Grande 2002. Here White can be more ambitious with 20 ltJdS l:!.dS 21 ltJf4 .i.f5 22 l:tc1 , but Black gets excellent counterplay due to the weakness of the light squares - 22 ... .i.e4 23 'tlixa6 gS (Nisipeanu & Stoica) . 1 7 f4 White prevents the thematic ... e6-eS break. 1 7 ltJa4 llbS doesn't force Black to surrender the b-ftle, while after 1 7 I:tfc1 eS I S dxeS 'iVxeS 1 9 1hb6 cxb6 20 'iVxc6 'iVaS Black has a good position. 1 7 . . . a5 1 8 �a4 �aS 1 9 J:!.fd 1 c 5 ! ? 1 9 ...n fb S 2 0 'it'xaS .i.c4 can also be considered. The position remains unclear. 20 ttJe4

20 . . . �d5 21 J:!.xbS �xe4 22 J:!.xaS Black forces a draw by repetition.

Y2 - Y2

4 . . . dx c 4 5 iL g 2 tD c 6

Summary After 6 0-0 l:[b8 White's strongest seems to be 7 lbc3. Then after 7 ... i.b4 White can claim a slight advantage after both 8 i.gs and 8 'ii'c 2. In reply to 7 ... bS, reacting with 8 e4 is not in White's interest, but Black should probably avoid the unclear 8 ... b4 and develop naturally with 8 ... i.e7. It is interesting that the game can transpose to the 'normal' variation S ... i.e7 6 0-0 0-0 7 lbc3 lbc6 8 e4 .l:tb8 followed by 9 ... bS, as seen in the notes to Game 27 in Chapter 4. Instead the best way to meet the early push of the b-pawn is undoubtedly 8 lbeS. Sufficient research has been carried out on the line 8 ...lbxeS 9 dxeS lbd7 10 i.xc6 a6 1 1 'ir'xd4 i.b7 12 i.xb7 l:txb7 1 3 l:[dl . Here it is not so easy for White to demonstrate compensation for the pawn, so it is necessary to pay attention to 10 iVd4!? (Game 66) . Now w e turn t o 6 'ii'a4. The strategy beginning with 6 ... i.d7 appears not t o offer Black equality. In the main variation 7 'ii'x c4 lbaS 8 'ii'd 3 cS 9 0-0 i.c6 1 0 lbc3 Black can choose between the plan with ... cS-c4 (immediately or - as is seen more often - after 10 .. .11c8 I I l:tdl ) o r the less ambitious exchange o n d4. In the former case White i s given the pawn majority in the centre and often develops an initiative through the positional breakthrough d4-dS (see, for example, Ftacnik-Lechtinsky in Game 67, note to Black's 1 0th move). After 1 0 ... cxd4 the game tends to be simplified and it is not so easy for Black to neutralise the slight but often enduring superiority enjoyed by White. After 6 ... lbd7 7 'it'xc4 lbb6 8 'ii'd 3 eS 9 i.e3 White cannot realistically claim a genuine ad­ vantage, but Black must be careful to avoid quickly running into trouble (see Raetsky-S.Ivanov, Game 68 - including Filippov- Short) . In the case of 9 lbxeS lbb4 the queen moves to dl and b3 fail to furnish White with anything positive, and nor does 1 0 'it'c3 if Black plays 1O ...'ir'xd4 1 1 0-0 'ii'x c3 1 2 lbxc3. As a whole the variation starting with 6 ... lbd7 contains numerous areas for new discoveries. For example 9 ...lbd4 has never been tested despite the fact that we do not see how White should achieve any appreciable advantage. Another deviation from the main line (which, of course, is 6 ...i.b4+), namely 6 ... i.d6, is quite playable. After 7 0-0 0-0 8 lldl Black can exploit the absence of White's bishop from d2 (as opposed to the modern line 6 ... i.b4+ 7 i.d2 i.d6 8 0-0 0-0) by playing 8 ...lbb4, illustrating the downside to the posting of White's queen on a4 (polgar-Benjamin, Game 70). If White is not suited to these complex positions it is better to allow ... e6-eS with 8 'ii'xc4, or to play an­ other line starting with 7 lbbd2. The theory of the line with 6 ... i.b4+ 7 i.d2 i.d6 has only just begun to develop. It is particularly difficult to evaluate how the chances are divided after 8 lbeS. A fter 8 'ir'xc4 0-0 9 0-0 eS 1 0 dxeS almost all games have ended in a draw, and White, if he is looking to achieve something in the opening, should go for 1 0 dS lbe7 I l lbc3. After 6 'ii'a4 i.b4+ 7 �2 lbdS the manoeuvre 8 'ii'b s loses time and is not dangerous for Black, who should simply continue with 8 ... 0-0. For example in Grabliauskas-Benjamin (Game 72) an important feature is the potentially precarious situation of White's king, which is stuck in the centre. In the very sharp line 8 i.xb4 lbxb4 9 a3 bS 1 0 'ii'x bS lbc2 1 1 'itd2 several recommenda­ tions can be made for Black. First there is 1 1 ...i.d7!? 1 2 'itxc2 lbxd4+ 1 3 lbxd4 i.xbS 1 4 lbxbS .l:tb8 (Chetverik-Grabliauskas in Game 73, note t o Black's 1 1 t h move), which i s not often quoted, although Black's queen does not seem to be inferior to White's three pieces. Secondly, Black no longer plays l 1 ...lbxal 12 'ii'x c6+ i.d7 13 'it'xc4 1Ib8 14 b4 cS!, but all the possibilities have not been refuted here. Finally, the modern line 1 3 ... cS 14 'ii'a2 'ii'a S 1 s lbc3 1 49

Th e Ca t a l a n

cxd4 1 6 ttJxd4 l:td8 also gives Black sufficient counterplay (according t o analysis b y Timo­ schenko) due to White's vulnerable king. In the solid variation with 9 0-0 we have seen considerable changes in recent fashion. The position after 9 ... l:tb8 1 0 ttJc3 a6 1 1 ttJe5 0-0 1 2 i.. x c6 ttJxc6 1 3 ttJxc6 i.. x c6 1 4. 'ii'xc4 'ii'd 6!? has become critical: Black prevents the enemy queen from taking up a dominating position on c5. If White does give up the b2-pawn Black's rook will have other qualities than those of a savage killer on the b-file (1 5 b3 i:tb4 and 1 5 ttJe4 'iVd5 1 6 'iVc2 "it'xd4 17 b3 i:tb5 should be noted) . We do not see anything special for White here. More attention might be given to Wojtkiewicz' 10 ttJa3, which sees the knight monitor both c4 and b5. If Black doesn't play 1 0 ... a6 White regains his pawn with the familiar manoeuvre 'itb5-c4.

1 d4 d5 2 c4 e6 3 lLJf3 lLJf6 4 g3 dxc4 5 �g2 lLJc6 6 �a4 6 0-0 l:!.b8 (D) 7 e3 Game 6 1 7 a4 7 ... a6 Game 51 (Chapter 5) ; 7 ... b6 Game 62 7 ttJc3 7 ... i.. b4 8 e3 Game 63; 8 'it'c2 Game 64 7 ... b5 8 e4 Game 65; 8 ttJe5 Game 66 6 . . . �b4+ 6 ... iLd7 Game 67 6 ... i..d 6 Game 70 6 ... ttJd7 7 "it'xc4 ttJb6 8 �d3 e5 (D) 9 iLe3 Game 68; 9 ttJxe5 Game 69 7 �d2 1LJd5 7 ... i.. d6 Game 71 8 �xb4 8 'itb5 Game 72 8 . . . lLJdxb4 (D) 9 a3 Game 73; 9 ttJe5 Game 74; 9 0-0 Game 75 -

-

-

-

-

-

-

-

-

-

-

-

-

-

-

6. . 1:.bB .

1 50

-

B . e5 .

.

B. . . lLJdxb4

CHA PTER SEVEN

I

4 . . . dxc4 5 .192 c 5

1 d4 d 5 2 c4 e6 3 4Jf3 4Jf6 4 g 3 dxc4 5 i.g2 c5 This is a rather logical continuation. 4 ... dxc4 has already given Black sight of the d4-pawn and what would be more natural than for Black to immediately contest the centre in this manner? Consequendy it is not surprising that the 5 ... c5 system has main­ tained its popularity since its use during the early days of the Catalan. In reply to 5 ... c5 White occasionally con­ tinues 6 'ilVa4+ (Game 76), and of course this can transpose to 5 'iVa4+. The variation with 6 �a4+ SLd7 7 'iVxc4 .1i.c6 was tested twice in the Kasparov-Korchnoi encounters in London in 1 983. Black solved the problems with his queen's bishop and managed to equalise with accurate play. Therefore 6 �a4+ has lost ground to the modern 6 0-0. Here Black is not obliged to resolve matters in centre at once with 6 ... cxd4 (Game 77), when White gets an e"-luring, albeit tiny ad­ vantage after both 7 'tli'xd4 and 7 ttJxd4. With this in mind Black almost always pre­ fers 6 ... ttJc6, increasing the pressure on the d4-pawn. White must act quickly if he is to achieve anything positive. The exchange in the centre with 7 dxc5 is harmless for Black as he has good development, but the direct 7 ttJe5 presents Black with serious problems to

address. After 7 ...ttJxe5 8 dxe5 'iNxdl 9 l:txdl ttJd5 (or 9 ... ttJd7) the manoeuvre ttJa3xc4 regains the pawn and secures a small but steady advantage, while even 8 ... ttJd5 (Game 78) favours White. The main answer to 7 ttJe5 is 7 ... SLd7. Now White must choose between 8 ttJc6, 8 ttJc4 and the main move 8 ttJa3 (Games 79 & 80) . Another common continuation in the dia­ gram position above is the standard attack with the queen - 7 'iVa4. Then we have a branch with 7 ... cxd4 (Game 81) and 7 ... .1i.d7 (Games 82 & 83) . After 7 ... cxd4 8 ttJxd4 'ilVxd4 9 .1i.xc6+ .1i.d7 10 l:tdl the ending that results from 1O ... .1i.c6 1 1 'iYxc6 bxc6 1 2 l:td4 tends to be inferior for Black, which explains why there is a preference for sacrificing the queen for rook, bishop and pawn with 1 O ...�xd 1 1 1 l:txd 1 .1i.xc6. In the case of 7 �a4 the reply 7 ... .1i.d7 has become increasingly popular. It has been established that 8 dxc5 does not confer an advantage due to 8 ... ttJa5 9 �c2 .1i.xc5, which brings us to the main continuation, 8 �c4, when Black has two possibilities 8 ... cxd4 (Game 82) and 8 ... b5 (Game 83). Portisch-Radulov and Ribli-Ljubojevic re­ ceived much attention and must have helped in promoting the popularity of 7 'iVa4. 151

Th e C a t a l a n

White plays 6 'i'a4+

Game 76 I n kiov-Pinter

Zagreb 1987 1 d4 d 5 2 c4 e6 3 lLlf3 lLlf6 4 g3 dxc4 5 �g2 c5 6 'i'a4+

6 . . . �d7 Of independent relevance is 6 . ..tLlc6. Then 7 0-0 i.d7 transposes to 6 0-0 ltJc6 7 'ii'a4 i.d7 below (Game 82) . 7 dxcS i.xcs 8 0-0 0-0 9 'ii'x c4 'ii'dS to ltJfd2 �xc4 1 1 ltJxc4 ltJd4 12 ltJc3 i.b4 1 3 e3 i.xc3 14 bxc3 ltJe2+ 1 5 �h 1 ltJxc1 1 6 .l:i.fxc 1 .l:!.b8 17 e4 ltJd7 1 8 f4 b6 1 9 ltJd6 i.b7 20 .l:i.dl ltJcs achieved nothing for White in Balashov­ Beliavsky, Vilnius 1 980. This leaves 7 ltJeS ltJdS (7 ... i.d7 8 ltJxc6 i.xc6 9 i.xc6+ bxc6 to 0-0 transposes to 6 0-0 ltJc6 7 ltJeS i.d7 8 ltJxc6 i.xc6 9 i.xc6+ bxc6 10 'iWa4, dealt with in Game 79) 8 ltJxc6 bxc6 (8 ... �d7 9 'ii'xc4 'iVxc6 10 ltJc3 cxd4 1 1 'ii'xd4 'itb6 1 2 'iVxb6 ltJxb6 1 3 0-0 favours White because Black is behind in develop­ ment and is under pressure on the long di­ agonal) 9 dxcS i.b7 to 'iVxc4 'iY'aS+ 1 1 i.d2 'iVxcs 12 'iVxcs i.xcs 1 3 ltJc3 and White has a slight but enduring edge. 7 'i'xc4 �c6 S 0-0 Another possibility is 8 dxcS, when 8 ...'iVdS 9 'iVxdS ltJxdS to ltJeS should fa­ vour White. Indeed 1 O ... ltJb4 1 1 ltJxc6 1 52

ltJ8xc6 1 2 ltJa3 i.xcs 1 3 i.d2 0-0 1 4 0-0 i.d4 1 5 ltJc4 l:1ab8 1 6 a3 ltJdS 1 7 .l:!.ac1 i.f6 1 8 e3 was rather pleasant in Keres-Klovan, Parnu 1 960. Black needs to be active, and a good try is 8 ... ltJbd7 9 i.e3 i.dS to 'iVb4 as, which was the subject of Sosonko-Farago, Wijk aan Zee 1 988. The game continued 1 1 'iVh4 i.xcs 1 2 i.xcs ltJxcs 1 3 ltJc3 i.xB 1 4 i. x B 0-0 1 5 0-0 'iVb6 1 6 ltJa4 ltJxa4 1 7 'iWxa4 l:tad8 1 8 .l:i.adl 'ii'xb2 1 9 'iVxaS l:ta8 20 l:td2 .l:!.xaS 21 .l:!.xb2 ltJdS and the simplifica­ tions and the pressure against a2 left Black with no problems and should have resulted in a draw. This is certainly an improvement for Black on to ... 'tic8 1 1 ltJc3 i.xcs 1 2 i.xcs "iYxcs 1 3 ltJxdS ltJxdS 1 4 'iVd2 l:tc8 1 5 0-0 0-0 1 6 .l:i.ac1 ifb6 1 7 'iVd4 .l:i.fd8 1 8 l:i.fdl �xd4 1 9 ltJxd4, which led to White's stan­ dard Catalan advantage of bishop versus knight in Korchnoi-Kasparov, World Cham­ pionship Candidates Match (game 8), Lon­ don 1 983. s . . . lLlbd7 9 �g5 9 ltJc3 is rendered hannless by 9 ... bS!, the point being that 1 0 ltJxbS?? ltJb6 1 1 'ifb3 c4 sees White lose the knight. After to 'iVd3 c4 1 1 'iWc2 b4 1 2 ltJd l ltJb6 1 3 i.gS i.e7 1 4 ltJe3 .l:!.c8 only Black can claim t o b e better. 9 . . .l:!.cS Black can also equalise with 9 ... ltJb6 10 'ii'd 3 cxd4 11 ltJbd2 i.e7 1 2 ltJxd4 i.xg2 13 'it'xg2 0-0 etc. 1 0 �xf6 .

4 . . . dx c 4 5 i.. g 2 c 5

1 0 . . . gxf6! Perhaps too inventive for some, but Black does not want to recapture with a piece. In fact after 10 . ..tbxf6 White is guaranteed a slight edge with 1 1 dxc5 iLxf3 1 2 iLxf3 iLxc5 13 'ifb5+ (1 3 e3!?), when Kasparov­ Korchnoi, World Championship Candidates Match (game 7), London 1 983 went 1 3. .. 'it'd7 14 tiJc3 'it'xb5 1 5 tiJxb5 �e7 1 6 b4 iLxb4 1 7 tiJxa7. Then Black's best i s to keep his disad­ vantage to a minimum with 1 7 ...l:!.a8 1 8 tiJb5 .l:'ta5. 1O ...'ii'xf6 looks the safest. Korchnoi gives 1 1 tiJc3 iLe7 1 2 e4 0-0 1 3 d5 tiJb6 1 4 'ii'd 3 exd5 1 5 exd5 ':'fd8 16 tiJd2 c4 17 tiJxc4 tiJxc4 1 8 'ii'x c4 iLxd5 1 9 tiJxd5 .l:'txc4 20 tiJxf6+ iLxf6 when the draw is imminent. 1 1 lLlc3 1 1 dxc5?! iLxc5 12 'it'g4 'ii'b 6 is awkward for White as the b2-pawn is under fIre. 1 1 . . . b5 1 2 'ifd3 c4 1 3 'ifc2 b4 1 4 lLle4 lLlb6 1 5 g4? ! This move is hard t o justify and should land White in trouble if Black were to react accurately. 1 5 e3 f5 1 6 tiJed2 c3 also gives Black a good game. 1 5 . . . h5?! Overlooking 15 ...l:.g8! 1 6 h3 (1 6 g5 fxg5 1 7 tiJe5 iLa4! and White is in a very bad way) 1 6 ... h5 1 7 gxh5 f5 with a decisive attack. 1 6 g5 fxg5?! 16 ... iLxe4 1 7 'ii'x e4 fxg5 18 tiJe5 iLg7! is preferable, after which the situation is far from clear. 1 7 lLlexg5 .tg7 17 ... iLe7? runs into 18 tiJxf7! �xf7 1 9 tiJe5+ etc. 1 8 h4 i..f6 1 8 ... iLxd4? 1 9 l:tadl iLa4 20 b3 cxb3 21 'it'e4! gives White a strong attack (Inkiov) . 1 9 e4 i.. x g5 19 ... c3!? is a sound alternative, 20 bxc3 bxc3 21 a4 leaving both sides with much to play for. 20 hxg5 .l:!.g8 21 .l:!.fe 1 i.. a4 22 'ifd2 c3 23 bxc3 bxc3 24 'iff4 'ifc7

Black needs to be careful. 24 ... tiJc4 25 .l:'tac1 c2 26 d5 'iVd6 27 'ii' f6 is double-edged. 25 'ti'h4 lLlc4 26 d 5 2 6 'it'xh5?! succeeds only i n providing Black with a large helping of counterplay after 26 ...'it'f4! (the h-pawn is practically worthless) . 26 . . . i.. d 7

27 dxe6? ! 27 e5! exd5 28 tiJd4! is stronger. After 28 ... iLe6 (28 ...'ii'c 5 29 e6! puts Black's king in danger) 29 'ii'x h5 'iWb6 30 tiJxe6 'iWxe6 3 1 iLh3 'it'e7 32 � h l White has a wonderful attack. 27 . . . i.. x e6 28 lLld4 i.. g 4 29 lLlf5 lLle5 30 'ti'g3 .l:!.d8?! Black's play seems to be a little planless, an appropriate course being 30 ... iLf3! 3 1 l:te3 c2 32 .i:!.c1 in order to generate counterplay. The most obvious move is then 32 ... iLd 1 , when the c-pawn is certainly a factor, but after 33 .l:tb3! with the idea of 34 :b7! the situation remains a complete mess. A possi­ ble conclusion to the game is 33 ... l:tg6 (pro­ tecting d6) 34 .l:'tb7 'ii'x b7 35 'ii'x e5+ �d8 36 'ii'h8+ �c7 37 'ii'e5+ with a draw by perpet­ ual check. 31 .l:tac 1 .l:!.d3 31 ... c2? 32 Itxc2! is simply a terrible blun­ der, while after 3 1 ...h4 32 'iWxh4 l:td2 33 tiJh6! l:th8 34 'iWg3 Black is also in dire straits. Perhaps it is already too late to save the game. 1 53

Th e C a t a l a n

1 O ... .ixd6 1 1 .ixd6 .ixf3! 1 2 .ixf3 'iib 6 1 3 .ia3 O-O-O! a s i n Bogdanovsky-Smagin, Prilep 1 992. Smagin gives 1 4 l:tc1+ 'it>b8 1 5 J:.c4 lLle5 1 6 J:.b4 lLlxf3+ 1 7 exf3 �c6 1 8 J:.xd4 l:!.d5 as level. 7 lLlxd4 White has a decent alternative in 7 'tWxd4 'tWxd4 8 lLlxd4 a6 9 lLld2 i.c5 1 0 lLlc2 c3 1 1 Black meets 6 0-0 with 6 . . . cxd4 bxc3 lLlbd7 1 2 lLlb3! .ie7 1 3 J:.bl , when the open ftle is perfect compensation for the Game ll isolated pawns. The subsequent 1 3 ... lLld5 14 Chetverik -Ivan .id2 b5 15 lLlb4 lLl7b6 16 lLla5 i.d7 17 lLlxd5 exd5 18 c4! .ifS 1 9 e4! .ixe4 20 i.xe4 Harka'!J' 1996 --------.. dxe4 21 cxb5 leaves Black somewhat worse in view of the ed pawn and his slightly 1 d4 d5 2 c4 e6 3 lLlf3 lLlf6 4 g3 dxc4 5 inferior development, while Balashov­ i.g2 c5 6 0-0 A.Petrosian, Yerevan 1 986 favoured White after 1 7 ... lLlxd5 1 8 c4 bxc4 1 9 lLlxc4 .ib5 20 l:!.fc1 0-0 21 a4! .ixc4 22 J:.xc4 l:i.ac8 23 nbc 1 . 7 . . .lLla6 ! ? 7...�6 8 lLlc3 .id7 9 .ie3! looks danger­ ous for Black, who should respond with 9 ... lLlg4!? or 9 ... lLla6!? rather than follow the example of V.Mikhalevsky-Murey, Israel 1 997, when the experienced GM went for the bait with 9 ... .ic5?! and ran into 1 0 b4! cxb3 1 1 J:.bl ! i.xd4 1 2 .ixd4 'iVc7 1 3 �xb3 i.c6 1 4 'iVa3!, the further 1 4 ... .ixg2?! 1 5 lLlb5 'iVd7 1 6 lLld6+ 'it>f8 1 7 'it>xg2 'iWc6+ 18 6 . . . cxd4 'iitg l �a6 19 �xa6 bxa6 20 .ixf6! gxf6 21 Black has more colourful alternatives. Ra­ l:!.b7 resulting in a decisive lead for White. zuvaev-Feller, Poland 1 988 continued 6 ... a6 7 S lLla3! ? dxc5 'iflxdl 8 .l:txdl .ixc5 9 lLle5 lLlbd7 1 0 8 �a4+ .id7 9 �xc4 l:!.c8 followed by ttJxc4 l:!.a7 1 1 ttJc3 with an edge for White. ...lLlc5 is no worse for Black. There followed 1 1 ...b5?! 12 ttJd6+, when S . . . i.xa3 9 bxa3 0-0 1 0 i.b2 'ii'b6 1 1 Black could have minimised the damage 'ii'c 2 i.d7 caused by the weakening thrust of the b­ 1 1 ...l:!.d8 12 lLlf3! and White has a strong pawn with 1 2 ... i.xd6 1 3 l:!.xd6 'it>e7 1 4 l:!.c6 position. Black needs to get his development lLlb8 1 5 .l:tc5 l:!.d7 1 6 .if4 .ib7 1 7 a4, al­ gOing. though this is nevertheless rather unpleasant. 1 2 l:tab 1 'ii'c 7 1 3 lLlf3 l:tacS Instead after 12 ... 'it>e7? 1 3 ttJce4 lLlxe4 1 4 13 ... c3 14 .ixc3!? l:!.fc8 1 5 l:!.fc1 ttJd5 lLlxe4 l:!.c7 1 5 .id2! White was already win­ looks strong but White has a counter with 1 6 ning. lLlg5 lLlxc3 1 7 'ifxh 7+ 'it' f8 1 8 'ifh8+ �e 7 1 9 6 ... .id7 is more interesting for Black, e.g. 'iflxg7 and Black i s forced into 1 9 ... ttJxe2+ 20 7 lLla3 .ic6 8 ttJxc4 lLlbd7 9 .if4 cxd4 1 0 �f1 ttJxc1 21 'iflxf7+ 'iitd 6, when White has ttJd6+ ( 1 0 ttJxd4!? might b e an improvement)

32 l:te3 i.e2?! 32 ...l:!.xe3 33 'iVxe3 c2 34 f3! .ixfS 35 exfS 'it>f8 36 f6 wins for White. 33 l:txd3 i.xd3 34 f4 h4 34 ... 'iflb6+ 35 'iKe3 lLlg4 36 'iflxb6 axb6 37 l:!.xc3 and White wins. 35 'ii'e 3 lLlg4 36 'ii'x d3 1 -0

1 54

4 . . . dx c 4 5 i.. g 2 c 5

no more than perpetual check - 22 iff4+ rl;e7 23 'iWf7+ 'iitd 6 24 tDe4+ 'it>dS! 25 tDd2+ 'iYtd6 26 tDe4+ 'it>dS and White cannot make progress. 14 i..e 5 'i'c5 1 5 l:tfd 1 ttJg4 Both the lines l S.J:tfdS 1 6 lId4 and lS ... i.c6 16 �xf6 gxf6 17 �c3 give White good compensation for the pawn on the kingside. 1 6 i.. d 4 �a5 1 7 1j'e4 17 i.c3!? .l1.a4 1 8 "iVc 1 is a possible im­ provement. 1 7 . . .f5? ! 1 7 ... eS is necessary. After 1 8 i.xa7 the game can swing either way. 1 S 1j'xb 7 .l:!.c 7 1 9 1j'b2 i.. a4 20 .l:!.dc 1 l:tbS

Allowing a nice queen sacrifice, but by now Black is already walking on thin ice. 20 ... h6 21 h3 tDf6 22 tDeS illustrates White's complete control over the centre. 21 1j'xbS + ! ttJxbS 22 l:txbS + �f7 23 ttJg5 + �e7 24 i.. c 3 1j'a6 25 J:!.gS e5? Black can still put up some resistance with 2s .. :iNb6 26 lIxg7+ 'it>e8 27 .l::!.g8+ rJJ e 7, al­ though White can look forward to a power­ ful attack after 28 e3. 26 .l:!.xg 7 + �f6 26 ... �d8 27 �aS! and White wins. 27 l:txc7 �xg5 2S l:txc4! Now Black can no longer defend. 2S . . . i.. b 5 29 .l:!.c7 'ilfb6 30 i.. d 2+ �h5 3 1 e 3 h 6 32 i.. b4 i.. a4 3 3 i.. d 5 i.. e S 3 4 h3 ttJf6 35 i.. f3 + �g5 36 l:tg7 + 1 -0

White plays 6 0-0 ttJc6 7 ttJe5

Game 78 Ksieski-Enders

Bad Harzburg 2000 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 i.. g 2 c5 6 0-0 ttJc6

7 ttJe5 This is probably the most ambitious choice. Indeed 7 dxcS 'iWxd1 8 lIxd1 �xcS appears insufficient for an advantage, e.g. 9 tDbd2 c3 1 0 bxc3 0-0 1 1 tDb3 i.e7 1 2 tDfd4 �d7 1 3 �gS .l::!. a c8 1 4 i.xf6 .l1.xf6 1 5 tDcs tDxd4 1 6 cxd4 �bS with equality in Speel­ man-Van der Sterren, Yerevan 1 996. 7 �a4 features in Games 81 -83 below. 7 " 'ttJxe5 This is, of course, not forced. 7 ... i.d7, which is considered in the next two main games, should be the strongest alternative. 7 . . . tDxd4??, on the other hand, spells disaster for Black due to 8 e3 tDf5 9 'iNxd8+ 'it>xd8 10 tDxf7+ and White wins. 7 ...tDdS 8 tDxc6 bxc6 9 dxcS �xcS 10 ifa4 'iVb6 1 1 tDd2 favours White, e.g. 1 1 ...�a6 1 2 tDxc4 'iVbs 1 3 �xbS cxbS 1 4 tDeS b4 l s lId1 �b7 1 6 �d2 i..e7 1 7 lIac 1 as 1 8 e4 tDf6 1 9 �gS i.d8 20 tDc4 ()-() 21 tDd6 with a pleasant game for \x'hite in Savchenko-Galliamova, Rostoy 1 996. S dxe5 ttJd5 Black cannot equalise here. 8 . . .tDd7 9 f4 l:tb8 1 0 a4 i.e7 1 1 tiJa3 b6 1 2 tiJxc4 i.b7 1 3 1 55

Th e C a t a la n

b 3 .Jtxg2 1 4 �xg2 'iYc7 1 5 .Jtb2 'iVc6+ 1 6 �g1 0-0 1 7 'it'c2 .l:!.bd8 1 8 e4 was better for White in ljngnau-Klovans, Berlin 1 996, and 8 ... 'it'xd1 9 J:txd 1 ttJd7 1 0 f4 l:tb8 1 1 a4 .Jte7 12 ttJa3 0-0 1 3 ttJxc4 is also very pleasant for White, who was successful in Hajtun-Gecsei, Budapest 1 9 5 1 : 13 ... b6 14 ttJd6 l:td8 1 5 .Jte3 .Jtxd6 16 ':xd6 .Jtb7 17 l:tad 1 .Jtxg2 1 8 �xg2 lIb 7 1 9 fS h6 20 g4 � f8 2 1 g5! exfS 22 gxh6 gxh6 23 l:txh6 l:tdb8 24 l:[xd7! and White won. 9 ttJa3 Another strong option is 9 'it'a4+!? .Jtd7 1 0 'ii'xc4, when Black needs to find an im­ provement on Vaulin-Koc, Koszalin 1 997, which went 1O ... .Jtc6 1 1 ttJc3 'it'd7 12 lId1 b5 1 3 'iVg4 'iVb7 1 4 ttJxd5 .Jtxd5 1 5 .Jtxd5 exd5 16 e6 l:[d8 17 .Jtg5 .Jte7 1 8 .Jtxe7 'iVxe7 1 9 'it'xg7 and White was already close to wmrung. 9 . . . i.d7 Black does not have time for 9 ... c3 1 0 bxc3 ttJxc3 1 1 'iVc2 ttJd5 1 2 l:td1 .Jtd7 1 3 ttJc4 which put the defender under too much pressure in Andersson-Guyot, 1 993, when 13 ...'it'c8 14 .Jtb2 .Jte7 15 ttJd6+ .Jtxd6 16 exd6 0-0 17 .Jtxd5 exd5 1 8 .l::lx d5 left White with a clear advantage. However, Black's capitulation soon followed as 1 8 ... b6? met with immediate punishment in the form of 1 9 .Jtxg 7! �xg7 20 llg5+ �h8 21 l:th5 etc. 1 0 ttJxe4 i.e6

1 1 e4! 1 56

White should play with energy. After the quieter 1 1 b3 White achieved little in Raet­ sky-Enders, Forchtenberg 1 997, when 1 1 ...ttJb4 1 2 .Jtb2 .Jtxg2 1 3 �xg2 'it'd5+ 14 'iVxd5 exd5 15 ttJd6+ .Jtxd6 16 exd6 �d7 17 .Jtxg7 l:thg8 18 .Jtc3 ttJc6 saw the game drift into equality . 1 1 . . . ttJb6 1 2 ttJd6 + i.xd6 1 3 exd6 This pawn can easily be underestimated. Karpov and Kramnik are known masters in exploiting such circumstances. 1 3 . . . ttJe4? ! A lesser evil is 1 3 ... 0-0 1 4 .Jtd ttJd7 1 5 'iVc2 with a clear advantage for White. 1 4 'We2! Black will soon find that winning the pawn is not much fun. 1 4 . . . ttJxd6 1 5 .l:I.d 1 'fie 7 1 6 i.e3 b6 1 7 i.xe5! bxe5 1 8 'fIxe5 'fIb7 ? ! 1 8 ....Jt b7 1 9 J:lxd6 i s more accurate, al­ though White still has a pawn for nothing. 1 9 'fIxd6 1 -0 Black has had enough.

Game 79 Shipov-Volzhin

Hastings 1997/98 1 d4 d5 2 e4 e6 3 ttJf3 ttJf6 4 g3 dxe4 5 i.g2 e5 6 0-0 ttJe6 7 ttJe5 i.d7 8 ttJa3 ! This has become the modem way t o han­ dle the position, taking over from continua­ tions such as 8 ttJxc6 .Jtxc6 9 .i.xc6+ bxc6 10 'it'a4, e.g. 1 O ... cxd4 (1O ... 'iib 6 1 1 dxc5 .Jtxc5 12 'i'xc4 l:td8 1 3 ttJd2 'iVb5 14 'i'xb5 cxb5 1 5 ttJb3 .Jtb6 was even in Belichev­ Mukhametov, Yalta 1 995) 1 1 'iVxc6+ ttJd7 12 'i'xc4 .Jtc5. Then 13 ttJd2 0-0 1 4 ttJe4 .Jtb6 1 5 .Jtg5 ttJe5 1 6 'i'b5 might look like Black is under pressure but after 1 6 ... 'iYd5! 1 7 'it'xd5 exd5 1 8 ttJd2 f6 1 9 .Jtf4 ttJc4 20 a4 l:He8 it was actually the other way around in Pav­ lovic-Byrne, New York 1 997. Salov­ A.Sokolov, Sochi 1 982 went 1 3 b4 .i.b6 14 .Jtb2 l:tc8 15 'iVb3 0-0 1 6 ttJd2 ttJe5 17 .l:lad 1 'i'f6 1 8 a4 a6 and Black was no worse.

4 . . . dx c 4 5 iL g 2 c 5

Worth further investigation is 8 ttJxc4 cxd4 9 .i.f4 ttJd5 10 ttJd6+ .i.xd6 1 1 .i.xd6. Gelfand-Dokhoian, Sverdlovsk 1 987 contin­ ued 1 1 ...'fib6 12 'iVc1 f6 1 3 ttJd2 �f7 1 4 ttJb3 ttJa5! 1 5 ttJxa5 it'xa5 1 6 'it'c4 and White emerged with good play for the pawn, while 1 1 ...ttJde7 1 2 ttJd2 0-0 1 3 'fib3 .i.c8 1 4 .i.a3 e 5 1 5 .l:tac1 it'c7 1 6 ttJc4 .l:.d8 1 7 .i.c5 .i.g4 18 l:tfe 1 .l:!.d7 19 'ili'a3 l:tad8 20 b4 also resulted in compensation in Romanishin­ A.Sokolov, Cannes 1 998. 8 cxd4 Black can also try 8 ... ttJd5!?, when 9 ttJaxc4 seems harmless, e.g. 9 ... ttJdb4 1 0 ttJxc6 .i.xc6 1 1 .i.xc6+ ttJxc6 1 2 dxc5 .i.xc5 13 .i.f4 0-0 14 it'xd8 .l:!.fxd8 15 ttfd1 f6 1 6 .i.d6 .i.xd6 1 7 lIxd6 � f7 1 8 l:tad 1 � e7 1 9 .l:.xd8 l:txd8 with a draw in Raetsky­ ASokolov, Biel 1 996. Instead White should consider 9 ttJxd7 it'xd7 10 e4 ttJdb4 1 1 d5!? with possibilities of stepping up a gear. 9 ttJaxc4 ttJxe5 For 9 ... .i.e7 see the next main game, which also covers Black's alternatives. 1 0 ttJxe5 'ifb6 . . .

.l:tb4! 'ilif6 1 9 l:hb7+ �c8 20 J:!.c1+! and White wins) 1 8 .l:f.d1 .i.xe3 1 9 .l:tb4 ttJc5 20 J:!.xb6 ttJxa4 21 lIxd8+ .l:txd8 22 .l:[xb7+ �c8 23 fxe3 with a clear plus according to Moro­ zevich. A more sober approach was essayed in Cvitan-Gyimesi, Baden 1 999, when 1 2 ... .i.c5 1 3 .i.f4 .l:.d8 1 4 ttac1 0-0 1 5 b4 'iWxb4 1 6 'it'xb4 .i.xb4 1 7 .i.c7 l:tde8 1 8 .i.xb7 .l:.e7 1 9 i:tb 1 ttJb6 20 .i.xb6 axb6 21 J:!.xb4 l:hb7 22 ':xd4 ended in a draw. There is an interesting alternative in 1 1 b4!?, White making no secret of his quest for the initiative, e.g. 1 1 ...it'xb4?! 1 2 .i.f4 .i.a4 1 3 it'c1 .i.e7 1 4 .l:i.b1 it'c3 1 5 .i.xb7 .l:td8 1 6 ttJc6 .i.xc6 1 7 .i.xc6+ � f8 1 8 it'xc3 dxc3 1 9 i:tfc1 and the endgame indeed looks like a grim ordeal for Black. Brodsky-Ulibin, Bydgoszcz 2001 went 1 1 ...l:td8 1 2 a3 .i.e7 1 3 .i.b2 it'a6 1 4 .i.xd4 0-0 1 5 e 3 .i.b5 1 6 l:1e1 .i.d6 17 a4 .i.xe5 18 axb5 it'd6, when Ulibin gives 1 9 l::t xa7 it'xb4 20 l:txb7 .i.xd4 21 exd4 l::t xd4 22 'iWb 1 ttJd5 23 b6 with a pull for White. 1 1 iLb5 1 2 a4 More challenging might be 1 2 b4!? .l:!.d8 1 3 a4 .i.c6 1 4 .i.xc6+! bxc6 1 5 .l:tc1 nc8 1 6 b5! ttJd5 17 ttJxc6 ttJc3 18 'it'd3 ttJxa4 19 .i.e5 with play for the pawn in Tukmakov­ ASokolov, Lenk 2001 . 1 2 i.a6 1 3 b4! ? Nothing more than equality i s achieved af­ ter 1 3 as 'fibs 1 4 it'xd4 .i.c5 1 5 'iWc3 0-0. 1 3 i.xb4 1 4 a5 . . .

. . .

. . .

1 1 i.f4 Not surprisingly this is not White's only opportunity. In the event of 1 1 ttJxd7 ttJxd7 1 2 it'a4 Black should avoid the risky 1 2 ... 0-0-0?! 1 3 .l:.d1 .i.e7 1 4 .l:.xd4 .i.c5 1 5 l:!.c4 �b8 1 6 '.t> f1 e5, which was seen in To­ palov-Morozevich, Madrid 1 996. Now White could have played 1 7 .i.e3! f5 (17 ... .i.xe3 1 8 157

Th e Ca t a l a n

14 .. :iVb5! After 1 4 ... 'iii'c 5 15 .i.xb7! .i.xb7 1 6 'it'a4+ 'it'ffi 1 7 l:tab 1 White is given some attacking possibilities. 1 5 'it'b3 0-0 1 5 ... .i.c3 1 6 'it'xb5+ .i.xb5 17 .i.xb 7 l:tb8 1 8 .i.c6+ .i.xc6 19 ttJxc6 l:tc8 20 l:tab I ! and White has good compensation for the pawn. 1 6 l:tab 1 g5 1 7 'it'xb4 In the case of 1 7 ttJd3 gxf4 1 8 ttJxb4 l:tab8 White cannot fully justify the sacrifice of a pawn. 1 7 . . . 'it'xb4 1 8 l:txb4 gxf4 1 9 il.xb 7 il.xe2 Also fine is 19 ... .l:!.ab8 20 l:tfb 1 .i.xe2 21 l:txd4 with equality. 20 il.xa8 il.xf 1 21 il.b 7 White should not lose his sense of reality. For example 21 'it'xfl lIxa8 22 gxf4 lld8 even allows Black to play for the win. 21 . . . il.e2 22 l:txd4 fxg3 23 hxg3 il.b5 24 l:tb4 a6 A last try? 25 l:txb5! ? axb5 26 a6 lLld5 27 a7 lLlb6 28 lLlc6 f6 29 lLle 7 + �f7 30 lLlc8 l:txc8 31 il.xc8 b4 32 �f1 b3 33 il.a6 Y:z - Y:z

Game 80 Piket-Van Wely

Monte Carlo 1997 1 d4 d5 2 c4 e6 3 lLlf3 lLlf6 4 g3 dxc4 5 il.g2 c5 6 0-0 lLlc6 7 lLle5 il.d7 8 lLla3 cxd4 9 lLlaxc4 il.e 7 This sound developing move is the most popular these days, but over time a number of alternatives have been tested. 9 ... ttJd5?! does not look sound, and after 10 ttJxc6 .i.xc6 1 1 'iVxd4 ttJb4 12 .i.xc6+ ttJxc6 1 3 'iVc3! Black has a difficult job bringing his pieces into the game. Kasparov-Andersson, Belgrade 1 985 continued 1 3. .. f6 14 .i.e3 .i.e 7 1 5 l:tfdl 'iVc7 1 6 'fib3! g5 1 7 l:tac1 l:td8 1 8 nxd8+ .i.xd8 1 9 'iVxb7! and White was on his way to winning the game. 9 ... l:tc8 10 .i.f4 ttJd5 1 1 ttJxd7 ttJxf4 1 2 .i.xc6 bxc6 1 3 ttJxffi ttJh3+ 1 4 'ittg2 'iVd5+ 1 58

looks natural but 1 5 e4! 'iVxc4 1 6 lIc1 'it'b4 1 7 ttJxe6! fxe6 1 8 'iVh5+ clearly favours White according to Glek. Instead of the check on h3, Ljubojevic-Yusupov, Belfort 1 988 went 1 3. .. 'it'd5 1 4 gxf4 'iYxc4 1 5 lIc1 'it'd5 16 'iVa4! g5 17 fS! exfS 1 8 l:tfdl �xffi 19 l:txd4 'iii'e6 20 l:tc5 �g7 21 'iVas h6 22 l:txfS with a clear lead for White. 9 ... .i.c5 10 'it'b3 0-0 1 1 'iVxb7! does not work out for Black, as was witnessed in Kas­ parov-Deep Blue, Philadelphia (game 2) 1 996: l 1 ...ttJxe5 1 2 ttJxe5 l:tb8 1 3 'iVf3 .i.d6 (13. . .'�Jd5 14 .i.g5! f6 1 5 'ii'g4! gives White all the chances - Sulava) 1 4 ttJc6 .i.xc6 1 5 'ii'xc6 e5 1 6 l:tb I ! and White had the better chances. The computer went on to prove that posi­ tional decision making is not a strength, re­ sulting in a clear advantage for White after 1 6 ...!1b6?! 1 7 'iVa4 'it'b8 1 8 .i.g5 .i.e7 1 9 b4! .i.xb4 20 .i.xf6 gxf6 21 'ilVd7!, when both f7 and h7 proved seriously weak. Black is ad­ vised to turn to the more circumspect 1O .. :iWc8. Rashkovsky-A.Sokolov, Cap d'Agde 1 994 was harmless after 1 1 .i.f4 0-0 1 2 l:tac1 ttJd5 1 3 ttJd3 ttJxf4 14 gxf4 .i.e7 1 5 ttJce5 l:tb8 1 6 l:tc4 .i.f6! 1 7 ttJxc6 1i.xc6 1 8 l:tfc 1 , when the game should end i n a draw after 1 8 ...'iii'd7 1 9 1i.xc6 bxc6 20 'iWa4 l:tb6 21 l:txc6 e5! 22 fxe5 'iVg4+ 23 �fl 'iVh3+ (Rashkovsky). Kharlov-A.Sokolov, St. Pe­ tersburg 1 993 saw multiple exchanges: 1 3 ttJxd7 'it'xd7 1 4 ttJe5 ttJxe5 1 5 1i.xe5 1i.b6 1 6 l:tfd 1 l:tac8 1 7 l:txc8 �xc8 1 8 .i.xd5 'iVxd5 1 9 'iVxd5 exd5 20 .i.xd4 .i.xd4 21 l:txd4 l:td8! 22 e4 �ffi 23 exd5 'it'e7 24 �g2 �d6 25 �f3 l:tc8 26 l:td2 l:te8! and the rook endgame can be drawn but still requires a little effort. 1 0 'it'b3 'tlic8 1 1 il.f4 0-0 1 2 l:tac 1 lLld5 1 3 l:tfd 1 White has more chance of an advantage with 1 3 ttJxc6!? 1i.xc6 1 4 .i.e5, which was successful in Salov-Van Wely, Wijk aan Zee 1 998 after 14 ....i.f6 1 5 l:tfd 1 .i.g5 1 6 f4 .i.e7 1 7 .i.xd4 'iVe8 1 8 l:tfl .i.b5 1 9 e4. 1 3 . . . lLlxf4 After 1 3 ...l:td8 White has the tactical 1 4

4 . . . dx c 4 5 iL g 2 c 5

.ixdS exdS 1 5 4JaS 4JxaS 1 6 it'xdS .ie6 1 7 'i'xaS b 6 1 8 it'a4 'ifb 7 1 9 4Jc6 with a slight plus. 14 gxf4 l:tbS 1 5 iLe4 �c7 1 6 �f3 1 6 4JaS i.d6 1 7 'i!fh3 h6 1 8 .ixc6 bxc6 1 9 ttJaxc6 i.xc6 20 l:txc6 'iYe7 and Black holds his gtound. 1 6 . . . iLf6 1 7 e3 ! ? g6 1 8 exd4 l:tfd8 1 9 tiJe3 If White is too optimistic and plays 1 9 dS?! Black should respond 1 9 ... exdS 20 .ixdS .ie8 21 iLxc6 iLxc6 22 4Jxc6 l:txd1+ 23 l:txdl bxc6 with play against both the b2and f4-squares. 19 . . . i.e8 20 l2lxc6 Not 20 dS? which runs into 20 ... i.xeS 21 fxeS exdS 22 4JxdS 'iVxeS when Black nets a pawn. 20 . . . bxc6 2 1 iLxc6 'iiVa 5 In the case of 21 ...iLxc6 22 l:txc6 'ifb7 White must push with 23 dS!? in order to keep the game flowing. 22 iLxe8 l:txe8 23 a3 l:ted8 24 �e4 l:tb3 25 d5 exd5 26 l2lxd5 iLxb2 27 l:tc7? ! Too optimistic. White should settle for the 'ive' 27 l:tb 1 with the point that after 27 ... l:tbS 28 4Je7+ <:J;;g7 29 l:txd8 'i/t'xd8 30 l:txb2 l:txb2 31 'iWeS+ f6 32 'iWxb2 Black is forced to take perpetual check. 27 . . . 'iiVx c7 ! Black can resist anything - except tempta­ tion. 28 l2lxc7 l:txd 1 + 29 �g2 l:tb8 Black cannot win with 29 ... iLxa3?! because after 30 'iYe8+ iLfB White has 31 4Je6! and the game is drawn after 3 1 ...l:tb8! 32 'it'xb8 fxe6 33 it'xa7 l:tdS, when Black will put the rook on f5 and erect a solid fortress. He would never defend the e-pawn - why should he? 30 'iiVa4 l:tbd8? ! But here there i s a n improvement in 30 ... l:tdd8 31 'iYxa7 iLd4 with play against White's king. The text leads to a draw. 3 1 l2lb5 .l:t 1 d7 32 l2lxa7 iLxa3 33 l2lc6 iLfS 34 tiJxd8 l::t x d8 % - %

White plays 6 0-0 l2lc6 7 �a4

Game 8 1 H . Olafsson-Hjartarsson

Reykjavik 1984 1 d4 d5 2 c4 e6 3 l2lf3 l2lf6 4 g3 dxc4 5 iLg2 c5 6 0-0 l2lc6 7 �a4

7 . . . cxd4 The modern 7 ... iLd7 is considered in the next two games. 8 l2lxd4 �xd4 9 iLxc6 + iLd7 9 ... bxc6?! is best avoided, 10 'ii'xc6+ 'i'd7 1 1 'i'xa8 iLcs 1 2 4Jc3 0-0 1 3 l:tdl �c7 1 4 'iY f3 leaving Black with insufficient compen­ sation for the exchange in Christiansen­ Lhagva, Lucerne 1 982. Indeed after 14 ... iLb7 1 5 .if4 'iYb6 1 6 e4 4Jxe4 1 7 4Jxe4 fS 1 8 b4! iLd4 1 9 iLc7 'iYxc7 20 l:txd4 fxe4 21 'iVe2 White was in the driving seat. 1 0 l::td 1 After 1 0 iLxd7+ 'iVxd7 1 1 'irYxc4 l:tc8 1 2 "iVb3 iLcs Black i s doing fine, but 1 0 .ie3!? looks good. In Gulko-Zsu.Polgar, Bid 1 987 Black failed to solve her opening problems after 10 ... iLxc6 1 1 'iixc6+ 'iYd7 12 'it'xc4 iLe7 1 3 4Jc3 0-0 1 4 l:tfdl 'iVc8 1 5 'iYbs a6 1 6 it'b6 iLd8 1 7 'ilVb4. Also possible is 1O ... 'iYxb2 1 1 iLxd7+ 4Jxd7 12 l:tdl bS 1 3 'iYa6 when White has a strong initiative. However, Black also has two pawns, and after 1 3 . ..l:!.d8 the situation is far from clear. 1 0 . . . 'iiVx d 1 + 1 59

Th e C a t a la n

1 O. . .�xc6 1 1 'iixc6+ bxc6 1 2 J.:txd4 �e7 1 3 .l:!.xc4 cS 14 �f4 0-0 1 5 ttJd2 ttJd7 1 6 ttJb3 a s 1 7 l:ld 1 ttJb6 1 8 l:tc2 left White with the easier ending in Vladimirov-Ghaem Maghami, Kelamabakham 2000. 1 1 'ifxd 1 iLxeS

1 2 liJd2 White has two important alternatives. 1 2 �gS seems only to give Black a tempo, but in fact White gets in f2-f3 for free. Neverthe­ less, there is no advantage: 12 ... ttJe4 1 3 �e3 hS 1 4 f3 ttJf6 1 5 ttJd2 l:td8 1 6 �c1 h4 1 7 ttJxc4 hxg3 1 8 hxg3 lIdS! and Black was no worse in Mochalov-Sturua, Moscow 1 979. The game continued 19 g4 �e7 20 �xa7 ttJxg4! 21 fxg4 .l:th1+! 22 'it>xh1 l:thS+ 23 'it>g 1 !Ih 1+ 24 'it>f2 �h4+ 25 'it>e3 �gS+ 26 'it>d3 Ihc1 27 ':xc1 �xc1 and a draw was agreed. 1 2 'iVc2 demonstrates that Black cannot really hold on to the pawn. Polugaevsky­ Andersson, Moscow 1 98 1 went 1 2 ... �e7 1 3 'ilVxc4 0-0 1 4 ttJc3 l:tfd8 1 5 �e3 ttJdS 1 6 ttJxdS �xdS 1 7 �g4 J::td c8 1 8 �d4 �ffi 1 9 e4 �c6 20 h 4 !Id8 21 �c3 ':d3 2 2 h S h 6 23 ne1 .l:tad8 24 'it>h2 a6 with a slight plus for White. But Andersson has some affection for these ive positions, so he was probably not too uncomfortable here. 1 2 . . . e3?! This helps only White because it wastes a tempo and opens the b-ftle. However, Black anyway has difficulties equalising. In Kas1 60

parov-Andersson, Niksic 1 983 White earned good chances of gaining the full point after 1 2 ... bS 1 3 a4 �e7 1 4 axbS �xbS 1 5 ttJxc4 0-0 1 6 b3 l:[fd8 1 7 'ilVc2 l:tdc8 1 8 �a3 �xa3 1 9 J::txa3 h6 20 'iVc3 �xc4 21 bxc4 l:tc7 22 'it'd4 l:tac8. There is an interesting line in 12 ... hS!? 1 3 ttJxc4 h 4 1 4 �f4, when White should be somewhat better. However, Black managed to do well in Vanheste-Blauert, Groningen 1 989 after 1 4 ... l:td8 1 5 �3 hxg3 1 6 �xg3 ttJe4 1 7 ttJeS ttJxg3 1 8 'iVxg3 �d6 1 9 'it'xg7 �xeS 20 'iVxeS l:!.g8+ 21 'it>f1 �g2+ 22 'it>e1 i.c6 with a draw by repetition. 1 3 bxe3 0-0-0 1 4 'ifb3 iLe5 1 5 liJf3 liJe4 1 S ... ttJg4 1 6 ttJd4 i.dS 1 7 'iVbs i.b6 1 8 a4 is excellent for White. 1 S liJd4 �xd4 Black essays an exchange sacrifice, but it is not enough. In the long-run the queen will reign. Unfortunately for Black 1 6 ... �dS 1 7 'ii'c 2 e S 1 8 i.e3! i s unpleasant anyway ac­ cording to H.Olafsson. 1 7 exd4 iLxd4 1 8 �b1 iLxf2+ 18 ... ttJxf2? loses to 19 �e3. 1 9 >tof 1 h5 19 ...l:td8 20 'iVc2 i.cs 21 l:.b3 .l::tdS 22 i.e3 and White is much better. 20 iLf4 White also has 20 �e3 h4 21 g4 �g3 22 l:tc1 ! �xh2 23 l:txc6+ bxc6 24 'iVa4 with a clear plus in Hjartarsson-Hardarson, Ne­ skaupsstadur 1 984. 20 . . . g5?! 20... .l:td8 21 nc1 �cS is necessary, al­ though White is still doing very well after 22 �e3. 21 iLxg5 h4 22 gxh4 iLxh4 23 'ii'b2 f6 24 iLxh4 �xh4 25 �e 1 >tod7 2s .. Jhh2 loses to 26 l:Ixc6+! bxc6 27 'ilVc2 and White will pick up all the enemy pawns. 2S 't'ia3 Black is lost. 2S . . . liJdS 27 'ifg3 �h5 28 �d 1 iLd5 29 'ii'g 7 + >toeS 30 �e 1 + iLe4 31 'ifg4 �f5+ 32 >toe 1 b5 33 h4 e5 34 h5 �f4 35

4 . . . dx c 4 5 iL g 2 c 5

'i'g2 + 'it>b6 36 h6 ttJe4 37 h7 l:th4 38 h8'i' 1 -0

Game 82 Portisch-Radulov

Buenos Aires 1978 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 iLg2 c5 6 0-0 ttJc6 7 'i'a4 iLd7 This looks like the most appropriate move, after which Black has good chances of equalising. 8 'i'xc4 White needs to play this as he achieves nothing after 8 dxc5 ttJa5! 9 'ilVc2 .ltxc5 1 0 ttJe5 l:tc8 1 1 ttJc3 ttJc6 1 2 ttJxc4 0-0 1 3 t[ d 1 'ife7 1 4 .lte3 .ltxe3 1 5 ttJxe3 .l:i.fd8 1 6 .l:.d2 and a draw was rightly agreed in Razuvaev­ A.Sokolov, Riga 1 985. In Fominyh­ Morozevich, Alushta 1 993 Black was even better after 1 1 i.g5 h6 1 2 .ltxf6 gxf6 1 3 ttJxd7 'iWxd7 1 4 ttJd2 .lte7 1 5 ttJe4 'iWc7 1 6 f4 f5 1 7 ..wc3 J:tg8 1 8 ttJf2 .ltc5. 8 . . . cxd4 This move lets White develop a bit too freely, and Black is under some pressure for a long time. 8 ... .l:!.c8 9 dxc5 'ilVa5 1 0 ttJbd2 'ifxc5 1 1 b3 i.e7 1 2 .ltb2 'iWb6 1 3 'iWf4 0-0 14 ttJc4 "iic 7 1 5 l::tfd l 'iWxf4 1 6 gxf4 l:tfd8 1 7 ttJfe5 i.e8 1 8 l:txd8 ttJxd8 1 9 lId 1 gave Black some problems to solve in Ftacnik­ Unzicker, 1 995. These were minor, but problems all the same. The energetic 8 ... b5 is considered in the next game. 9 ttJxd4 l:tc8 1 0 ttJc3 ttJxd4 Korchnoi-Karpov, Candidates Match (game 1 7) , Moscow 1 974 also proved awk­ ward for Black after 1 0 ...'it'a5 1 1 .l:!.dl i.e7 1 2 ttJb3 'ifc7 1 3 ttJb5 (1 3 .ltf4!? e 5 1 4 .ltg5 i.e6 1 5 'it'a4 is interesting and looks a shade bet­ ter for White) 13 ... 'iib 8 14 ttJc5 a6 1 5 ttJxd7 ttJxd7 16 ttJc3 ttJde5, when 17 'ifb3 0-0 1 8 .ltf4 favours White according to Botvinnik. 1 1 'i'xd4 iLc5 1 2 'i'h4 iLc6 1 3 l:td 1 ! 1 3 i.g5 invites 1 3.. .'�d4! (the equaliser), when Karpov-Piket, Monaco (blindfold)

1 998 continued 1 4 i.xc6+ l:txc6 1 5 .ltxf6 'ilVxf6 1 6 'ii'a4 0-0 1 7 l:tadl a6 1 8 nd7 b5 1 9 'iie 4 .tIfc8 2 0 e 3 h 6 21 l:tfdl i.b6 with a level game. 1 3 . . . 'i'b6? A tactical error that is punished severely. 1 3 .. .'ilVa5 is the alternative. Benko-Peters, USA, 1 979 continued 14 .ltxc6+ ':'xc6 1 5 i.g5 .lte7 1 6 ttJe4 'ili'e5 1 7 ttJxf6+ .ltxf6 1 8 i.xf6 gxf6 1 9 'ifb4 l:tb6 20 'it'a4+ 'itt f8 21 l:td2 and Black still had much work to do in order to feel more secure. 1 4 i.h6 works less well here than in the main game for after 1 4 ... 0-0 1 5 .ltxc6 l:!.xc6 1 6 i.xg7 Black has 1 6 ... i.xf2+! 17 'itt x f2 'itt xg7 1 8 l::t d 3 :ctc5 with equaliry, as In Csom-Peters, Hastings 1 978/79. 14 iLxc6 + l:txc6 1 5 iLh6!

Breaking open Black's kingside. 1 5 . . . gxh6 15 ... .ltxf2+? 16 'ittg2 0-0 17 i.xg7! gives White a winning attack. Alternatively 1 5 ... i.f8 1 6 ':d2 e5 1 7 i.e3 'it'a6 1 8 .l:!.ad l i.e7 1 9 'it'g5! 0-0 20 'iWxe5 won White a pawn in Ribli-Ljubojevic, Buenos Aires 1 978. 1 6 'i'xf6 0-0 1 7 ttJe4 'i'b4 1 8 'ii'e 5 iLe 7 1 9 a3 'ii'b 6 20 .l:!.d7 iLg5 20 .. .f6?! accelerates defeat, White winning with 21 'i¥h5 'iWxb2 22 l:txe7 'it'xa1+ 23 'ittg2 when Black is mated - 23 ... f5 24 'iWxh6 and 1i'xh7 decides. 21 b3 ! Accurate. After 21 ttJxg5 ':c5 22 'ife4 161

Th e Ca t a l a n

hxg5 2 3 lIxb 7 White i s a pawn up, but no more. 21 . . :ii'x b3 After 21 ...f6 22 �2 lId8 23 1Ixd8+ 'it'xd8 24 h4 f5 25 hxg5 fxe4 26 gxh6 Black is in serious trouble. 22 tDxg5 hxg5 23 'iYxg 5 + �h8 24 l:tad 1 'iYc2 25 'iYf6 + �g8 26 l:t 1 d4 "ii'g 6 26 ...lIc4 27 lIxf7! and Black is mated. 27 "ii'f 3! h5 28 l:txb7 e5 29 l:th4 l:td8 30 'ii'x h5 'ii'x h5 31 l:txh5 f6 32 l:txa7 l:td2 33 e3 l:tc 1 + 34 �g2 l:tdd 1 35 l:tf5 1 -0

Game 83 Korchnoi-Ivanchuk

Istanbul 2000 1 d4 d5 2 c4 e6 3 tDf3 tDf6 4 93 dxc4 5 i.g2 c5 6 0-0 tDc6 7 'ii'a4 i.d7 8 'iYxc4 b5!?

The most energetic move. 9 'ii'd 3 Black should be advised against 9 'iVxb5? tLlxd4 10 tLlxd4!? (1 0 �d3 i.b5 clearly fa­ vours Black) 1 O ... �xb5 1 1 tLlxb5 tLld5 1 2 tLl l c3 a 6 1 3 tLlxd5 axb5 1 4 tLlc3 b 4 1 5 �c6+ We7 1 6 tLle4 .l:tc8 and there was clearly not enough compensation for the queen in Hodgson-Karolyi, Kecskemet 1 988. 9 . . . l:tc8 Completing queens ide development and stepping off the long diagonal. With 9 ... c4 Black loses options: 10 'it'd l lIc8 1 1 tLle5 1 62

i.d6 1 2 tLlxd7 'ifxd7 1 3 tLlc3 a6 1 4 d5 exd5 1 5 tLlxd5 tLlxd5 1 6 'it'xd5 'it'e6 1 7 lIdl ..te5 18 a4 gave White a slight plus in Stefanova­ Zhu Chen, Groningen 1 999. 10 tDc3 After this move Ivanchuk successfully tries for the initiative. It seems more natural to play 1 0 dxc5 �xc5 1 1 tLlc3, when Black has more than one path to follow. 1 1 ...0-0 12 �g5 (1 2 tLlxb5? tLlb4 1 3 'ifc4 ..txf2+ and Black wins) 1 2 ... tLlb4 1 3 i.xf6 gxf6 1 4 'Yid2 �c6 1 5 'it'h6 .l:te8 1 6 lIadl ..tfS 1 7 'iVh5 'ike7 1 8 a3 �xf3 (1 8 ... tLld5 1 9 tLld4) 1 9 .i.xf3 with a winning position for White in Anasta­ sian-Kaidanov, Lucerne 1 997. l 1 ...tLlb4 12 'iVd2!? 'iVb6 (1 2 ... tLlbd5 13 tLlxd5 tLlxd5 14 a3 0-0 15 b4 �b6 16 .i.b2 and White is better placed thanks to the co­ ordination of his bishops) 1 3 tLle5 l:td8 1 4 a3 tLlc6 1 5 tLld3 0-0 1 6 'ifg5 e5 1 7 tLlxc5 'ikxc5 1 8 'iVh4 favoured White (..tg5 is coming) in Vladimirov-De La Villa, Marchena 1 990. l 1 ...b4 puts the question to the knight. In Hansen-Van Der Sterren, Kerteminde 1 991 White chose 12 tLle4 tLlxe4 13 �xe4, when there followed 1 3. .. tLle7 (13. .. 0-0 1 4 l:tdl 'ife7 1 5 �g5 f6 1 6 �e3 gives White a little something as Black's queenside pawns are potentially weak) 1 4 tLle5 ..tb5 1 5 ..te3 0-0 1 6 lIfd 1 'iVb6 1 7 ..txc5 lIxc5 1 8 tLld7 �xd7 1 9 lIxd7 tLld5 20 �h4 h6 and Black was no worse. After 21 l:!.dl l:tc2 22 �xd5 l:txe2!? 23 'iVd4 exd5 24 l:!.xa7 'it'xd4 25 l:txd4 l:!.xb2 the game was drawn. Khalifman-Dokhoian, 1 99 1 saw instead 1 2 tLlb5 0-0 1 3 tLld6 tLld4! 1 4 ..tf4 tLld5 1 5 �e5 tLlxf3+ 16 i.xf3 'ifg5 1 7 tLlc4 ..tb5 18 h4 'ike7 1 9 b3 lIfd8 and Black had been very successful in the opening. White has problems getting his pieces into play, and the queen is struggling to contribute. 1 0 . . . cxd4 1O ... c4 1 1 �1 (1 1 'tidl ! b4 12 tLla4 'ifa5 results in an interesting position with chances for both sides) 1 1 ...b4 1 2 tLle4 tLld5 1 3 tLleg5 h6 1 4 e4 tLlf6 1 5 tLlh3 �e7 1 6 l:!.dl 'ifa5 was

4 . . . dx c 4 5 iL g 2 c 5

unclear in Kozul-Bunazovic, Ljubljana 1 993. 1 1 ttJxd4 ttJe5 Black presses for the win. After the sim­ pler 1 1 ...ttJxd4 12 'iWxd4 i.c5 1 3 'iVd3 b4 the situation is comfortable, the subsequent 1 4 ttJe4 ttJxe4 1 5 i.xe4 'iVb6 1 6 i.f4 i.b5 1 7 'i'B 0-0 (1 7 .. .f5!? i s messy) 1 8 l:tad1 i.d4 1 9 l:td2 e 5 being fine for Black in Inkiov­ Adamski, Tbilisi 1 986. 1 2 'i!Vd 1 b4 Black needs to maintain the high pace. Af­ ter 1 2 ... 'iYb6 1 3 a4 b4 1 4 ttJe4 White might be a little bit better. 1 3 ttJe4 ttJd5 Black can equalise (again) with 1 3 ... ttJxe4 14 i.xe4 i.e7 1 5 i.f4 ttJc4 1 6 b3 ttJd6 1 7 ..It B 0-0 etc. 14 ttJg5 Perhaps this is careless. After the more natural 14 i.f4 ttJxf4 1 5 gxf4 ttJg6 16 e3 'iVb6 the position is level. 1 4 . . . iLe7 1 5 iLxd5?! A decision that White will come to regret. After 1 5 e4 ttJb6 1 6 h4 h6 1 7 ttJgB ttJxf3+ 1 8 'iWxB .i.f6 1 9 nd1 i.a4 20 ttJb3 'fie7 Black is doing well. In the event of 1 5 ttJf5!? it is tempting to go for 15 ... l:t.xc1 !? 16 l:txc1 i.xg5, when 1 7 f4 'iVb6+ 1 8 'it>h 1 ttJg4 1 9 fxg5 ttJge3 20 'iVb3 ttJxfl 21 ltxfl 0-0 seems fine for Black. Meanwhile there is 1 5 ... exf5 1 6 'iYxd5 l:tc5 1 7 'ii'd4 0-0 1 8 i.f4 .i.c6 1 9 ..Itxe5 .i.xg2 20 'it>xg2 l:td5 21 'iWf4 i.xg5 22 'ii'x fS l:te8 23 f4 i.f6 24 .i.xf6 gxf6 25 �g4+ 'it>h8 with compensation for the pawn ac­ cording to Ivanchuk, and Black can also play the position in another fashion - with 23 .. .f6!?, for example. 1 5 . . . exd5 1 6 ttJgf3 ttJc4 1 7 'ifd3 0-0 1 8 J:td 1 J:te8 1 9 J:tb 1 ? ! White should not tolerate the knight on c4: 1 9 b3 ttJd6 20 ttJe5 (20 a3 ttJe4 21 axb4 i.xb4 22 lha7 l:tc3 23 'iVb1 .i.c5 clearly favours Black, 24 l:ta2 l:txB! 25 ttJxB i.xf2+ 26 'it>g2 ttJc3 completely ruining White's structure) 20 ... ttJe4 21 ttJxd7 'fixd7 22 .i.b2 i.f6 and White has good chances of main-

taining the balance. 1 9 . . .iLf6 20 b3 ttJd6 21 iLd2 'it'b6 22 J:tbc 1 ttJe4 Black is better. 23 iLe 1 h6 24 Wg2 a5

25 h3 J:tc5 26 J:txc5 'it'xc5 27 e3 J:tc8 28 g4 'ifb6 29 ttJg 1 iLxd4 30 exd4 A better try is 30 'iWxd4 'iVxd4 31 l:txd4 i.e6, although the endgame is quite grim. 30 . . . iLb5 31 'ifb 1 'ifc6 32 iLd2 �f6 33 iLe3 J:tc3 ! Eliminating the dark square defender. 34 'iWb2 J:txe3! 35 fxe3 �h4 36 J:td2 'it'g3 + 37 Wh 1 'ifxe3 38 �c2 ttJg3 + 39 Wg2 iLf1 + 40 Wh2 ttJe4 4 1 �c 1 'ifg3 + 4 2 Wh 1 iLd3 4 3 J:tc8 + W h 7 44 J:th8 + ! ?

44 . . . Wg6! 44 ... 'It>xh8? 45 �c8+ allows White to es­ cape with a draw. 45 'iWc6 + ttJf6 46 J:te8 Wh7 0-1 1 63

Th e Ca t a l a n

Summary The position after S ... cS 6 0-0 ttJc6 7 'it'a4 cxd4 8 ttJxd4 'it'xd4 9 .i.xc6+ .i.d7 10 l:txdl 'it'xdl + 1 1 'it'xdl .i.xc6 remains very important for the overall evaluation of the S ... cS system. It used to be thought that Black has two plans of approximately equal value - to play for sim­ plification in the hope of achieving a draw or to strive for an initiative on the kingside by ad­ vancing the h-pawn. Now it has been established that the aggressive approach fails. However, the 'drawing' fortress seems quite feasible, which partly undermines the reputation of 7 'it'a4. The positive contribution 7 'ii'a4 received from Portisch-Radulov (Game 82) was only tempo­ rary as after 7 ... .i.d7 8 'it'xc4 cxd4 9 ttJxd4 J:Ic8 10 ttJc3 ttJxd4 1 1 'ii'xd4 .i.cs 12 'ii'h4 .i.c6 1 3 l:tdl Black made a n error with 1 3. . .'iVb6?; 1 3 . . .'iIi'aS steers the game towards equality. In the 1 990's the following variation became fashionable for Black: 8 ... bS 9 'iVd3 l:tc8, when Korchnoi-Ivanchuk (Game 83) illustrates that 1 0 ttJc3 is hannless for Black. After 1 0 dxcS .i.xcs 1 1 ttJc3 both 1 1 ...0-0 and 1 1 ...ttJb4 are not quite satisfactorily for Black, but 1 1 ...b4 se­ cures a level game after 12 ttJbS and 12 ttJe4. Consequently 7 'ili'a4 is no longer considered dangerous for Black. Therefore White's attention has turned to 7 ttJeS. Now on 7 ... .i.d7 White should not hurry with 8 ttJxc6 .i.xc6 9 .i.xc6 bxc6 1 0 'iVa4 because Black has time to get his game going despite the ruined pawn structure. Also not so popular these days is the complex continuation 8 ttJxc4 cxd4 9 .i.f4 ttJdS 10 ttJxd6 .i.xd6 1 1 .i.xd6. The main line is 8 ttJa3 cxd4 9 ttJaxc4, when White has achieved a lead in development. Nevertheless the (extra) d4-pawn restricts White a little. In the modern line 9 ... ttJxeS 1 0 ttJxeS 'iVb6 White must consider 1 1 .i.f4 .JibS 1 2 b4 (fukmakov) and 1 1 b4 (Brodsky), and in the case of 9 ... .i.cS 1 0 'iVb3 0-0 he can probably grab the b7-pawn. Thus it is preferable for Black to play 1 0 ...'iIi'c8 with chances of gradually equalising, as is demonstrated in Piket-Van Wely (Game 80). 1 d4 d 5 2 c4 e6 3 t'Llf3 t'Llf6 4 g3 dxc4 5 iLg2 c5 (D) 6 0-0 6 'iVa4+ .i.d7 7 'ili'xc4 i.c6 - Game 76 6 . . . t'Llc6 6 ... cxd4 - Game 77 7 t'Lle5 7 'iVa4 7 ... cxd4 - Game 8 1 7 . . ..i.d7 8 'iVxc4 (D): 8. . .cxd4 - Game 82; 8. . .b S - Game 83 7 . . . iLd7 7 ... ttJxeS - Game 78 8 t'Lla3 cxd4 9 t'Llaxc4 (D): 9 ... ttJxeS - Game 79; 9 ... i..e 7 - Game 80

S. . . cS 1 64

8 'iix c4

9 t'Llaxc4

CHA PTER EIGHT

I

4 . . . dxc4 5 .,tg2 b 5

1 d4 d 5 2 c4 e 6 3 ttJf3 ttJf6 4 g 3 dxc4 5 �g2 b5 This is a genuine gambit, Black immedi­ ately protecting his extra pawn despite it not yet being attacked. In return for the pawn White has a space advantage in the centre and a lead in development. S ... bS has always been considered a very provocative continuation, being seldom seen in international practice. However, it is impor­ tant to note that other systems often transpose to S ... bS lines. For example if, in the S ... c6 variation, White plays 6 0-0, then the best answer is considered to be 6 ... bS with a trans­ position to S ... bS. And S ... a6 and S ... bS are also closely related - the only real difference being that S ... a6 is more flexible and allows Black to avoid the S ... bS lines if he so desires. After S ... bS there are two main lines. The first is the tactical line introduced by Gulko 6 a4 c6 7 axbS cxbS 8 ttJeS ttJdS 9 ttJc3 ii.b4 1 0 0-0 .i.xc3 1 1 e4! (Game 84) . In the 1 980's White achieved a number of brilliant victo­ ries with this system, mainly in Soviet tour­ naments. But the games somehow failed to follow standard, rational paths and did not feature precise defence and accurate calcula­ tion. Not surprisingly Black began to find ways in which the defence could be strength­ ened and prospects improved, prompting an

evaluation of the position as - at best - un­ clear. Thus in the struggle between the sword and the armour it would seem that the ar­ mour should triumph, and this forcing line has practically vanished from international practice. Now we are more likely to see the albeit less inventive 6 a4 c6 7 0-0 .i.b7 8 ttJeS (Games 8S & 86), holding back the b 1 knight, a s i n the S. . .a 6 lines.

Game 84 Balashov -Beliavsky

Kiev 1986 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 �g2 b5 6 a4 c6 7 axb5 cxb5 8 ttJe5 ttJd5 9 ttJc3

1 65

Th e Ca t a l a n 9 . . i.b4 .

9 ... �b7?! is a dubious alternative as 1 0 ttJxb5 ttJc3 1 1 ttJxc3 .i.xg2 1 2 l:.g 1 (polo­ vodin) might net Black the light-squared bishop but the price of time is a more impor­ tant factor. White exerts pressure on the queenside and is clearly better. 9 .. .f6 looks risky. Indeed Black was pun­ ished in Hackel-Rausch, 1 996, which continued 10 e4!? ttJxc3 1 1 'Yi'h5+ g6 1 2 ttJxg6 hxg6 1 3 'iix h8 'iVxd4 1 4 1t.h6 ttJd7 1 5 0-0 f5?! 1 6 'Yi'g8 ttJxe4 17 'Yi'xe6+ Wd8 1 8 l:i.ad 1 'Yi'c 5 1 9 1t.xe4 fxe4 20 l:i.d5 'Yi'b6 21 1t.g5+ Wc7 22 1t.f4+ 'it>d8 23 'iVe5 1t.b7 24 .i.g5+ and White won. An improvement is 15 ... ttJe2+ 16 'it'h1 e5 17 .i.xfS ttJxfS 1 8 'tWxf6 when Black has two minor pieces for a rook but White enjoys more activity and has the safer king. 1 0 0-0 i.xc3 1 1 e4!

Gulko's contribution, which is preferable to 1 1 bxc3 ttJxc3 1 2 'iWd2 ttJd5 1 3 1t.a3 f6 \vith an edge for Black as White's knight is driven back and there are no tricks with 'iVh5. 1 1 . . . i.xb2! The best of the candidate replies. 1 1 ...�b4? is the worst in view of 12 exd5 exd5 13 ttJxf7! Wxf7 14 'iVh5+ and White wins. 1 1 ...ttJf4 doesn't look too convincing either, Gulko giving 1 2 �xf4 1t.xb2 1 3 l:i.b 1 .i.xd4 1 4 ttJxc4! bxc4 1 5 e5 etc. 1 1 ...ttJe7 proved disastrous in Gulko1 66

Mikhalchishin, Volgodonsk 1 98 1 , Black re­ sponding to 1 2 bxc3 with the poor 1 2 ... f6?, running into serious trouble after 1 3 'iVh5+! g6 1 4 ttJxg6! ttJxg6 1 5 e5 ttJc6 1 6 .i.xc6+ .i.d7 1 7 .i.xa8 'iYxa8 1 8 �h6 Wf7 1 9 exf6. Of course this can be avoided with 1 2 ... 0-0, but then 1 3 1t.a3 f6 14 ttJxc4! bxc4 1 5 e5 ttJd5 16 .i.xfS 'iWxfS 17 it'a4 is pleasant only for White who, despite having a rook for two minor pieces, has the open flies on the queenside with which to generate play. In the event of 1 1 ...ttJf6 1 2 bxc3 1t.b7 1 3 1t.a3 'iVc7 1 4 'iVe2 ttJbd7 1 5 ttJxd7 'iVxd7 1 6 d 5 White has compensation for the pawn due to the dark-squared bishop. 1 2 exd5 The alternative is 1 2 .i.xb2 ttJe7 1 3 d5! 0-0 1 4 1t.a3, when 1 4... ttJd7? 1 5 ttJxf7 �xf7 1 6 dxe6+ �xe6 1 7 'iih 5 g6 1 8 'iVxh 7 ttJe5 1 9 l:i.ad 1 ttJd3 saw Black walking o n thin ice in Polovodin-Zhelnin, Moscow 1 983. In fact the game ended 20 .i.h3+ We5 21 'iVg7+ Wxe4 22 ..ig2+ Wf5 23 ..ixe7 "iVe8 24 .i.h3+ 'it'e4 25 l:i.fe H ttJxe1 26 'iVd4+ and Black was mated. Instead of the faulty 1 4 ... ttJd7?, 011 assessed the situation after 14 .. .f6 1 5 d6 ttJec6 1 6 ttJxc6 ttJxc6 1 7 d 7 b4! 1 8 dxc8'iV l:i.xc8 as unclear, thanks in no small part to Black's has three connected ed pawns. 1 2 . . . i.xa 1 1 3 i.a3

1 3 . . . a5 13 ... exd5?! is dangerous in view of 1 4 'iWh5, e.g. 1 4. . .g6 (1 4. . ...ie6 i s punished by

4 . . . dx c 4 5 iL g 2 b 5

Chernin's 1 5 liJxf7! g6 1 6 �f3! l:tfB 1 7 l:tel Wd7 1 8 l::t x e6, when White wins) 15 �f3 and White has a strong initiative, e.g. 1 5 .. .f6 1 6 J:tel ! 1Le6 1 7 1Lh3! f5 1 8 liJxg6 hxg6 1 9 J:txe6+ <j;} f7 20 .ixf5 .ixd4 21 1Lxg6+! and Black is crushed. 14 'iVg4 Also possible is 1 4 dxe6 1Lxe6 1 5 .ixa8, when Kengis-Mejster, Toljatti 1 985 contin­ ued 1 5 ... .ixd4? 1 6 liJc6! .ixf2+ 17 'itxf2 'i'c7 1 8 .id6! 'iVb6+ 1 9 'iVd4 'iixd4+ 20 tL'lxd4 b4 21 l:tal ! 'it>d7 22 .if4 l:tc8 23 .ib7! J:td8 24 l:txa5 'ite8 25 liJxe6 fxe6 26 l:b5 and White won. Black's queenside pawns were inferior to White's active pieces. Dunnington gives 1 5 ... b4 1 6 'iVxal 0-0 1 7 1Lel with at­ tack, while 1 5 ...�xd4 1 6 �xal �xal 1 7 J:txa 1 b4 1 8 .iel 0-0 1 9 ':xa5 .l:td8! i s analysis by Kengis. 14 . . . b4 After 1 4 ...g6!? 1 5 'i!kf4 f6 1 6 .l:hal l:ta6 1 7 tL'lg4 e 5 1 8 dxe5 .ixg4 1 9 'iVxg4 b 4 20 1L c 1 c3 21 d6 (011) the siruation is highly complex. 1 5 'iVxg7 �f8 1 6 .l:!.xa 1 .l:!.a6 16 ... bxa3 17 dxe6 .ixe6 18 .ixa8 'iixd4 19 l:tb 1 liJd7 20 .ic6 and now Chernin­ Yudasin, Sverdlovsk 1 984 continued 20 .. :ifd6? 21 .ixd7+ .ixd7 22 liJxc4 'iic 5 23 J:tb8+ 1Lc8 24 l:.xc8+ and White won. Black should play 20 ... a2! 21 .ixd7+ .ixd7!, sacri­ ficing and relying on the dangerous a2-pawn. Then after 22 .l:tb8+ 'it>e7 23 �xfB+ 'iit e 6 24 'i'xf7+ 'it>xe5 25 �e7+ 1Le6 26 nbS+ <j;}e4 27 'i'xe6+ c;t>d3 the battle rages on with chances for both sides. 17 iLc 1 exd5 1 7 .. .f6 is also possible, Balashov giving 1 8 tL'lxc4 l:. f7 1 9 'iih 6 exdS 20 liJe3 1Le6 2 1 .to. White is the exchange down but has attacking prospects against Black's exposed king, while Black has rwo dangerous con­ nected ed pawns. 18 iLh6! �xh6?! Better is 18 ...�e7 1 9 l:tel .ie6! 20 'iixh7 tL'ld7 21 liJxd7 'iitx d7 22 .ixdS when White has clear compensation for the exchange in

this open position but Black's dangerous pawns contribute to an evaluation of unclear. 1 9 'ilfxh6 iLe6 20 .l:!.e 1 'ilfe7 21 ttJxc4 21 liJxf7!? 'iVxf7 22 .l:txe6+ 'iit d7 23 .l:td6+ 'iit c 7 24 f4 and Black's slightly troubled king confers White the more pleasant game. Again Black's pawns are dangerous! 21 . . . dxc4 22 d5 �d7 23 "ii'e 3! White has a strong attack. 23 . . . 'ilfd6 24 'ilfa7 + 'ilfc7 25 dxe6 + fxe6 26 'iVd4+ �c8? Better is 26 ... 'iid6 27 'iig7+! 'iie 7 28 lieS but White still has a strong initiative. 27 .l:!.xe6

White is winning. 27 . . . c3 28 iLh3 �b7 29 'iVd5 + �a7 Or 29 ... liJc6 30 1Lg2 l:tc8 3 1 l:txc6 �xc6 32 'iVbS+ and White wins. 30 iLg2 ttJa6 31 .l:!.c6 ! 1 -0

Game 85 Mochalov-Spirin

Correspondence 1986 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 iLg2 b5 6 a4 c6 7 0-0 iLb7 8 ttJe5 After 8 liJc3 b4 9 liJb 1 cS 10 liJbd2 c3 1 1 bxc3 bxc3 1 2 liJc4 cxd4 1 3 �xd4 lixd4 1 4 liJxd4 .ixg2 1 5 'it>xg2 1Lb4 Black had no problems in Grinshpun-Rabinovich, Tel Aviv 1 997. Moiseenko-Kharlov, St. Peters­ burg 1 998 took a different course: 8 ... a6 9 e4 liJbd7 10 e5 liJd5 1 1 liJgs liJ7b6 1 2 �h5 167

Th e Ca t a l a n

�d7 1 3 a s 4Jxc3 1 4 bxc3 4JdS 1 5 jixdS cxdS 16 4Jxh7 0-0-0 17 4Jxffi l:tdxffi with chances for both sides. S . . . ltJd5 Also possible is 8 ...'iVb6 9 b3 cxb3 1 0 'it'xb3, offering the d4-pawn. I n fact after 1 0 ... 'ii'xd4 1 1 jib2 �6 1 2 axbS cxbS 1 3 jid4 'iixd4 1 4 jixb7 �xal 1 5 �xbS+ 4Jbd7 16 4Jxd7 4Jxd7 17 jixa8 White has compen­ sation for the pawn due to his lead in devel­ opment. Khalifman-Sveshnikov, Elista 1 996 went 1O ... 4Jbd7 1 1 jie3! cS, prompting White to open up the centre with 1 2 4Jxd7 4Jxq.7 1 3 dS!, when there followed 1 3 ... bxa4 1 4 'iVxa4 exdS (1 4 ... jixdS 1 5 4Jc3 jixg2 1 6 �xg2 'it'b7+ 1 7 �gl .l:!.d8 1 8 .l:tabl 'iVc8 1 9 4JbS! a6 20 4Ja7 'iVc7 21 I:.fd l ! 'iVxa7 22 'iVc6 'iVa8 23 I:.b7 'ii'c 8 24 nc7 is clearly better for White - Khalifman) 1 5 4Jc3 with an excel­ lent game for White. After I S ... d4 1 6 4JdS jixdS 1 7 jixdS .l:!.d8 1 8 jif4 'iVf6 1 9 'iVbs jid6 20 l:ta6 l:tb8 21 jib7! this had been translated to a decisive advantage. Finally there is 1 0 ... a6 1 1 lldl 4Jbd7 1 2 4Jc3 4JxeS 1 3 dxeS 4JdS 1 4 jixdS!? exdS 1 5 axbS axbS a s in Raetsky-Z.Yarga, ZUrich 2000, when White's development lead was balanced out by Black's consolidation in the centre. After 1 6 jie3 �d8?! 1 7 jia7! 'iVc7?! 18 e4 b4 19 exdS! bxc3 20 d6 �d7 21 jicS! White stood better despite being a piece down. Black should play 1 6 ... jicS 1 7 .l:!.xa8+ jixa8 1 8 l:tal jib7 1 9 4Ja4 bxa4 20 �xb6 jixb6 21 jixb6 �d7 22 �xa4 l:ta8 with equality. 9 b3 9 4Jc3 a6 transposes to 8 ... a6 9 4Jc3 4JdS 9 e4 4Jf6 1 0 dS!? (1 0 4Jc3 leads to equality after 1 O ... b4 1 1 4Je2 cS) l O ... cxdS 1 1 exdS 4JxdS, when Raetsky-Sofrigin, Leukerbad 2002 continued 1 2 4Jc3!? b4 1 3 4JxdS jixdS 14 jigS! jie7 (14 ... 'ii'xgS 1 5 jixdS iVxeS 1 6 jixa8 'iVxb2 1 7 .l:!.c1 favours White) 1 5 jixe7 �xe7 16 'it'hS g6 17 'ii'h4+ f6 1 8 jixdS exdS 1 9 nfe1 ! gS 20 �h6 and White had a devas1 68

tating attack: 20 ... fxeS 21 �xeS+ �f7 22 l::t f5+ and White won. In Vladirnirov­ Gedevanishvili, Daugavpils 1 978 the queen­ side saw all the action: 1 2 axbS jie7 1 3 4Jxc4 0-0 1 4 4JaS 'it'b6 1 5 4Jxb7 'ii'x b7 1 6 4Jc3 l:td8 1 7 �3 4Jc6 1 8 4JxdS exdS 1 9 l::td l 4Jb4 20 jid2 'it'xbS 21 jixb4 'ii'xb4 22 'ii'xb4 jixb4 23 jixdS nab8 24 .l:txa7 and in this opposite-coloured bishop ending White had good winning chances. 9 . . . cxb3

1 0 'i'xb3 Also possible is 10 axbS cxbS 1 1 'ii'x b3 a6 12 e4 4Jf6 13 dS!, e.g. 13 ... exdS 14 exdS �xdS 1 5 �e3 �e7 1 6 jixdS 4JxdS 1 7 'iVe4 'ii'e 6 1 8 l:tdl with comfortably enough for the two pawns. 1 3 ... �cS!? is interesting but appears inadequate in view of Agzamov's 14 dxe6 fxe6 15 'iixe6+ �e7 16 �xe7+ �xe7 1 7 4Jc3 0-0, when White is slightly better because the centre pawns are more mobile than Black's a- and b-pawns. 1 0 . . . a6 1 1 ltJc3 ltJd7 1 2 axb5 12 4JxdS exdS 13 e4 4JxeS 14 exdS!? cxdS! 1 5 dxeS �cS 16 axbS 0-0 17 b6!? with a slight advantage for White. 1 2 . . . axb5 1 3 J:txaS .ixaS 1 4 ltJxd5 exd5 1 5 e4 ltJxe5 1 6 dxe5 d4? ! 1 6 ... dxe4 1 7 �xe4 �e 7 1 8 �d 1 gives White compensation by way of the usual development lead. 1 7 e6 f6 Mochalov gives 1 7 .. .fxe6 1 8 'it'xe6+ 'ii'e7

4 . . . dx c 4 5 iL. g 2 b 5

1 9 �a2! .1b7 20 e5! c5 21 .1g5! 'iixg5 22 .1xb 7 with an excellent game for White. 1 8 e5!

After 26 ... 'ii'e 7 27 'it'xe5 'ii'xe6 28 'iix e6 l:txe6 29 i.fS White wins a piece and the game. 27 'ii'g 7 + 'ii'e 7 28 'ii'x e5 + Wb6 29 'ii'x d4+ Wc7 30 'ii'e 5+ Wb6 31 l:tb8 + 1 -0

Game 86 Filippov-Pridorozhni

Ekaten'nburg 1997

1 8 . . . fxe5 18 ... i.e 7 19 exf6 .1xf6 20 .1a3 .1e7? 21 'it'f3 l::t f8 22 'it'xc6+! is a nice little trick. 1 9 iL.e4! iL.e7 1 9 ...'it'd6 20 .1g5 .1e7 21 l::t a 1 gives White the initiative as Black has no time to castle: 2l ...i.b7 22 l:ta7 .1xg5 23 .l:!.xb7 etc. 20 'iff3 'ifd6 20 ... .1f6 21 .1xc6+ <3;e7 22 .1a3+ <3;xe6 23 l::t c 1 and White is better. Black's king has seen better days. 2 1 'iff7 + Wd8 22 iL.g5! iL.b7 23 l:ta 1 iL.c8 23 ... l::te 8 24 l:ta7 'it'c7 25 .1xe7+ l:txe7 26 llxb7! l::tx f7 27 exf7 'iVd6 28 .1xc6 and White wins. 24 J:ta8 iL.xg5 25 'ifxg7

25 .. .1:1e8 26 'ifxg5 + Wc7

1 d4 d5 2 c4 e6 3 ltJf3 ltJf6 4 g3 dxc4 5 iL.g2 b5 6 a4 c6 7 0-0 iL.b7 8 ltJe5 a6 9 ltJc3 Also possible is 9 b3 cxb3 10 i.b2. Then 1O ... �6 1 1 �xb3 ttJbd7 12 ttJxd7 ttJxd7 1 3 tDd2 .1e7 1 4 d5! saw White exploit his supe­ rior development after 14 ... cxd5 1 5 i.xg7 l::tg8 in Tukmakov-Korchnoi, Leningrad 1 973, Tukmakov giving 1 6 .1h6!? l::tg6 1 7 .1e3 .1c5 1 8 .1xc5 ttJxc5 1 9 'itb2 as unclear due to Black's weak dark squares. Alterna­ tively, Black sorted out his kingside in Orlov­ Akopian, Minsk 1 990, although after 1 O ... .1e7 1 1 'iNxb3 0-0 1 2 ttJd2 l:ta7 1 3 l:tfe1 'iWe8 14 l::t ac1 as 1 5 e4 .1a6 1 6 axb5 .1xb5 1 7 'iVe3 lla6 1 8 g4 White started an offensive in this sector, the subsequent 1 8 ... a4 1 9 g5 ttJfd7 20 ttJec4 f6 21 gxf6 l::t xf6 22 e5 l::tg6 23 'it'h 1 resulting in a balanced game. 9 . . . ltJd5 Black can try 9 ... 'itb6, e.g. 10 e4 ttJbd7 1 1 ttJxd7 ttJxd7 1 2 d5 i.c5 1 3 dxe6 fxe6 1 4 i.h3 ttJ f8 1 5 'ikh5+ g6 1 6 'iih 6 and White has considerable play for the pawn, although this is preferable for Black to Shabalov­ Martynov, Moscow 1 987, which went 1 2 ... cxd5 1 3 exd5 e5 14 axb5 axb5 1 5 �xa8+ i.xa8 1 6 i.e3 i.c5 1 7 'it'g4 g6 1 8 i.xc5 'iNxc5 19 J:ta 1 i.b 7 20 lla5. In Filippov-Akhmadeev, Kolontaevo 1 994 Black opted for the quieter 1O ... .1e7 1 1 .1e3 0-0 12 b3 cxb3 13 'iWxb3 'it'd8 (13. .. c5 is too early in view of 14 ttJc4! with an edge for White), inviting White to gradually build up his forces with 1 4 llfd 1 ttJbd7 1 5 ttJd3 .l:!.e8 16 l::t ac 1 l::t c 8 17 h3 h6 1 8 g4 .ta8 19 f4 etc. 1 69

Th e C a t a l a n

1 0 e4 White concentrates on the centre. 1 0 ltJe4 ltJd7 1 1 i.g5 f6 12 ltJxd7 'iVxd7 1 3 i.d2 i.e7 1 4 ltJc5 i.xc5 1 5 dxc5 "fie7 16 e4 ltJc7 17 i.a5 e5 1 8 "fid6 resulted in the usual dark square theme in Szabo-Shamkovich, Con­ stanza 1 969. 10 axb5 must also be investi­ gated. Then 10 ... cxb5 transposes to 5 ... a6 6 0-0 b5 7 ltJe5 ltJd5 8 a4 c6 9 axb5 cxb5 1 0 ltJc3 i.b7 (Chapter 5), which leaves 1 O . . . axb5 1 1 �xa8 i.xa8 12 i.xd5 cxd5 1 3 ltJxb5 as in Piket-Lputian, Montecatini Terme 2000. White wins back the pawn but now Black is able to smoothly bring his forces into the game, and Lputian secured equal chances: 1 3. . .'it'a5 1 4 ltJc3 ltJc6 15 i.f4 i.b4 16 e4 0-0 17 'iVg4 h5 18 ltJxc6 i.xc6 19 �xh5 i.xc3 20 bxc3 'iYxc3 21 i.e5 "fid2 22 nd1 'iYh6 23 'iVxh6 gxh6. 1 0 . . . tDxc3 1 1 bxc3 tDd7 1 2 f4 tDf6?! 1 2 ...i.e7 is appropriate, when 1 3 ltJxf7?! is dubious in view of 13 ...'it>xf7 1 4 f5 e5! and Black keeps the position closed. G.Gross­ M.Gurevich, 1 998 is a more useful model for White: 13 'i!Vg4 0-0 14 f5 ltJf6 1 5 'i!Ve2 exfS 1 6 nxfS ltJd7 1 7 ltJg4 ltJb6

Now White found the dangerous looking 1 8 ltJh6+!?, both players having a fight on their hands after 1 8 ...gxh6 1 9 i.xh6 �h8 20 e5 ltJd5. 1 3 f5 exf5 14 .l:!.xf5 i.e 7 1 5 'iVf3 c5 1 6 d5 0-0 1 7 i.e3 The forcing 17 ltJc6 i.xc6 18 dxc6 prom1 70

ises White less than the game continuation. 1 7 . . . g6? ! 1 7 ... 'iYa5 1 8 i.g5 h6?! 1 9 i.xf6 i.xf6 20 nxf6! gxf6 21 ltJg4 with good chances of success on the kingside. 1 8 i.h6 ! !

A brilliant sacrifice! 1 8 . . . gxf5 1 9 'iiVx f5 'it>h8 20 .l:!.f 1 ! White allows his forces full participation. Black is helpless. 20 .. :�e8 After 20 ... 'i!Vd6 21 "figS ltJe8 22 i.xfS i.xg5 23 ltJxf7+ Black avoids mate but not defeat. 21 d6 i.d8 22 d7 'iVe7

23 'iiVg 5 tDh5 23 ... ng8 doesn't help Black 24 'i!Vxf6+ 'i!Vxf6 25 nxf6 ng6 26 ltJxg6+ hxg6 27 nxf7. 24 'iiVx h5 f6 25 i.xf8 'ii'x e5 26 'iiVe 8! i.c6 27 i.e7 + 1 -0

4 . . . dx c 4 5 iL g 2 b 5

Summary After 6 a4 c6 7 axb5 cxb5 8 ttJe5 ttJd5 9 ttJc3 �b4 1 0 0-0 �xc3 1 1 e4 it is difficult for White to achieve an advantage as Black is provided good counterplay by his queenside pawns. Con­ seljuently the line with 6 a4 c6 7 0-0 �b 7 8 ttJe5 holds more potential for White. The prob­ lems with 8 ... 'ifb6 (such as reducing Black's influence over d5) were pointed out in Khalifman­ Sveshnikov, while after 8 .. ttJd5 White can successfully generate an with either 9 e4 or 9 b3. Perhaps Black should play 8 a6, when White has control over the centre, while 9 b3 is less effective - prompting the conclusion that there is no guarantee of an advantage. .

...

1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 iLg2 b5 6 a4 c6 (D) 7 0-0 7 axb5 cxb5 (D) 8 ttJe5 ttJd5 - Game 84 7 . . . iLb7 8 ttJe5 (D) 8 .. ttJd5 Game 85; 8 . . a6 Game 86 .

-

6. . . c6

.

-

7 . . . cxb5

8 ttJe5

171

I CHA PTER NINE I 4 . . . dxc4 5 Ji.g2 c6

1 d 4 d 5 2 c4 e 6 3 ttJf3 ttJf6 4 g3 dxc4 5 �g2 c6 S ... c6 is rarely seen. White can choose to head for the S ... bS lines with 6 0-0 bS 7 a4 but there is also the more challenge with 6 ttJeS!?, which has a downside in that 6 ... ii.b4+ can disrupt the first player's devel­ opment. White can block the check with 7 ttJc3 with decent chances of achieving some­ thing concrete from the opening phase, or he can try the interesting gambit with 7 ii.d2, as seen in both games in this chapter. In the first game Black accepts the second pawn with 7 .. .'ihd4 and faces a difficult task de­ fending against White's threats (something in which he does not succeed). In the second game Black seeks to exploit the temporarily poor posting of the bishop on d2 with 7 ... ii.e7, adopting a more positional ap­ proach.

Game 87 Chetverik -Burma kin

Nagykanizsa 1993 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 �g2 c6 6 ttJe5 �b4+ 7 �d2 Let us look at 7 ttJc3 ttJdS. Then 8 'il'c2? ttJxc3! 9 bxc3 'it'xd4! 10 ii.b2 ..ItaS 1 1 ttJf] 'it'f6 12 ttJd2 0-0 13 ttJxc4 ..Itc7 1 4 0-0 ttJd7 1 72

1 5 ttJd2 �g6 1 6 e4 eS saw Black keep the pawn at no cost to himself in Castillo-Sadler, Iinares 1 995. After 8 .i.d2 the ball is in Black's court. In Bondarevsky-Keres, Moscow 1 947 White emerged with a slight edge after 8 ...ttJb6 9 ttJe4 as 1 0 e3 0-0 1 1 �c2 ttJ8d7 1 2 .i.xb4 axb4 1 3 ttJxc4 ttJxc4 14 'it'xc4 thanks to his easier piece play. This leaves 8 ... ..Itxc3 9 bxc3 bS 10 a4 when White's activity compensates fully for the pawn deficit. Tkachiev­ Kazhgaleev, Cannes 1 999 continued 1 O .. .f6 1 1 ttJf3 0-0 1 2 e4 ttJe7 1 3 ..Ite l !? a6 1 4 i..a3 ttJd7 1 5 0-0 ttJb6 1 6 as ttJd7?! (1 6 ... ttJa4!?) 1 7 i.. h3 'it> f7 1 8 dS!? (1 8 eS!?) 1 8 . . .exdS 1 9 exdS ttJxdS, and now White could have played 20 ..Ite6+! �xe6 21 ttJd4+ 'it>f7 22 ttJxc6 ttJeS! 23 ttJxd8+ llxd8. Black has only two knights for the queen but the semi-closed position and the great outposts on d3 and dS mean something, although White has the better chances. 7 .. :,'bd4 8 �xb4 'iii'x e5 9 ttJa3 ! b5 Forced. In the case of 9 ...�xb2? 10 llbl �eS 1 1 ttJxc4 White wins after ttJd6+, and 9 ... cS?! is not to be recommended in view of 10 ttJxc4 �c7 1 1 �d6! (Neistadt) . 1 0 f4 'iii'c 7 1 1 ttJxb5 cxb5 1 2 �xa8 iLb 7 1 3 �xb7 'i!Vxb7 1 4 0-0 This is the most natural move. Another

4 . . . dx c 4 5 i. g 2 c 6

try is 1 4 lUI !? tLld5 1 5 i.c5 tLld7 1 6 'iVd4 f6, when Pankratov-Kharlov, Moscow 1 998 saw 1 7 e4 tLlxc5 1 8 ir'xc5 'iVb6! 1 9 'iVxb6 tLlxb6 20 0-0-0 cJ;;e 7 21 .l::td4 l:I.c8 22 l:I.fdl with a very pleasant ending for White. At the mo­ ment only the d-ftle is open so it is difficult for White to break through. After a further opening up of the position the difference in strength between the rook and knight will show. Instead of 1 7 e4 White should con­ sider castling long, as 17 O-O-O! tLlxc5 1 8 'iWxc5 fS 1 9 e4! fxe4 20 fS ! creates a formida­ ble attack against Black's stranded king. 1 4 . . . 'iWb6 + In Shipov-Maljutin, Moscow 1 99 1 Black's knights were busy after 14 ... tLlc6 1 5 i.c3 tLle4 16 a4 (1 6 i.xg7 l:I.g8 17 i.d4 h5 gives Black counterplay) 1 6 ... b4?! (1 6 ...tLlxc3 1 7 bxc3 b 4 1 8 cxb4 'iVxb4 1 9 e 3 is a lesser evil) 1 7 i.e 1 0-0 1 8 'iVc2 tLld6 1 9 i::td 1 tLlfS 20 i.f2 but were nevertheless short of outposts, resulting in insufficient compensation for the exchange. 1 5 'it;h 1

After 1 6 ... tLlc6 1 7 i.c3 tLle4 1 8 i.xg7 l:1.g8 1 9 i.c3 tLlf2+ 20 l:I.xf2 'iWxf2 21 'iWd6 Black restores the points score but the weakness of his dark squares spells serious trouble. 1 7 f5! ttJa6 1 7 ...l:th6 1 8 fxe6 .l::t x e6 1 9 'iWfl favours White. 1 8 fxe6? ! More clear-cut is 1 8 i.c3 0-0 (1 8. . .tLlf2+ 1 9 l:txf2 'iWxf2 20 i.xg7 l:tg8 21 i.f6) 1 9 'iWd4 with a superior endgame for White. 1 8 . . . 'iWxe6 1 9 a3 In the event of 1 9 i.c3 Black can consider 1 9 ... h4!? because 1 9 ...0-0 20 'iWd4 'iWg6 21 �xa7 is preferable for White. 1 9 . . . ttJxb4 20 axb4 ttJg5 21 J:l.f2 21 ::!.xa7 'iWc6 is equal. 21 . . . 0-0 22 'iWd4 'iWc6 + 23 J:l.g2 a6 _24 J:l.d 1 24 'iVh4!? is worth a look. 24 . . . ttJe4 Black's forces enjoy good co-ordination. The knight stands well on e4 and White has pawn weaknesses, thus affording Black com­ pensation for the exchange. 25 �d7 �f6 Also good is 25 ... 'iVxd7 26 l:I.xd7 l:I.e8 but Black prefers to keep the queens on the board. This makes sense as the white king is susceptible to attack. 26 'iWd4 'iWe7 27 e3 J:l.e8 28 'iWd7 'iWe5 29 J:l.d5 �b8 30 J:l.d4 ttJf6 31 'iWd6 'iWb7 32 h3 J:l.xe3 33 'it;h2 J:l.e6

1 5 . . . h5?! Black should prefer 15 ...tLlc6 16 i.c3 tLlg4 1 7 'iWd6 tLle3 1 8 l:I.f3 (not 1 8 .l::t fdl ?? 'iVb7, when White resigned in Pasman-Stean, Beer­ Sheva 1 978) 1 8 ... tLlf5 1 9 'iVd2 b4 20 i.e5 0-0 and White found his bishop stuck on e5 and his centre weakened, leaving the situation unclear. 1 6 J:l.f3 ttJe4 1 73

Th e Ca t a l a n

3 4 �d8 + ? 3 4 �c5 minimises Black's lead to a slight advantage. 34 . . .'.i.>h7 35 'ii'f8 .:te1 36 'ii'c 5 �f3 37 z:tf4 'iVd 1 38 g4? This loses, but 38 ttgl Uxgl 39 �xgl �d2+ 40 �f2 �xb4 does not help White. 38 . . . l:!.h 1 + 39 'it.'g3 �d3 + 0-1

Game 88 Stefanova-Ku rajica

Benasque 1997 1 d4 d5 2 c4 e6 3 4Jf3 4Jf6 4 g3 dxc4 5 i.g2 c6 6 4Je5 i.b4+ 7 i.d2 i.e7

8 i.c3 \,('hite can also play 8 e3, when 8 ... c5!? 9 'ii'a4+ .ltd7 1 0 'ii'x c4 ttJc6 1 1 ttJxc6 .ltxc6 1 2 .ltxc6+ bxc6 1 3 dxc5 �d5 1 4 �xd5 cxd5 1 5 .lt d .ltxc5 1 6 i.. xf6 gxf6 brings about an approximately level ending and 1 1 0-0 offers White an edge. After the text Black has tried 8 ... ttJd5 9 ttJxc4 b5?!, but this proved good for White in Chetverik-Matras, Prague 2002 when the bishop landed on as: 10 i.. a 5 i..b4+ 1 1 i.. xb4 ttJxb4 12 ttJe3 (1 2 ttJe5?! runs into 1 2...�xd4! 13 �xd4 ttJc2+ 14 'it>d2 ttJxd4 note that from e3 the knight protects c2) 12 ... i..b 7 1 3 0-0 and White was in control of the vital c5-square. In order to avoid this

1 74

situation Black now elects to begin his queenside play by first pushing the a-pawn. 8 . . . a5 9 a4? ! The natural looking 9 0-0 ttJd5 1 0 ttJxc4 b5 1 1 ttJe5 secures White a modest edge. 9 . . . 4Jd5 1 0 4Jxc4 b5 1 1 axb5 cxb5 1 2 4Je5 Again 12 ttJe3!? is an interesting move, e.g. 12 ... ttJxe3 13 fxe3 tta6 with chances for both sides, or 1 2 ... i..b7 1 3 �d3 b4 14 i..d2 0-0 with equality. 1 2 . . . 0-0 1 3 0-0 i.b7 1 4 'ilib3 14 e4?! ttJf6! 1 5 �d3 b4 1 6 .ltd2 .lta6 1 7 ttJc4 ttJc6! and Black wins material because d4 is under fire and there is the threat of ... ttJe5! etc. 14 .ltd2 a4 is slightly better for Black. 1 4 . . . b4 1 5 i.d2 4Jc6 1 6 4Jxc6 i.xc6 1 7 e4 4Jb6 1 8 i.e3?! Or 18 l:tc 1 �b7 1 9 �e3 a4 and Black is making considerable progress on the queen­ side. 1 8 . . . a4 1 9 'ii'd 3 'ilid7 20 d5? The lesser evil is 20 l:tdl .ltb5 21 �d2 ttJc4. 20 . . . i.b5 Black wins material. 21 �d4 4Jc4 22 i.f4 e5!

23 i.xe5 4Jxe5 24 �xe5 i.xf 1 25 i.xf 1 i.f6 26 �h5 i.xb2 27 '!:!'a2 g6 28 �f3 i.g7 0-1

4 . . . dx c 4 5 Ji. g 2 c 6

Summary This line is a bit too obscure and lacks something in flexibility for today's top players. If White wants he can transpose to the sharp S ... bS line with 6 0-0 bS 7 a4, while another option is a guaranteed advantage after 6 ttJeS .i.b4+ 7 .i.d2. It is then very dangerous for Black to take the pawns in the centre, and White enjoys an enduring lead after 7 ... .i.e7 8 .i.c3 followed by ttJxc4. 1 d4 d5 2 c4 e6 3 tLlf3 tLlf6 4 g3 dxc4 5 Ji.g2 c6 (D) 6 tLle5 Ji.b4+ 7 Ji.d2 7...'ii'xd4 (D) Game 87; 7 ... .i.e7 (D) Game 88 -

5 . c6 . .

-

7 . . flixd4 .

•

7 . . Ji.e 7 .

1 75

CHA PTER TEN

I

4 . . . dxc4 5 i.g2 i.b4 +

1 d4 d5 2 c4 e6 3 4:Jf3 4:Jf6 4 g3 dxc4 5 i.g2 i.b4+ This check is designed to throw a (small) spanner in the works as far as \'Vhite's co­ ordination is concerned, as well as the de­ fence of the d4-pawn. However, the idea is seen rather seldom with this particular move order. The positions in this line can usually be compared to those of the Bogo-Indian De­ fence (1 d4 ttJf6 2 c4 e6 3 ttJf3 .i.b4+), while 6 ttJc3 0-0 7 0-0 transposes to the Nimzo­ Indian Defence (1 d4 ttJf6 2 c4 e6 3 ttJc3 .i.b4 4 g3 0-0 S ttJf3 dS 6 .i.g2 dxc4 7 0-0) in this book these positions can be found via the move order S ... ttJc6 6 0-0 in Chapter 6. The continuations 6 ttJbd2 and 6 �d2 are of approximately level theoretical impor­ tance. The position after 6 ttJbd2 often arises through the move order 1 d4 ttJf6 2 c4 e6 3 ttJf3 .i.b4+ 4 ttJbd2 dS S g3 dxc4 6 i.g2 when, in the case of 6 ... c3 7 bxc3 i.xc3 8 �b1 , \'Vhite's pressure on the b- and c-files and his lead in development compensate for the pawn. Therefore we are more likely to see 6 ... ttJc6 or 6 ... 0-0, and these rwo options often overlap (Game 89) . In reply to 6 i.d2 Black rarely trades bish­ ops, and after 6 ...i.xd2+ 7 �xd2 Black strug­ gles if he tries to hang on to the extra pawn

pawn with 7 ... bS?! thanks to 8 �gS with a double attack. However, alternatives also leave Black ive and without counterplay, which explains why avoiding the exchange with 6 ... �e7, 6 ... cS and 6...aS (Game 90) is preferable.

Game 89 Kozul-Lju bojevic

Belgrade 1989 1 d4 d5 2 c4 e6 3 4:Jf3 4:Jf6 4 g3 dxc4 5 i.. g 2 i.b4+ 6 4:Jbd2

-

1 76

6

0-0 After 6 ... c3 7 bxc3 �xc3 Black wins a pawn. Then 8 �b 1 ttJe4 9 0-0 �xd2 1 0 i.xd2 ttJxd2 1 1 'Yiixd2 ttJd7 1 2 �fc1 left . . .

4 . . . dx c 4 5 iL g 2 iL b 4 +

White with the usual 'development' compen­ sation in Haba-Hiibner, 1 993, which was soon unclear after 12 ... 0-0 13 a4 as 14 'iVc2 c6 15 e4 ltJf6 1 6 l:td1 . Another option for Black is 6 ... ltJc6 7 0-0 i.xd2. Tukmakov-Lputian, Novosibirsk 1 986 continued 8 i.xd2 ltJxd4 9 ltJe5!? 0-0 10 .l:r.c1 c5 1 1 llxc4 ltJd7 1 2 ltJd3, White winning back the invested material with the better game, retaining his lead after 12 ... e5 1 3 ltJxc5 ltJxc5 1 4 :xc5 'iVe7 1 5 SLb4 i.g4 1 6 f3 SLe6 1 7 e 3 ltJc6 1 8 SLa3 etc. White can adopt a different set-up by re­ capturing with the queen, e.g. 8 'iVxd2 l:tb8 9 b3!? cxb3 1 0 i.a3, when White is trying to keep Black's king stranded in the centre. This plan was used in Kozul-Vukovic, Banja Vru­ cica 1 99 1 which went 1 0 ... ltJe7 1 1 axb3 i.d7 1 2 ltJe5 i.b5 1 3 l:tfc1 a6 1 4 g4 0-0 1 5 g5 ltJe8 1 6 Uc3 and White's kingside offensive was up and running. Instead of 8...1:!.b8 Black has 8 ... 0-0 9 'iVc3 b5, when after 10 ltJe5 ltJxd4 1 1 l:!.el l:!.b8 12 e3 White hits the piece that defends against the fork on c6. Conse­ quendy Black should maintain the momen­ tum by parting with the knight, although 1 2 ... c5 1 3 exd4 cxd4 1 4 'iVa3 anyway favours White. 7 0-0 Creating a genuine threat against c4 by lift­ ing the pin. 7 'iVc2 is the alternative means to hit c4, when Black is not obliged to react. 7 ... b5 8 a4 bxa4 sees Black voluntarily weaken his pawn structure in return for piece play, Haba-Hiibner, 1 99 1 continu­ ing 9 0-0 SLb7 1 0 ltJxc4 c5 1 1 i.g5 ltJc6 1 2 dxc5 i.xc5 1 3 'iVxa4 'iVe7 1 4 �ac1 l:!ac8 1 5 SLe3 i.xe3 when a draw was agreed, al­ though White is slighdy better in this posi­ tion. Black can also play 7 ... c5 8 dxc5 c3 9 bxc3 i.xc5 1 0 0-0 ltJc6 1 1 ltJb3 i.e7 1 2 l:i.dl 'iVc7 13 i.f4 e5 14 i.e3, this time White accepting a litde structural damage in order to generate pressure on the queenside. This seemed to work in D.Gurevich-Tiviakov, New York 1 998 after 1 4... i.g4 1 5 h3 i.h5 1 6

ltJh4 l:tfd8 1 7 l:[xd8+ l:lxd8 1 8 lIdl with an edge for White. A simple, albeit rather unambitious ap­ proach is to anticipate the coming capture on c4 and prepare a trade of queens with 7 ...ltJc6 8 'iVxc4 'ii'd 5, although 9 0-0 i.xd2 1 0 'iVxd5 exd5 1 1 i.xd2 i.g4 1 2 lIfc1 l:lac8 1 3 e3 ltJe4 1 4 i.el resulted in White having the more comfortable game in D.Gurevich­ Young, Reno 1 992 thanks to the bishop pair. There followed 1 4 ...lIfe8 1 5 b4 ltJd6 1 6 a4 a6 1 7 nab 1 i.f5 1 8 lIb3 ltJa7 1 9 ltJe5 c6 20 g4 i.e6 21 f3 f5 22 h3 l:tcd8 23 SLh4 l:!.a8 24 i.g3 .l:!ac8 25 :tbc3 g6 26 i.f1 and Black had failed to drum up worthy counterplay, being reduced to the unenviable policy of waiting to see how White chose to continue the bat­ de. 7 c3 Raetsky-Gipslis, Senden 1 998 went 7 ... b5 8 a4 c6 9 'it'c2 as 10 b3 cxb3 1 1 ltJxb3 bxa4 1 2 l:ha4 i.a6 1 3 SLg5 SLb5 1 4 J:ta2 h6 1 5 SLxf6 'ii'xf6 1 6 ltJe5 l:td8 1 7 l:tb 1 a4 1 8 ltJc5 when, for the pawn, White had managed to generate a formidable initiative. S lLlc4 lLlc6 . . .

9 fid3 Quite different is 9 b3, White hoping to win the c3-pawn. Haba-Ostrowski, Czehia 1 999 developed as follows: 9 ....l:tb8 10 a3 i.e7 1 1 'iVc2 b5 1 2 ltJce5 i.b7 1 3 ltJxc6 i.xc6 1 4 'iVxc3 i.e4 1 5 i.b2 ltJd5 1 6 'iVc1 'iVc8 1 7 ltJe5 and White was ready to ad1 77

Th e Ca t a l a n

vance the e-pawn. 9 . . . cxb2 1 0 .ll. x b2 .ll. e 7 1 1 �ac 1 as 1 2 'ilVb 1 1 2 l:tfd1 a4 1 3 e4 a3 1 4 �a1 �d7 1 5 'ifb1 l:ta4 1 6 d5 exd5 1 7 exd5 tt:Ja5 was the course of Raetsky-Rashkovsky, Cappelle la Grande 2000. Now White could have had chances for an advantage with 1 8 d6! .l:!.xc4 1 9 dxe 7 'iVxe7 20 tt:Jg5 (20 l:txc4 tt:Jxc4 21 ii.xf6 h'Xf6 22 'iVxb7 with compensation for the mate­ rial) 20 ...g6 21 l:txc4 tt:Jxc4 22 ii.xf6 'iVxf6 23 tt:Jxh7! 'it>xh7 24 J::!. xd7 etc. 1 2 . . . a4 1 3 e4 a3 1 4 .ll. a 1 ttJb4 1 5 �fd 1 �a4 1 6 ttJe3 c6 1 7 h4 h6 1 8 .ll.f 1 .ll. d 7 An interesting option is 18 ... b5!? 19 tt:Je5 'ifb6 with chances for both sides. 19 .ll. c4 b5 20 .ll. b 3 �a7 2 1 ttJe5 White has better in 21 �c3! tt:Ja6 22 �d2!, redeploying the bishop for an attack on the kingside, e.g. 22 ...'ifb6 23 e5 tt:Jd5 24 tt:Jg4. 21 . . . .ll. e 8 22 f4? ! Dubious. The correct treatment is 22 tt:J5g4 tt:Jxg4 23 tt:Jxg4 Wib6 24 e5! (Kozul). 22 . . . ttJh5! 23 ttJf 1 23 'it>g2 .i.xh4! helps Black. 23 .. :ii'b 6 24 Wh2 c5 25 dxc5 .ll. x c5 26 �xc5 'ilVxc5 27 .ll. d 4 'ilVe7 28 �c 1 ttJa6 29 .ll. x a7 'iVxa7 30 ttJg4 ttJc5

3 1 .ll. d 1 Perhaps White might try 31 e5!? �c6 32 ii.d 1 , thus avoiding Black's next, which opens up the centre. 31 . . . f5! 32 exf5 exf5 33 ttJe5 ttJe4 34 1 78

�c2 'ilVd4 35 .ll. f 3? 35 �xh5 �xh5 36 'iVxb5 'il'd 1 IS the lesser evil. 35 . . . b4 Black has a decisive lead. 36 'ii'b 3+ Wh7 37 .ll. xe4 'it'xe4 38 �c7 .ll. b 5 39 ttJe3 ttJxf4 40 �c5 40 l:te 7 .i.a4! etc. 40 . . . �d8! 4 1 gxf4 'ii'xf4+ 42 Wh3 .ll. f 1 + ! 0-1

Game 90 Karpov-Piket

Monaco 1999 1 d4 d 5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 .ll. g 2 .ll. b4+ 6 .ll. d 2

6 . . . a5 Of course Black has other options. After 6 ... .i.xd2+ 7 'iVxd2 a mistake is 7 ... b5?! in view of 8 'iVg5 with simultaneous threats against b5 and g7. Instead 7...0-0 8 tt:Ja3 ii.d7 9 tt:Jxc4 �c6 1 0 0-0 tt:Jbd7 1 1 l:tfc1 �d5 1 2 b 4 left White i n charge o n the queenside in Portisch-P.Nikolic, Ijnares 1 988. 6 ... ii.e7 7 0-0 leads only to equality after 7 ... 1Ld7! 8 'il'c2 ii.c6 9 'iVxc4. Unander-Plato, Sweden 1 995 saw the more combative 7 'iVc2 .i.d7 8 tt:Je5!? tt:Jc6 9 'iVxc4 tt:Jxe5 10 dxe5 tt:Jd5 1 1 0-0 0-0 12 tt:Jc3 c6 1 3 a4 'tIVb8 14 f4 as 1 5 'YWd3 'iVc7 1 6 l:tac 1 and White enjoyed more space. An interesting situation arises after 6 ... c5 7

4 . . . dx c 4 5 s... g 2 s... b 4 +

jLxb4 cxb4 8 4Je5 0-0 9 4Jxc4 4Jc6 1 0 e3, when Chetverik-Ziabari, Prague 2002 con­ tinued 10 ... e5!?, the point being that after 1 1 jLxc6 bxc6 both 1 2 dxe5 'YWxd1+ 1 3 'it'xdl 4Je4! 1 4 'it'e2 �a6 1 5 l:tc1 l:tad8 and 1 2 4Jxe5 jLh3 1 3 'i¥ f3 l:tc8 Black has compen­ sation. Consequently White chose 1 1 d5, when the counter l 1 ...b5! 12 dxc6 'iVxd1+ 1 3 'i.t>xd 1 bxc4 1 4 4Jd2 c 3 1 5 bxc3 bxc3 1 6 4Jc4 e4 17 'i.t>c2 i.e6 1 8 'it>xc3 i.d5 19 4Ja5 i.xc6 would have led to equality. 7 0-0 0-0 After 7 ... b5?! 8 a4 c6 9 axb5 i.xd2 1 0 'YWxd2 cxb5 White again has 1 1 iVg5 with the now familiar double threat against b5 and g7. Voloshin-Fiodorov, Polanica Zdroj 1 997 saw 1 1 ...4Jd5 1 2 'it'xg7 �f6 1 3 �xf6 4Jxf6 1 4 4Je5 4Jd5 1 5 4Jc3 f6 1 6 4Jxd5 exd5 1 7 4Jf7! i.e6 18 4Jd6+ 'it>d7 1 9 4Jxb5. 8 'iVc2 b5?! Babula-Jirovsky, Zlin 1 997 witnessed an­ other exchange of queens: 8 ... 4Jc6 9 'iVxc4 �d5 1 0 l:tc1 l:td8 1 1 �xd5 exd5 1 2 �f4 i.g4 1 3 e3 l:td7 1 4 4Jc3 4Jh5 1 5 4Jb5 4Jxf4 1 6 gxf4 f6 1 7 a3 jLfS 1 8 4Je 1 a4 1 9 jLfl l:ta5 20 jLd3 g6 21 4Jc3 4Ja7 22 jLc2 b5 23 4Jd3 and White's pull on the queenside was enough for an advantage. 9 a4

9 . . . bxa4 1 0 .l:!.xa4 White has an interesting idea in picking up the a4-pawn with the knight, an approach

tried in Gulko-Makarychev, Moscow 1 974, when 10 4Jc3!? jLb7 1 1 4Jxa4 �xd2 1 2 �xd2 4Jbd7 1 3 l:tfc 1 4Jb6 1 4 4Jc5 4Je4 1 5 �e 1 4Jxc5 1 6 dxc5 4Jd5 1 7 4Je5 c3 1 8 bxc3 f6 19 l:tcb 1 favoured White. The advantage remained intact after the subsequent 1 9 ...l:ta7 20 c6 jLa8 21 4Jd7 l:te8 22 c4 i.xc6 23 4Jb8! jLb7 24 l:txb7! l:txb7 25 4Jc6 'iVd6 26 cxd5 exd5 27 'it'xa5 l:tb2 28 'iVxd5+ 'it>fS 29 e3, the two minor pieces being stronger than the rook (although Black can still fight on). 10 . . . s... b 7 1 1 s... x b4 axb4 1 2 lha8 s... x a8 1 3 ttJbd2 1 3 �xc4 leads only to equality after 1 3 ... �d5 1 4 'YWxb4 4Jc6 1 5 'iVc5 jLxf3 1 6 jLxf3 4Jxd4 1 7 4Jc3 4Jxf3+ 1 8 exf3. 1 3 . . . c5 Black can clear the queenside with 13 ... c3 14 bxc3 bxc3 1 5 iVxc3 4Je4 16 4Jxe4 jLxe4 17 4Je5 jLxg2 1 8 'it>xg2 'iVd5+ 1 9 'it>g1 (piket), when White can claim a lead in view of his superior pawn structure. 1 4 dxc5 'iVa5 1 5 �xc4 s... d 5 1 6 �d4 1 6 iVc2 l:tc8 1 7 e4 l:txc5 1 8 iVd3 gives White something to bite on. 1 6 . . . �d8 1 7 ttJc4 s... x c4 1 8 'iVxc4 .l:!.c8 1 9 ttJd4?! Better is 1 9 l:td 1 h6 (1 9 ...l:txc5?! 20 iVxb4!) 20 h3 with a minimal advantage to White. 1 9 . . . .l:!.xc5 20 'iVd3 .l:!.c8 21 .l:!.d 1 h6 21 ...l:td8 22 iVc2 g6 23 e3 e5 is level. 22 h3 'iVa2 23 'iVb3 �xb3 24 ttJxb3 ttJfd7 25 f4 �f8 26 s... e4 �e7 The king comes to the centre. Black has equal chances in this endgame. 27 .l:!.d4 f5 28 s... d 3 ttJc6 29 .l:!.c4 ttJb6 30 .l:!.c 1 �d6 31 �f2 g5 32 e4 32 fxg5 hxg5 33 jLb5 4Je5 34 l:txc8 4Jxc8 35 h4 with approximate equality. 32 . . . gxf4 33 gxf4 fxe4 34 s... xe4 ttJe7 35 .l:!.xc8 ttJexc8 36 ttJd4 ttJa4 37 b3 ttJc5 38 s... c 2 ttJe 7 39 �f3 e5 40 ttJe2 exf4 41 ttJxf4 �e5 42 ttJd3 + ttJxd3 43 s... x d3 ttJc6 Y:z - Y:z

1 79

Th e Ca t a la n

Summary In the variation with 6 lbbd2 lbc6 Black usually exchanges on d2 and holds on to the gambit pawn. In return White enjoys better development and has definite compensation. After 6 ... 0-0 White has a choice between taking direct aim at c4 with 7 'ifc2, with a slight edge, or seeking an initiative with 7 0-0, allowing Black to hang on to the material (after 7 ... bS or 7 ... c3) . In reply t o 6 .Jtd2 a ll of 6 ... .Jte7, 6 ... cS and 6 ...a S are reasonable for Black. Note than after 6 ... iJ... e7 7 0-0 Black can bring his other bishop (via d7) to c6 because 7 ... .Jtd7 S lbeS?! lbc6 makes the bishop look misplaced on d2. White has a modest space advantage after 7 'iVc2 .Jtd7 S lbeS. In the variation with 6 ... cS 7 .Jtxb4 cxb4 S lbeS 0-0 9 lbxc4 lbc6 10 e3 it was thought that 1 0 ... eS was dubious due to 1 1 .Jtxc6 bxc6 1 2 dxeS 'ii'x d1 1 3 'iit x d1 lbg4 1 4 'iit e 1 (Bareev), but Black can certainly improve on 1 3 ... lbg4? with 1 3 ... lbe4, guaranteeing compensation for the pawn. Perhaps White should prefer 7 dxcS, with a slight lead in a standard Catalan position, rather than 7 .Jtxb4. After 6 ... aS 7 0-0 0-0 8 'iVc2 lbc6 9 'ir'xc4 iLd6 we have a position which belongs in the line starting with S ... lbc6 6 'it'a4 iJ... b4+ 7 .Jtd2 .Jtd6 S lfxc4 0-0 9 0-0 but with the additional ... a7as. The continuation 9 'iVxc4 'iVdS differs from the well known variation in the Bogo-Indian Defence - namely 1 d4 lbf6 2 c4 e6 3 lbf3 .Jtb4+ 4 .Jtd2 as S g3 dS 6 'iVc2 lbc6 7 .Jtg2 dxc4 S 'iVxc4 'it'dS - in that neither player has castled in the Bogo. Here, apart from 1 0 'iVxdS and 1 0 'iVd3, White has the additional possibility o f 1 0 l:!.c1 with a small but steady advantage. 1 d4 d5 2 c4 e6 3 �f3 �f6 4 g3 dxc4 5 .i.g2 .i.b4+ (D) 6 lbc3 0-0 7 0-0 lbc6 S e3 .l:!.bS - Game 63 (Chapter 6) ; S 'iVc2 l:tbS - Game 64 (Chapter 6) 6 lbbd2 0-0 7 0-0 c3 (D) - Game 89 6 .Jtd2 as (D) - Game 90

5

1 80

. . .

.i.b4 +

7

. . .

c3

6

. . .

a5

CHA PTER ElEVEN

I

4 . . . dxc4 5 �g2 et:Jbd 7

1 d4 d5 2 c4 e6 3 4:lf3 4:lf6 4 g3 dxc4 5 i.g2 4:lbd7 This natural posting of the knight is simi­ lar to Black's set-up in the Queen's Gambit Declined. Black addresses queenside devel­ opment and prepares the liberating ... c7-c5 and ... e6-e5, but the system's peak in popu­ larity has long gone. This is because White tends to find himself with an advantage and in little danger of facing counterplay from Black. Note that both 6 1t'a4 and 6 'i¥c2 transpose to 5 1t'a4+ ttJbd7 in Chapter 3, so here we shall consider only the consequences of 6 0-0. Then 6 ... �e7 7 it'c2 will again steer us to 5 'iVa4+, but White stands better after 7 ttJbd2 and 7 ttJfd2. 6 ... c5 is discussed In Game 91 and 6 ... J::tb 8 in Game 92 . ...---

Game 9 1 Vasilchenko-Meszaros

Kecskemet 199 1 1 d 4 d5 2 c4 e 6 3 4:lf3 4:lf6 4 g 3 dxc4 5 i.g2 4:lbd7 6 0-0 c5 Immediately contesting the centre. 6 ... ttJb6 7 ttJbd2 c5 8 ttJxc4 transposes to 6 ... c5 7 ttJa3 ttJb6 8 ttJxc4 in the note to Black's 7th move below, but an alternative is 6 ... �e7. Then 7 �c2 0-0 8 'YWxc4 transposes

to 5 'it'a4+ ttJbd7 6 �g2 iLe7 7 'iVxc4 0-0 8 0-0 (Chapter 3) . 7 ttJfd2 ttJb6 8 ttJxc4 ttJxc4 9 'iVa4+ ttJd7 10 'ii'x c4 ttJb6 1 1 'ikd3 0-0 12 ttJc3 ttJd5 1 3 l:td1 b6 1 4 e4 earns White the initiative. In Krasenkow-Hubner, Polanica Zdroj 1 996 Black chose the inferior 14 ... ttJb4?! 1 5 'itb 1 l:tb8 1 6 iLf4 �g5 1 7 a3 ttJa6 1 8 iLxg5 'ii'xg5 1 9 b4 c5 20 f4 'iYh6 21 ttJb5 and found him­ self looking at a misplaced knight on a6. Hubner proposes an improvement in 14 ...ttJxc3 1 5 bxc3 iLb 7 1 6 c4. 7 ttJbd2 also hits c4, A.Zaitsev-Hubner, Buesum 1 969 continuing 7 ... ttJb6 8 'ii'c2 a6 9 a4 0-0 1 0 as ttJbd5 1 1 ttJxc4 b5 1 2 axb6 cxb6 1 3 'iVb3 iLd7 14 ttJfe5 iLb5 1 5 iLd2 ttJe8 1 6 e4 ttJdf6 1 7 J::t fd 1 ttJd6 1 8 ttJxd6 iLxd6 1 9 iLg5 J::tc 8 2 0 iLfl ! iLxfl 21 'it>xfl � e7 2 2 d5! with a nice attack for White. 7 ... b5 has been tried here, e.g. 8 a4 c6 9 axb5 cxb5 10 ttJe5 ttJxe5 1 1 iLxa8 'ii'xd4 12 ttJO ttJxO+ 1 3 .i.xO 'iNb6 1 4 b 3 0-0 1 5 bxc4 bxc4 1 6 iLe3 .i.c5 17 �xc5 it'xc5 1 8 1t'a4 e5 19 J::t fc1 iLe6 20 iLb 7 and White was in the driving seat in Ni Hua-Jakubowski, Oropesa 1 999. 6 ... c6 s the advance of the b-pawn, which soon saw action in Pine-Euwe, New York 1 95 1 after 7 a4 b5 8 axb5 cxb5 9 ttJc3 'iVb6 10 b3 �b4 1 1 ttJa2 iLe7 12 bxc4 bxc4 13 ttJd2 ttJd5 14 ttJxc4, when White stood 181

Th e Ca t a l a n

better thanks t o his healthier pawns and queenside pressure. In Budnikov-Thesing, 1 992 Black sent his knight on a tour: 7 ... tDd5 8 e4 tDb4 9 tDa3 tDd3 10 .ig5 f6 1 1 tDxc4 b5 1 2 axb5 cxb5 1 3 tDa5 tDxb2 14 'W!Vb3 tDc4, and now after 1 5 .id2 tDdb6 16 �xb5+ �d7 17 'iVh5+ g6 1 8 'iVh4 .ie7 1 9 �h6 White was in charge, in no small part thanks to Black's poor king. 7 lLla3 a6 7 ... cxd4 8 tDxc4 'iVc7 9 �xd4 b5 10 tDce5 .2i.c5 1 1 �d3 looks nice for White, while Raetsky-Lechtynsky, Pardubice 1 992 contin­ ued 7 ... 1I.. e7 8 tDxc4 0-0 9 1I.. f4 tDd5 10 .l:i.c1 tDxf4 1 1 gxf4 .l:i.b8 12 dxc5 �xc5 13 �c2 tDf6 1 4 .l:i.fdl �e7 1 5 tDfe5 tDd5 1 6 e3 a6 1 7 a3, again with a queenside pull for White. After 7 ... tDb6 8 1I..g5 Atalik gives 8 ... c3! 9 bxc3 cxd4 1 0 tDb5 �d7 1 1 'iVbl ! a6 1 2 tDbxd4 h 6 1 3 1I.. xf6 !,,"X f6 as unclear in view of the pawn weaknesses on both sides. Korchnoi-Miles, Rotterdam 1 984 went 8 tDxc4 tDxc4 9 1Wa4+ .2i.d7 1 0 �xc4 �6 (1 0 ... b5 is a bit too early, as was demon­ strated in Quinteros-Schweber, Buenos Aires 1 998, which continued 1 1 it'd3 c4 1 2 �c2 .2i.c6 1 3 �g5 �e7 14 .2i.xf6 �xf6 1 5 e4 �b7 16 .l:1ad 1 0-0 17 e5 .2i.e7 1 8 tDg5 .2i.xg5 1 9 .2i.xb7 nb8 20 .ie4 g6 21 h4 .ih6 22 d5 exd5 23 .ixd5 'iVe7 24 .l:i.fe l and Black was left to contend with e5-e6) 1 1 dxc5 .ixc5 1 2 b4! 'iYxb4 1 3 it'xb4 .ixb4 1 4 .l:i. b1 .id6 1 5 .l:i.dl �e7 1 6 tDg5 tDd5. Korchnoi then gives 1 7 tDxf7!? �xf7 1 8 .l:i.xb7 We8 1 9 e4 .ia4 (1 9 ... .ic6 20 .l:lxg7 .2i.e5?! 21 exd5 .2i.a4 22 llg5! is excellent for White) 20 .l:i.d3 .l:i.c8 21 .ib2 .ic2 22 .l:i.d2 �b4 23 .l:i.xc2 .l:i.xc2 24 �xg7 tDe3 25 fxe3 as a route to advantage for White. 8 dxc5 i.xc5 9 lLlxc4 0-0 9 ... b5? runs into 10 tDce5 .ib7 1 1 tDxf7! <j;xf7 12 tDg5+ �e7 13 .ixb7 and White was close to winning in Raetsky-Andersen, Bad Raga? 1 994. 1 0 'iWc2 b5?! 1 O ...�e7 11 .ig5 h6 1 2 .ixf6 tDxf6 1 3 1 82

.i:!.ac1 with a slight advantage for White due to his lead in development. 1 1 lLlce5 �b6 Another try is 1 1 ...'iVc7 12 .ie3 �b6 (1 2 ... .ib7? 13 tDxd7 tDxd7 14 tDg5 and White wins) 1 3 it'xc7 1I.. xc7 1 4 .i:!.ac1 with a considerable lead for White. Again Black has problems on the h l -a8 diagonal. 1 2 b4! i.xb4 1 3 i.e3 'iWd6 1 4 .l:.ad 1 1 4 tDg5?! is tempting but Black pounces with 14 ... tDxe5 1 5 11.. xa8 tDc4 . 1 4 . . . lLld5

1 5 lLlg5! f5 15 ... tD7f6 16 tDg4! and White wins at least a piece. 1 5 ...g6 is more stubborn, although White wins after 1 6 tDexf7! .l:i.xf7 1 7 tDxf7 Wxf7 1 8 .if4 �c5 19 �xc5 tDxc5 20 .l:i.xd5! etc. 1 6 lLlxe6! lLlxe3 1 7 fxe3 'iWxe5 1 8 lLlxf8 'iWxe3+ 1 9 �h 1 1 -0

Game 92 Kasparov-Korchnoi

Candidates Semifinal (9), London 1983 1 d4 d5 2 c4 e6 3 lLlf3 lLlf6 4 g3 dxc4 5 i.g2 lLlbd7 6 0-0 J:tb8 7 a4 b6 Not Black's only configuration on the queenside. After 7 ... a6 8 as b5 9 axb6 cxb6 10 .2i.f4 .l:i.b7 1 1 .l:i.xa6 tDd5 12 tDc3 Black played 12 ... .2i.e7?! in Tukmakov-O.Rodriguez, New York 1 987, but following 1 3 tDb5! 0-0 1 4 it'c 1 tDxf4 1 5 gxf4 .i:!.b8 16 .i:!.a7 .2i.b 7 1 7

4 . . . dx c 4 5 i.g 2 0, b d 7

'iVxc4 emerged a pawn down for nothing, his situation taking a turn for the worse after 1 7 ... ctJf6 1 8 ctJe5 ctJd5 1 9 f5! .i.g5 20 ctJxf7! Itxf7 21 fxe6 l:.f5 22 l:.xb7 .l::t xb 7 23 e4, when White won. Black can consider 12 ... ctJxf4, when 1 3 gxf4 b5 1 4 l:ta8! ctJb6 1 5 l:.a5 favours White, the b5-pawn being weak and Black lagging behind in development. In Gulko-Shapiro, New York 1 987 Black man­ aged to use the b5-square (temporarily) for his own knight after 8 ... ctJe4 9 'ilVc2 ctJd6 1 0 ctJe5 ctJb5 1 1 l:.d 1 ctJxe5 1 2 dxe5 .i.d7 1 3 'iix c4 c 5 1 4 ctJc3, when White collected the gambit pawn with a more active position. Indeed after 14 ... 'it'c7 1 5 ctJe4 .i.e7 1 6 .i.f4 0-0 1 7 ctJf6+! .i.xf6 1 8 exf6 e5 1 9 .i.e3 ctJd4 20 lhd4! exd4 21 .i.f4 White won material. S 0,fd2 8 ctJbd2 .i.a6 9 'iVc2 c3 10 bxc3 .i.xe2 1 1 lie 1 .i.a6 was tried in Lutikov-Klovans, Vladivostok 1 978, when White could have obtained the better chances after 1 2 c4! .i.e7 1 3 d5 ctJc5 1 4 ctJe5 .i.b7 1 5 dxe6 ctJxe6 1 6 .i.c6+ .i.xc6 (Neistadt gives 1 6 . ..'.ti> f8 1 7 ctJdf3 with initiative) 1 7 ctJxc6 ctJd4 1 8 l:.xe7+! 'it'xe7 1 9 ctJxd4 'ilVe1+ 20 'it'g2, with two pieces for a rook. S . . . e5 8 . . . .i.b7 9 .i.xb7 l:.xb7 10 ctJxc4 .i.e7 1 1 ctJc3 ctJd5 1 2 e4 ctJxc3 1 3 bxc3 ctJf6 1 4 'it'e2 is an edge for White, while 14 e5?! ctJd5 1 5 'it'g4 ctJxc3!? 1 6 .i.b2 h5 1 7 'iWxg7 ctJe2+ 1 8 �g2 'ilVd5+ 1 9 f3 l:.f8 is unclear. 9 0,xc4! 9 d5?! as! 10 ctJc3 .i.a6 1 1 ctJb5 .i.xb5 1 2 axb5 .i.d6 1 3 ctJxc4 0-0 i s equal. 9 . . . exd4 1 0 'iWxd4 i.c5 1 1 'it'd3 0-0 1 2 0,c3 i.b 7 1 3 i.xb 7 l:txb 7 1 4 'liVf3 ! 1 4 .i.f4 .i.b4! 1 5 'if f3 ctJc5 with only a small advantage for White. 1 4 . . :ii'a S 1 5 i.f4 a6 1 5 .. J:tbb8 16 'it'xa8 t!.xa8 17 l:.ad 1 c6 1 8 e4 11fd8 1 9 .i.c7 l:.e8 20 .i.d6 .i.xd6 21 ctJxd6 l:.e7 22 f4 favours White, whose proud knight is most unwelcome on d6. 1 6 e4 l:ta7 1 7 0,d5

Also possible is 1 7 'iitg2!? 11e8 1 8 e5 'iixf3+ 19 'it'xf3 ctJh5 20 l:tad1 ctJxf4 21 gxf4 with a better endgame. 1 7 . . . b5 1 7...'iVc6 1 8 b3 ctJxd5 1 9 exd5 'iVf6 20 .uac 1 is a lesser evil. 1 S 0,a5!

18 ctJe5 forces too many exchanges and results in equality after 1 8 ... ctJxe5 19 .i.xe5 ctJxd5 20 exd5 .i.d6 21 .i.d4 l:.b7 when, sud­ denly, Black is no longer cramped on the queenside. 1 S . . . bxa4? ! Far better is 1 8 ... l:.e8! 1 9 ctJxc7 'iVxe4 20 'it'xe4 1he4 21 axb5 axb5 22 ctJxb5 l:.a8 with chances to fight on despite White's extra pawn. 1 9 l:tfc 1 ! i.d4 Or 1 9 ... .i.d6 20 ctJc6 .i.xf4 21 ctJxf4 l:.b7 22 e5 ctJe8 23 11xa4 with a decisive lead. 20 l:txa4 i.xb2? Losing the exchange. After the more accu­ rate 20 ... ctJxd5 21 exd5 .i.xb2 22 11c2 .i.f6 23 .i.e3! ctJb6 24 11g4 White stays on top thanks to his attacking chances and Black's poor co­ ordination. 21 0,e7 + ! 'it>hS 22 .l:tc2 'liVeS Or 22 ... ctJe5 23 .i.xe5 .i.xe5 24 ctJac6 'iVe8 25 11a5 .i.d6 26 e5 .i.xe 7 27 ctJxa7 ctJd7 28 ctJc6 and White wins. 23 l:txb2 'it'xe7 24 0,c6 'iWc5 25 0,xa7 'liVxa7 26 e5 0,9S 27 i.e3 'iiVa S 2S 'ii'x aS �xaS 29 f4 0,e7 30 l:td2 1 -0 1 83

Th e Ca t a l a n

Summary Practice shows that the S ... ttJbd7 system is difficult for Black, which is why it is seldom seen on the tournament circuit these days. After 6 ... cS 7 ttJa3 (Game 91) 7 ... ttJb6 our attention is drawn to the standard collection of the pawn with 8 ttJxc4! ttJxc4 9 'ifa4+ .i.d7 10 'ifxc4 (note that this is possible after 6 ...ttJb6 7 ttJbd2 cS 8 ttJxc4) . After 6 ....l:tb8 7 a4 b6 (Game 92) Black's defensive task seems sufficiently difficult to prefer 7 ... a6 8 as ttJe4. 1 d4 d5 2 c4 e6 3 liJf3 liJf6 4 g3 dxc4 5 i.g2 liJbd7 (D) 6 'iVa4 a6 7 'iVxc4 Games 25 & 26 (Chapter 3) 6 0-0 6 ... cS (D) Game 9 1 ; 6 ... l:tb8 (D) Game 92 -

-

5. . . liJbd7

1 84

-

6. . . c5

6. . . 'D.b8

I CHA PTER TWEl VE I 4 . . . dxc4 5 �g2 i.d7

1 d 4 d 5 2 c4 e 6 3 tDf3 tDf6 4 g 3 dxc4 5 i.g2 i.d7 This is a logical attempt to provide the traditional 'problem' bishop with a good outpost, Black planning to place the piece on c6 in order to challenge the Catalan bishop. The point is that, compared to accommodat­ ing the bishop with ... b7-b6 or ... b7-bS, in this way Black avoids compromising the queenside strucrure and denies White targets in the form of weakened light squares. S ... .i.d7 was reintroduced into modern practice by Korchnoi, notably in the 7th game of his match with Kasparov in London in 1 983. Not surprisingly this provided a significant contribution to the theory of the line. White can react in two ways, 6 ttJbd2 or 6 'iVc2, both with the intention of regaining the gambit pawn. We should note that the ma­ jority of variations after 6 'iic2 have no inde­ pendent value. Usually there is a transposi­ tion to S it'a4+ .i.d7 6 it'xc4 .i.c6 7 .i.g2, which was discussed in Chapter 3. It makes sense to prevent the arrival of Black's bishop on c6. Consequently 6 ttJeS is the main continuation, and the position after 6 ... .i.c6 7 ttJxc6 ttJxc6 forms the tabia of the S ... .i.d7 system. We then have a typical Catalan gambit siruation where White's initiative

(initially in the centre) compensates for the pawn.

Game 93 Kozul-H . Olafsson

Wijk aan Zee 199 1 1 d 4 d 5 2 c 4 e 6 3 tDf3 tDf6 4 g 3 dxc4 5 i.g2 i.d7 6 tDbd2 6 'ii'c 2 .i.c6 transposes to S 'iVa4+ .i.d7 6 'iVxc4 .i.c6, which leaves 6 ... cS. Then 7 'iixc4 .i.c6 is familiar, while 7 0-0 .i.c6 8 'fixc4 transposes to S ... cS 6 'i*'a4+ .i.d7 in Chapter 7. Vaganian-Portisch, Saint John 1 988 con­ tinued 7 ttJeS ttJc6 8 ttJxc6 .i.xc6 9 .i.xc6+ bxc6 10 dxcS .i.xcs 1 1 0-0 'fidS with the better strucrure for White, although Black was active. After 1 2 ttJc3 'fihs 1 3 'it'g2 0-0 1 4 ttJa4 'ilVdS+ I S f3 .i.e7 1 6 .J:t d l 'it'hS 1 7 l::td4 ttJdS 1 8 l::t xc4 l::t fd8 1 9 a3 l::t a c8 20 b3 .i.f6 21 l::t b l ttJb6 both sides had chances. 6 . . .i.b4 7 "i!Vc2 7 0-0 c3!? 8 bxc3 .i.xc3 9 nb l O-O?! fa­ voured White in Sigulski-Yuferov, Naleczow 1 984 after 1 0 .i.a3 l:te8 1 1 l::t xb7 .ic6 1 2 l::tb 3 .i.xd4 1 3 l::t d 3 e S 1 4 ttJxd4 exd4 I S ttJ f3 'fid S 1 6 'ilVc 2 'it'a s 1 7 ttJxd4 .i.xg2 1 8 'it'xg2 ttJbd7 1 9 ttJc6 'ii' fS 20 l::t d2 but the immedi­ ate 9 ... .i.c6 improves. Then 10 .i.a3 as 1 1 'iic2 .i.b4 1 2 .ixb4 axb4 1 3 l::t xb4 0-0 re1 85

Th e Ca t a l a n

stores material equity with a n unclear posi­ tion. 7 " .b5 7".i.xd2+ 8 i.xd2 i.c6 9 'iVxc4 i.dS 10 'it'd3 ttJc6 1 1 0-0 leaves White with the dark­ squared bishop. However, 7".i.bS! looks best, when play has continued 8 0-0 ttJc6 9 lId1 . One game went 9".ttJaS 10 e4 0-0 1 1 ttJeS i.xd2 1 2 i.xd2 'it'xd4 1 3 i.c3 'iVb6 1 4 'iYd2 ttJc6 I S 'it'gS h 6 1 6 'iVh4 ttJxeS 1 7 i.xeS ttJd7 1 8 Si.d4 c S 1 9 i.c3 and the bishop pair gave White compensation in Kozul-Liang Jinrong, Novi Sad 1 990. In Sosonko-Korchnoi, Wijk aan Zee 1 984 Black accepted the invitation to take on d4: 9".ttJxd4 10 ttJxd4 1lVxd4 1 1 ttJe4 �b6 1 2 Si.e3 1lVa6 1 3 i.d4 i.e7 1 4 i.xf6 gxf6 (1 4".i.xf6 I S ttJxf6+ gxf6 1 6 1lVc3 eS 1 7 '€Vf3 <j;;e 7 1 8 1lVxb7 �xb7 1 9 i.xb7 lIab8 20 i.f3 with equality) IS 'iYc3 eS and White had the trademark Catalan activity in return for the material investment. 8 a4 0-0

.bg5 'iVe8 1 7 i.e4! :!:tf5 1 7".gxhS 18 'it>hl �f7 19 l1g1 and White wins. 1 8 'it'h 1 ! e5 1 9 .l:!.g 1 �xg5 20 hxg6 Also good is 20 'iixgS i.e7 21 �d2. 20 . . . l:!.xg 1 + 21 .l:!.xg 1 'it'g7 21 ...'iie7 22 'iYh7+ does not help Black. 22 'ii'h 7 + 'it'f6 23 g7

23 . . . ttJa6 24 l:!.g6 + 'ilixg6 25 '(/ixg6 + 'it'e7 2 6 dxe5 1 -0

Game 94 Sulava-Farago

Vinkovci 1993

9 0-0 \,{'hite achieves an advantage after 9 axbS i.xbS 10 0-0 ttJc6 1 1 ttJxc4 in view of Black's somewhat weakened queenside. 9 . . . c6 1 0 ttJe4 ttJxe4 1 1 'ii'xe4 bxa4 1 2 ttJg5 g6? ! Black can try 1 2".f5 1 3 'iYc2 iLe7 1 4 ttJf3 ttJa6 although it leaves White with a clear lead after I S 'iYxc4 ttJc 7 1 6 ttJeS. 1 3 'iWh4 h5 1 4 g4! f6 1 5 gxh5! fxg5 1 6 1 86

1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 i.g2 i.d7 6 ttJe5 i.c6 7 ttJxc6 ttJxc6 8 0-0 White elects to keep his Catalan bishop in play rather than surrender it with 8 i.xc6+ bxc6 9 'iYa4 (9 0-0 cS 10 'iYa4+ 'iYd7 1 1 'iYxd7+ ttJxd7 gives Black a good endgame) 9,,:YWd7 1 0 'iYxc4 'iYdS 1 1 'it'xdS cxdS with a level ending. Slightly different is the solid 8 e3 eS 9 iLxc6+ bxc6 1 0 'iYa4 1lVdS 1 1 0-0, e.g. l 1 ...iLe7 1 2 ttJc3 'iYe6 1 3 dxeS ttJd7 14 f4 0-0 I S e4 iLcs+ 1 6 'it>g2 i.d4 \\�th Black's inferior structure no more significant than his healthy counterplay (White's king is a bit exposed here) . Another try is 8 'iYa4, e.g. 8".i.b4+ 9 iLd2 iLxd2+ 10 ttJxd2 0-0 1 1 i.xc6 bxc6 12 'iVxc4

4

�d5 1 3 0-0 l:tfd8 1 4 'ilfxd5 l:txd5 1 5 ttJb3 as and a draw was agreed in Dorfman-Psakhis, Lvov 1 984. Black's doubled c-pawns are weak and White has fIrm control of the c5square, but Black is very active and has pres­ sure both against d4 and down the b-flie. Also possible is 8 ... �d7 9 �xc4 ttJxd4 1 0 ..txb7 l:tb8 1 1 i.g2 ..tb4+ 1 2 ttJd2 0-0 1 3 e3 ttJf5 14 ..tc6!? 'ilVd6 1 5 a3 ttJg4 1 6 it'c2, which was unclear in Wojtkiewicz­ Timoshenko, Fredericksburg 1 999. 8 . . . ii.. e 7 Black's 8 ... ttJxd4?! in Gulko-Korchnoi, Amsterdam 1 989 shouldn't be repeated. Af­ ter 9 ..txb7 l:tb8 1 0 i.g2 i.c5 1 1 ttJd2 c3 1 2 bxc3 ttJb5 1 3 'ilfc2 0-0 1 4 a 4 ttJxc3 1 5 'iVxc3 .i.d4 1 6 'iVa3 i.xa1 1 7 �xa1 the rook was no match for White's bishops. 9 'i'a4 9 e3 has also been played, when Black must choose his next step carefully. Zsu.Polgar-A.Maric, Tilburg 1 994 went 9 ...'ilfd7 10 'it'a4 ttJb4 1 1 'iVxd7+ ttJxd7 1 2 ttJd2 ttJb6 1 3 i.xb7 l:tb8 1 4 i.f3 c 5 1 5 dxc5 .i.xc5 16 i.e2 0-0 17 a3! ttJd3 1 8 b4! i.e7 1 9 ttJxc4 i.f6 20 .l:ta2 ttJxc1 21 l:txc1 l:tfc8 22 l:'Iac2 ttJxc4 23 l:!.xc4 .uxc4 24 l:txc4 with good winning chances for White. Krasenkow-Nei, Juvaskyla 1 99 1 saw 9 ... ttJd5 10 'iVe2 ttJb6 1 1 ttJd2 ttJa5 12 ttJe4 0-0 1 3 ttJc5 'ilVc 8 1 4 i.d2 ttJc6 1 5 l:tac1 l:td8 1 6 l:tfd1 a s 1 7 b3 a4 1 8 b4 l:ta7 1 9 a3 ttJb8 20 b5 c6 21 l:tb1 cxb5 22 l:txb5 and White had the initiative. Tratar-Wells, Bled 1 995 fea­ tured Black's best response: 9 ... e5!? 1 0 i.xc6+ bxc6 1 1 dxe5 'iVxd 1 1 2 ':'xd1 ttJd7 1 3 f4 0-0-0 1 4 ..td2 f6 1 5 e 6 ttJb6 1 6 ttJc3 f5 1 7 e4 i.c5+ 1 8 Wg2 g6 1 9 a4 a6 and the situa­ tion can justifIably be evaluated as unclear. 9 . . . 0-0 1 0 li!.d 1 ? ! A more solid approach is 1 0 e3 ttJb4 1 1 a3 ttJbd5 12 'iWxc4 c6 1 3 l:td1 with a small but solid advantage for White due to his centre, extra space and bishop pair. 1 0 . . . b5! 1 1 'i!Vxb5 ltJxd4 1 2 'ilia4 Forced ( 1 2 'ilVxc4? ttJf3+) .

.

.

.

dx c 4 5 ii.. g 2 ii.. d 7

1 2 . . . ltJxe2 + 1 3 'it>f 1 ltJd4 1 4 ii.. e 3 Not to be recommended is 14 i.xa8?! �xa8 15 l:lxd4 'iVhH 16 'ite2 'iVxc1 (Farago), when White will pay for his cramped queenside and exposed king. 1 4 . . . c5 1 5 ltJc3? ! White should prefer 1 5 ttJa3 l:tb8 1 6 �xa7 ttJd5, e.g. 1 7 ttJxc4 (1 7 ..txd4 cxd4 1 8 ttJxc4 ..tf6 is unclear - White has a danger­ ous queenside majority but Black has central control) 1 7 ...l:ta8 1 8 'iVb7 .l:tb8 and a draw was agreed in Raetsky-Sax, Munich 1 994. 1 5 . . . .l:!.b8 1 6 'ilixa7 '!:!xb2 1 7 ii.. x d4 cxd4 1 8 '!:!xd4 ii.. d 6 1 9 a4 ltJd5!

20 ii.. x d5 After 20 ttJxd5 exd5 21 l:txd5 �f6 Black again has the better prospects. The c-pawn is more dangerous than the a-pawn and it is easier for Black to create threats against his opponent's dark squares. 20 . . . ii.. e 5! 2 1 .l:!.dd 1 ii.. x c3 22 ii.. x c4 'ilif6 Black has sacrifIced his c-pawn but the at­ tack is very strong. 23 .l:!.a2? No better is 23 l:tab 1 ? i.d4!, but White could try 23 l:tac1 .i.a5 24 'iVd4 'iVxd4 25 l:txd4 ..tb6 (25 ... l:txf2+ 26 'iit x f2 .i.b6 27 We3 e5 28 We4 ..txd4 29 l:tb 1 is less con­ vincing - Black should keep as many pieces on the board as possible) 26 l:tf4 g5 27 l:tf3 l:tc8, when White's co-ordination is very poor but at least he can still fIght. 23 . . . 'i!Vf3 0-1 187

Th e Ca t a la n

Game 95 Beliavsky-Karpov

Moscow 1988 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 g3 dxc4 5 .ltg2 .ltd7 6 ttJe5 .ltc6 7 ttJxc6 ttJxc6 8 0-0 'i!Vd7 9 e3

9 . . . �b8 Black can try to hold on to the c4-pawn with 9...ltJdS 10 'iVe2 ltJb6 1 1 ltJd2 ltJaS. Slipak-Adla, Buenos Aires, 1 990 continued 12 ltJf3 i.d6 1 3 i.d2 ltJc6 14 i.c3 ltJe7 1 5 e4 c6 1 6 a4 0-0 1 7 as ltJbc8 1 8 �xc4 and White regained the pawn with a much better game thanks to his space advantage - 1 8 ... a6 19 ltJd2 ltJa7 20 b4 ltJbS 21 i.b2 l:tad8 22 ltJb3 l:tfe8 23 i.h3 ltJg6 24 f4 etc. 9 ... eS 10 'iVe2 and now Hjartasson-Piket, Tilburg 1 989 went 10 ... 0-0-0 1 1 'iYxc4 exd4 12 exd4 ltJxd4 1 3 ltJc3 (1 3 i.f4!? ltJe6 1 4 i.e3 i s interesting, with good play) 1 3. ..�e6 14 'iVa4 'iVa6 1 5 'iVxa6 bxa6 1 6 i.gs i.e7 1 7 l:tac 1 with White's compensation for the pawn coming in the shape of Black's dam­ aged structure. The queens stayed on in Glavina-Toth, Mar del Plata 1 990: 1O ... exd4 1 1 exd4+ i.e7 12 'iixc4 ltJxd4 1 3 i.xb7 l:td8 14 i.g2 cS 1 5 ltJd2 0-0 1 6 ltJf3 'iVrs 1 7 ltJxd4 cxd4 1 8 i.f4 i.cs 1 9 l:tae 1 l:tfe8 20 l:txe8+ l:txe8 21 b4 i.b6 22 a4 and the queenside majority and bishop pair secured White an advantage. 1 88

1 0 'i!Ve2 b5 1 1 b3 cxb3 1 2 axb3 �b6 1 3 .ltb2 13 i.xc6 �xc6 1 4 1ha7 releases much of the tension. 1 3 . . . a6 1 3 ... i.b4 gave White a strong attack in Gleizerov-Cs.Horvath, Budapest 1 989 after 1 4 l:td 1 0-0 1 5 dS! exdS 1 6 i.xf6 gxf6 1 7 l:txdS 'it'e6 1 8 'iYhS. 14 e4 14 l:tdl i.e7 15 ltJc3 ltJb4 16 e4 0-0 1 7 i.h3 (petursson suggests 1 7 f4!?) 1 7...�e8 1 8 ltJa2 ltJd7 1 9 i.c3 ltJxa2 2 0 l:ha2 and White had the initiative in Petursson-S.Polgar, Aar­ hus 1 993, eventually winning the game. 1 4 . . . ttJxd4! The quieter 1 4 ... i.e7 presents White with an opportunity to strike an attacking stance with 1 5 dS ltJb4 16 i.h3 0-0 17 l:td 1 . 1 5 .ltxd4 'i!Vxd4 1 6 �xa6 �xa6 1 7 'i!Vxb5 + ttJd7 1 8 'ii'x a6 .lte7 1 9 'i!Va8 + .ltd8 20 .ltf3 20 'YWa4 �d3 21 b4 0-0 22 l:tdl ltJb6 23 l:txd3 ltJxa4 24 ltJc3 ltJxc3 25 l:txc3 i.e 7 leads to a level endgame. 20 . . . 'i!Vd3 21 'it'g2 ttJe5 22 .lth5 0-0 Not 22 ... c6? 23 ftdl 'iVxe4+ 24 'it'gl 0-0 25 ltJc3 and Black loses. 23 �d 1 .lte7 24 'ii'xf8 + 'it'xf8 25 �xd3 ttJxd3 26 .lte2 ttJb4 27 'it'f 1 ttJc2 28 ttJc3 ttJd4 29 .ltd 1 .ltb4 30 ttJa4 g5 31 ttJb2 f6 32 ttJd3 .ltd2 33 'it'g2 'it'e7 34 'it'f 1 'it'd6 Black is slightly better but is faced with the age-old problem of t1J�ng to find a way through the enemy defences. 35 f3 f5 36 'it'g 1 h6 37 ttJb2 'it'e7 38 ttJc4 .ltc1 39 ttJe5 .ltd2 40 ttJc4 .ltc 1 4 1 ttJe5 h 5 4 2 exf5 exf5 4 3 g 4 h 4 44 gxf5 'it'f6 45 ttJd3 .ltd2 46 'it'g2 'it'xf5 47 'it'h3 .ltf4 48 'it'g2 .lte5 49 'it'h3 .ltd6 50 'it'g2 'it'e6 51 'it'h3 ttJf5 52 .lte2 'it'f6 53 ttJf2 ttJd4 54 .ltd 1 'it'f5 55 ttJd3 .lte7 56 'it'g2 .ltf8 57 'it'h3 .ltd6 58 'it'g2 'it'e6 59 'it'h3 ttJf5 60 .lte2 ttJe3 61 f4 gxf4 62 'it'xh4 'it'f5 63 ttJe 1 'it'e5 64 'it'g5 .lte7 + 65 'it'g6 'it'e4 66 .It f3 + 'it'd4 67 .ltc6 Yz - Yz

4 . . . dx c 4 5 � g 2 � d 7

Summary In reply to S ... .td7 the continuation beginning with 6 ttJbd2 has been played by Catalan ex­ perts grandmasters Sosonko and Kozul. Nevertheless White fails to achieve an advantage here - with 6 ... .tb4 7 'iic 2 .tbS! (as in Sosonko-Korchnoi) Black managed to retain the extra pawn, while it seems White is struggling to fmd sufficient compensation. Consequently ttJbd2 has left tournament practice at the highest level. After 6 ttJeS .tc6 7 ttJxc6 ttJxc6 White achieves nothing from 8 'iia4. Black can aim for simplifications and equal chances with 8 ... .tb4+ or play the more complex positions that oc­ cur after 8 ... 'iid 7. The positions after 8 e3 and 8 0-0 will often simply transpose but, after 8 e3, Black's 8 ... eS is not bad, either. In the event of 8 0-0 .te7 the reputation of 9 'iia4 was harmed by the game Sulava-Farago. However, after 9 'ii'a4 0-0 White is not obliged to play 10 l::td 1 ?! (there is 1 0 e3!?, for example) which, incidentally, was partly rehabilitated in Raetsky-Sax. Finally, after the important 8 0-0 'it'd7 9 e3 l::tb 8 10 'iVe2 bS it would appear that White's ini­ tiative more than compensates the absence of a pawn. 1 d4 d5 2 c4 e6 3 ltJf3 ltJf6 4 g3 dxc4 5 �g2 �d7 (D) 6 ltJe5 6 'iic 2 6 ... cS 7 0-0 .tc6 8 'iixc4 - Game 76 (Chapter 7) 6 ... .tc6 7 'ii'xc4 (D) - Games 21 & 22 (Chapter 3) 6 ttJbd2 - Game 93 6 . . . �c6 7 ltJxc6 ltJxc6 8 0-0 (D) 8 ... .te7 - Game 94; 8 ... 'ii'd7 - Game 95

7 flxc4

8 0-0

1 89

I

INDEX OF COMPLETE GAMES

Alekhine-Junge, Pra,l!,ue 1 942 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Alekhine-Rabar, Afunid; 1 942 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Andersson-Sokolov.A, Bar 1 997

.......................................................

.. . . . .

..

..

.......

.

..................

.

50

Balashov-Beliavsky, Kin} 1 986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 65 Bauer-Ippolito, 1\/eJI' ) 'ork

2000

. .

.

.........

.

.............................

.

.....

.

. . . . .

.. . .

.......

..

............

.

. . . . . .

.

140

... ..

Beliavsky-Karpov, Mo.rcolI! 1 988. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 88 Beliavsky-Mitkov, Panormo

2001 ................................. ................................... ........................ 27

Chetverik-Bunnakin, Na!!),kanizra 1 99 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 72 Chetverik-Ivan, Harkany 1 996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 54 Chiburdanidze-Vaganian, Hie/ 1 994

...............

.

. . .. . ...........................................................

..

.

. 20

Comas Fabrego-Nogueiras, Havana 1 999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Dautov-Tiviakov, Ven/o

2000 .

....

.

......

.

...

. . . .

.....

.

.....

.

.........

. . ..

...........

.

...........

.

......

.

..............

.

......

Dizdar-Sadler, Pula 1 997

..........................................................................................................

24 77

Filippov-Pridorozhni, I ikaterinbulJ!, 1 997. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 69 Filippov-Rausis, I)ubai 1 999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4 7 Filippov-Sulskis, Poland 1 999. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Fominyh-Sveshnikov, I :/i.rta 1 99 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Ftacnik-Dutreeuw, HatulJli 1 999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Gelfand-Aseev, Klajpeda 1 988 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Gelfand-Lutz, l )orilJlund 2002 ...................................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............................. 90 Gimza-Linder, 1 998 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 04 Gleizerov-Barua, Kolkata

2002

. . . .

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.

........

.

...................

.

........................

17

Gleizerov-Ivanov.S, Poland 1 992 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 5 Gleizerov-Raetsky, Riazall 1 99 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 1 Grabliauskas-Benj amin, Nell' York

2000 . .

.........

.

.......

. . ..

.........

.

..........

.

..............

.

. . . . . . . . . . . . . . . .

142

Haba-Gorin, Pardubice 1 999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 09 Hansen.Cu-Van Wely, I.rtanbul 2000

...............

.

........

.

..................

.

.......................

..

....

.

........

.

..

55

Horvath.Jo-Wells, Odorfieill .\·emie.rc 1 993 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

1 90

Index o f Comple te G a m e s

Tilburg 1982 ............................................................................................... 52

Hiibner-Smyslov, Ilincic-Djuric,

Vrnjacka Banja 1999...................................................................................... 1 1 6

Inkiov-Pinter,

Zagreb 1987 ..................................................................................................... 152

Ivanchuk-Korchnoi,

Tilburg 1989 ........................................................................................ 144

Ivanchuk-Kuporosov,

Tallinn 1986 ..................................................................................... 1 0 1

Karpov-Milos,

Buenos Aires 2000............................................................................................. 8 1

Karpov-Piket,

Monaco 1999 .................................................................................................... 1 78

Kasparov-Korchnoi,

Candidates '12jinal Match If!,ame 9), L ondon 1983 .............................. 182

Kengis-Gurevich.M, Jurmala Khalifman-Ivanchuk, Khalifman-Portisch,

1985 ......................................................................................... 23

Minsk 1986 ....................................................................................... 1 13

Rtykjavik 199 1 ...................................................................................... 63

Dubai 2002 ............................................................................................... 88

Kobalij a-Kiriakov,

Korchnoi-Ivanchuk,

IstanbuI2000 ....................................................................................... 162

Korchnoi-Nogueiras,

Moscow 1994 ....................................................................................... 19

Belgrade 1989 ........................................................................................... 1 76

Kozul-Ljubojevic,

Lucerne 1997 ................................................................................................... 25

Kozul-Lputian,

Wijk aan 7,ee 199 1 ................................................................................ 185

Kozul-Olafsson.H,

Kramnik-Gelfand, Astana 200 1 ..............................................................................................

Bad Harzburg 2000 ..................................................................................... 155

Ksieski-Enders,

Wijk aan Zee 2000 ............................... ...................................................... 79

Lputian-Polgar.J, Manor-Anand,

lJlndon 1987 .................................................................................................... 73

Marin-Berescu, Mednis-Prie,

Iasi 1999

.................................................................................. .......................

Correspondence 1986 .................................................................................. 167

Olafsson.H-Geller.E,

Reykjavik 1986...................................................................................... 9

Olafsson.H-Hjartarsson,

Reykjavik 1984 ........................................................................... 1 59

Turnov 1996 ........................................................................................................ 4 1

Panov-Makogonov.M, Pigusov-Aseev,

Kiev 1 938........................................................................................... 40

SevastopoI 1986

Pigusov-Goldin, Piket-Adams,

32

Cannes 2000 ........................................................................................................ 45

Mochalov-Spirin,

Orsag- Haba,

83

unares 1998 ................................................................................................ 86

Kramnik-Svidler,

................................

.

..............

.

.....

.

.........

.

.....

.

.......

.

..........

. . ..

....

92

Irkutsk 1986 ................................................................................................. 1 1

Wijk aan Zee 2000 ............................................................................................ 69

Piket-Van Wely,

A10nte Carlo 1997................................................. ....................................... 1 58

Polgar.Zsu-Benjamin,

Dortmund 1985 ............................................................................... 139

Polovodin-Zviagintsev, Poluljahov-Nikolaev, Portisch-Radulov,

Buenos Aires 1978 ................................................................................... 1 6 1

Psakhis-Stefansson, Raetsky-Ekstrom,

Winnipeg 1997 ....................................................................................... 65

Zurich 1998 ............................................................................................. 1 1 1

Raetsky-Gattenloehner, Raetsky-Ivanov.S,

St. Petersbury, 1994 ........................................................................ 137

Belorechensk 1988 ................................................................................ 6 1

Winterthur 2002 ............................................................................. 34

SimJeropoI 1989 .................... ................... ................................................. 136

Raetsky-Kelecevic,

Silvaplana 1997................................... ............................ ....................... 133

19 1

Th e Ca t a la n Raetsky-Landenbergue, Raetsky-Naiditsch,

Scuol 200 1 ...................................................................................... 30

Dortmund 2000 ..................................................................................... 1 02

Raetsky-Sveshnikov,

Kolontaevo 1994 .................................................................................... 46

Rashkovsky-Grigorian.K, Reti-Bogoljubov, Ribli-Bonsch,

Kishinev 1975............................................................................... 98

New York 1924 .......................................................................................... 38

Thessafoniki 1 988 .............................................................................................. 71

Ricardi-Smyslov,

Buenos Aires 1990 ..................................................................................... 129

Rogers-Chandler,

Wellington 1986 ........................................................................................ 1 19

Romanishin-Ribli, Romero-Antunes, Rustemov-Sax,

Polanica Zdrrj 1993 .................................................................................. 15

Havana 1 99 1 ............................................................................................ 121

Germatry 2000 ............................................................................................... 126

Sandner-Luther,

Bad Zwesten 1 999 ....................................................................................... 128

Shipov-Volzhin,

Hastings 1997/98 ....................................................................................... 156

Stefanova-Kurajica, Sulava-Farago,

Benasque 1997....................................................................................... 1 74

Vinkovci 1993 ............................................................................................... 186

Tkachiev-Solozhenkin,

2000.................................................................................... 135

Topalov-Kramnik,

Linares 1997 ............................................................................................. 13

Tukmakov-Hulak,

Croatia 1999........................................................................................... 1 14

Vajnerman-Novikov,

Lvov 1984 .......................................................................................... 1 1 7

Vakhidov-Ziatdinov,

Tashkent 1987 .................................................................................... 145

Vasilchenko-Meszaros,

Kecskemet 199 1 .............................................................................. 1 8 1

Vila-Sky,

Castrop-Rauxel 1990 .......................................................................................... 48

Wells-Barsov,

York 2000

........................................................................................................

Wojtkiewicz-Dzhandzhgava, Yevseev-Goldin,

1 92

107

Hastings 1989/90 ................................................................ 59

St. Petersburg 1998 ...................................................................................... 132

·

.... .

..

ISBN 1 -85744-346-2

EVE RYMAN C H ESS

9 £ 1 4.99

$ 1 9.95