Anatoly Karpov - Anatoly Karpov's Best Games.pdf

June 19, 2018 | Author: Mi Otro Yo | Category: World Chess Championships, Traditional Board Games, Abstract Strategy Games, Board Games Competitions, Competitive Games
Report this link


Description

Anatoly Karpov'sBest Games Anatoly Karpov Translated by Sarah J. Young B. T. Batsford Ltd, London First published 1 996 © Anatoly Karpov 1996 ISBN 0 7 1 34 7843 8 British Library Cataloguing-in-Publication Data. A catalogue record for this book is available from the British Library. All rights reserved. No part of this book may be reproduced, by any means, without prior pennission of the publisher. Typeset by Petra Nunn and printed in Great Britain by Redwood Books, Trowbridge, Wilts for the publishers, B. T. Batsford Ltd, 4 Fitzhardinge Street, London W l H OAH A BATSFORD CHESS BOOK Editorial Panel: Mark Dvoretsky, Jon Speelman General Adviser: Raymond Keene OBE Specialist Adviser: Dr John Nunn Commissionin Editor: Graham Bur ess Contents Symbols Introduction 1 Karpov - Gik, Moscow University Ch 1969 2 Karpov - Hort, Moscow Alekhine mem 1971 -\Vsmejkal - Karpov, Leningrad IZ 1973 arpov - Polugaevsky, Moscow Ct (6) 1974 5 Karpov - Spassky, Leningrad Ct (9) 1974 6 Karpov - Korchnoi, Moscow Ct (2) 1974 7 Karpov - Vaganian, Skopje 1976 8 Tatai - Karpov, Las Palmas 1977 9 Karpov - Korchnoi, Baguio City Wch (14) 1978 1 0 Karpov - Korchnoi, Baguio City Wch (32) 1978 1 1 Timman - Karpov, Montreal 1979 1 2 Korchnoi - Karpov, Merano Wch (9) 1981 1 3 Karpov - Korchnoi, Merano Wch (18) 1981 14 Karpov - Yusupov, USSR Ch (Moscow) 1983 1 5 Karpov - Kasparov, Moscow Wch (9) 1984/5 1 6 Karpov - Kasparov, Moscow Wch (27) 1984/5 1 7 Karpov- Kasparov, Moscow Wch (4) 1985 1 8 Kasparov - Karpov, Moscow Wch (5) 1985 ,!.2_ Karpov - Kasparov, Moscow Wch (22) 1985 �Q·· Karpov - Beliavsky, Moscow tt 1986 2 1 Karpov - Kasparov, London/Leningrad Wch (5) 1986 22 Karpov - Kasparov, London/Leningrad Wch (1 7) 1986 2} Karpov - Kasparov, London/Leningrad Wch (19) 1986 241Karpov - Sznapik, Dubai OL (14) 1986 25 Karpov - A.Soko1ov, Linares Ct (10) 1987 26 Kasparov - Karpov, Seville Wch (2) 1987 CZ1';, Karpov - Farago, Wijk aan Zee 1988 2f Karpov - Timman, Brussels World Cup 1988 29 Karpov - Kasparov, Belfort World Cup 1 988 ®> Karpov - M.Gurevich, USSR Ch (Moscow) 1988 -€}K 5 6 9 13 18 25 29 33 37 41 45 49 52 55 60 65 68 73 78 83 86 88 91 94 98 1 02 105 1 10 1 13 1 16 1 20 1 25 3J Karpov- Yusupov, USSR Ch ( Moscow) 1988 @ Karpov - Malaniuk, USSR Ch ( Moscow) 1988 3 3 Karpov - Hjartarson, Tilburg 1988 34 Hjartarson - Karpov, Seattle Ct (3 ) 1989 35 Karpov - P.Nikolic, Skelleftea World Cup 1989 Karpov - Yusupov, London Ct ( 8) 1989 37,, Karpov - Timman, Kuala Lumpur Ct ( 4) 1990 38 Karpov - Anand, Brussels Ct ( 4) 1991 39 Karpov- Short, Linares Ct ( 7) 1992 40 Karpov - Kamsky, Moscow Alekhine mem / 992 (�l Kamsky - Karpov, Dortmund 1993 42 Karpov - Kramnik, Linares 1994 - ::41' Karpov - Beliavsky, Linares 1994 � 1 28 131 1 33 1 36 1 38 142 145 1 50 153 1 57 1 63 167 1 72 Symbols 1-0 0� 1 lf2_lf2 + ++ # ! ? !! ?? !? ?! Ch Cht Echt tt jr worn rpd Web z IZ Ct OL Corr (D) (n) White wins Black wins Draw Check Double check Checkmate Good move Bad move Outstanding move Blunder Interesting move Dubious move Charnpionship Team championship European team championship Team tournament Junior Event Women's event Rapidplay World Championship Zonal Interzonal Candidates Olympiad Postal game Diagram follows nth match game Introduction In 1968 I started at Moscow State University as a seventeen-year-old, and in 1969 I won my first tournament in Moscow. You could say that the Mos­ cow University Championship was my first step in the struggle to reach the top in chess. Immediately after it I was selected for the World Junior Championship, won it, and six years later, in 1975, became 'adult' World Champion. A quarter of a century has passed since 1 969, and quite re­ cently, in 1994, I won the ' super-tournament' at Linares - the 'World Championship Tournament' , and what is more, I can say without false modesty that I had a unique result - 1 1 points out of 13, and two and a half points clear of Kasparov. So, this anniversary of a quarter of a century in top-level chess, and also my success at Linares, gave me the idea of compiling a collection of my best games 1969- 1994. In my already fairly long chess life I have taken part in around two hun­ dred tournaments and matches (not counting team competitions, rapid­ plays and blitz tournaments) and won well over a hundred of them. Although it would be difficult to count up precisely, I think that all in all I have won about a thousand games. Obviously, out of a quantity like that it is not at all easy to pick out the very best games. I hope that all of them are models, which the reader can judge, of the clearest examples of my work, and many of them are also part of the true art of chess. It stands to reason that this book includes all my most interesting victo­ ries from matches for the chess crown, from Candidates battles of various years, and from prominent international tournaments. Many of the victo­ ries I have chosen won prizes for their beauty, best game prizes, and most important theoretical game prizes in one of the most popular world chess publications, lnformator. Incidentally, the author of this book has the leading number of these lnformator prizes, way ahead of other grandmas­ ters. Of course, over many years chess players remember fewer and fewer of their old games, as they are displaced by fresher and more important du­ els. That is why this book in the main concentrates on my victories over the last ten years. Introduction 7 I have also written new commentaries for the older games, or at least re­ worked the opening part, as theory has advanced so much since then. In an overwhelming number of games in this collection I am playing White. This is not surprising, as beautiful victories are most often gained with the white pieces. Let us quickly run through the composition of this collection. Which of my victories have got into it? The first game, dated 1969, was played when I was still a master, but the next one is taken from a strong grandmaster tournament. Then my pro­ gress towards the Mount Olympus of chess is illustrated - here the reader will find one game from the Interzonal tournament and all the Candidates' matches. My match with Fischer did not take place, through no fault of mine, and in 1975 I became the 1 2th World Champion. Then there are some games from prominent international tournaments, and then two en­ counters from my duels with Korchnoi in Baguio and Merano. You could say that my victory in our anniversary national championship ends the first part of the book. In 1984 a new era began in chess, linked with Karpov-Kasparov con­ frontation. Here the reader will find ten of my victories from my first four duels with Kasparov. I also bring you some games played in between these 'epoch-making' matches. After our battle in Seville, our dispute was interrupted for three years. In this part of the book you will find my victo­ ries in the World Cup, our 55th national championships, and other con­ tests. After 15 years I again had to climb the Candidates ladder, and most of the steps are illustrated with a victory, as well as games from my next duel with Kasparov. This is where you could say the second part of my book ends. Soon I was joining the regular cycle of battles for the crown, which also means regular games from new Candidates matches. Unfortunately, my encounter with Short did not end happily for me, and, alas, my possible sixth meeting with Kasparov did not take place. Therefore there are sev­ eral wins from tournaments in the early 1990s. In 1993 a schism occurred in the chess world (it is scarcely worth dwelling on that question in this book), as a result of which I was able to play another match for the chess crown. Having won my match against Timman, I was again declared FIDE World Champion (Kasparov and Short had played for the title of PCA World Champion). 8 Introduction I end the book with two victories from the 'super-tournament' at Li­ nares in 1994. Although after that I have won other interesting tournament games, I think that the full stop in this book should be placed precisely af­ ter Linares, where I achieved one of the most convincing victories in the entire history of chess. In conclusion it remains for me to thank the chess master and writer Evgeny Gik for his help in preparing the manuscript. Anatoly Karpov Karpov Game 1 Karpov - Gik Moscow University Ch 1969 Sicilian, Dragon The Moscow University Champi­ onship was my first tournament in the capital, and I really wanted to win it. The champion's title was decided by this very game, and who would have thought that my opponent in my first tournament in the capital would later become co­ author for some of my books? c5 1 e4 d6 2 lLlf3 cxd4 3 d4 lLlf6 4 lLlxd4 g6 5 lLlc3 �g7 6 �e3 lLlc6 7 f3 0-0 8 �c4 'it'aS 9 ifd2 �d7 (D) 10 0-0-0 w - Gik 9 The Dragon Variation, one of the sharpest and most fascinating in chess theory. Both sides' aims are the same - getting to the enemy king as quickly as possible. Be­ cause of this, White attacks on the kingside, stopping at nothing, while Black in his turn attacks on the queenside. Games played in this variation are almost always played in an open fighting spirit, and are often awarded prizes for beauty. lLle5 1 1 h4 l:r.fc8 12 �b3 Nowadays the queen's rook is most often placed on c8 (putting the queen on a5 in that case is not obligatory), but at that time mov­ ing the king's rook to c8 was more popular. lLlxh5 13 h5 14 �h6 At the end of the 1 960s this posi­ tion was subjected to thorough analysis, and moreover there was a lively discussion about the knight check 14 . . . lLld3+. In this game I was preparing to test one of my own ideas on this theme, but my opponent managed to avoid my preparation. �xh6 14 ... .:xc3 (D) 15 1Wxh6 The standard exchange sacrifice in the Dragon - on the one hand Black protects himself from the knight lunging onto d5, and on the 10 Moscow University Ch 1969 other he shatters the enemy king's fortress. B w 16 bxc3 'if:xc3 The black queen here is occupy­ ing an ideal position to generate threats to the white king, and it is difficult to believe that this move can already be a decisive mistake. Either 16 . . .lLlf6 or 1 6 ....:r.c8 was necessary. 17 lLle2! (D) In the event of 1 7 ..t>b1 a5 Black obtains fair counterchances. How­ ever, the modest knight retreat to e2 is extremely unpleasant for Black. The knight deals with the problem of ousting the queen beau­ tifully, and simultaneously joins in the attack on the kingside. 'iWcS 17 Alas, after 1 7 ... lLld3+ 1 8 .:r.xd3 'iWa1+ 19 ..t>d2 'iVxh1 20 g4 lLlg3 2 1 'iVxh 1 lLlxh 1 2 2 �e3 ! and 2 3 .:r.d 1 Black's knight is lost. ••• lLlf6 lLlhS (D) 18 g4 19 gS w 20 .:r.xhS! There is no time to lose. 20 lLlg3, which I almost played at the board, had to be rejected because at the last moment I spotted the effective 20... .i.g4 ! , which would exclude the white queen from the game. 20 gxhS 21 .:r.h1 'iWe3+ 22 �bl! ••• Karpov In the Dragon even a slight inac­ curacy can spoil the whole game. Thus, for example, 22 �b2? would have given Black at least a draw: 22 . . . �d3+ 23 cxd3 (23 �bl loses after 23 . . .'ifxf3 ! ) 23 ... 'ifxe2+ 24 �al 'ifxd3 and Black is guaran­ teed a perpetual check. 'i'xf3 22 The white knight is untouch­ able: 22... 'ifxe2 23 'fi'xh5 e6 24 'ifxh7+� 25 'fi'h8+ �e7 26 'iff6+ �e8 27 .::t h8#. 22 ... e6 also does not work: 23 'ifxh5 'ifxf3 (23 . . .�g6 24 'ifxh7+ �f8 25 �g3 and 26 �f5) 24 'ifxh7+ �f8 25 �d4. 23 .::txhS (D) ••• - Gik 1 1 for White here was only revealed ten years after this game. The key move for White was ignored by the world press, as many commentar­ ies only examined, after 24 "ii'xh7+ 'ittf8, 25 .::th6 and 25 'fi'h6+: a) 25 .::t h6 e6 26 .::t xg6 fxg6 27 it'xd7 "ii'xe2 28 "ii'xd6+ �g7 29 "ii'e7+ 'itth8 30 "ii'f6+ �h7 31 "ii'f7+ �h8 32 "ii'xg6 "ii'd l + 33 �b2 'ii'd4+ 34 c3 "ii'd2+ 35 .i.c2 'it'd7 ! and Black maintains equality. b) 25 il'h6+ was also suggested in my notes, with the following variation: 25 . . .�e8 26 il'h8+ �f8 27 .::th7 and now: bl) 27 . . . e6 turns out to be inac­ curate, as the bishop should be placed on e6, not the pawn; after 28 g6 fxg6 29 �d4 'ifxe4 30 'ii'g 8 it'xd4 31 'iff7 + 'it>d8 32 'ifxf8+ 'it>c7 33 il'xa8 the white king easily steals away from the 'perpetual' check: 33 ...'ii'dl+ 34 'it>b2 'ii'd4+ 35 c3 'ii'f2+ 36 .i.c2 "ii'b6+ 37 �al 'ifgl+ 38 .i.bl . b2) 27 ....i.e6 ! 28 g6 (28 �d4 is not dangerous for Black either, as after 28 . . .'ii'xe4 29 �xe6 fxe6 30 g6 .::tc 8 31 g7 il'e l + 32 �b2 White is forced to submit to the perpetual check; or 28 .i.xe6 fxe6 29 g6 'ii'xe2 and then . . . "ii'el+) 28 . . . fxg6 (28 . . . "ii'xe2 29 g7) 29 .i.xe6 "Yi'xe2 30 .:n 'ii'el+ 31 �b2 'ifb4+ 32 .i.b3 'i!fxb3+! (32 . . .�d7 33 .::t xf8 .::t xf8 34 'fi'xf8 'ifxe4 with three 12 Moscow University Ch 1969 pawns for the piece) 33 axb3 �xf7 with very unclear play. Moreover, maybe Black does not have to give up his queen. It is amusing that old and, as we shall see, not particularly precise variations for some time caused experts to change their attitude to­ wards 16 ...'ikxc3, and in some theoretical publications this move has been considered quite play­ able. But all the same, microscopic analysis allowed me to return, and I think definitively, to the old assess­ ment: taking on c3 loses for Black. If after 23 ...lLlg6 24 'i!t'xh7+ �f8, instead of 25 'ii'h6+ or 25 .:h6, White should make the quiet rook move, 25 l:hl ! ! , then Black would be completely helpless. Having thought of this rook ma­ noeuvre, it was easy to find sup­ porting variations: c) 25 .:hi ! ! (D) and then: B c l ) 25 . . ..i.e6 26 lLld4 (but not the continuation 26 .i.xe6 fxe6 27 lbd4 'ii'f7 !) 26 ... 'ii'xe4 27 lLlxe6+ fxe6 28 :Cl+ 'iti>e8 29 'ii'g8+. c2) 25 ...e5 26 lLlg3 ! 'ifte8 27 .:n 'itxg3 28 'i!t'g8+ lLlf8 29 'ii'xf7+ �d8 30 'ilixf8+ .i.e8 31 'ii'xd6+ and Black is forced to lay down his arms. c3) 25 . . . e6 26 ltJd4 'iixe4 (al­ ternatively, 26 .. .'iff4 27 lLlf5) 27 .:n .i.e8 28 lLlxe6+ 'ifte7 29 lLlc7. 24 g6! The courageous pawn sacri­ fices itself, guaranteeing a breach in the enemy defences. The hasty 24 ifxh7+ would have left the black king at large: 24 . . . �f8 and now not 25 ii'h8+ (nor 25 lLld4 ow­ ing to 25 . . .'ilidl + 26 'iti'b2 'ii'xd4+) 25 .. .ri;e7 26 'ii'xa8 due to 26...'ii'xh5. lLlxg6 24 Not 24 ... fxg6 (24 . . . hxg6 is even worse: 25 'ilih8#) 25 'ii'xh7+ 'ifilf8 26 'ii'h 8+ �e7 27 .:h7+ lLlf7 28 'iVxa8. However, if, as in the game, Black takes on g6 with the knight, it seems that everything is in order for him: White's attack has been repulsed, and furthermore he is two pawns down. 25 'i!t'xh7+ 'itf8 26 .:rs ! ! (DJ This unexpected rook move was like a thunderclap in a clear sky for Black ! The elegant geometrical .•• Karpov - Hort 13 � � � � � ­ .I � � � � � w • � w� �' � ... .� ' � �\Wr � B idea immediately decides the fate of the game. Two lines - the a2-g8 diagonal and the f-file - intersect on one critical square - f7. The threat is 27 'ifxf7#, and moreover, the rook is supporting the queen along the file, while the bishop - in the event of 26 . . .exf5 - is support­ ing her on the diagonal. Because of this Black is forced to part with his own queen. 'ilxb3+ 26 exfS 27 axb3 28 lbf4! (D) Another elegant manoeuvre. The aS-rook is not defended, and thanks to this White can com­ pletely destroys his opponent's de­ fences. 28 ... .:d8 29 'ii'h6+ A last finesse - the g6-pawn will drop with check. �e8 29 30 lbxg6 fxg6 . . -�· • • ••• � . �� � . � � ·�· . . -�· . . .<it?. • • B 31 'ii'xg6+ 32 'ilg5+! After 32 exf5 .:fs still resist. 32 33 exfS 34 'ii'g8+ 35 'ii'g7+ ri;e7 Black could ri;e8 .:c8 q;e7 1-0 Game 2 Karpov - Hort Moscow Alekhine mem 1971 Sicilian, Keres Attack cS , 1 e4 d6 2 lbf3 cxd4 3 d4 lbf6 4 lbxd4 5 lbc3 e6 6 g4 (D) The Keres Attack - my favourite weapon against the Scheveningen in the 1970s and 1980s. This vic­ tory is perhaps one of the clearest I gained in this variation, although 14 Moscow Alekhine mem 1971 8 f4 8 .ie3 a6 (two other popular lines are 8 ...4Jb6 and 8 ....ie7) 9 h4 (theory has also minutely studied 9 a4, 9'tlfd2 and 9 :g1) is an impor­ tant alternative ; 8 4Jdb5 4Jb6 9 .if44Je5 10 'ilih5 has been played in its time; now 10 .. 4Jg6 ! (after 1o... g6 1 1 it'h3 .id7 12 o-o-o it'b8 13 it'g3 Black is defenceless) 11 .ig3 ( 1 1 .ixd6 .ixd6 1 2 :d1, inci­ dentally, does not work, as after the cold-blooded 12 . 0-0 13 4Jxd6 'ii'e 7 ! , the initiative is on Black's side in spite of his lack of pawns) l l . . .a6 1 2 4Jd4 .ie7 (12 . . . d5 ! is also interesting) 13 0-0-0 .ixg5+ 14 'it>b1 0-0 is unclear. a6 8 9 .ie3 .ie7 9 . . . h6? is very risky: 1 0 4Jxe 6 ! fxe6 1 1 'ii'h5+ �e7 1 2 .ic4 'ii'e 8 13 1li'h3. However, 9 ...'tlfc7 de­ serves attention. 10 :g1 (D) 10 'tlfd2 4Jxd4 11 .ixd4 e5 1 2 .ie3 exf4 1 3 .ixf44Je5 1 4 0-0-0 .ie6 15 h4 'i!i'a5 16 a3 :c8 174Jd5 led to an even game in Tarjan-Hu­ lak, Indonesia 1983. 10 4Jxd4 Black does not usually hurry with this exchange , but here it is not easy for him to find another move : IO. .'i'c7 looks too sluggish; deciding on 10 ...0-0 when one is directly under attack is not easy; . B almost a quarter of a century has passed since . 6 4Jc6 It is interesting that I used the Keres Attack in the very first game of my long marathon against Kas­ parov (Moscow 1984/5). True, we did not play it again, except a modified form of it in game 14 of our second match, when I chose an unusual move-order: 2 . . .e6 3 d4 cxd4 4 4Jxd4 4Jc6 5 4Jc3 d6 6 g4 !?, not then expecting the knight to go to f6. After a sharp battle the matter ended peacefully. In the first game of our first match Kasparov chose 6 . . . h6 in the diagram posi­ tion, a more popular continuation in recent years (6. . . a6 is a third pos­ sibility). After 7 h4 4Jc6 8 :g 1 h5 ! ? 9 gxh5 4Jxh5 10 .ig54Jf6 1 1 'ilfd2 I gained somewhat the better chances, but that encounter also ended in a draw. 4Jd7 7 gS ... .. ••• ••• . Karpov - Hort 15 B 10 . . . lDc5 1 1 lDxc6 bxc6 1 2 �xc5 gives White a clear advantage. 1 1 'ilt'xd4 e5 exf4 12 'ilt'd2 13 �xf4 lDe5 1 3 . . . 'ilt'b6 14 llg3 'ilt'xb2 1 5 llb1 Wa3 1 6 lDd5 is obviously bad for Black. 14 �e2 �e6 Black could not generate any ac­ tivity with 1 4 . . . 'ilt'a5 as the end­ game which follows is clearly in White's favour: 1 5 lDd5 'ilt'xd2+ 16 'itxd2 �d8 17 llad 1 �e6 18 �c 1 . 15 lDd5 Immediately ! Otherwise Black's queen could take up an active posi­ tion, e.g. 15 0-0-0 'ilt'a5 ! . 1 5 ... �xd5 16 exd5 (D) As a rule, you should try to oc­ cupy a blockading square with a piece. Actually, 16 'ilt'xd5 would also have left me with an advan­ tage, as the d6-square demands B constant defence by Black. But then my e4-pawn would also in some cases need watching, which might restrict my light-squared bishop. Now this bishop is free for its own action, all the more so as its black counterpart is no longer on the board. lbg6 16 Black's position is appreciably worse, so Hort seeks a tactical so­ lution to his problems. Castling on either side ( 16 0-0 or 16 ...'ilt'c7 followed by 16 ...0-0-0) would condemn Black to passive defence. 17 �e3 h6!? It is difficult to give a move like this a symbol. It shows a fighting mood and a desire to carry out the battle in a tactical key. But it also represents a precise assessment of the position with a clear under­ standing of its disadvantages. In short, it is risky for both sides ! 18 gxh6 �h4+ ••• ... 16 Moscow Alekhine mem 197 1 1 9 �d1 (D) B The white king is not troubled by the loss of his castling rights. At the same time, the black king maintains that possibility right to the end of the game, but he does not manage to make use of it. 19 ... gxh6 �f6 20 �xh6 20 ...Wf6 would probably not have worked out for Black in view of the fact that the h4-bishop has perceptibly lost its mobility. Hort was intending idealistically to re­ build his forces, but in chess there are two sides, and I in my turn took some necessary countermeasures. 21 c3 �eS (D) Black seems to have achieved his aim - he is threatening to play 22 ...Wh4+, while after 22 �g5 "ilfb6 23 �e3 Wc7 my achievements will not be that great. But White finds a very strong retort. 22 l:.g4! The h2-pawn is disturbing no­ one (for the time being !). Now the main thing is not allowing the en­ emy queen an active position on the kingside, where my pieces are rather precariously arranged. 22 ... 'iVf6 22 ...�xh2 would have been a relatively better choice, restoring the material balance. But I can un­ derstand my opponent; he wanted to develop. 23 h4! Now not 23 ... li)xh4? because of 24 �g7. In general it will be diffi­ cult for Black to win back this until recently so helpless pawn, which is now gradually finding its strength. 23 . ..0-0-0 was also impossible in view of 24 �g5. 23 ... 'ii'fS Defending against the threat­ ened 24 h5, Black prepares to cas­ tle long. Karpov - Hon 17 24 l:tb4! (D) B A beautiful place for the rook ! In the first place 24 0-0-0 25 .ig4 is impossible, and in any case the b7 -pawn is under threat of capture. 24 .tf6 25 h5 t[je7 Of course, this retreat is less val­ iant than 25 . . . �5. but then his pieces would not be able to gener­ ate any real activity (after 26 l:tf4) . I should note in passing that during all this time it has not been feasible for Black to sacrifice the exchange with ...l:txh6. 'iWe5 (D) 26 l:tf4 27 l:tf3 The rook, which so often lacks agility, in this position shows re­ markable manoeuvrability. It cre­ ates one threat after another, and moreover is acting productively not only in attack, but also in de­ fence. Thus, Black will now gain ... .•• w nothing after 27 ......xh5 28 l:txf6 'iVh l + 29 .tfl (the rook defends both bishops) 29 ... t[jg8 30 •e l+ and White wins. 27 ... t[jxd5 28 l:td3 l:txb6 Nothing better is apparent - if 28 ... t[je7, then 29 .if4 is decisive. 29 l:txd5 "ii'e4 (D) , 18 Leningrad IZ 1973 A hymn to the rook! The chess player's consciousness is used to the working potential of all his pieces, but, you will agree, with rooks this mainly applies to the endgame. 'ii'h 1+ 30 Black is forced to go into this by no means equal exchange. 31 �c2 'iVxa1 i.eS 32 'ifxh6 33 'iWgS Preventing queenside castling and creating irresistible threats to the exposed king. In this desperate position Hort lost on time. 1-0 The game was nominated as the best at the Alekhine Memorial, and then lnformator included it in their best games. Although many years have passed, I will confess that I still love this game. ••• Game{!) Smejkal Karpov Leningrad IZ 1973 Sicilian, Paulsen - 1 2 3 4 5 6 7 8 e4 ltlf3 d4 lbxd4 lbc3 i.e2 0-0 i.e3 c5 e6 cxd4 lbf6 a6 'fic7 lbc6 i.b4 (D) w 9 lba4 One of the most popular posi­ tions in the Paulsen System, which was established long ago. The b6square is weakened, and White dis­ patches his knight to it. As well as this he makes use of the fact that taking the e4-pawn is dangerous: 9 . . . lbxe4 1 0 lbxc6 'iVxc6 1 1 ltlb6 .:.bs 1 2 'i!Vd4 i.f8 1 3 i.f3 d5 (or 1 3 ...f5 14 .:.ad 1 ) 14 c4. 0-0 9 . Black has a wide choice: 9 ...b5 ? or 9 .. d5?; retreating the bishop to d6 or e7, or even 9 . . lbe7. But cas­ Hing is the most reliable way of fin­ ishing his development. bxc6 10 lbxc6 Theory shows a clear-cut way for White to gain an advantage af­ ter either 10...dxc6 or 10...'iVxc6. 1 1 lbb6 The most logical, although 1 1 c4 and 1 1 f4 are also known. 1 1 ... .:.bs (DJ .. . . Smejkal - Karpov 19 -�· ••• �� �..... - · ·�···- . . . . •• . · �····· - · �-···- . · . - �· - . . " ��,� -�U U � �"� . . -��-� � . . . . �"� "�rf.� U · � U " , w 12 tbxc8 l:l.fxc8 Twenty-odd years after this game I decided to test another capture1 2 ... 'ifxc8 (Salov-Karpov, Buenos Aires, Sicilian theme tournament 1 994) 1 3 e5 ltJd5 14 i.c 1 i.c5 and now White, instead of the theoretical recommendations 1 5 i.d3 or 1 5 c4, decided on an aggressive policy - bringing his queen over to the kingside with the standard Sicilian manoeuvre 'ii'd 1-d3-g3: 1 5 'ii'd 3 0-0 16 'it'g3l:l.e8!? 17l:l.dl aS 1 8 b3 a4 1 9 i.h6 i.f8 20 h4 'f!lc7. Thus, pressure on the g7-point is parried, and the ending arising is highly favourable for Black. 13 i.xa6 l:l.d8! (D) At the time when this game was played, the rook move to d8 was presented as a novelty, and moreover a very important one. It looks as though the rook is no worse on e8, but run ahead for a second and have a look at the situation i n the w game after 24 'ifxf7+. As you can see, if the rook were on e8, Black would have had to resign immed.iately. 14 i.d3 14 i.g5 has apparently not been tested in practice, for example, 14 .. .'ii'e5 1 5 i.h4 'ii'xe4 1 6 .i.g3 l:l.a8 17 i.d3 leads to a complex po­ sition. 14 ... i.d6 15 �h1 15 f4 has also been seen more than once. If 1 5 ... e5 16 f5l:l.xb2 17 g4, then 1 7 . ..'ifa5 18 �h 1 i.c5 19 i.c 1 l:l.xa2 20 l:l.xa2 "ii'xa2 21 g5 ltJe8 22 'ifh5 d5 ! 23 g6 lLlf6 is good for Black. .i.e5 15 l:l.xb2 16 c3 ltJg4! (D) 17 'it'cl 18 f4 Of course, 1 8 'W'xb2 loses after 1 8 ...i.xc3 ! . ltJxe3 18 • • - � - • • � ��� � :� � 'I • • � - -�� � :� � ••• • • 20 Leningrad IZ 1973 21 eS Later it was established that here the best move is 2 l . .. g5 !? with roughly the following variation: 22 g3 'ifd6 ! (an important intermedi­ ary manoeuvre) 23 i.e2 i.e5 24 'ifxf7+ Wh8 25 l:td l 'it'c7 26 'ife7 (26 .tg4 d5 ! 27 'ifxe6 i.xc3 28 exd5 cxd5 29 i.f3 We5 30 Wxe5+ i.xe5 with a draw) 26 ... i.xc3 27 i.g4 W'c8 ! 28 i.xe6 l:te8 29 Wc5 l:txe6 30 Wxc3+ �g8 3 1 'ii'c4 'ii'b 8 32 a4 'ife5 Razuvaev-Matulovic, Tbilisi 1 973. 22 g3 'ifd6! i.gS 23 i.e2 24 'ilfxf7+ �h8 25 a4 (D) Incidentally, 25 i.g4 is also a promising coninuation for White, for example, 25 ...i.f6 26 �g2 l:tf8 27 'ii'xd7 'ii'c5 28 �h3 Wc4 29 l:tf2 preserving an advantage for White; Adorjan-Matulovic, Novi Sad 1 973. ..• w 19 'li'xbl 1 9 fxe5 is no good in view of 1 9 'ifb6 with a clear advantage. i.xf4 19 ... But not 19 ... lLlxfl ? 20 fxe5 l2Je3 2 1 W'e2 'it'b6 22 :tel . 20 W'f2 He has to return the exchange, because after 20 l:tf3 lLlg4 Black's chances are better, as 2 1 h3 is im­ possible due to 2 1 ...i.c 1 ! !. 20 lLlxfl 21 :xn (DJ ... .•. B Smejkal - Karpov 21 Strangely enough, I had this po­ sition on the board when I was pre­ paring for the game. True, it turns out to be weak for Black. I had to apply more than a little force to ex­ tricate myself from it, and then I managed to swindle my opponent. 25 Jle7 l:.f8 26 aS 27 'i'c4 l:.xfi+ 28 Jlxfi 'ili'f6! The only way for Black to or­ ganize his position. The bishop be­ longs on c5, and so needs support from the queen on f8. 29 �gl 'ili'f8 30 Jle2 Jlc5 31 Jlg4 'iif2+ d6 (D) 32 �h3 w The critical moment. 33 Jld7 33 'ii'e6 g6 (Black cannot move into an ending with opposite-col­ oured bishops, since 33 ... 'iif l+ 34 Wh4 'ii'f6+ 35 'ii'xf6 gxf6 immedi­ ately loses a pawn to 36 Jld7 and the white king will easily steal his way through the holes in the pawn barrier) 34 'i!fe7 ! Jle3 ! (the strongest continuation because co-operative variations such as 34 ... h6 35 i.e6 \i'fl + 36 �h4 lead to victory for White), and now Black, by defend­ ing against the direct threat, creates counterplay: a) 35 Jle6 'ii'f l + 36 �g4 'i'e2+ (but on no account 36 ... h5+?? be­ cause of 37 �h4) 37 �h3 'ii'f l + 38 Wh4 g5+ 39 �g4 (after 39 'itth 5? Black forces mate: 39...'ii'e2+ 40 Jlg4 'fi'xh2+ 4 1 Jlh3 'ii'xh3#) 39 ...'ii'e2+ 4(1 'itth 3 'iWfl +, etc., is a perpetual check. b) White cannot even advance his passed pawn with 35 a6 since after 35 . ..h6 threatening ...'iWfl + and ...Jlg5+, he would be forced to give perpetual check himself. c) 35 Jle2! (not 35 ...'ii'xe2?? 36 ir'f8#). d) 35 'i!fxd6 also leads to per­ petual check after 35 ... 'iWfl+ 36 �h4 g5+ 37 �h5 'iWf7+ 38 'itth6 ii'g7+ 39 �h5 'i'f7+. Smejkal was counting on win­ ning the c6-pawn and quietly going about attaining an advantage. The move played is perhaps the correct practical decision. g6 33 34 Jlxc6 'ittg7 22 Leningrad IZ 1973 'ifb2! 35 .i.bS A fine move, the idea of which Smejkal had apparently over­ looked. A threat to the h2-pawn has suddenly arisen, and both the c3pawn and the b5-bishop find them­ selves under attack. White has to part with the c3-pawn, but in the first place he has a very dangerous passed pawn. The misfortune for Black lies in the fact that he can never transfer play into an oppo­ site-coloured bishop endgame, because White, by threatening an exchange, can at any moment 'question' the black queen from almost any square. 36 a6 .i.gl 37 We2 Wxc3 (D) • • • • � �-� �-� -� � :� -� .,. �- . . • • • .i. . • • . ·�· . • � • . �� . �tlli • � �..wrll.� " • �,@-� u � . �-� � � . � � w Precise play by Black has al­ most equalized the position, but the presence of queens on the board makes the situation dangerous for both sides. Wet 38 .i.c4 39 Wn ? 39 �g2 was correct. Now Black achieves an advantage. 39 Wh6+ Wxh2+ 40 �g2 41 �f3 Wh5+ 42 �g2 (D) Here the game was adjourned. Analysis shows that with precise play White has chances for a draw, but to achieve this he will have to overcome great difficulties. B Wh2+ 42 .i.d4 43 �f3 44 .i.dS .i.cS .i.d4 45 .i.c6 46 .i.b7? (D) White has fallen into the trap I set for him and made my problems easier. 46 .i.d5 ! was correct. I do not at all want to say that after this move everything is in order for White, because that is really not Smejkal - Karpov 23 the case. In the adjourned position Black can try various ways of achieving victory, but White has defensive resources for every one of them. • • • • �· .i.. -· �• � � �·· . . � . • • • • A �-� • -L:l· • •�o � � . • . • mu � . . . .. . B The position which has arisen on the board is a kind of zugzwang for White. The king cannot go to the only free square, g4, in view of the mate. The white queen cannot move to either g2 or e2, as in that case he will be forced to advance the g3-pawn by means of ...'iVhS+, and definitively compromise the position of the king. In fact, only the bishop can move. In his turn, Black has to find a plan to strengthen his position. As we saw earlier, simple checks here achieve nothing. I chose a plan linked with advancing my kingside pawns. Analysis revealed that this con­ tinuation is not the only one, but it is effective enough. It is better for Black to begin his pawn movement with his bishop on c5 or d4, while it is better for White to keep his own bishop on c6 or dS. As you can see, a system of corresponding squares has arisen. Smejkal feared this and accordingly spent more time ana­ lysing a different continuation, so he did not find this area of corre­ spondence, and took his bishop too far away. I will show the difference in later notes . 46 gS! 47 �g4 hS+ 48 �rs The only move. After 48 �xg5 'iVxg3+ 49 �xhS (49 �fS 'ii'g4#) 49 . . ..i.f2! White has no defence against mate. 48 'iVxg3 49 �e6 'iif2 ! (D) . 24 Leningrad IZ 1973 his advantage: 50 'ii'xf2 .i.xf2 5 1 �xd6 g4 ! and the bishop does not manage to stop the pawn from reaching f1 , whilst in the event of 52 .i.c8 �f6 Black advances the king. The difference lies in the fact that if the bishop were not on b7, but on d5 instead, White could have continued 50 "iixf2 .txf2 5 1 �f5 ! g4 52 .tc4 ! and the passed pawn is held back, while the black king cannot move forward. 'ii'f6+ 50 51 �dS g4 "fle7 52 .tc8 53 .trs �h6 "ilc7 54 "ii'n 'ii'cS+ SS 'ii'el 56 �e6 (D) Black has successfully arranged his forces and easily parries all White's attempts to create counter­ play. 61 'ii'al 'ii'fl 62 'iVbl g3 63 .th3 (D) • • • • . . - . - . �� � . -� . 0. ?&l.& • w;@ A. • • • . .0 • • • • • .i. .W%:-?1 .0:?;!; • • . . 56 57 58 59 60 wn 'ii'el 'ii'dl+ 'ii'aS �gS 'ii'a3 .tcs 'ii'e3 .tb6 " � "'• mu - • � • . B'ii'B • B B White has occupied the last line of defence, but this blockade is only temporary. �h4 63 'iVgl 64 .tgl Black forces an exchange of queens, after which the road to vic­ tory becomes quite short. 65 'ii'xgl .i.xgl 66 �d6 Black would have achieved an effective finale after 66 �f6 .td4 67 .tfl g2! 68 .txg2 �g3 fol­ lowed by ...h4. .i.d4 66 .txa7 67 a7 �g4 68 �xeS .•• B ,__ ' Karpov - Polugaevsky 25 h4 �dS h3 eS �xh3 .txh3+ .tcS! e6 0-1 This encounter won the regular competition in lnformator for the most important theoretical game. 69 70 71 72 Game@ Karpov - Polugaevsky Moscow Ct (6) 1974 Sicilian, Najdorf cS 1 e4 2 tbf3 d6 cxd4 3 d4 4 tbxd4 tbf6 5 tbc3 a6 eS 6 .te2 .te7 7 lbb3 .te6 8 0-0 'fic7 9 f4 lbbd7 10 a4 0-0 1 1 �h1 12 .te3 exf4 tbeS (D) 13 .:txf4 14 aS Fourteen moves have gone and I have not yet made a single com­ ment. This is not surprising - this position is one of the standard pat­ terns in the Najdorf Variation, and it has been seen in practice many times. In the fourth game of the match I played the weaker 14 tbd4, and after 14 ....:tad8 1 5 'ii'g 1 .:td7 1 6 •• • ••• -� ·� �.�. � - •• ·-*-- • . . - . � • • �a • � � � ­ !@�� ��� � �� � " " u �· . i. � u a ••• •<att w .:td 1 .:te8 17 tbf5 .td8 1 8 tbd4 tbg6 1 9 .:tffl tbe5, Black had a lovely game. Later on both 14 'ii'd2 and 14 tbd5 were tested several times. The second of these moves deserves special attention, for ex­ ample in the encounter Hulak-Por­ tisch, Indonesia 1 983, where after 14 tbd5 .txd5 1 5 exd5 tbfd7 1 6 .:tb4 .:tfe8 1 7 a5 .tf6 1 8 .t g1 .tg5 19 tbd2 tbf6 20 tbfl g6 White could have gained an advantage by attacking on the queenside: 2 1 c3 .:te7 22 ifb3 h5 23 h3 lbh7 (pre­ paring ... f5) 24 .:tb6 followed by 25 .:ta4. tbfd7 14 ... It was scarcely worth moving this knight away from the centre. The back rank is free of pieces, and an initial manoeuvre by one of Black's rooks suggests itself ....:tfe8, ....:ae8, or ....:ac8. For ex­ ample, after 14 . . . .:tac8 1 5 tbd4 tbfd7 16 'ii'd2 .:tfe8 17 tbf5 .tf8 1 8 26 Moscow Ct (6) 1974 .:tf2 �h8 in Matanovic-Polugaev­ sky, Moscow 1 977 a tense, com­ plex struggle arose. 1s .:r.n ..tf6 16 lbdS! ..txd5 17 •xd5! ? After the quiet 17 exd5 Black should continue 17 ...lbc4 1 8 ..txc4 'ifxc4, and White's opening initia­ tive is exhausted. In order to pre­ serve it, an idea came to me during the game of sacrificing two pawns straight away. 17 "ikxc2 (D) w Polugaevsky spent more than an hour wondering whether to take or not. He might have been thinking that if I was prepared to sacrifice, it meant that I had analysed every­ thing thoroughly at home, but on the other hand, he could not find the forced loss at all, so he kept search­ ing through it again and again. If Black declines the sacrifice he ends up in a difficult positional bind. 'ji'xb2 18 lbd4 Not going the whole hog (even if that is what Black wanted to do) would lead to a position with material equality and a noticeable advantage to White. For example, after 1 8 . . ...c5 1 9 lbf5 ! 'ji'xd5 20 exd5 the black pawn on d6 cannot be saved. 'ji'cJ 19 .:tabl 1 9 . ....a3 is worse - the black queen finds herself on the edge of the board and totally out of the game. 20 lbfS (D) The aim of this move is to de­ fend the bishop and not allow the queen to c5, which might happen in the variation 20 .:tb3 'ji'cS 2 1 lbf5 'ji'xd5 2 2 exd5 lbc5, when Black can defend himself. Karpov - Polugaevsky 27 Polugaevsky is manoeuvring magnificently. If Black procrasti­ nates, White can improve the inter­ action of his pieces: 2 1 J.d4 'ii'd2 22 l:.b2 or 2 1 l:.fc l Wa3 22 l:.al 'i!i'b2 23 l:.a2 'i!i'b4 24 J.d2 ! trap­ ping the queen. With the move in the game Black avoids this. 21 l:.be1 Imprecise. After 2 1 l:.fe 1 White would have gained a big advan­ tage. Now Black can construct a defence. 21 t:hc5 22 t:hxd6 lhcd3 23 J. xd3 t:hxd3 (D) .1.• �·· �$! · • ..... ··-· . � � . !?L.X ... • . � � � � " •••. . -�· . . ··� . ••• •�n • • a.:•� w 24 l:.ed1 t:hb4 This is the essential difference. If on my 2 1 st move I had played l:.fel, the white rooks would now be on bl and d l. Consequently, Black would only have two reason­ able possibilities: 24. . . lhf2+ 25 J.xf2 'i!i'xf2 26 e5 J.e7 27 l:.fl , and 24 ...t:he5 25 lhxb7, but in both cases White has the initiative. 25 'ii'xb7 (D) I was short of first two, and then one pawn, and perhaps I was there­ fore unconsciously trying to liqui­ date my 'material disadvantage' . However, 25 'i!i'h5 would have cre­ ated the threat of 26 l:.xf6 gxf6 27 'ili'g4+ c;i;lh8 28 J.h6, which would not. have been easy to repulse (25 ...g6 26 'ii'c 5). •• • ••• .'ii' . • ....... .... � . . � u . . • . • • - .,q� .0. • • • • m • •v• •�o . •.:• : •� B 25 l:.ab8 26 'ii'a7 Wc6 The decisive mistake. 26 ...t:hc6 is no good either: 27 'i!i'c7t:hb4 28 'ti'xc2 lhxc2 29 J.b6 with advan­ tage to White. But 26 ...'ii'e2 would have given Black definite counter­ play, although here too after 27 J.b6 White has an advantage. 27 J.f4 Later I discovered that in the press centre Furman was at this .•• 28 Moscow Ct (6) 1974 very moment proving that an ex­ change sacrifice leads to victory for White: 27 llxf6 ! gxf6 28 .i.h6 (threatening 29 'ii'e 3) 28 . . . 'iic2 (or 28 . .. tbd3 29 .i.xf8) 29 .U.c l 'ii'd3 30 1t'c5 ! . But when I played 27 .i.f4 (threatening 28 e5) my trainer con­ tentedly commented, 'and that is good as well' . In fact, the threat of e4-e5 contains Black's activity, while the b4-knight finds itself out of the game. 27 .:.as 28 1ifl .U.ad8 (D) Otherwise e4-e5 with a decisive attack. However, the attack has not been stopped. - �W% • ta'i - • ,� /,ffl i . i •• m • • " · �}A iQ)i ,W£i �,«, R ··· ,..."- • . • . r� � '·' �;� . 8 ��1W illti ���� .\Ulr. · " i{�:y,, � u . . • :•� --"� w 32 .i.h6 The fire is now directed at two points at once, f7 and g7. tbc6 32 In the hope of complicating the game after 33 tbxf7li'c4 ! [editor's note: but then 34 .i.xg7 .i.xg7 (af­ ter 34 . . .'Wxfl + 35 .U.xfl .i.xg7 36 tbh6+ �h8 37 .U.xf8+ .U.xf8 38 tbf5 White has a decisive material ad­ vantage) 35 1ixg7+ rJilxg7 36 .U.g3# is mate]. 33 tiJfS! (D) 33 'ii'b2 The game would have had a cu­ rious finish in the event of 33 ....i.e5 34 .i.xg7 .i.xg3 35 .U.xg3, when Black cannot prevent mate. 34 .i.cl! White frees h6 for the knight with tempo. There is no defence. .•• w 29 'Wg3! The threat is stronger than its execution ! It is a paradoxical prin­ ciple, but it is valid surprisingly often. White does not hurry with the advance e4-e5, and increases the pressure. 29 'ii'c3 30 .U.f3 'fi'c2 31 .U.dfi .i.d4 (D) ••• - ••• Karpov - Spassky 29 B ti'b5 34 Wh8 35 lbh6+ 36 lbxf7+ l:.xf7 Or 36 ...�g8 37 lbh6+ �h8 38 .:.Xf8+ and mate. ..tf6 37 l:.xf7 Wg8 38 ti'f2 39 l:.xf6 gxf6 40 ti'xf6 1-0 Game 5 Karpov Spassky Leningrad Ct (9) 1974 Sicilian, Scheveningen - 1 2 3 4 5 6 7 8 9 9 e4 lbf3 d4 lbxd4 lbc3 ..te2 0-0 f4 ..te3 (D) c5 e6 cxd4 lDf6 d6 ..te7 0-0 lbc6 ..td7 B The so-called Modern Schev­ eningen is seen these days only rarely, and this match with Spassky did not play the final role in this. In my duels with Kasparov in 1 984 and 1 985, where the Sicilian De­ fence arose regularly, my opponent preferred the Classical Scheven­ ingen with a more flexible move­ order - first ... a7-a6 and ...ti'd8-c7, then ... lbb8-c6 and ...l:.f8-e8, and only then .....tc8-d7. Premature de­ velopment of this bishop is playing into White's hands. The alterna­ tives are 9 ...ti'c7 or 9 ...e5. 10 lbb3 Black was preparing to ex­ change knights in the centre and advance the bishop with .....tc6. Thus this knight retreat is quite ac­ ceptable and more precise than the other well-known moves 10 �h1 and 10 'fi'el . With the bishop on d7, Black's counterplay on the queenside falls 30 Leningrad Ct (9) 1974 behind in comparison with White's threatening pawn storm on the kingside, all the more so as the cen­ tre is fully under his control. aS?! (D) 10 Although White also has the initiative after the waiting moves 1 0 ... a6 or 10 ...'ifc7, playing like this is even less dangerous. The a­ pawn's march is superficially ac­ tive, but in strategic terms it is dubious. White gains the excellent b5-square for the knight, and hin­ ders freeing manoeuvres by the black pieces. This game may have been played twenty years ago, but it is still a good model of how White should play in situations like this. ••• w lbb4 1 1 a4! It is more precise to play 1 l ...e5 1 2 'ith1 and only then 1 2 ... lbb4, al­ though 1 3 .tb5 ! .tc6 14 .txc6 bxc6 15 fxe5 dxe5 16 'ii'e2 gives White a clear positional plus in any case. 12 .tf3 .tc6 (D) Returning the knight to d4 is not really to Black's liking, but 1 2 ... e5, with the passive bishop on d7, is not very attractive either. w 13 t2Jd4 g6 Or 1 3 . . . 'ii'b 8 14 'ili'e2 l:l.e8 1 5 l:l.ad 1 '1th8 1 6 lbdb5 l:l.a6 1 7 l:l.d2 l:l.d8 18 l:l.fd 1 d5 19 e5 lbd7 20 'iif2 ! and White had a clear advan­ tage in Klovans-Vasiukov, Mos­ cow 1968. 14 l:l.f2 e5 15 tbxc6 bxc6 dxe5 (D) 16 fxe5 17 'ii'f l! The battle is flaring up around the c4-square, which White is counting on occupying with one of his pieces. If Black managed to prevent this, things would not look at all bleak for him. It makes no Karpov - Spassky 31 w sense to move the rook from the f­ file ( 17 l:td2), as it is still unclear which file will be the centre of ac­ tivity. 'ii'c8 17 18 h3 (D) Of course, there was no point al­ lowing an exchange of knight for bishop ( 1 8 ... lLlg4). with 19 l:tcl (or 19 g4 followed by g5 and .i.g4) 1 9 . . . l:tfd8 20 .i.e2 l:td4 21 b3 (in order to establish the bishop on c4). h5 19 .i.g4 This severely weakens the posi­ tion of the black king. He should have chosen the simple 19 ...'i'c7, escaping from the pin and connect­ ing the rooks along the back rank. 'i'xd7 20 .txd7 21 'i'c4 (D) So, the queen has arrived at her appointed destination. ••• 8 .th4 21 An unpleasant endgame awaits Black after 2l ...'i'e6 22 1fxe6 fxe6 23 l:taf l . 22 l:td2 'i'e7 23 l:tO! I was not tempted by the chance of winning the exchange; after 23 .i.c5 'ii'g 5 24 l:td7 lLlxc2 25 i..xf8 l:txf8 Black has active counterplay. ••• 8 lbd7 18 In the event of 1 8 . . . 'Mi'e6, White could have developed the initiative ••• 32 Leningrad Ct (9) 1974 The d-file does not matter; the d8-square is securely covered. Therefore I have to generate ac­ tivity in other directions. l:l.fd8 (D) 23 The time has come to invite the black knight to move away from its familiar place. 26 llJa6 27 l:l.e2! (D) w 24 llJbl! The craftiness of this move lies in the fact that I have managed to choose the most suitable moment to move the knight into a more ac­ tive position (this can happen par­ ticularly quickly if Black swaps rooks). It is amusing that conse­ quently this unexpected knight retreat has become an almost char­ acteristic illustration of my work. 'iiib7 24 25 'itth 2! It is a rare occurrence when, in the middlegame, the mobility of your king limits the activity of an enemy bishop. 25 �g7 26 c3 White does not intend to ex­ change rooks yet; the major pieces are needed for an attack on the f­ file. This also frees a square for the knight transfer llJd2-f3-g5, and moreover threatens 28 g3 �f6 29 l:l.ef2 l:l.d6 30 �g5. .:r.rs 27 28 llJd2 �d8 29 llJf3 f6 In defending the e5-pawn, Black is simultaneously trying to cover the f-file. But White's attack is al­ ready unstoppable. 30 l:l.d2! The white rook's 'hesitation' might seem illogical. First he occu­ pied the d-file, then abandoned it, and now returns in decisive fashion ••. Karpov - Korchnoi 33 (that is what is important!) onto the open file. 30 �e7 (D) After 30 ... o!Db8 3 l.!Dg5 ! I would have won straight away. .•. w When Spassky made his move, I lost my head a little. At first I thought that I was winning in all variations, but then I looked and could not find a win. Fortunately, my confusion only lasted for a mo­ ment. 31 W'e6 l:tad8 This loses by force. Dragging out resistance is only possible by means of 3 1 ....!Db8, as if setting up the pieces for a new game. 32 .:.Xd8 �xd8 If 32...l:txd8, then 33.!Dxe5 'ilc7 34 'ilt7+ �h8 and 35 'ilxe7 'ilxe5+ 36 'ilxe5 fxe5 37 l:tf6. 33 lld1 (D) Material is level on the board. Black's king has avoided White's . - � . d - d .\Ull • d • � dwd • •••••••• - . � •• �· ·�- . � ��-� . � d -�d d �-"�� u . � � ."•.:• . B immediate threats, but Spassky's position is worsening with every move. The problem is that the black pieces are disconnected and cannot come to each other's help. Now, for example, it is impossible to defend the seventh rank by means of 33 . l:tt7 due to the hang­ ing bishop on d8. 33 o!Db8 34 �cS llh8 35 l:txd8! 1·0 After 35 . . .l:txd8, 36 �e7 ! is de­ cisive. .. Game 6 Karpov - Korchnoi Moscow Ct (2) 1974 Sicilian, Dragon 1 2 3 4 s e4 o!Df3 d4 o!Dxd4 o!Dc3 cS d6 cxd4 .!Df6 g6 34 Moscow Ct (2) 1974 In my long chess career the Dragon Variation has been used against me about 20 times, and I think only once has my opponent managed to gain a draw. But al­ though many years have passed, to this day this encounter with Korch­ noi remains one of the clearest ex­ amples in the whole history of the variation. Possibly the level at which his game was played had a part in that. .tg7 6 .te3 lh c6 7 f3 0-0 8 'ifd2 .td7 (D) 9 .tc4 the symmetrical . . .h5 (now or on the next move). The basic develop­ ment of contemporary theory in this variation has also gone in pre­ cisely that direction. 1 1 .tb3 lheS 12 0-0-0 The immediate 1 2 h5 is also possible. lhc4 12 ... This was Black's last chance of averting the dangerous opening of the h-file by means of 12 . . .h5. .:txc4 13 .txc4 14 hS 14 g4 also leads to a sharp game. lhxhS 14 lh f6 (D) 15 g4 w 10 h4 .:cs In Game 1 examined above, the then fashionable line .. .'i!i'a5 and ....:tfc8 was played. Transferring the f8- rook to c8 was later totally displaced by putting the queen's rook there. Besides, in recent years Black has tended to meet h4 with w A very topical position in those days. Here White has many possi­ bilities: 1 6 .:tdg1, 16 lhd5, 16 .th6, or 16 e5. However, I chose the modest knight retreat to e2, as I had in mind one new idea which I had Karpov - Korchnoi 35 prepared specially for this match. I should note that a detailed investi­ gation of all the finesses of even one of the above-mentioned lines of the Dragon would take many, many pages. 16 lbde2! ? The idea o f this move i s under­ standable - to reinforce the c3square, as exchange sacrifices such as . . Jhc3 often let B lack gain a powerful attack; indeed the Dragon Variation is underpinned by this blow. However, from e2 the knight, when the occasion arises, can eas­ ily be transferred over for a direct attack against the enemy king. 16 ... 'iia5 1 6 . . . lte8 seems safer for Black, as 1 7 �h6 ( 1 7 eS? is no good be­ cause of 17 . . .lbxg4 ! 18 fxg4 �xg4 19 'ii'd3 't!l'c8 with advantage for Black; Bemai-A.Schneider, Hun­ gary 1 976) is met by the retreat 17 ...�h8. The game Klovans-Beli­ avsky, USSR 1 977 continued 1 8 eS lbxg4 1 9 fxg4 �xeS 20 �f4 'ifaS 2 1 �xeS 't!l'xe5 22 lbd5 ltxg4 with roughly even chances. 17 �h6 �xh6 Here as well it was possible to play 1 7 . . . �h8, in the spirit of GM Simagin, sacrificing the exchange but preserving Black's long-range bishop, which proves useful for Black in both attack and defence . On the other hand, 17 . . . ltfc8 leads merely to a transposition of moves after the continuation 1 8 �xg7 rj;xg7 19 'ii'h6+ rj;gS. 18 'iixh6 ltfc8 19 ltdJ! (D) B Only this move at that time was a novelty - by overprotecting c3, White frees the e2-knight to par­ ticipate in the attack. The theoreti­ cal paths 19 ltdS or 1 9 g5 bring White no great advantage, and fur­ thermore Korchnoi was certainly ready for them. But after the mod­ est rook transfer he plunged deep into thought. 19 l14c5 This move apparently leads to a forced win for White. However, af­ ter 1 9 ... �e6 20 gS lbh5 2 1 lbg3 'it'eS 22 lbxh5 gxh5 23 'ii'xhS as well, he has an appreciable ad­ vantage. Botvinnik's continuation 19 ...'ili'd8 was the most stubborn. 20 g5 ... 36 Moscow Ct (2) 1 974 The knights on c3 and f6 are both defending their kings, and therefore it is precisely these pieces which are exposed to the greatest danger (if the black knight moved from f6, it would immedi­ ately invite a white invasion on the square d5). llxgS (D) 20 ••• - • • ••• - .. �� � .. � .. � .. -·� illit �� )";-. • �;l _m - .e.z.J. • • · A� � .?:.'. • -0 !!!f}� •.t� . - � ·· � · - . " � � j� � • �C w��� • • w 21 lidS! Not, of course, 21 lt:Jd5 llxd5 and the main guardian of Black's for­ tress, the f6-knight, remains alive. llxdS 21 ... lle8 22 lt:JxdS Now the queen can no longer manage to return to her camp: 22 . . .'ifd8 23 lt:Jef4 'ii'f8 24 lt:Jxf6+ exf6 25 'fi'xh7#; if 22 ... lbh5, then White wins by 23 lbxe7+ �h8 24 lt:Jxc8. 23 lt:Jef4 ..tc6 (D) It is necessary to control the d5square, as otherwise lt:Jxf6+ and - .. �� .. � m � • • . .... lW� � w lt:Jd5, mating. After the alternative 23 . . . ..te6, I had prepared 24 lt:Jxe6 fxe6 25 lt:Jxf6+ exf6 26 ifxh7+ r.fi>f8 27 ifxb7 'fi'g5+ 28 �bl lle7 29 'ii'b8+ lle8 30 'ii'xa7 (but not 30 llh8+?? <l;g7 and Black even wins, due to the threat of 3 1 . . .'ii'g l#) 30 . . .lle7 3 1 'ili'b8+ lle8 32 'ili'xd6+ with a form of 'windmill' . 24 eS! Cutting off everything on the fifth rank. I was almost dazzled by the wealth of apparently effective possibilities, but only this continu­ ation appears to be decisive. The straightforward 24 lt:Jxf6+ exf6 25 lt:Jh5 'ili'g5+ (the point !) 26 'ifxg5 fxg5 27 lt:Jf6+ �g7 28 lt:Jxe8+ ..txe8 would not have won. 24 ... ..txd5 After 24 ... dxe5 25 lt:Jxf6+ exf6 26 lt:Jh5 mate is inevitable. exf6 25 exf6 The main thing is to be strong to the finish ! It is still not too late to Karpov - Vaganian 37 lose: if 26 lLlh5 (to meet 26 . . . gxh5 by 27 l:tg1 + and 28 "ilg7#) then the sobering 26 ...l:te1+. 26 "ii'xh7+ <iitf8 27 "ilh8+ After 27 .. .rt;e7, 28 lLlxd5+ ii'xd5 29 l:te1+ would be decisive. 1-0 In the second half of 1 974 this game won the best game prize in lnformator, with 89 marks out of 90 ! In the almost thirty-year his­ tory of this contest, never have the grandmaster jury been so unani­ mous! support, . . . b7-b5, has already been prepared. But this rook's pawn move has one essential defect - it does little to assist piece develop­ ment. 6 dxcS Apparently the simplest solu­ tion to the problem. However, 6 .te2 also gives White reasonable prospects. .txcS 6 .tb6 (D) 7 lLlb3 The bishop has many possible retreats - besides this, it could also go to e7, d6 and a7. It is difficult to say which one of these is the safest. •.. Game 7 Karpov - Vaganian Skopje 1976 French, Tarrasch e6 1 e4 2 d4 dS cS 3 lLld2 exdS 4 exdS a6 5 iLlgf3 In those days I almost always chose the variation with 3 lLld2 in this opening. And my opponent, for whom the French Defence was then and is to this day a fundamen­ tal weapon for Black in reply to 1 e4, had prepared a quite rare continu­ ation. The aim of it is to prevent the bishop appearing on b5. Besides that, when the occasion arises he is prepared to play . . . c5-c4, and its w 8 .td3 8 .tg5 gives White nothing: 8 ... lLle7 9 'ii'd2 lLlbc6 10 .te2 0-0 1 1 0-0 h6 12 .te3 l:te8 1 3 l:tad1 lLlf5 ! 14 .txb6 'ii'xb6 15 l:tfe 1 .te6. lLle7 8 lLlbc6 9 0-0 38 Skopje 1976 J. g4 10 :et Both sides are making their own moves, seemingly with little regard for the other. The battle is being fought around the critical d4square. If Black manages to ad­ vance his central pawn to d4, he will have a reasonable game. If, on the other hand, White can guaran­ tee control of this important square for his pieces, his advantage from the opening will not be in doubt. In connection with this, 1 1 J.e3 would now be a mistake because of 1 l . ..d4. 1 1 c3 (D) Immediately pushing back the enemy light-squared bishop by 1 1 h3 makes no sense, as Black can­ not castle at the moment in view of the sacrifice on h7. B 11 12 h3 13 J.e3 h6 J.hS 0-0 Exchanging dark-squared bish­ ops is in principle not to Black's advantage, although, for example, in the event of 1 3 . . . J.c7 he is even further behind in development. 14 J.xb6 'ifxb6 15 'ife2 This does not look very logical ­ the queen is standing in front of the rook on an open file. Nevertheless, Black does not have time to make use of this fact: after 1 5 . . .l:tfe8 his rook is insufficiently defended. 15 l:tfd8 16 l:tad1 aS (D) w Vaganian is trying to become ac­ tive on the queenside by forcing the white knight from b3. He prob­ ably underestimated the ensuing pawn sacrifice, after which the game takes on a forcing character. 17 J.b1 It may seem superfluous to re­ peat that White's main idea is to Karpov - Vaganian 39 control d4, but he has to watch it constantly. Therefore the move 'it'e3, which enters into White's plans, would in this specific case be unsuccessful: 1 7 'it'e3 ti'xe3 1 8 l:.xe3 i.xf3 1 9 l:.xf3 a4, and Black has everything under control. i.xf3 17 18 'it'xf3 a4 19 lbd4 'it'x b2 Not 19 ...lbxd4 20 l:.xd4 'it'xb2, as after 2 1 l:.b4 Black loses a piece (the rook not only defends the bishop, but also blocks the a3-f8 diagonal). 20 lbxc6 Black's final piece is diverted away from defending its king. 20 lbxc6 (D) c3-pawn, and keeping the major black pieces out of the game, but is also creating the threat of a bishop sacrifice on g6, with a mating at­ tack). 22 l:.d7 After 22...l:.e8 I was planning to continue 23 i.xg6 fxg6 24 ti'xg6+ 'it>f8 25 ti'xh6+ 'it>g8 26 l:.e6. 23 i.fS! (D) The point! ... B l:.e7 23 After 23 . . .l:.c7 White would continue with the simple 24 l:.xd5; 23 ... gxf5 loses due to 24 l:.d3 (it must be precisely this rook, cover­ ing the b1 -h7 diagonal, as the other white rook has to control the e-file) 24. . .f4 (24 . . . lbe7 25 l:.xe7) 25 'ii'xf4 ti'c2 (White was threatening 26 'ilfg4+, but 25 . . .f6 is probably more stubborn) 26 l:.g3+ 'it>h7 27 'ilff6 l:.g8 28 l:.xg8. 24 l:.xe7 ..• w g6 21 'ii'fS 22 'it'f6 The white pieces are very har­ moniously placed (in particular, the queen is not only defending the 40 Skopje 1976 I could also have gained an ad­ vantage by means of 24 .i.xg6 fxg6 25 llxe7 li:Jxe7 26 'iixe7, but I wanted more. li:Jxe7 24 ... li:JfS 25 .i.d3 The only move, as there are no other possibilities for the knight (25 . . . li:Jc6 26 .txg6); 25 ....:te8 26 .:tel 'iia 3 27 .i.b5 is no good either. Defending the knight with the king is also unsuccessful: 25 ... �f8 26 llbl 'iixa2 27 .:txb7, and now both 27 ... lle8 and 27 ...1\Va3 should be met by 28 .txg6. 26 .i.xfS gxfS (D) 27 .:tel! The black rook must not be al­ lowed onto his third rank, a s a fter ....:ta6-g6 he could gain reasonable counterchances linked with the dangerous passed a-pawn, by giv­ ing up some pawns. 'iixa2 27 ... 28 'it'xh6 The immediate 28 lle3 sug­ gested itself, but then Black would have had an interesting way of gaining a draw due to his cunning trap: 28 .. .f4 29 'iixf4 'iVbl + 30 �h2 a3 ! 3 1 llg3+ 'tig6 32 .:txg6+ fxg6, and here White has nothing better than a perpetual check, as his queen cannot even approach the black rooks. 28 ... a3 29 'figS+ The queen is heading for f6 with tempo. 29 'i1i>f8 30 'iif6 'it>g8 31 'iixfS 'ifd2 (D) w 32 .:te7! A final finesse. The black rook is now diverted from the a-file. .:tf8 32 'i1i>h7 33 'ifg4+ 'li'h6 34 .:teS Tatai - Karpov 41 35 36 37 38 39 40 l:th5 'ii'f5+ l:txh6 'iff6+ 'ii'xf7+ Wxb7 :as <j;g7 <j;xh6 <j;b7 <j;bS 1-0 sufficient compensation for the pawn, as in the game Kaiszauri­ T.Georgadze, Sukhumi 1977. 8 e6 9 ltJge4 ltJb6! (D) It is ridiculous to defend the pawn with the bishop - 9 . . i..f8, while 9 'ii'e7 loses straight away in view of 10 ltJxdS exd5 1 1 ltJc3. ... . Game S Tatai Karpov Las Palmas 1 977 English Opening ... - 1 ltJf3 c5 2 c4 ltJf6 d5 3 ltJc3 4 cxd5 ltJxd5 g6 5 g3 6 ..tg2 ..tg7 The so-called Modem Variation of the English Opening. The stand­ ard continuation here is 7 0-0, but. 7 ii'a4+ An entertaining idea. White strives to make immediate use of his small advantage in develop­ ment and the lack of harmony among the black pieces. ltJc6 7 After 7 .....td7 8 Wc4 ltJb4 9 0-0 ltJ8c6 10 'ii'xc5 it is not easy for Black to prove that he has compen­ sation for the pawn. 8 ltJg5 In the event of 8 Wc4 ltJdb4 ! 9 0-0 'ilfaS ! 1 0 ltJe4 ii'a6 ! 1 1 'ilfxc5 b6 1 2 'ile3 0-0 1 3 ltJe1 ..te6 14 ltJc3 l:tac8 B lack has more than .. ••• w c4 10 'ii'b5 1 1 ltJa4 A logical continuation of the plan which was begun on move 7 . 1 1 'iWcS is repulsed by means of 1 l . . . ..tf8, while 1 1 ltJcS 0-0 1 2 ..txc6 bxc6 1 3 'ii'xc6 e5 would give Black a dangerous initiative. 0-0 11 12 ltJxb6 axb6 13 ii'xc4 (D) The critical position. White has got his way, having won the c­ pawn. Black meanwhile has man­ aged to remove his king from the 42 Las Palmas 1 977 B centre and is ready to get down to action. Incidentally, the fact that the a-file has been opened is un­ doubtedly to his advantage (the weakness of the doubled pawns will only tell in the endgame, but that is still a long way off). Be­ sides, the white queen has made an early exit from the centre, and now one of Black's problems is how to use this fact wisely, and this, in conjunction with other threats, should guarantee him a definite ad­ vantage. e5!? 13 1 3 . . . .i.d7 also looked tempting, in order to move the rook to the c­ file as quickly as possible. In this case White has to be on the look­ out for danger with every step. However, he does have an interest­ ing resource: a) Not 14 'ii'c 2? lbd4 15 'ii'b 1 .i.a4 16 b3 .i.xb3 17 axb3 l:txa1 1 8 'ii'x a1 lbc2+. ... b) 14 0-0 lbd4 1 5 lbc3 (the only defence against 1 5 ....i.b5) 15 . . . b5 1 6 'ii'd3 b4 17 e3 lbb3 is no use to White either. c) White can try to take the d4square away from the knight by means of 14 e3, but then 1 4...e5 is better for Black. d) 14 lbc3 ! lbd4 15 'ii'd3 ! .i.c6 ( 1 5 . . .b5 looks promising, but here as well there is a satisfactory de­ fence 16 e3 ! lbb3 17 l:[bl lbxcl 1 8 :xc 1 b4 19 lbe2 lha2 20 'ii'b3 fol­ lowed by 2 1 d4) 1 6 .i.xc6 ( 1 6 0-0 is bad due to 16 . . .lbb3 ! 17 'ii'xd8 l:[fxd8 1 8 l:[b1 lbxd2 19 .i.xd2 :xd2 with an advantage for Black) 1 6 . . . bxc6 1 7 e3 ! - the point! 17 . . . lbb3 (White has removed the important knight, without drop­ ping his the queen: 1 7 . . . lbf3+ 1 8 �e2), and 1 8 'ii'xd8 l:[fxd8 19 l:[b1 lbxc 1 20 :xc 1 .i.xc3 2 1 bxc3 :xa2 22 d4 gives rise to an equal ending. It is not surprising that it took me an hour to decide between 1 3 . . ..i.d7 and the move played in the game. 14 'ii'c2 The only move. 14 0-0 is bad due to 14....i.e6, when the queen has no good square. 14 lbc3 is also weak in view of 14 ... .i.e6 15 'ii'e4 ( 1 5 .i.d5 is answered by 15 ... b5 !) 1 5 ....i.f5 followed by 16...lbd4 . 14 ... lbd4 Tatai - Karpov 43 15 ii'b1 fS e4 (D) 16 ltlc3 1 6 . . .b5 allows White the ade­ quate response 17 e3, but perhaps I should have paid more attention to the preparatory 16 ... -te6, threat­ ening 17 .. .-txb3. 21 'it'xd4 (D) ••• w w 17 d3 1 7 e3 gives White a hopele ssly weak position after either 17 ... ltlc6 followed by 1 8 . . . ltle5 or 1 8 ... ltlb4, or 17 ... ltlf3+ 1 8 -txf3 exf3. 17 bS 18 -te3 . Again the only move. 18 e3 ltlf3+ 1 9 -txf3 exf3 20 ltlxb5 �a5+ 2 1 ltlc3 b5 serves no purpose at all. b4 18 19 ltld 1 .:.es fxe4 20 dxe4 21 -txd4 But not 21 -txe4? in view of the reply 2 1 . . . .:.xe4 22 'it'xe4 -tf5 and ... ltlc2+. ••• 22 a3 22 0-0 allows Black a pleasant choice between the two possibili­ tie s 22 ... -tg4 and 22 . . . ii'd2. The latter is more promising, as White is obliged to go for the variation 23 -txe4 -th3 24 -tg2 -txg2 25 <iPxg2 .:.xe2 26 'ilfc1 'ii'd5+ 27 'ittg 1 -td4 with good attacking prospects for Black. 22 -tg4 Black guesses his opponent's crafty little idea: 22 ... bxa3 23 0-0, making use of the undefended po­ sition of the aS-rook. 23 �c2 (D) Ruling out the threat 23 . . .-tf3, because of 24 -txf3 exf3 25 �b3+ and 26 �xf3 . 'it'd3! 23 It is rare that one can place the queen under attack from a pawn . 24 exd3 ••• ••• 44 Las Palmas 1977 �c4 l:lxc2+ 29 �xd5 followed by 29 ....i.f3+. 27 l::txc2+ 28 �xb4 l:lcd2 (D) - - -·­ • •• • • • - - • •• B Allowing the attack to develop elegantly. 24 lbe3 would have lost immediately to 24...'ilfxc2 25 lbxc2 .i.xb2. 24 .:le 1 would not have brought White any relief due to the simple 24 ...bxa3, but even the su­ perior 24 'ii'd2 would have left B lack a number of promising pos­ sibilities, e.g. 24 ... 'ii'xd2+ 25 �xd2 .:ladS+ 26 �e 1 , and now either 26 . . J::tc 8 threatening 27 ...l:lc2, or the immediate 26 ... .i.f3 27 l:lgl (both 27 exf3 exf3+ 28 �fl fxg2+ 29 �xg2 b3, and 27 .i.xf3 exf3 28 e3 lead to a difficult endgame for White) 27 ....i.xg2 28 l::txg2. 24 exd3+ 25 �d2 l:le2+ It is possible that White was not expecting this check when he was considering his own 24th move. 26 �xd3 l:ld8+ 27 �c4 Black emerges with an extra piece after 27 .i.d5+ l:lxd5+ 28 ••• B B B B . . ... ... . . � � . � � �f . �g� . - ;,- -� �� � � ��-� --- D�.i.r,;� ;;;_ a •ttJ• a : ,, � /, 0 , � � ' / r:' !;;;: � " � w Again White's knight is the key piece ! It contains the action of all the black pieces, and must be at­ tacked immediately. 29 f3 Forced, as otherwise the b2pawn will be lost. .i.f8+ 29 30 �aS After a retreat to any other square the g4-bishop jumps aside with check and the g2-bishop is lost. .i.d7! 30 The white bishop is under at­ tack, and if 3 1 .i.fl (or 3 1 lbe3), then 3 1 ... .i.c5, and there is no way of saving White from mate by the rook. 0-1 ••• ••• Karpov - Korchnoi 45 Garne 9 Karpov - Korchnoi Baguio City Wch (14) 1978 Spanish, Open Variation 1 e4 e5 2 .!iJf3 .!lJc6 3 .i.b5 a6 4 .i.a4 .!iJf6 5 0-0 .!lJxe4 The Open Variation of the Span­ ish, one of the most popular open­ ings in both my matches with Korchnoi for the world chess crown. One could almost write a whole book about the games we played with this opening. In this encoun­ ter, for the first time in our theoreti­ cal dispute I managed to use an extremely valuable novelty in this variation. 6 d4 b5 7 .i.b3 d5 8 dxeS .i.e6 9 c3 .i.cS 10 .!iJbd2 0-0 1 1 .i.c2 .i.f5 12 .!iJb3 .i.g4 1 2 . .i.g6 has the same value. 13 h3 .i.h5 14 g4! ? .i.g6 (D) In opening manuals at the time this position was said to be totally in Black's favour. In fact, if you are referring to the rniddlegame, the open position of the white king is the basis for this assessment. But . w I thought of an interesting idea linked with transferring into the sort of endgame where opposite­ coloured bishops do not diminish, but on the contrary even increase, White's advantage. dxe4 15 .i.xe4 exf3 16 .!lJxc5 17 .i.f4! (D) . B If White exchanges queens im­ mediately by 17 Wxd8, then he loses a tempo and Black has time 46 Baguio City Wch (14) 1978 to organise counterplay against the e5-pawn. The point of my play lies in the fact that after the move played Black is forced to exchange queens, which gives White two ex­ tra tempi. 17 'ii'x d1 17 ...'Wh4 is insufficient in view of 1 8 'ti'xf3, defending the h3pawn and attacking the c6-knight. After 17 ..."ile7 White wins a pawn: 1 8 'ii'd5 .!Lla5 1 9 b4 .!Llc4 20 'ii'xf3, and 20 . . ..!Llxe5 is impossible due to 2 1 .i.xe5 'ti'xeS 22 .!Lld7. 18 .:.axd1 .!Lld8! (D) Korchnoi finds the best response to White's unexpected move. Ob­ viously, Black cannot tolerate the beautifully placed white knight on cS. Exchanging off into a rook end­ ing with opposite-coloured bish­ ops promises him some chances for defence. Naturally, I had fore­ seen this possibility. ••• w .!Lle6 19 .:.d7 fxe6 20 .!Llxe6 .:.ac8 21 .i.e3 Exchanging one pair of rooks (2 l ... .:.f7) only weakens the possi­ bilities for defence. After 22 .:.fd 1 White has enough squares to main­ tain the rook on the seventh rank. 22 .:.rd1 (DJ 22 .i.cS immediately deserved attention, so that after 22 . . . .:.fe8 (if 22....:.fd8 { or 22 . . ..:.f7 } . then 23 .:.fd 1 ; 22 ....:.f4 leads to unclear consequences) White can play 23 .:et . • • • • •• . � �1 -;;; : • ,� � -. •• . . .... . � • �- - . • 0%f; �" QJ� -� � • E ·�· . " � ·� �� �;,: - • • �� � :«�� ;r�"' � m1 'I'll' . . � p • . t=. . � B 22 ... .i.e4 Moving the bishop to d5 in an attempt to disturb the co-ordina­ tion of the white rooks is Black's only realistic defensive possibil­ ity. If 22 . . .h5, then 23 gxhS .i.xh5 24 .:.e7, invading with the second rook. .:res 23 .i.cS Karpov - Korchnoi 47 Here is the whole point of B lack's play; he does not put his rook on either f4 or f7, offering to exchange, but takes the e7 -square under his control, and the white rook becomes short of space on the seventh rank. 24 ll7d4 .i.dS 25 b3 aS Again the only possible attempt to create counterplay, as otherwise nothing will oppose White's gen­ eral advance on the kingside. lla8 26 �h2 If 26 ... a4, then 27 c4 bxc4 28 bxc4 .i.c6 29 a3 . lla6 27 �g3 27 ... a4 would have been more logical, and in reply to 28 c4 both 28 ...bxc4 29 bxc4 .i.c6 (now White does not have time for 30 a3 in view of 30 ... lla5 when the e5pawn perishes), and 28 ... .i.c6 are possible, in order to exchange a few more pawns, although in this case as well White has a clear posi­ tional advantage. Korchnoi 's move is part of an in­ correct plan. 28 h4 llc6 (D) Continuing the same mistaken plan. Korchnoi clearly underesti­ mated his opponent's idea, which is described after the next move. 29 llxdS! White immediately gains a bishop and pawn for the rook, and besides, the beleaguered black f3-pawn is already under attack by the king, and the queenside pawns are also becoming objects for at­ tack. 29 exdS 30 llxdS llce6 31 .i.d4 c6 32 lieS (D) Korchnoi may have underesti­ mated the power of this move. If he had played 32 lld7, Black would have had the very strong reply 32 . . . c5 at his disposal, exchanging off one of White's fundamental trump cards - the e5-pawn. Now the white rook is occupying a very powerful position. White wants to isolate the queenside pawns by means of a2-a4, and the f3-pawn is already under attack. llf8 32 In the game it would have been hard to decide on the continuation 32 ...lld8 33 �xf3 lld5 34 .:.Xd5 (or ... 48 Baguio City Wch (14) 1978 B 34 �f4 lhc5 35 .txc5) 34 . . . cxd5. There have been many arguments about this position. I suggest that the mobility of the white king side pawns should in the final analysis lead him to victory, although it is not as simple as it appears at first glance. In particular, after 35 �e3 35 ....:h6 36 h5 g6 the tempting 37 �f4 does not work in view of 37 ...gxh5 38 �g5 h4 ! . 35 a3 was my intended continu­ ation, in order to create a fortress on the queenside with b3-b4, and only afterwards advance the kingside pawns, for example 35 ...g6 36 �g3 .:e8 37 f4 and if 37 ....:f8, then 38 e6 with the plan of h5, and if ... gxh5, then f5, creating two connected passed pawns. Nevertheless, trying to open up the game on the kingside was Black's only chance of survival, since the passive defence to which Korchnoi condemned himself on his last move gives him no hopes of saving himself. 33 a4! In the first place this fixes the objects for attack on the queenside, in order to chain the black rooks down to their defence and divert at­ tention from operations on the op­ posite side of the board. 33 bxa4 34 bxa4 g6 35 .:xa5 .:ee8 In the event of 35 . . . h5 the move 36 .:a6 ties Black down to defend­ ing the c6-pawn and prepares the advance of the a-pawn. 36 .:a7 .:r7 37 .:a6 A cunning little move, forcing the black rook to occupy a passive position. 37 .:c7 37 ...c5 does not work because of 38 .txc5 .:xe5 39 .:a8+ �g7 40 .td4. 38 .tc5 (D) After the bishop has reached d6, the position should be considered won. .:cc8 38 .:as 39 .td6 40 .:xc6 .:Xa4 h5 41 �xf3 42 gxhS gxh5 .:a2 43 c4 44 .:b6 �f7 .:a4 45 c5 ..• Karpov - Korchnoi 49 �e6 46 c6 �d7 47 c7 l:c8 48 l:b8 49 �e3 Hoping for 49 e6+? �xe6 50 l:xc8 �xd6 and the possibility of a rook ending with f- and h-pawns. 49 l:xh4 50 e6+! 1-0 Since 50 . . .'it>xe6 (50 ... �xd6 5 1 l:xc8 l:c4 5 2 l:d8+ 'it>xc7 5 3 e7) 51 .i.g3 is decisive. ••• Game 10 Karpov - Korchnoi Baguio City Wch (32) 1978 Pirc Defence The battle in Baguio lasted for more than two months, and things started rather successfully. After game 27 the score was 5-2, and it looked as though the match was coming to an end. But excessive confidence weakened me, and aided by the fact that he had nothing to lose, my opponent equalled the score at 5-5. In four games I lost everything I had gained in the pre­ vious 27 . I was shattered. All the same I managed to cast off the bur­ den of painful blunders and made up my mind that the 32nd game was to be the decisive one. I played quietly, confidently, and, having gained an overwhelming position, instilled myself with the thought, 'just do not hurry ! ' , as I knew that the desire to realize my advantage had already let me down more than once. 1 e4 d6 2 d4 lbf6 3 lbc3 g6 4 lbf3 .i.g7 5 .i.e2 0-0 c5 6 0-0 In the 1 8th game when the same opening occurred, play proceeded along the main variation 6....i.g4 7 .i.e3 ltlc6, and here I introduced the novelty 8 'ilfd3 ? ! . Wishing to avoid another surprise, Korchnoi this time chose another plan. In principle the position after 6 . . . c5 7 dxc5 dxc5 is thought by theory to be slightly better for White, but I thought my challenger might have prepared an improvement, so I avoided the early exchange of pawns. 7 d5 lba6 50 Baguio City Wch (32) 1978 8 i.f4 liJc7 9 a4 b6 10 l:le1 i.b7 1 1 i.c4 The prophylactic 1 1 h3, with the aim of maintaining the bishop on the h2-b8 diagonal, would have led to a double-edged game after 1 1 ...'i'd7 12 i.c4 l:lad8 1 3 W'd3 e5. liJh5? (D) 11 ... a) 19 liJxd5 is simple and good: 19 ... i.xd5 ( 1 9 ... 'ii'xd5 20 'i'g3 'ifa2 2 1 i.c4 W'xb2 22 lbxt7) 20 c4 i.a8 (or 20 . . .i.e6 2 1 liJc6 'ifxd3 22 lbxe7+) 21 liJd7. b) 19 liJd7 liJxc3 20 bxc3 liJxd7 21 i.xe7 and Black suffers a loss of material. In any other competitive situ­ ation I would certainly have played like that, but in this game I did not want to take risks. liJd7 15 Preventing a break in the centre and hoping for counterplay with 16 ...b5 17 axb5 liJb6. 16 'ife3 i.a8 b5 17 i.h6 18 i.xg7 �xg7 ttJr6 19 i.n axb5 (D) 20 axb5 A pawn attack without a fi­ anchettoed bishop on g7 is like an infantry attack without a prelimi­ nary bombardment. ..• w Losing valuable time. 1 1 . . .iid7 was necessary, trying at any cost to provoke tactical complications: 1 2 e5 dxe5 1 3 lbxe5 'i'f5 14 lbxg6 hxg6 1 5 i.xc7 lbg4. However, af­ ter 12 'i'd3 White has the better chances. 12 i.g5 liJf6 13 'i'd3 a6 l:lb8 14 l:lad1 15 h3 15 e5 ! looks logical. 15 ...dxe5 16 lbxe5 b5 ! 1 7 axb5 axb5 1 8 i.xb5 lbcxd5 and now: w Karpov - Korchnoi 51 21 t:De2 ..tb7 Black's misfortune lies in the poor communication between his pieces. If he had managed to carry out the manoeuvre ...e5, his pieces would have had an easier passage from one side of the board to the other. However, 2 1 ...e5 22 dxe6 t:Dxe6 23 lDg3 would have weak­ ened his pawn structure, and Black would have been forced to move, all but crawling, from one flank to the other along the 8th rank. :as 22 lDgJ %:ta4 23 c3 24 ..tdJ 'iia8 (D) White's knife blow has a firm tactical basis: 25 ...lDfxd5 26 lDh5+ (or 26 lDf5+) 26 ... gxh5 (26 ...�h8 27 'fih6 %:tg8 28 lDg5) 27 'ilfg5+ �h8 28 'ilff5. t:DcxdS 26 'ilfxeS %:ta7 27 ..txbS 28 lDb4 (D) B Now not only do threats hang over the black king, but the prosaic c3-c4 is also unpleasant. Thus, af­ ter 28 ... i.c6 there is the possibility of 29 i.xc6 'ilfxc6 30 c4 lDb4 3 1 %:td6 exd6 32 lDh5+ gxh5 3 3 'ii'g5+ �h8 34 'ii'xf6+ �g8 35 lDf5, whilst 28 . . .'ii'b 8 29 c4 'ii'xe5 30 %:txe5 leads to a hopeless ending. [Editor's note: Black then has the surprising resource 30 ...l:.a5, when it is not even clear that White is better, since 3 1 cxd5? l:.xb5 leaves White with some very weak pawns, while freeing the b5-bishop is not possible by other means. Af­ ter 28 . . .'ii'b 8, White might do bet­ ter to try 29 'iig 5.] 28 ..tc8 29 i.e2! You must agree that there is something attractive about this bishop move. ••• 52 Montreal l979 29 30 c4 31 'ifxcS 32 �n �e6 lbb4 'ifbs l:r.c8 White has an extra pawn and a menacing initiative. l:r.g8 35 36 lbf3 'iff8 37 'ilfe3 �g7?! 37 ...l:r.b7 would have prolonged the battle by preventing movement by the pawns. �d7 38 lbg5 ila8 39 b4 lba5 40 b5 41 b6 In this position Korchnoi sealed the move 4 1 . . .l:r.b7, but the follow­ ing day resigned the game and the match. 1-0 Game 1 1 Timman - Karpov Montreal l979 English, Four Knights A game from the 'tournament of stars' which is noteworthy because the Dutch grandmaster fell into an opening trap I had stored up for Korchnoi, but it was not played un­ til after the Baguio match. lbf6 1 c4 2 lbc3 e5 lbc6 3 lbf3 4 e3 4 g3 is played more often, but because in the Baguio match I had managed to achieve a good game in a variety of lines of the Four Knights English, Timman chose a rarer move. 4 �e7 Here the more popular continu­ ation is 4 .�b4. At the time this modest bishop move to e7 had dis­ appeared from the scene. However, this very game changed its assess­ ment and it again attracted the at­ tention of theoreticians. 5 d4 As will become clear, this natu­ ral move gives Black a wonderful game. Evidently, 5 �e2 0-0 6 0-0 is a better option for White, and if 6 . . d5, then 7 cxd5 lbxd5 8 d3 leads to a reversed Scheveningen Sicil­ ian with an extra tempo for White. ..• .. . Timman - Karpov 53 exd4 5 ... 6 �xd4 Or 6 exd4 d5 ! 7 cxd5 �xd5 8 .i.b5 0-0 with a good game for Black. 0-0 6 bxc6 7 �xc6 8 .i.e2 dS 9 0-0 .i.d6 'fie7 10 b3 1 1 .i.b2 (DJ B This position had been seen in practice before our game, and moreover it was thought to be in White's favour. Thus, 1 1 . . ..:r.d8 12 cxd5 'fie5 ( 1 2 . . . cxd5 1 3 �b5 .i.a6 14 �d4 ! .i.xe2 1 5 "ii'xe2 'ii'e5 16 g3 l:te8 1 7 l:tac 1 gives White a po­ sitional advantage) 1 3 g3 .i.h3 14 .:r.e1 .i.b4 1 5 'ifc2 .i.f5 1 6 'ilc l cxd5 17 .i.f3 'fie7 1 8 a3 .i.a5 19 b4 .i.b6 20 �xd5 ! is obviously to White's advantage; Keene-Jans­ son, Haifa OL 1976. 1 1 ... dxc4! A move which sharply changes the assessment of the position. This idea was also prepared for the Ba­ guio match. Black solves his fundamental problem by completely widening the action of his pieces, which are directed towards an attack on the kingside. 12 bxc4 If White had taken with the bishop all sorts of attacking ideas would have appeared for B lack, linked with the weakened control over g4, for example 1 2...'fie5 1 3 g 3 .i.h3, 1 2. . .�g4 1 3 g 3 �xh2 or 1 2 . . . .i.xh2+ !?. Now the queen's rook joins the battle with tempo. 12 ... l:tb8! 13 'i!fcl (DJ Not falling into the trap 13 .:r.b1 ? .:r.xb2 1 4 .:r.xb2 'ii'e5 ! . B 13 .. . �g4 54 Montreal l 979 Black's attack is developing easily and naturally. Not a shadow remains of White's opening advan­ tage, and on the contrary he is watching with alarm as events de­ velop on the kingside. 14 g3 After 14 �xg4 �xg4 1 5 l:r.e 1 Black increases the pressure with 1 5... l:r.b4 ! . 1 4 ... l:r.e8 Of course, I could have played 14 . . . lDxh2 1 5 �xh2 'ili'h4+, imme­ diately forcing a draw. But now the threat of the knight sacrifice on h2 is not so harmless, since the rook would be included in the game along the sixth rank with decisive effect: 15 . . .lDxh2 1 6 �xh2 'ifh4+ 1 7 �g2 'ifh3+ 1 8 �g1 �xg3 1 9 fxg3 'ifxg3+ 2 0 � h 1 l:r.e6 and the rook reaches h6. 15 lDd1 (D) After 1 5 �f3 there follows 15 ...'ili'f6 16 �xg4 ( 1 6 �g2 is bad: 1 6 ... 'ili'h6 1 7 h3 lDe5 ! ) 1 6 . . . �xg4 17 f3 �h3 1 8 l:r.f2 'ii'g6 with a pow­ erful initiative for Black. 15 lDxh2! The sacrifice which helps Black to increase his advantage. Timman, of course, saw this blow, but placed his hopes on the strength of his replying intervening move. How­ ever, when I began the combina­ tion, I had carefully taken into account all the tactical nuances. ••• ltt • • ••• Wi • �•-� �·-· �A• � � � � � � � % %, � • • • • · � · ••• . �• � � � u �� � - Bi.� B � a NI : �-��� , "'Z.J ��� : %= /. � %; " / / � w� " B 16 cS Not, of course, 1 6 �xh2 'ii'h4+ 17 �g2 'ii' h 3+ 1 8 �g1 �xg3 1 9 fxg3 'ili'xg3+ 20 �h1 l:r.e4 ! (or 20 . . . l:le6 2 1 �f6) 2 1 l:lf4 �h3 and everything is over for White. 16 lDxfl! 17 cxd6 (D) Korchnoi - Karpov 55 for White, he has absolutely no choice. 18 fxg3 'ii'xd6 19 � 'ii'h6 20 ..td4 'ii'h2+ 21 'it>e1 'ii'xg3+ 22 'it>d2 'ii'g2 23 ltlb2 ..ta6 24 ltld3 ..txd3 25 'it>xd3 l:.bd8 26 ..tn 'ii'e4+ cS! (D) 27 'it>c3 w Clearing the final approach to the white king. 28 ..txcS 'ii'c6 29 'it>b3 l:.b8+ 30 'it>a3 :es 31 ..tb4 'ii'b6 0-1 The merit of the novelty used in this duel was recognized - it won the lnformator prize for the most important theoretical game. Re­ markably, in the very same issue of lnformator it also won the prize for the best game ! In the almost thirty­ year history of these prizes, this double has only happened twice. Game 1 2 Korchnoi - Karpov Merano Wch (9) 1981 QGD, Orthodox 1 2 3 4 5 6 7 c4 ltlc3 d4 ltlf3 ..tgS ..th4 :et e6 dS ..te7 ltlf6 h6 0-0 dxc4! (D) w Amazingly, this simple move had never been seen in grandmas­ ter practice before. Chess really is inexhaustible - even in such a fa­ miliar opening as the Queen's Gambit it is still possible to think up something new on move seven. 56 Merano Wch (9) 1981 8 e3 After 8 e4 llk6 ! Black has quite a playable game. 8 cS 9 i.xc4 cxd4 10 exd4 The idea of an early exchange on c4 was used by Kasparov in game 23 of our first match. I cap­ tured on d4 with the knight, and it quickly ended peacefully after 1 0 lL!xd4 i.d7 1 1 0-0 lL!c6 1 2 lL!b3 %lc8 1 3 i.e2 lL!d5 ! with a mass of exchanges. lL!c6 10 1 1 0-0 (D) bishop on h4 directed me towards an unusual idea, so I abandoned the standard idea of playing ...lLlc6-b4d5. 11 lL!hS! Exchanging bishops is useful for Black, but the straightforward 1 1 . . .lL!d5 12 i.g3 would allow White to avoid it, and in the event of 1 1 ...lL!e4 1 2 i.xe7 lL!xc3 1 3 bxc3 White i s clearly better. 12 i.xe7 After 12 i.g3 lL!xg3 1 3 hxg3 i.f6 White has immediate difficul­ ties defending the d-pawn. lL!xe7 (D) 12 B A typical position with an iso­ lated d-pawn has arisen. But there is one important nuance which is favourable for Black. In situations like this, White's dark-squared bishop does not usually hurry to get into the game, but waits for the best moment. The position of the w An important link in Black's plan; one knight temporarily finds itself at the edge of the board, while the second has managed to take control of d5 . Here White could rid himself of the weak pawn: 1 3 d5 exd5 1 4 lL!xd5 lL!xd5 1 5 i.xd5 lL!f4 16 i.e4 'ii'xd 1 1 7 ••• ••• Korchnoi - Karpov 57 :cxd1 i.e6 with equality. Korch­ noi did not want to give up his ad­ vantage as White so easily, and as a result he fell into a difficult position. 13 i.b3 lbf6 14 lbe5 i.d7 15 'ilfe2 :cs 16 lbe4 White cannot find a clear-cut plan, and fails to take account of the fact that every exchange of mi­ nor pieces reduces the dynamic po­ tential of the d4-pawn. 1 6 :fe l looks expedient, inci­ dentally preventing 16 . . .i.c6 be­ cause of 1 7 lbxf7. In that case I would have preferred to reply with 1 6...:c7 or 1 6...i.e8. lbxe4 16 17 'ii'xe4 i.c6! (D) an isolated pawn. In that case his knight would be an able and reli­ able defender of his pawn, while also attacking the enemy d4-pawn. At the same time the function of the white bishop is limited. 18 lbxc6 :xc6! Very precisely seizing the c-file, in case White wants to move his rook to the kingside. After 1 8 ...lbxc6 1 9 d5 exd5 20 i.xd5 the initiative would have passed into Korchnoi's hands. 'Mid6 19 :c3 20 g3 It was not worth hurrying to open an escape hatch. Now White's prospects on the kingside have dis­ appeared. :ds 20 21 :d1 (D) 58 Merano Wch (9) 1981 the possibility of attacking on the kingside, it is important for Black to preserve his major pieces from exchange so that he can put pres­ sure on the isolated pawn. At­ tempting to build up on the d-file does not yet work ( 2 1 . . .'ifd7 22 .i.a4), and he did not want to make the move . . . a7-a6, as it is not yet clear what formation the queenside pawns will need to take on. The text move chains the bishop to cov­ ering the b2-pawn and leaves the possibility for an attack by ....:tb4. 'iid7 22 'ifel Consistently carrying out the fundamental plan. 22 . . ..:tb4 is par­ ried with 23 l:lc4. 23 .:tcd3 (D) Attempts to gain counterplay on the c-file are refuted by tactics: 23 .Uc5 .Ud6 24 .Udcl lLlc6 25 .i.a4 lLlxd4 ! 26 .i.xd7 lLlf3+ 27 �fl lLlxe1 28 .i.xe6 lLld3. B 23 ••• 24 'iVe4 .:td6 In this situation relieving one­ self of the weak pawn does not work: 24 d5 lLlxd5, and Black un­ ties himself by means of ...'iic 6, ... .:t6d7 and . . .lLlf6. Therefore he would have to exchange minor pieces and defend a difficult major piece ending with a pawn less. 'ifc6! 24 25 'iff4 Here the break 25 'ifxc6 lLlxc6 26 d5 just loses the pawn after 26 ...lLlb4. 25 ... lLld5 Forcing the queen to occupy an unfortunate position. Now 26 'ife4 is impossible because of 26...lLlb4 27 'ifxc6 lLlxc6 28 d5 lLlb4. 26 'iVd2 'ifb6 (D) ••• w The unpleasant threat of . . .lLlb4 has arisen. White stubbornly does not wish to play a2-a3, as he is Korchnoi - Karpov 59 afraid of weakening his bishop's position. Korchnoi's patience was by now exhausted, and further­ more he had very little time left, so he exchanged off minor pieces, depriving himself of any tactical chances whatsoever and condemn­ ing himself to passive defence. .:.xd5 27 .i.xd5 28 .:.b3 Beginning an unsuccessful op­ eration, as a result of which the white rook turns out to be 'offside' . If Korchnoi had planned to play f2f4, he should have done it straight away, without moving the rook from the d-file. 28 29 '6'c3 30 f4 (D) B The only way of preventing ... e6-e5, but at the price of weaken­ ing his own king. 30 b6! ..• 31 .:.b4 b5 32 a4 (D) 32 a5 was threatened, while af­ ter 32 .:.b3 Black should continue 32 ....:.cs and 33 ....:.c4. ... B bxa4 32 ... Not for the sake of winning a pawn, but to distract the white pieces from the kingside. aS 33 '6'a3 34 .:.xa4 Or 34 .:.c4 'ii'b5 35 'iWxa4 'iWxb2. '6'b5! 34 ... Chaining the queen to the de­ fence of the unfortunately placed white rook, and incidentally threat­ ening to invade on e2. I thought that White now had to choose 35 b3, in order to free the queen. This would probably have led to a win­ ning rook ending after 35 ....:.bs, al­ though I will not hide the fact that I wanted to finish the struggle in the middlegame. I was helped in this 60 Merano Wch (18) 1981 aim by Korchnoi's organic dislike of parting with material. 35 :d2 eS! Here Black had several tempting continuations (35 . . .:cs; 35 ... g5), but, in making the text move, it was necessary to have a sort of feel for the geometry of the position. At first I could not manage this, until I found the 37th move, which gave me personal aesthetic satisfaction and was said not to have been ex­ pected by the majority of those pre­ sent in the press centre. 36 fxe5 :xe5 37 ii'a1 'W'e8! ! (D) w Making full use of the queen's ability to strike on both the files and the diagonals. 38 dxeS :xd2 39 l:.xa5 Wc6 40 :aS+ <ith7 41 'W'b1+ g6 4 t ...:c2 is also good. 42 'ifn Not, of course, 42... Wxa8? be­ cause of 43 'ifxf7+, with perpetual check. 43 �h1 Wd5+ 0-1 The idea of the knight thrust to hS did not go unnoticed. This duel won the next lnformator competi­ tion for most important theoretical game. Game 13 Karpov - Korchnoi Merano Wch (18) 1981 Spanish, Open Variation My opponent's choice of opening somewhat surprised me, as the pre­ vious two games in the Ruy Lopez had turned out to be rather difficult for him, especially game 14. Obvi­ ously at the end of the match other openings seemed even more dan­ gerous to Korchnoi. e5 1 e4 tl:lc6 2 tl:lf3 a6 3 .tb5 4 .ta4 tl:lf6 tl:lxe4 5 0-0 b5 6 d4 d5 7 .tb3 i..e6 8 dxe5 tl:lc5 9 tl:lbd2 d4 10 c3 tl:lxe6 11 .txe6 12 cxd4 tl:lcxd4 (D) Karpov - Korchnoi 61 w The Open Spanish underwent fundamental tests in Baguio and Merano. In the final stage of the second of these matches a telling blow was inflicted on B lack. Now after 1 3 tbxd4 1Wxd4 14 1Wf3 l:d8 1 5 a4 White achieves only an even position, as in the old game Capa­ blanca-Lasker, St Petersburg 1914. In game 14 of this match I had al­ ready used a novelty here, taking Korchnoi unawares: 13 tbe4 ! . Hav­ ing thought for a record amount of time, 78 minutes, he answered 1 3 ....i.e7 14 .i.e3 tbxf3+ ( 1 4...tbf5 was the more accurate move my opponent played in game 1 6, when he came out of the opening with no losses), and after 1 5 it'xf3 0-0 1 6 l:fd 1 �e8 17 tbf6+ ! .i.xf6 1 8 exf6 it'c8 1 9 fxg7 l:d8 20 h4 ! he found himself in a critical position. I then gained my fifth victory. For this de­ cisive encounter I had prepared yet another dangerous surprise. 13 a4! The fact that Korchnoi spent 54 minutes thinking testifies to the strength of this novelty. The his­ tory of game 14 was being re­ peated. 13 ... .i.e7 (D) The rook move 1 3 . . . l:b8 loses a tempo, and after 14 axb5 axb5 1 5 tbe4 .i.e7 1 6 .i.e3 White has a seri­ ous initiative; however, 1 3 . . . .i.c5 14 tbe4 .i.b6 1 5 tbxd4 .i.xd4 1 6 tbg5 0-0 1 7 axb5 tbxg5 1 8 .i.xg5 .i.xf2+ 1 9 l:xf2 'ii'xg5 20 bxa6 �xe5 deserves attention, as in the game Tischbierek-Chekhov, Pots­ dam 1 985, when after 2 1 �h 1?, 21 ... �b5 ! would have led to an even game, although it is true that after 2 1 'iWfl ! ? White would have preserved slightly better chances. w 14 tbxd4 (D) The manoeuvre 14 tbe4 has no independent significance, as in this 62 Merano Wch (18) 1981 case play transposes to game 14 of the match (with 14 a4 instead of 14 �e3). Neither 1 4 h3 nor 14 lle l gives White many prospects. After 14 axb5 axb5 (a complex game would have arisen after 14 . . .l£lxb5) 15 llxa8 'iba8 1 6 lbxd4 l£Jxd4 1 7 'ifg4 l£le6 1 8 f4 White has an ap­ preciable advantage. B l£Jxd4 14 It must be said that after the Merano match various defensive resources after both 13 l£le4 and 1 3 a4 were found for Black. But find­ ing an antidote demands more than a little time. In particular, here cap­ turing with the queen, 14 . . . 'ir'xd4, was later tried several times. Here is one revealing example: 1 5 axb5 'ifxe5 16 bxa6 0-0 1 7 'ir'a4 llfb8 1 8 a7 llb7 1 9 l£lf3 'ir'b5 20 'ikxb5 llxb5 2 1 �e3, which gave White slightly the better chances in Hjar­ tarson-Smejkal, Bundesliga 1990. ... 15 l£le4 (D) Every move has many lines, not all of which have been exhausted; here for instance it is also possible to exchange on b5. Evidently, this is the primary cause of B lack's difficulties; the black knight should have remained in its place. 1 5 ... 0-0 is more pre­ cise, for example 1 6 axb5 l£lxb5 17 �e3 'ir'c8 18 'ir'c2 1i'e6 19 f4 llad8 20 l:ta4 l:td7 2 1 l:tfa1 1i'd5 22 h3 f6 23 exf6 �xf6 24 l£lxf6+ :Xf6 25 llxa6 llxa6 26 llxa6 l£Jd4 yielded Black some compensation for the pawn in Adarns-Yusupov, Hastings 1989/90. 0-0 16 �e3 17 f4 Threatening to play 1 8 f5, and if 1 7 ... g6, then 1 8 g4. In a difficult position Black finds the best de­ fence. Karpov - Korchnoi 63 17 18 lUxd1 19 l:[d7 (D) ••• 'iVxd1 l:[tbS B ..tfS 19 ... It was necessary to continue 1 9... ..td8, although after 20 a5 lDf8 2 1 l:[d3 Black's position is not the most pleasant. lbd8 20 fS 21 aS! Emphasizing the passive nature of the black pieces; 2 1 .z:r.xc7 bxa4 22 ..td4 l:[b4 23 .z:r.d 1 .z:r.ab8 is not so clear. lbc6 (D) 21 ... fxe6 22 e6! lbeS 23 f6! Defending against the unpleas­ ant threat of 24 f7+ and 25 tbg5. Simplifying by 23 ....z:r.d8 24 .z:r.xc7 .z:r.ac8 would have led to a hopeless endgame for Black after 25 f7+ �h8 26 .z:r.xc8 .z:r.xc8 27 .z:r.c t . .z:r.cs 24 .z:r.xc7 w 24 . . . lbc4 25 ..tc5 ..txc5+ 26 lbxc5 gxf6 27 lbd7 .z:r.cs 28 lbxf6+ is no good. .z:r.xc7 25 .:.act 26 .z:r.xc7 l:[dS (D) w 27 h3! There is no hurry, and this quiet move deprives Black of the chance of 27 ....z:r.d1 + 28 � lbg4+. h6 27 ... He must not allow the knight to invade g5. 64 M1rano Wch (18) 1981 28 :a7! The most clear-cut way of real­ izing the advantage. 28 et:'lcS gxf6 29 et:'lxe6 :d 1 + 30 'iltf2 i.b4 would have activated the black pieces. 28 ... et:'lc4 The rook is tied to the back rank: 28 ...:d1 + 29 'iPf2 :b1 30 i.d4 lLlc6 3 1 fl+ 'iPh7 32 :as et:'lxd4 33 :xf8 �g6 34 et:'ld6, and the f-pawn will become a queen. 29 i.b6 :bs In the event of29 ...et:'lxb6 30 fl+ �h7 3 1 axb6 Black cannot cope with two passed pawns. i.xcS+ 30 i.c5 gxf6 (D) 31 et:'lxc5 • • ••• a . • • � � W& � �·� ·· � � � "'� . · � u - ••• • • . . . ·� " -� · W& � � u � � � � � · � � � W& • W& • - w At this point Black even has an extra pawn. However, White's pos­ session of the seventh rank, the dif­ ference in possibilities of the knights, and the abundance of weak­ nesses in the black camp leave him no hopes of saving himself. 32 b4 :ds 33 :xa6 �f7 34 :a7+ Avoiding a trap: after 34 :xe6 et:'lxa5 35 bxa5 :cs the rook ending may turn out not to be winning. 34 ... �g6 35 :d7 Since exchanging rooks leads to the loss of a piece - 35 ...:xd7 36 et:'lxd7 et:'ld6 37 lbb6 - Black is forced to give up the open file and resign himself to the loss of his last queenside pawn. 35 ... :es :as 36 a6 37 :b7 'iPfS 38 :xbS 'iPeS 39 :b7 'iPdS 40 :r7! Precisely here; after 40 :e7 eS 41 :n f5 ! Black is still holding on. fS 40 ... 41 :r6 (DJ 1. . • • • • • • • �• ••a • . �·· · · �·· W& • u• � � � W& • • . . . ·� . . -�· • • • = B Karpov - Yusupov 65 In this position Black sealed the move 4 l ...e5. However, the next day Korchnoi resigned without resuming. The match, and together with it the era of our world championship rivalry, ended. 1 0 13 lDfd4 .txd4 14 cxd4 aS 15 .te3 a4 (D) Instead 1 5 ... lDb4 1 6 .i.b1 a4 1 7 lDd2 a 3 1 8 'it'c 1 ! gives White a big advantage. - Game 14 Karpov - Yusupov USSR Ch (Moscow) 1983 Spanish, Open Variation In this game from the 50th anniver­ sary of our national championships we see again the Open Variation, so popular in the World Champi­ onship matches at Baguio and Merano. Not long before this championship, at Linares, Yusupov had already chosen this against me, and after a difficult defence the game ended in a draw. This time that situation was not repeated. 1 e4 eS lDc6 2 lDf3 a6 3 .tbS 4 .ta4 lDf6 5 0-0 lDxe4 6 d4 bS 7 .tb3 dS 8 dxe5 .te6 .tcS 9 c3 10 lDbd2 0-0 11 .tc2 .trs 12 lDb3 .tg6 w 16 lDd2 At Merano (match game 6) I re­ treated the knight to c 1 , and as a result sustained a defeat. Now I de­ cided to return to a manoeuvre I had used in an old game against Savon (Moscow 197 1 ). 16 a3 axb2 17 lDxe4 .txe4 18 l:tb1 When I was commentating on my encounter with Savon, I esti­ mated the position after 1 8 ...dxe4 1 9 l:txb2 �7 20 l:txb5 l:txa2 2 1 'it'b1 1i'a8 22 l:tc l to be favourable for White. The text was introduced into practice by Yusupov. 19 l:txb2 'ii'd7 66 USSR Ch (Moscow) 1983 20 .id3 In Ivanov-Yusupov, USSR Ch (First League) 1979, 20 .ixe4 dxe4 2 1 lhb5 lllxd4 22 .l:.c5 .l:.fd8 led to an equal position. Three years after this game, Hiibner, in two battles with Korchnoi, used the new text move. 20 ... .ixd3 An exchange of bishops hap­ pens all the same. 20 ...b4 does not work because of 2 1 .ib5 .l:.fb8 22 .l:.xb4 .l:.xa2 23 .ixc6. 21 'i'xd3 (D) The pawn structure is more pleasant for White. end the game. However, returning material is not compulsory. 22 lUb1 22 .l:.xb5 .l:.xb5 23 'i'xb5 lllxe5 24 'i'b7 'i'c6 25 'i'xc6 lllxc6 26 .l:.c 1 .l:.a6 only leads to an equal po­ sition. b4 22 23 h3 Following 23 a3 bxa3 24 .l:.xb8+ .l:.xb8 25 .l:.xb8+ lllxb8 26 'i'xa3 'i'c6, Hiibner-Korchnoi, Chicago 1982 ended in a draw after 27 'i'e7 'i'd7 28 'i'a3 although according to Hiibner, by continuing 27 g4 h6 28 f4 llld7 29 f5 lllb6 30 .if2 White would have retained an advantage. Interestingly, this move was my first independent decision. Strictly speaking, this has nothing to do with the pawn move, but with White avoiding the simplification on the queenside, which occurred in the above-mentioned games. 23 h6 (D) Not, of course, 23 ....l:.a3 because of 24 'i'xa3 ! . However, a year later in a game against Popovic (Sara­ jevo 1 984), my opponent played more accurately: 23 ....l:.b6, saving tempi. White had no great desire to spend time trying to realize a mi­ croscopic advantage, and a couple of moves later, after 24 'i'c2 .l:.ab8 25 .l:.c l .l:.8b7 26 'i'c5, he had to ac­ cept a draw. 24 :et .l:.b6 ••• ••• B .l:.fb8 21 In the second Hiibner-Korchnoi encounter (Lucerne OL 1 982) an even position arose after 2 1 ...b4 22 .id2 .l:.fb8 23 .l:.fb 1 'i'g4 24 .ie3 .l:.b6 25 h3 'i'c8, although as a re­ sult of 26 .l:.c2? b3! 27 .l:.xb3 lllb4 Black won the exchange, and in the .•• Karpov - Yusupov 67 · - . . •• with a position which has possi­ bilities for both sides. w In the event of 24... lDa5 there is the unpleasant 25 'lfbl , whereupon 25 . . . lDc4 loses: 26 l:txb4 l:txb4 27 'iVxb4 l:txa2?? 28 'iVb8+ �h7 29 'iih l+. llab8 25 'ilbl After 25 ...l:ta7 too, B lack will suffer, as in the game Popovic­ Timman, Sarajevo 1984. lDd8 26 l:tc5 27 l:tcc2 lDc6 If 27 ... lDe6, then 28 f4, with a pawn storm; therefore Black re­ turns the knight to its place. 28 'ii'c t l:t8b7 29 l:tc5 li:J.e7 30 �h2 lDf5?! (D) White has a clear initiative on the queenside, besides which he can also send forward the kingside pawns, giving his opponent new problems. Black decides to give back the c-pawn to stir up some counterplay. All the same, 30...llb5 or 30 . . .c6 would have been better, w 31 llbc2 32 l:txc7 33 l:txc7 34 g4! 34 ...'iVe2 was threatened, and if 35 llc2, then 35 ...%lxg2+ ! 36 �xg2 lDxe3+. The defence 34 'iVc2 is insufficient in view of the reply 34 ... 'ii'f l ! . However, the move by the g-pawn saves White from all his problems. lDh4 34 After 34 ... lDxe3 35 'iVxe3 the black attack has run dry, and a cru­ cial pawn has gone missing. �h7 35 l:tc8+ 36 'ii'd l 'ii'a6 37 l:tc2 White's aim is to carry out the manoeuvre .i.f4-g3, securing his king position. f5 (D) 37 � �·� � --· � � � -�· . . �·� . . � -� u � . - u � --� � • .,.m • � � u R, � ���� � -� � � � � � � �:� � � � .•• ••• 68 Moscow Wch (9) 1984/5 40 �xh3 "ili'e6+ 4 1 �h2 'ilff5 threatening 42 ..."ili'e4 is dangerous. Now the rook comes to the aid of the king. 40 'ii'e6 41 "ili'hS! Not allowing 4 l ...h5. 41 'ii'e7+ 42 'itxh3 'ii'f7 43 l::thl ! Preventing 43 . . .l::tg 3+ 44 �h4 lth3+. 43 'ii'd7+ 1-0 44 fS It is interesting that Yusupov and I were jointly handed the prize for the most beautiful game in the championship. The founder of the prize for beauty understood that every work of art in chess has two authors ! Then, a little later, this victory won the next competition for the best game in lnformator. •.. w In the event of 37 ...1i'a3 38 ...e2 b3 39 axb3 "ili'xb3 the position can be strengthened by means of 40 l::tc 7. 38 �g3! The very same knight which was meant to bring havoc into White's camp is cold-bloodedly eaten by the king. fxg4 38 gxh3 (D) 39 �xh4 ••• Garne 1 5 Karpov - Kasparov Moscow Wch (9) 198415 QGD, Tarrasch w 40 f4 In his Candidates matches against Beliavsky, Korchnoi and Smyslov, my next opponent for the world chess crown had used the Tarrasch Defence successfully. However in our first duel I managed to find the key to Black's position. dS 1 d4 Karpov - Kasparov 69 2 3 4 5 6 7 8 9 10 11 12 c4 lbf3 cxd5 g3 .tgl 0-0 lbc3 .tg5 lbxd4 .teJ 'ili'b3 (D) e6 cS exd5 lbf6 .te7 0-0 ltJc6 cxd4 h6 .:r.es B Kasparov's opponents in the Candidates matches had played a variety of moves here - 12 1t'a4, 1 2 'j/fc2 and 12 a3, and Black achieved a good game every time. I instead used a rarer continuation, which I had specially prepared for the encounter. lbaS 12 .tg4 13 'ili'cl 14 lbfS GM Lajos Portisch thought of this move. It is better to dispatch the knight to f5 straight away than to play 14 h3 .th5 1 5 .:tad 1 .:tc8 1 6 g4 .tg6 1 7 lbf5 first, a s given i n earlier opening manuals . 14 ... .:r.cs In our first match the Tarrasch Defence was played twice - in this game and earlier in game 7. It was then that Black first chose to place his rook on c8, which was at that point considered a novelty. 15 .td4! (D) A strong move. In game 7 I had exchanged on e7 - 1 5 lbxe7+ .:txe7 1 6 .:tadl 'ii'e 8, and Black quickly equalized. He was defeated not be­ cause of his opening formation, but due to an inaccurate move he made in the endgame. B .tcs 15 .:txcS 16 .txcS 17 lbe3! Compensation for the isolated d-pawn usually takes the form of 70 Moscow Wch (9) 198415 active piece play, but in this case the black pieces are quite passive. Thus, the opening dispute has con­ cluded in White's favour. 17 .ie6 The fork 17 ...d4 is not danger­ ous because of 1 8 :ad 1 . Wc8 18 :ad1 White was already threatening 1 9 lbexd5 ! �xd5 20 e4. 19 Wa4 1 9 Wb1 :d8 20 :d3 deserved attention, when 20 ... d4 is bad due to 21 :Cd l �6 22 .ixc6. :ds 19 20 :d3 a6 21 :rd1 (DJ After 2 1 Wd l Black would have to reply 2 1 . . .11t'c6, as 2 1 . . .lLlc4 22 �exd5 lLlxd5 23 �xd5 .ixd5 24 .txd5 �xb2 25 .ixf7+ �xf7 26 :xd8 lLlxd I 27 :xc8 :xc8 28 lbdl would lead to a rook ending with an extra pawn for White. ••• Now the white rook and knight are very unusually positioned, in a letter T (it is as if the d3-rook is suspended between epaulettes on c3 and e3), and his pieces are moreover directed at the most vul­ nerable spot in Black's fortress. All the same, the d-pawn is withstand­ ing the pressure. 21 ... lLlc4 22 lLlxc4 22 lLlexd5 �xd5 23 lLlxd5 .txd5 24 .ixd5 gives nothing in view of 24 . . .:dxd5 ! 25 lbd5 :xd5 26 lbd5 �b6 27 'it'd4 lLlxd5 28 'it'xd5 'it'c 1 + 29 �g2 'it'xb2 with an level queen ending. :xc4 (D) 22 ... w 23 WaS The natural move 23 'ifb3 would probably have been stronger. After 23 . . . d4, not 24 'ii'xb7 'it'xb7 25 .txb7 :b8 26 .txa6 dxc3 27 .txc4 c2 28 :d8+ lbeS ! , but the quiet 24 Karpov - Kasparov 71 'ii' b 6! wins a pawn: 24 ...ltld7 25 .:txd4 ! . Black generally has to be very careful, so as not to lose his isolani. 23 .:tcS 24 'ii'b6 .:td7 25 .:td4 (D) Prophylactic measures such as 25 h3 or 25 a3 would probably have given Black more trouble. 32 Wg2 .:t7c5 (D) ••• w .:txc4 33 .:txc4 wrs 34 .:td4 35 .i.e2 .:txd4 We7 (D) 36 exd4 Insufficiently energetic play by White has almost allowed Black to stabilize the position; here 36 .. lbe4 37 ltla2 ltld6 38 ltlb4 a5 39 ltlc6 ltlc4 would have given him more chances to equalize. . B 25 ... 'ii'c7 .:tdxc7 26 'ii'xc7 27 b3 Exchanging queens turns out to be possible because the d5-pawn is invulnerable: 27 ltlxd5 ltlxd5 28 .i.xd5 .i.xd5 29 .:txd5 .:txd5 30 .:txd5 .:tc2 3 1 .:td8+ Wh7 32 .:td7 .:txb2 33 .:txf7 .:txe2 drawing. b5 27 ... g6 28 a3 Wg7 29 e3 30 Wh2 .:tc4 b5 31 .i.f3 72 Moscow Wch (9) 1984/5 37 38 39 40 41 42 lt::!a2 li)b4 f3 h4 'itf2 lt::Jc2 (D) .tc8 'itd6 lOgS lt::!h6 lt::!f5 • .i.. • • • • ••• �·� � · ·� � � � � � · · · · ·�· · � � � � �· � u � � u �8� �- � � u � � BlLl · ii.= • • • • • B Here the game was adjourned, and Black sealed the move 42 ... f6. After 42 ...lt::!g7 43 g4 f6 44 .td3 g5 45 .tg6 ! hxg4 46 h5 White has a dangerous passed h-pawn, but 42 ... .td7 would probably have been more reliable. f6 42 g5 43 .td3 .txf5 44 .txf5 .tbl 45 li)e3 46 b4 gxh4 (D) Black's bad bishop (his a6-, b5and d5-pawns are placed on the same coloured squares) determines White's stable advantage, but is it decisive? After 46....tg6, piercing a hole in the fortress would not w have been so easy. However, Black has taken the h4-pawn. Exchang­ ing pawns (if it happened) would make achieving a draw simpler. In fact, the break g3-g4 is harmless, and White only has the f4-square left through which he can penetrate the enemy camp; his pieces, the king and knight, cannot both pass through it at the same time. How­ ever, White has at his disposal a study-like route. 47 lt::!g2! ! Psychologically i t was impossi­ ble for Black to foresee such a move. It is all too easy to assume that White will automatically re­ capture the pawn on h4. The unex­ pected knight manoeuvre is based on a pawn sacrifice. By giving up one of his own pawns White clears a square which is needed for his king. The material balance is soon restored, after which the white pieces invade. Karpov - Kasparov 73 hxg3+ 47 ... 48 �xg3 �e6 49 lDf4+ �s 50 lbxh5 �e6 �d6 51 lbf4+ 52 �g4 .tc2 53 �hS .tdl 54 �g6 �e7 (DJ After 54... .txf3 55 �xf6 Black loses the d5-pawn after all. . - - . . . - . - • • - ·� · -···· - R . � � u u � 0 • -8· . . - . B B.t.B B w 55 lbxd5+ �e6 �d7 56 lbc7+ 57 lbxa6 .i.xf3 58 �xf6 �d6 �dS 59 �fS .i.hl 60 �f4 61 �e3 �c4 .i.c6 62 lbcS If 62 ... .tg2, then 63 lbd3 �b3 64 lbf4 .*.b7 65 �d3 �xa3 66 �c3 and 67 d5 is decisive. 63 lbd3 .*.g2 �c3 64 lbeS + 65 lbg6 �c4 (DJ w .*.b7 66 lbe7 66...�b3 67 d5 �xa3 68 d6 i.h3 69 lDd5 will not save Black either; the last chance of prolong­ ing the game lay in 66...i.h l 67 ltks �d5. .tg2 67 lbf5 �b3 68 lbd6+ 69 lbxbS �a4 1-0 70 lbd6 Game 1 6 Karpov - Kasparov Moscow Wch (27J 1984/5 QGD, Orthodox d5 1 lbf3 lbf6 2 d4 3 c4 e6 .te7 4 lbc3 h6 S .tg5 Here White has a choice be­ tween immediately taking on f6 (the Petrosian Variation) or retreat­ ing the bishop to h4, in order to 74 Moscow Wch (27) 198415 carry out the exchange at a more appropriate moment (usually after ...b7-b6). These two plans are seen alternately in practice, and this is how it was in my duels against Kasparov. 6 .i.xf6 .i.xf6 7 e3 0-0 8 'ircl cS 9 dxcS dxc4 (D) Kasparov and I often borrowed from each other's opening systems, which was inevitable, as in those days our repertoires were in many ways similar. In particular, this position was seen in Kasparov­ Timman, USSR v Rest of World, London 1984, which was played some months earlier. On that occa­ sion Kasparov gained a convincing victory after 9 ...'ira5 1 0 cxd5 exd5 1 1 0-0-0 .i.e6 1 2 lL!xdS llc8 1 3 �b1 ! . Now he had to fight against it himself. w 'iraS 10 .i.xc4 .i.xc3 1 1 0-0 In the event of 1 1 ...'ii'xc5 1 2 lL!e4 'fke7 1 3 lLlxf6+ 'ii'xf6 14 llfdl White has tangible pressure. 12 'irxc3 'irxc3 lL!d7 13 bxc3 14 c6 bxc6 lL!b6 15 llab1 16 .i.el cS 17 llfcl! (D) B Moving the rook to c 1 is the first in a long series of precise moves by White. 17 llfdl suggests itself, but in the long term this can only lead to exchanges on the d-file. On c l the rook is fulfilling a prophylactic function (defending the c-pawn), and is not taking the d 1 -square away from the bishop. The situ­ ation will only become completely clear in six moves time, and I will confess that I sensed rather than saw how I should place my pieces. Karpov - Kasparov 75 17 ... .i.b7 After the game this move pro­ voked unanimous criticism from commentators. In fact, 1 7 ....i.d7, controlling the b5-square, is more reliable, for example 1 8 �fl (there is another possibility linked with moving the bishop to a6 and the knight to e5) 1 8 .. .l:tfd8 1 9 l:.b3 l:.ac8 20 l:.a3 l:.c7 2 1 c4 .i.a4 22 l:.bl .i.e8 23 l:.a5 and White had a minimal advantage in Novikov­ Sturua, Lvov 1 985. But in the game itself White's advantage from the start was also scarcely noticeable. Therefore, glancing at this almost symmetri­ cal position, it is difficult to under­ stand immediately why the bishop is better placed on d7 than on b7. 1s �n .i.ds At the cost of a tempo it would have been possible at this point to prevent the rook from landing on b5 : 1 8 ...�c6, but after 1 9 lDe5 .i.a4 20 .i.b5 (20 .i.a6 l:.fd8 2 1 �e2 threatening 22 l2Jd3 i s also un­ pleasant for him) 20 ... �xb5+ 2 1 l:.xb5 l:.fc8 22 l2Jd3 c4 2 3 l2Jb2 Black has not escaped difficulties. 19 l:.bS (D) The white rook is headed for an ideal position - the a5-square. The a2-pawn cannot be taken in view of the reply 20 c4 and the bishop is trapped - the cl -rook is already in­ fluencing the course of events. B 19 l2Jd7 In the final analysis this loses a pawn; 1 9...l:.ac8 20 l:.a5 l:.c7 21 c4 .taB is more resilient. 20 l:.aS l:.tb8 21 c4 .i.c6 22 lbel! Paradoxically, before going on the attack, the white pieces first of all retreat. 22 l:.b4 23 .i.d1 ! (D) ..• 76 Moscow Wch (27) 198415 If 23 'Lld3 immediately, then 23 ...l:.a4, and the pawn is safe. Now, however, the march 'Llel­ d3xc5 cannot be prevented. The full extent of White's resources is now revealed, and the hidden strength of the quiet move 1 7 l:.fc I becomes clear. l:.b7 23 24 f3 l:.d8 25 'Lld3 g5 26 ..tb3 The final preparations; after the hasty 26 'Llxc5 'Llxc5 27 l:.xc5 l:.b2 28 l:.xc6 l:.dd2 all White's gains would have dissolved. �f8 . 26 'Llxc5 27 'Llxc5 28 l:.xc5 White still faces a large and la­ borious task, but the preliminary results of the engagement are not difficult to sum up: a pawn is a pawn. 28 l:.d6 �e7 29 �e2 l:.xd1 30 l:.d1 31 �xd1 �d6 f5 32 :as h5 (D) 33 �e2 Kasparov is trying with all his might to gain some counterplay. With his last move he has weak­ ened the g5-pawn, and I immedi­ ately make use of this. fxe4 34 e4 ..txe4 35 fxe4 . - - ­ ••• - R.t.li .t. R R a • •••• -�- - .ii.. B � R .� B W B � �'7}., �· -�- - - - w 36 l:.xgS ..trs 37 �e3 The black h-pawn will in the long run cause White distinct trouble, and it would have been worth fixing it with 37 h4. For ex­ ample, the following variation is possible: 37 . ..l:.f7 38 �e3 ..tg4 39 c5+ �c6 40 ..ta4+ �c7 41 ..te8 l:.h7 42 ..tg6 l:.h6 43 ..te4 with an easy win. h4 37 38 �d4 e5+ 39 �c3 ..tb1 40 a3 l:.e7 41 l:.g4 (D) White's final move was sealed. In the process of home analysis it became clear that the adjourned game was exceptionally sharp, and that Black could gain counterplay. Therefore, in order to find a clear­ cut route to victory, I had to solve various problems straight away. It is no accident that this ending has Karpov - Kasparov 77 B already found its way into mono­ graphs on endgame theory. 41 h3! In the event of passive defence with 4 1 ...1lh7, I would have fixed the h-pawn (but now on h4), after which Black would have had noth­ ing to hope for. If 4 l ...e4, then 42 �dl .:.n 43 'iit d4 llf2 44 c5+ 'iite7 45 �b3 lld2+ 46 �c3 lld3+ 47 'iitb2 e3 48 �a4 is decisive. 42 g3 Black was counting on my ac­ cepting the pawn sacrifice - 42 gxh3, and then 42 ...1lh7 !, notice­ ably widening the rook's sphere of influence. lle8 42 Preparing to transfer the rook to the second rank via the f-file. 42 ...1lf7 loses immediately to 43 c5+. .:.rs 43 llg7 44 llxa7 .:.n llxh2 (D) 45 �b4 White has a beautiful win after 45 ...1lb2. Here is the main line: 46 c5+ 'iitc6 47 �c4 �c2 48 lla6+ �c7 (48 ... 'iitb7 49 llb6+ �c7 50 'iitc 3) 49 �xc2 llxc2+ 50 �d5 llxh2 5 1 lla7+ �b8 (alternatively, 5 1 ...'iitc8 52 llh7 llh1 53 'iit d6 h2 54 llh8+ 'iitb7 55 c6+ 'iitb6 56 c7) 52 llh7 llh 1 53 'iite4 h2 54 �f3 e4+ (54 .. .1la1 55 llxh2 llxa3+ 56 �g4 llc3 57 lle2 llxc5 58 �f5, etc.) 55 �g2 llc l 56 �xh2 llxc5 57 llf7 .:r.c4 58 g4 e3 59 �g3 winning. ••. .•. w 'iitc6 46 c5+ �d5 47 �a4+ �e4 48 lld7+ 48 . .. �e6 49 c6 llb2+ 50 �b3+ llxb3+ 5 1 'iit xb3 �e4 52 lld8 �xc6 53 .:r.hs �g2 54 a4 �f5 55 llh4 will not save Black either. 49 c6 llb2+ llb8 50 'iita S! If 50 . . . h2, then 5 1 c7 !, and as soon as the black queen appears on 78 Moscow Wch (4) 1985 the board, she is lost: S l ...hl'ii 52 .i.c6+. l:.c8 51 c7 �e3 52 �b6 53 .i.c6! h2 54 g4! (D) Taking the f5-square away from the bishop. Now it is all over. S .i.gS h6 .i.xf6 6 .i.xf6 0-0 7 e3 8 'ii'c2 As the queen move had led me to victory in the game we have just seen, there seemed no point in avoiding this popular position a year later. However, Kasparov had prepared an interesting novelty. 8 tba6! (D) ••. B 54 ss l:.dl 56 57 58 59 l:.el+ l:.e4+ l:.xeS l:.e2 l:.h8 .i.a2 �f4 �g3 �xg4 1-0 Game 17 Karpov - Kasparov Moscow Wch (4) 1985 QGD, Orthodox 1 2 3 4 d4 c4 tbc3 tbf3 dS e6 .i.e7 l£lf6 w A curious manoeuvre, which allows Black to make use of the white queen's position on c2 in order to simplify - 9 cxd5 lbb4 ! 10 'ii'b3 tbxd5. cS 9 l:.dl 'iiaS 10 dxcS tbxcS 1 1 cxdS 12 'ii'd2 After 1 2 d6 .i.d7 followed by .. .l:.ac8, or 1 2 dxe6 .i.xe6, Black has sufficient compensation for the pawn. Karpov - Kasparov 79 lidS 12 Not, of course, 1 2...exd5 in view of 1 3 ltlxd5, while 1 2 ....i.xc3 1 3 1Wxc3 1Wxc3+ 1 4 bxc3 exd5 1 5 l:.xd5 is also favourable for White. 13 ltld4 The only attempt to fight for the initiative. After 1 3 .i.e2 .i.xc3 1 4 Wxc3 Wxc3+ 1 5 bxc3 l:.xd5 the time has come to agree a draw. exd5 13 ... 14 .i.e2 'ii'b6 After 14 ...lDe6 1 5 ltlb3 .i.xc3 1 6 bxc3 as well, White i s slightly better. 15 0-0 ltle4 (D) Definite problems remain for Black after 15 ... lDe6 16 ltlf3 d4 1 7 ltle4 dxe3 1 8 ltlxf6+ gxf6 1 9 Wxe3. .•• The queen retreats and occupies a safe place; when she is needed she can quickly return to the centre. llc7 19 ... 20 l:.d2 l:.dc8 After 20....i.xd4 2 1 llxd4 all the black pieces would be tied up de­ fending the d5-pawn. Possibly in this case the game would have ended in a draw, but passive play is not Kasparov's style. It was diffi­ cult to imagine that changing the pawn structure would lead in the future to serious problems for Black. 21 ltlxe6! If 2 1 llfd l , Black should con­ tinue 2 1 .. ..i.xd4 and then 22 llxd4 llc2 23 .:.4d2 .i.f5 . 21 ... fxe6 (D) 2 1 . . .Wxe6 22 llfd l is scarcely any better. ltlxc3 .i.e6 llac8 w 22 .i.g4 80 Moscow Wch (4) 1985 If 22 llfd 1 , then 22 . . .'iVb4 is pos­ sible, threatening 23 .. ."tlt'xd2 ! with complete equality. For the time be­ ing Black's position looks hopeful, but while his bishop is firing into thin air, White's bishop is capable of creating dangerous threats - a typical motif in opposite-coloured bishop endings. In the end White's light-squared strategy is trium­ phant in this game. As Grandmas­ ter (and pianist !) Taimanov noted, for the next 1 7 moves I was only playing on the 'white keys' . l:lc4 22 'ifc6 23 h3 24 'ii'd3 After 24 'it'g6, there is the retort 24...'it'e8, and White cannot yet in­ vade via the light squares. �h8 24 aS 25 l:lfdl l:lc3 26 b3 l:lf8 (D) 27 'ife2 w White would maintain a mini­ mal advantage after 27 ...l:lc l 28 l:lxc l 'ii'xc 1 + 29 �h2 'ii'c6 30 g3. But Kasparov, it seems, underesti­ mated White's threats. 28 .i.hS! The bishop bounces off the sides of the board like a billiard ball, changing direction and ending up on another, even more dangerous, diagonal. bS 28 28 ...i.d8 29 .i.g6 .i.c7 30 .i.d3 'it'd6 3 1 g3 'it'e5 32 ..g4 'ii'f6 was necessary, erecting a fortress. The text move loses time, and allows White to improve his position. 29 .i.g6 .i.d8 30 i.d3 b4 31 'ii'g4 'ife8 .i.gS (D) 32 e4! .•• w .:.Xc2 33 l:lc2 After either 33 ...'it'f7 34 l:le2 or 3 3 ...'it'c6 34 'ife2 l:lc8 35 l:lxc3 Karpov - Kasparov 81 bxc3 36 exd5 exd5 37 .i.c2 White's initiative has not been dampened down, but B lack should neverthe­ less feel calmer. 'ii'c6 34 .i.xcl 35 'ii'el 'ii'cS 36 .:0 (D) Defending against the threat of 36 ...l:txf2 37 'ii'xf2 .i.e3. The black queen's activity will be short-lived, and clouds are gathering over the black king. 36 B 'ii'c3 exdS 37 exdS 'ii'dl 38 .i.bl! 39 'ii'e5 39 'ii'e6 was an alternative, but then a white piece has to go onto a dark square at some point! Obvi­ ously there is no reason to pursue the pawn - 39 'ii'xd2 .i.xd2 40 l:tdl .i.g5 4 1 :Xd5 l:td8 and now the op­ posite-coloured bishops work in B lack's favour. 39 l:td8 39 . ..d4 40 'ife4 Wg8 4 1 .i.d3 is also bad, but 39 ....i.f6 is a more dogged defence. Wg8 (D) 40 'ii'fS ••. w Here the game was adjourned, and painstaking analysis showed that White has huge attacking re­ sources. 41 'ii'e6+ �h8 If 41 ...�f8. then 42 .i.g6 'ili'f4 43 l:tel with the deadly threat 'ife8+. Wg8 42 'ii'g6 43 'ii'e6+ �h8 1Wc3 44 .i.fS �g8 45 'ii'g6 46 .i.e6+ �h8 �g8 47 .i.fS Wf8 48 g3 'iff6 49 �gl 'ili'f7 so 'ifh7 5 1 f4 was threatened, winning a piece, and if 50...g6, then 5 1 .i.xg6 'ii'g7 52 f4 .i.f6 53 l:tdl . 12 Mo�eow Wcla (4) 1985 !1 h4 .td2 Other bishop retreats are no bet­ ter: 5 1 . . .i.f6 52 l:.e l 'ili'g8 53 'iWg6 'iWf7 54 'ii'g4, or 5 1 ...i.e7 52 l:tel with the threat of 53 i.e6 or 53 .tg6. i.c3 52 l:tdl 53 l:td3 l:td6 (D) If 53 ... 'iWg8, then 54 'ii'g 6 'ii'f7 55 'ifb6 is very strong. Now that the white rook has entered the game, events develop swiftly. w 54 l:tf3! One imprecision - 54 l:te3 would be enough to allow Black to free himself by 54 ...g5 ! . <j;e7 54 The king is forced to abandon his refuge. If the f-file is recovered, the appearance of the rook on e3 will quickly be decisive: 54 ...i.f6 55 l:te3 g5 56 'iWxh6+, and the rook's action is covered; 54 .. .l:tf6 55 l:te3 g5 (55 ...l:txf5 56 'ii'h 8+ ••• 'ii'g 8 57 l:te8+ <j;xe8 58 'ii'xg8+ <j;d7 59 f4) 56 'ii'h 8+ covering the action of the bishop. Problem mo­ tifs in abundance ! 55 'ii'h 8! 55 l:te3+ would have been hasty in view of 55 ...<j;d8 56 'ifh8+ <j;c7 57 'ii'c8+ <j;b6 58 'ii'b8+ <j;c5 !, and Black can defend himself with the help of ...l:tc6 and ...'ii'c7. 55 d4 (D) ... w If 55 ...i.e5, then 56 i.h3 l:tf6 57 l:te3 ! l:txf2+ 58 <j;gl is the end. l:tf6 56 'ifc8 57 'ifc5+ <j;eS 58 l:tf4 'iib7+ rj}f7 59 l:te4+ It looks as though Black would escape without a scratch after 59 ...l:te6, because of 60 i.xe6 'ii'xe4+, but again a problem ma­ noeuvre is decisive: 60 'ifc4! l:txe4 6 1 'ii'g 8+ <j;e7 62 'ifxg7+ and 63 1Wxb7. Kasparov - Karpov 83 60 'ii'c4+ �8 lU7 61 �h7! 'ii'd 7 62 'ii'e6 63 'ii'e 5! 1-0 I had prepared 63 . . .:e7 (if 63 . . .'iWd8, then 64 'ii'c 5+ :e7 65 :f4+ �e8 66 'iWc6+ 'ii'd7 67 ..tg6+; or 63 ...'ife7 64 'iWb8+) 64 'ii'f4+ :n 65 'iWb8+. 12 a4 'ifd7 (D) This queen move is relatively rare; the usual continuation here is 1 2 ... h6. Game 1 8 Kasparov - Karpov Moscow Wch (5) 1985 Spanish, Zaitsev 1 e4 e5 2 lLlf3 lLlc6 3 �b5 a6 4 �a4 lLlf6 5 0-0 �e7 b5 6 :et 7 �b3 d6 8 c3 0-0 9 h3 �b7 10 d4 :e8 This rook move was brought into practice by my long-time sec­ ond Igor Zaitsev, and the variation is therefore named after him. Black fortifies the centre, and does not spend time on the prophylactic ... h7 -h6. True, there is a danger that White will repeat moves with 1 1 lLlg5 :f8 1 2 lLlf3 :e8 1 3 lLlg5. Thus, if he needs a victory, Black should choose something else. i.f8 11 lLlbd2 w axb5 13 axb5 i.xa8 14 :Xa8 After 14 ... :xa8 1 5 lLlg5 the black knight is forced to retreat to the back rank allowing White to seize the centre: 1 5 ... lLld8 16 lLldf3 exd4 ( 1 6 . . . h6 1 7 lLlxf7 ! lLlxt7 1 8 dxe5; 1 6 ...c 5 17 dxe5 dxe5 1 8 'it'xd7 lLlxd7 1 9 lLlxf7 ! c4 20 lLlxd8 :xd8 2 1 i.a2 :e8 22 �e3 and White had a clear advantage in lvanchuk-Portisch, Linares 1990) 17 e5 !. 15 d5 Now, as he has taken on aS with the bishop, after 1 5 lLlg5 Black can reply 15 ... :e7, and then ...h7-h6. 15 ... lLla5 In game 46 of our first match I retreated with 15 . . .lLld8, and after B4 Mo1cow Wch (5) 1985 1 6 �fl h6 1 7 ltl3h2 ! ltlb7 1 8 i.c2 ltlc5 1 9 b4 ltla6 20 ltlg4 ltlh7 2 1 ltlg3 c 6 2 2 dxc6 i.xc6 2 3 i.b3 ltlc7 24 W'f3 ltle6 25 h4 W'd8 26 l:.dl White achieved an advantage. 16 i.a2 c6 17 b4 ltlb7 (D) .i.� � � � .I �·� · %% � � � ···'ii' · · · · � -- � � � - � �·� � �·· • � �8� � �8· n � - � u � %% � -� � � /";"\. . 8 � �"LJ� � �8. i.� � � � u � 7, !ii blocked in by its own b7-knight and c6-pawn, will quickly free it­ self and develop energetic activity, and precisely thanks to his bishop Black will gain the upper hand. 18 c4 The standard 1 8 ltlfl is better, and after 1 8 ... cxd5 ( 1 8 ... c5 19 i.g5 i.e7 20 ltlg3 g6 21 'ii'd2, preparing ltlh2 and f4) 1 9 exd5 h6 20 ltlg3 l:.c8 21 W'd3 ltld8 ! 22 ltlh2 g6 23 h4 White maintains the initiative. 18 l:.c8! (D) Before retreating the knight to d8, it makes sense to place the rook on a more promising file. ••• . � .� �-� - - � w After 17 ...ltlc4 1 8 ltlxc4 bxc4 1 9 i.g5 ! cxd5 20 i.xf6 dxe4 2 1 ltlxe5 W'a7 22 ltlg4 'ii'xa2 23 ltlh6+ �h8 24 ifh5 gxf6 25 1li'xf7 i.xh6 26 '1Vxe8+ �g7 27 'ii'e7+ White has the advantage, since the enemy bishop has no prospects. However, now it seems that Black is all but suffocating from a lack of vital space. But this impression is de­ ceptive. Nimzowitsch once said that a piece which has been in cap­ tivity for a long time finds incred­ ible strength when it has broken free. In fact, in this game my aS­ bishop, hiding in the corner of the board, and for the time being 19 dxc6 Handing over the centre allows Black immediately to gain a play­ able game. Fascinating events arise after 19 W'e2 ltld8 ! 20 i.b2 bxc4 2 1 ltlxc4 W'a7 ! (2 l . ..l:.b8 ! also gives him good prospects) 22 l:.a1 cxd5 23 exd5 i.xd5 24 ltlxd6 Kasparov - Karpov 85 �xd6 25 �xd5 1fxa1 + 26 �xa1 l:tcl + 27 �h2 l:txa1 and Black has enough compensation for his small material loss. '1Vxc6 19 20 c5? Instead of this highly risky ven­ ture, 20 �b2 straight away would have been better. As it is, the threat to f7 (20 ... dxc5 2 1 �xf7+) is eas­ ily repulsed, and Black seizes the initiative. 20 lbd8 21 �b2 dxcS! 22 bxc5 Black also has good chances after 22 lbxe5 1fa6 ! 23 ..,al c4 24 �c3 1fa3 25 l:te3 .i.xb4 26 lbexc4 bxc4 27 �xf6 c3 ! 28 l:tg3 lbe6, or alternatively 22 .i.xe5 lbd7 23 .i.b2 c4. 22 •xc5 23 .i.xeS lbd7 �4! (D) 24 .i.b2 ••. ••. ••. More precise than 24... ..,c2 25 1fa1 ...a4 26 lbd4 lbc5 27 l:te3 with a sharp game. 25 lbb3? After 25 1fb1 White's position would still have been defensible. The desire to generate some activ­ ity has led him into a difficult position. lbcS! 25 . 26 .i.a1 A forced pawn sacrifice. After 26 lbxc5 it'xb2 ! 27 l:te2 it'a3 28 lbd3 .i.xe4 29 .i.xf7+ lbxf7 30 l:txe4 b4 Black has a clear advan­ tage. .i.xe4 26 ... 27 lbfd4 (D) Black would also have won ef­ fectively after 27 lbg5 .i.c2! 28 'ii'xc2 'ii'xe 1 + 29 �h2 .i.d6+ 30 g3 .. lbe4. 86 Moscow Wch (22) 1985 29 30 31 32 33 34 35 36 37 38 39 40 41 lL:!xcS 'ii'g4 .:.dt 'ii'f4 'ii'c l .:.et ..tb3 'ii'b2 'ii'xcS .:.e8 ..tg6 'ii'b 4! ..te4 'ii'a5 'ii'a8 b4 ..tg6 'ii'xe8 lbe4 lDc5 lDd3 Game 1 9 Karpov - Kasparov Moscow Wch (22) 1985 QGD, Exchange In game 2 1 , when Kasparov had White, after 5 ... c6 6 e3 ..tf5 7 g4 .te6 8 h4 lDd7 9 h5, I introduced a novelty, 9 ... lDh6, but did not man­ age to solve my opening difficulties. 6 e3 In game 20 I stopped the bishop coming out to f5 with 6 'i!fc2 but the time I lost in doing this led to rapid equality: 6 ...0-0 7 e3 c5 ! 8 dxc5 ..txc5 9 lDf3 lDc6 10 .te2 d4! . 6 0-0 ..trs 1 lDf3 8 h3 Preparing to carry out the stand­ ard manoeuvre in this position, g2-g4. This rook's pawn move had until that time not been used with this move order. 8 c6 9 g4 .tg6 10 lDe5 lDfd7 (D) For the third time in a row in the match a sharp form of the Queen's Gambit Exchange Variation arose. In the first two games White had the initiative, but they ended in draws. This time, however, I man­ aged to win. d5 1 d4 e6 2 c4 3 lDc3 ..te7 4 cxd5 exd5 5 ..tr4 lDr6 w .:.eJ .:.xe8 'ii'c l ..td5 lDb3 0-1 The final move was sealed. In the adjourned position Black' s passed pawn guarantees him vic­ tory and Kasparov considered that there was no point in resuming. Karpov - Kasparov 87 After 10 ... lL'lbd7 1 1 h4 Black feels rather hemmed in. 1 1 tt:'Jxg6 fxg6 Following the banal 1 1 .. .hxg6 Black is deprived of all possible counterplay. tt:'Jb6 12 ..tg2 �h8 13 0-0 1 3 . . .tt:'Ja6 is not sufficient for equality because of 14 'ikb3 ! ; nor is 1 3 . . . g5 satisfactory, due to the re­ ply 14 ..tg3 i.d6 1 5 ..txd6 1i'xd6 1 6 e4 ! . 1 4 tt:'Je2 (D) Both 1 4 'ikb3 and 1 4 ..tg3 fol­ lowed by e3-e4 would have been promising for White. ( 1 8 ...gxf4 19 tt:'Jxf4) 19 fxg5 ..txg3 20 tt:'Jxg3 'iixg5 21 e4. 18 'iixd6 19 f4 White's kingside pawn advance guarantees him a definite initiative. gxf4 19 .:Z.ae8 20 exf4 21 f5 (D) 2 1 .:Z.f3 tt:'Jd7 22 g5 followed by h4 and lL'lg3 is probably more accu­ rate. ••• ••• B B 14 g5 ..td6 15 ..tg3 tt:'Ja6 16 'iid3 'iie7 17 b3 18 ..txd6 The immediate 1 8 f4 would have led to a sharp struggle: 1 8 ....:Z.ae8 21 22 23 24 25 26 27 28 29 30 31 .:Z.f2 g5 h4 .:Z.d1 'iixe3 �h2! tt:'Jg3 tt:'Jn l:td3 tt:'Jg3 tt:'Jc7 tt:'Jd7 'ike7 'iie3 tt:'Jb5 .:Z.xe3 tt:'Jb6 tt:'Jc8 .:Z.e7 tt:'Jcd6 BB Moscow tt 19B6 Here I was preparing to move my king to g4 and force back the knight with a2-a4, maintaining the pressure. Obviously, Black should also bring his king up to the cen­ tre with 3 l ...�g8, or activate the rook by means of 3 1 ...llel . Instead of this Kasparov, in severe time trouble, made a serious mistake. 31 t:be4? (D) ... . . . .,. � �. ··- . . ··· · · 8 �-� . ��� � � � � •8•:• � 8 • • ail.= • • • • w dxe4 32 �xe4! 33 lle3 t:bxd4 34 �h3 By 34 f6 ! gxf6 35 gxf6 White achieves a strong position, for ex­ ample 35 ...lle6 36 llxe4. But be­ fore moving into decisive action, I decided to activate the king. 34 •.. lieS 35 �g4 h5+? Kasparov gives an impulsive time-trouble check; 35 . ..�g8 or 35 . . .llfe8 was necessary, defend­ ing obstinately. 36 �xh5 Black was possibly counting on 36 �f4 llexf5+ 37 t:bxf5 lbxf5 with a probable draw (but not 3 8 �xe4 due to 3 8 ...liJd6+ 39 �e5 llxf2). t:bxf5 36 37 :xf5 llfxf5 38 t:bxf5 llxf5 �h7 39 :Xe4 40 lle7 b5 In the event of 40...llb5, White can achieve his goal with 4 1 a4 llxb3 42 g6+. b4 41 llxa7 1-0 42 �g4 The adjourned rook ending is easily winning for White, as the mobility of the h-pawn is decisive. Game 20 Karpov - BeUavsky Moscow tt 1 9B6 QGD, Orthodox 1 2 3 4 5 6 7 8 9 10 11 12 d4 c4 m t:bc3 i.g5 i.xf6 e3 :et i.d3 0-0 i.xc4 h3! (D) t:bf6 e6 d5 �e7 h6 i.xf6 0-0 c6 t:bd7 dxc4 e5 Karpov - Beliavsky 89 B w Both 12 .i.b3 and 1 2 li)e4 used to be played, but in both cases con­ vincing routes to equality have been found for Black. This modest, but poisonous peripheral pawn move was first used by Kasparov in the 23rd game of our second match. The idea behind it is that the light-squared bishop for the time being stays where it is, and so after 1 2 ... l:leS he still has the possibility of 1 3 'iib 3. Besides this, the g4square is taken under control, and after opening the centre the cS­ bishop will have problems devel­ oping. Thus Black finds himself in a peculiar zugzwang. 1 S l:le1 l:ldS 19 'iif4 li)d5 20 li)xd5 cxd5 2 1 li)e5 .i.xe5 22 l:lxe5 .i.e6 23 'it'e3 White gained an advan­ tage. 12 13 exd4 14 .i.b3 exd4 li)b6 .i.f5 (D) In the above-mentioned game against Kasparov I continued with 14 ... l:leS and brought the bishop out on the next move. After 1 5 l:le1 .i.f5 16 l:lxeS+ 'fixeS 17 'iid2 'iid7 15 l:le1 aS In the event of 15 ... l:leS 1 6 l:lxeS+ 'fixeS 17 'iid2 the game fol­ lows the variation examined above. 16 a3 'iid7 In game 22 of the return match with Kasparov I replied 1 6 . . .l:le8, and after 17 l:lxeS+ ..xe8 1 8 'iid2 I tried a new idea - 1 S ... li)d7. How­ ever after 1 9 1i'f4 .i.g6 20 h4 'ifdS 2 1 li)a4 I ran into difficulties. An important novelty was used in Gurevich-Van der Sterren, Balcu 1 9S6: 1S ...'ifd7 19 l:le 1 l:leS ! 20 l:lxeS+ 'fixeS 2 1 ..f4 .i.e6 ! , and Black equalized. 17 li)eS 18 l:lxe5 19 1i'e2 .i.xe5 l:lfe8 l:lad8 Black has decided that there is no hurry to push his rook's pawn. 90 Moscow tt 1 986 AxeS a4 (D) w w It looks as though Black's delay in pushing his a-pawn makes no difference, and White must remove the bishop to a2, when the variation 22 ll:!xa4 ll:!xa4 23 .i.xa4 'ibd4 24 'it'xf5 'it'xa4 cannot worry him. However, White finds an attractive intermediary manoeuvre. axb3 22 'it'cS 23 :e7 'it'd6 Yet another zwischenzug. Before he gives back the piece, the white rook penetrates the seventh rank. Here 23 . . .'it'xd4 loses because of 24 Ae8+! �h7 25 'it'xf5+, while if 23 ...'it'c8, then 24 'it'xb6 with an obvious advantage for White. lhe8 24 Ae8+ 25 26 27 28 'it'xd6 'it'b4 dS! ll:!xd5 ll:!c4 b5 cxd5 .i.c2 (D) When I played 27 d5, I envis­ aged 29 ll:!e7+ here, but while I was checking the variation I turned my attention to the possibility of 29 .. .llxe7 ! 30 'it'xe7 lbxb2 3 1 'it'e8+ �h7 32 'it'xb5 ll:ldl 33 'it'c4 f5 with a study-like draw: the knight leaps between b2 and d l , and Black saves himself. In the event of 34 a4 lDb2 35 'it'c6 ll:!dl the queen has to return to c4, as after 36 aS .i..e 4! and then 37 . . . b2 Black actually takes the upper hand. :ds 29 'it'c3 30 lDf4 Adl+ 31 �b2 ll:!d2 After 3 1 ...Abl 32 a4 ll:!xb2 33 axb5 the pawn cannot be stopped. 32 b4 Preventing perpetual check by the knight. 32 33 f3 34 bxg5 35 ll:!e2 (D) .i.f5 g5 bxg5 Karpov - Kasparov Avoiding the final trap - 35 lbb5 t'Dfl+ 36 Wgl t'Dg3+ and White's knight is lost. 8 cxd5 9 'ilt'd2 10 bxc3 (D) 91 t'Dxc3 'ilt'xal B Game 2 1 Karpov - Kasparov London/Leningrad Wch (5) 1986 Griinfeld Defence This variation came into prac­ tice after the famous game Petro­ sian-Fischer, Buenos Aires Ct (2) 1 97 1 , which continued 10 ...'1i'a5 1 1 .i.c4 t'Dd7 12 t'De2 �5 1 3 .i.a2 .i.f5 14 .i.xe5 ! .i.xe5 1 5 .!Dd4 tWxc5 1 6 .!Dxf5 gxf5 17 0-0 with a dangerous attack for White. Later various lines were proposed for Black, the most effective being 1 2 ... .!Dxc5 1 3 0-0 0-0 14 f3 e5 ! 1 5 .i.g3 b 5 1 6 .i.a2 'li'b6! 17 Wh 1 a5 with sufficient counterplay. 10 1 2 3 4 5 6 7 d4 c4 lDc3 .i.f4 e3 dxc5 .:.et t'Df6 g6 d5 .i.g7 c5 'it'aS t'De4 ... 'ilt'xd2+ Kasparov prefers to exchange queens. Until now this move had resided somewhat in oblivion, and only recently have new ideas been found for Black. 1 1 Wxd2 12 .i.b5 lbd7 t2 LDndol'lllAnlngrad Wch (5) 1986 After 1 2 c6 bxc6 1 3 dxc6 lZ1b6 (or 1 3 . . .le!f6) White has achieved nothing. 0-0 12 13 -txd7 If 1 3 c6, then 1 3 ...lZ1c5. Now Black gains the advantage of the bishop pair, but as will soon be­ come clear, both will be crippled. -txd7 (D) 13 has completed his development, maintaining powerful pressure. .•• ••• z • • ••• -·· -� i.� -i� -i • • ••• • o�• • � � � � · � � • � � - • - � " u �� • �� mi � � -� A � -� � � L.l � � � . � .: . w f5 14 e4 e6! ? 15 e5 A n important moment. Why did Kasparov reject the previously tested 1 5 ...l:l.ac8 1 6 c6 bxc6 17 d6 exd6 1 8 exd6 l:l.f6 with an advan­ tage to Black (Schmidt-Gross, Nal­ eczow 1 984)? The answer is that 1 5 . . . l:l.ac8 would be met by 16 c4! l:l.xcS 17 -te3, as in Seirawan­ Adorjan, New York 1 987. 16 c4 l:l.fc8 (D) After 1 6 ... gS 1 7 -txgS -txe5 1 8 lbf3 -tg7 1 9 l:l.b 1 and l:l.he 1 White w 17 c6! He does not keep the extra pawn, but in returning it, White will elicit maximum profit by cre­ ating a passed pawn and limiting the movement of Black's rook and light-squared bishop. 17 bxc6 18 d6 c5 Now the light-squared bishop finds some freedom, but the dark­ squared one begins to suffocate. Perhaps the fate of the bishops should have been dealt with differ­ ently - 1 8 .. . g5 1 9 -txg5 -txeS 20 cS l:l.cb8, and Black maintains hopes of counterplay. h6 19 h4! 20 le!h3! (D) After the more natural move 20 lZ1f3, the reply 20 . . .-tc6 ! would have led to a complex game with ••• Karpov - Kasparov B B chances for both sides. However, I managed to find a most logical solution to the position. The white knight is concentrating on the only true trajectory towards its optimal square, d3 ! As soon as it has reached its goal, the g7-bishop will be caught in a trap constructed by the white e5- and d6-pawns and the f4-bishop. Incidentally, White pre­ vents the freeing advance ... g6-g5. 20 ••• 93 23 24 25 26 27 lbf2 lbd3 llal hxg5 .i.xg5 (D) a2 lla3 g5 hxg5 aS The passed a-pawn is not dan­ gerous, so Black should have im­ mediately occupied another file with 20...llcb8. 21 f3 a4 22 llhel ! (D) Overprotecting the e5-pawn. After the hasty 22 t:Df2 Black can muddy the waters with 22 ... g5 ! 23 hxg5 hxg5 24 .i.h2 f4 25 lbd3 .i.e8 26 .i.g1 .i.g6 27 lbxc5 .i.xe5 28 lbxe6 .i.xd6 29 lbxg5. 22 ••• a3 w 27 ••• rM7 After 27 . . . llb8 the duel might conclude with a study-like draw: 28 .i.f4 llbb3 29 lbxc5 llb2+ 30 <itc1 llxg2 3 1 .i.d2 .i.h6 ! 32 .i.xh6 llc3+ 33 �d 1 lld3+! 34 lbxd3 .i.a4+ 35 �c 1 llc2+ and Black � I.IMtlll£lrtlrtii'IMI Wch (17) 1986 diUwn perpetual check. But a ample refutation to this 'study' can bt found: 28 �e2! l:tbb3 29 lLlxc5 �fl and White has the upper hand. lb2+ 30 28 29 30 31 32 .i.f4 l:tecl l:tc3 l:tcl lLlcl l:tb8 .i.c6 .:taS l:tba8 1-0 Game 22 w Karpov - Kasparov London/Leningrad Wch (1 7) 1986 The critical position in the Smyslov Variation. Almost every­ thing has been tried here, includ­ ing 12 'it'xd6, 1 2 lLlb5, 1 2 d5, 1 2 h3 and 12 0-0. In the key game Bot­ vinnik-Fischer, Vama OL 1962, af­ ter 12 h3 .i.xf3 1 3 gxf3 .:tfd8 14 d5 lLle5 1 5 lLlb5 'it'f6 1 6 f4 lLled7 17 e5, Fischer gave Botvinnik an un­ pleasant surprise with 17 ...'ifxf4 ! . However, after mistakes b y both sides this fascinating game eventu­ ally ended in a draw. However, it turns out that White has at his dis­ posal one other interesting move, which has not been tested in prac­ tice until now. Grilnfeld Defence This miniature is a classic example of a battle which was decided en­ tirely in the process of home prepa­ ration. White's novelty cannot be called unusual - it consists of a simple one-square advance by the h-pawn, but careful analysis has shown that after this move Black's defence becomes noticeably more difficult. 1 d4 lLlf6 2 3 4 5 6 7 8 9 10 11 c4 lLlc3 lt:lf3 'iWb3 'iWxc4 e4 .i.e3 .:tdl .i.el 'iVcS g6 dS .i.g7 dxc4 0-0 .i.g4 lLlfd7 lLlc6 lLlb6 'iWd6 (D) 12 eS! At first glance this looks para­ doxical. White not only hopelessly weakens his e5-pawn, but also al­ lows an exchange of queens. How­ ever, things are not so simple as that. 12 ... 'iWxcS Karpov - Kasparov 95 13 dxcS �c8 (D) In reply to the more natural 1 3 ... �d7 White has 14 h3 ! .txf3 1 5 gxf3 ! and the 'doomed' e5pawn is invulnerable, as if it is cap­ tured by either knight, 1 6 f4 ! wins a piece. Otherwise, White will strengthen the e5-pawn with its neighbour, and gain a clear advan­ tage. resurrecting his idea. Meanwhile, when I again set up the position on the board, I managed to spot that the c8-knight's prospects can be significantly limited. 14 .t:xf3 ••• If 14....te6, then 1 5 �g5 looks unpleasant: 1 5 ...�xe5 1 6 �xe6 fxe6 1 7 f4. 15 .txf3 .txeS (D) After the alternative capture Black also runs into difficulties: 15 ...�xe5 16 .txb7 :bs 17 c6 �4 1 8 :d7. 14 h3! In game 1 5 , where the novelty 1 2 e5 was used for the first time, I played 14 �b5, and after 14 ... :b8 ! 1 5 �xc7 e6 ! Black managed to equalize completely. In fact, the threat to surround the knight by means of 1 6...a6 forces White to lose a tempo by playing 1 6 �b5, and the black c8-knight quickly transfers to a comfortable position. The result of that encounter obvi­ ously satisfied my opponent, who thought that White had no way of w 16 .txc6! 17 .td4 bxc6 .tf4 (D) After an exchange of bishops with 17 ....txd4 1 8 :xd4 :bs 1 9 b 3 a5 the white rook can penetrate the seventh rank, but after 20 :d7 (20 :a4 is stronger: 20 ...:as 2 1 �e2 :ds 2 2 :d 1 :xd 1 23 �xd 1 followed by the transfer of the knight to c4) 20... �a7 21 :xc7 96 London/Leningrad Wch (17) 1986 lllb� 22 ltlxb5 .:.xb5 23 .:.xc6 a4! Black has counterplay. w aS?! 18 0-0 It was necessary to move the e­ pawn. My preparation showed that after 18 ...e5 19 .i.e3 (19 lbe2 lbe7!) 19 . . ..i.xe3 ( 1 9 ... lbe7 is also possi­ ble) 20 fxe3 White has a definite advantage. However, soon after the match this position was seen in Karpov-Timman, Tilburg 1 986, and the Dutch grandmaster proved that Black has chances to hold the position. The game continued thus: 20 ... lbe7 21 .:.d7 lbf5 22 .:.xc7 :res ! 23 .:.d7 :adS 24 .:.fd 1 .:.xd7 25 .:.xd7 lbxe3 26 .:.c7 .:.bs 27 b3 :dB with a swift draw. True, the feeling remains that somewhere White could have played some­ thing stronger. 19 :rei White plans to dominate the centre. The move g2-g3 will chase away the bishop, whereupon .i.e5 and .:.d7 will make inroads into Black's position. In view of the atrocious position of the knight on c8, Black has great difficulty op­ posing this plan. 19 ... a4?! The pawn continues its march, but this is fiddling while Rome bums. 1 9 ...e6 would not help in view of 20 lbe4 when White's pieces co-operate superbly. It is hard to see what Black should actu­ ally play here - maybe 1 9 .. .f5 to deny White's pieces access to e4, but to commit oneself to such an extreme measure would be diffi­ cult at the board. What is not in doubt, is that the pawn advance chosen by Kasparov carries no real threat, and consti­ tutes a fatal waste of time. 20 .:.e4! Originally I had intended 20 a3, to fix the pawn on a4 so as to gang up on it. However, it occurred to me that I could ignore the pawn; especially since if (following 20 a3) the white knight captures on a4, his black counterpart could spring out to b5 via a7, with some chances of real resistance. Clearly there would be no point complicat­ ing matters in this way, when sim­ ply centralising my pieces is so strong. .i.h6 20 ... Karpov - Kasparov 97 21 i.e5 Black's planned . . .liJa7 is ren­ dered rather ineffectual by this move. 21 a3 22 b3 Clearly the best move. !lJa7 22 ... Although this leads nowhere, there were already no sensible moves for Black. 23 l:td7! (D) 23 i.xc7 would regain the pawn and hit e7, but then 23 ...e6 or the more precise 23 ... i.g7 would pro­ vide counterchances. Thus I pre­ ferred instead to keep full control of the long diagonal. ••• Black seems to want to justify his a-pawn's moves by using the b2-square. However, White can evict the bishop. 24 l:txc7 The e7-pawn is the next item on the menu. i.b2 24 ... Black would do no better with 24 . . .!iJb5 25 l:txc6 or 24 . . .e6 25 l:ta4. Now White must be precise. 25 !lJa4! This move guarantees decisive material gains, and is the clearest way to win. Instead the careless 25 l:txe7 would allow Black counterplay with 25 ...i.xc3 26 i.xc3 !iJb5 and then . . .l:td8, eyeing the a2pawn. Although White is still bet­ ter, Black is in the fight again. !iJbS 25 ... After 25 ...i.xe5 26 l:txe5 e6 White wins an exchange at least. 26 l:txc6 (D) 98 London/Leningrad Wch (19) 1986 White simultaneously creates a passed pawn and knocks the sup­ port away from the b5-knight. 26 lbe7 would be less logical, since a knight firmly planted on b5 would guarantee counterplay for Black. l:lfd8 26 Black's moves are forced. 27 l:lb6! This is the main point of White's plan. Black's knight has had an un­ fortunate career in this game, and it is now hounded out of the one good square it has found. l:ld5 27 28 .i.g3 I avoided a last, though transpar­ ent, snare here: 28 lL!xb2? :Xe5 29 l:lxe5 axb2 would even win for Black, e.g. 30 l:le 1 l:lxa2 followed by 3 1 ...l:la l . Now the advance of the c-pawn decides the game quickly. lL!c3 28 .i.xc3 29 lL!xc3 30 c6 .i.d4 31 l:lb7 1-0 ••. ••• to the system brought into practice by Ragozin, which Najdorf then worked on in detail. But here as well an opening surprise awaited Black. 1 d4 lL!f6 g6 2 c4 3 lL!c3 d5 4 lL!f3 i.g7 dxc4 5 ..,3 6 Wxc4 0-0 lL!a6 7 e4 In the two previous odd-num­ bered games in this match Black preferred 7 ...i.g4. 8 J.e2 cS 9 d5 e6 10 0-0 (D) 10 i.g5 exd5 1 l lL!xd5 .i.e6 1 2 0-0-0 .i.xd5 1 3 l:lxd5 'it'b6 i s not dangerous for Black. Game 23 Karpov - Kasparov London/Leningrad Wch (19) 1986 Griinfeld Defence Despite his disappointment in the Smyslov Variation (game 17), Kas­ parov did not abandon the Griin­ feld Defence. This time he turned B 10 11 exdS 12 i.f4 (D) exd5 .i.£5 Karpov - Kasparov In the Seville match I twice played 1 2 l:td1 l:te8 1 3 d6 h6 and then 1 4 h3 (game 1 5), or 1 4 .i.f4 (game 2 1 ). Both encounters ended in a draw. 99 I. B . I. B*B � -·� � - �·· � � -· �B B B i B BlD. 8 B.t.B B•B�fl B B B BlDB � .t� 6 8 � �� �8� � � � �:� � � � .: � � B B :es 12 1 2 . . ....b6 and then 1 3 h3 (or 1 3 .i.e5) has independent interest. Here a few practical examples have accumulated, but it is too early to make a definitive judgement. l£le4 13 l:tad1 Here 1 3 . . .,..b6 has also been tried, and then 14 ,..b5 (or 14 l£lh4) with chances for both sides. 14 ltJbS! (D) The leap by the knight to b5 was prepared at home. Until this point only 14 .i.e3 and 14 .i.d3 had been tested, and White had achieved nothing. The basic strategic conflict in the middlegame is centred around the d-pawn. If White manages to ••• use its potential energy, he will have the initiative, while if Black manages to set up a reliable block­ ade, his position will become more promising. Having led the knight to b5, I was ready to part with the b-pawn in order to develop pres­ sure in the centre after 14 . . . .i.xb2 15 d6 .i.f6 16 .td3 ! and 17 l:tfe2 . Kasparov refused the offer. ...f6! 14 15 .td3 l£Jb4 After 15 ....tg4 16 .te5 l:txe5 17 l£lxe5 .txd1 1 8 .txe4 irxe5 19 l:txd 1 White has gained the initia­ tive. Not wishing to condemn him­ self to passive defence, Black risks everything, but the exchange sacri­ fice is not justified. The capture 1 5 ......xb2 is definitely risky, al­ though later practice in the vari­ ation showed that Black can gain counterplay by continuing with either 15 ...l:tad8 or 15 ....td7. l£lxd3 16 l£lc7! ••• lOO LondonllAralragrad Wch ( 19) 1986 l:.xe8 17 lllxe8 18 'ii'x d3! (D) 18 ... 'ii'xb2 Black's misfortune is that nei­ ther 1 8...lllxf2 nor 1 8 ...lllg3 works because of 1 9 'ii'b 5, attacking the undefended rook on e8. 19 l:.de1 The straightforward 19 d6 (inci­ dentally threatening to play 20 g4) 19 ...lld8 ! 20 We3 h5 2 1 llb 1 'ilfxa2 22 l:.xb7 'it'd5 23 l:.xa7 lbxd6 24 .i.xd6 'ilfxd6 25 lbg5 is unpleasant for Black, and it is therefore prob­ ably worth playing. However, I de­ cided to pin the knight, limiting the enemy pieces' activity. Of course 19 l:.fe1 does not work in view of 1 9 . . .'ilfxf2+ 20 �h1 'ilfxel + ! 21 l:.xel lbf2+ 22 �g1 l:.xe1+ and 23 ... lbxd3. 'i!fb4? (D) 19 This is the decisive mistake. Af­ ter 1 9 . . . lbf6 20 l:.xe8+ lbxe8 2 1 ..• 'ii'e 3 realizing the material advan­ tage is not difficult, but in the event of 1 9 .. .'ii'xa2 20 Wb5 l:.d8 2 1 'ii'xb7 'ii'xd5 Black holds on: 22 'ilfxd5 .:Xd5 23 g4 lbf6!. w 20 lbd2! Playing for the pin - the main motif of White's idea. 'ii'a4 20 21 'ilfc4 Forcing an exchange of queens; 21 lbxe4 is not so clear: 2 1 . . .l:.xe4 22 l:.xe4 .i.xe4 23 'i!fd2 c4 with counterplay. 21 'ilfxc4 .i.c3 22 lbxc4 22 ...b5 23 lbd2 lbf6 24 l:.xe8+ lLlxe8 25 l:e1 , etc., loses immedi­ ately. 23 liJd2 .i.xd2 .i.d7 (D) 24 .i.xd2 During the game many people thought that Black had got himself out of his mess: the d2-bishop is ••. ••• Karpov - Kasparov 101 under attack, and . . . .i.b5 is threat­ ened. - - �. -·­ · · --*- - · - · - - -·- ·8- - - -·- "" �A � � A � . � -%m·;; -"'B LJ LJ � B� % � 0+ � , y � /� '""' !.;.; % -""' � .: � . . . . � � %- w , !/;::" 37 �e2 38 ..td3 39 �c4 w 25 .i.f4! 26 f3! .i.b5 g5 After 26 . . . .i.xfl 27 'itxfl lLlf6 28 l:lxe8+ lLlxe8 29 .i.e5 ! the d­ pawn finally shows what it is capa­ ble of: 29 . . . f6 30 d6 ! and he has to give back the piece. 27 .i.xg5 .i.xn 27 . . . lLlxg5 is no better after 28 .:xe8+ .i.xe8 29 h4 ! , and White wins the knight. 28 �xn 29 .i.e7 lLJd6 lLlc8 (D) 29 . . . lLlc4 30 d6 lLlb6 3 1 .:bl lLld7 32 .:xb7 is also bad. 30 31 32 33 34 35 36 .i.xc5 .:eS .:r5 .i.d4 .i.xf6 .:g5+ .i.xg5 .:d8 f6 b6 lLle7 .:xd5 .:xg5 lLlc6 �f7 �e6 lLle5+ (D) - - - . . •• � •.• � • . . . . - � - -<&it- - • - .8. " 8. - • - • -8u - - - . w lLlc6+ 40 �d4 41 �c4 Here the game was adjourned, and Black resigned without resum­ ing. After 4 1 . ..lLle5+, both 42 'itb5 and 42 ..td4 lLlc6+ 43 ..te4 lead to victory, as the white pawns are ir­ repressible. 1-0 102 DNIHJI OL (14) 1986 Oame 24 Karpov - Sznapik Dubai OL (14) 1986 Sicilian Dragon This encounter took place in the last round of the Olympiad; our team could not afford to drop even half a point, so this game was deci­ sive. By winning the match 4-0, we won the Olympiad, ahead of Eng­ land by only half a point. 1 e4 c5 d6 2 lDf3 3 d4 cxd4 lDf6 4 lDxd4 g6 5 lDc3 6 ..te3 ..tg7 0-0 7 f3 lDc6 8 Wd2 9 ..tc4 ..td7 lDeS 10 0-0-0 l:tc8 1 1 ..tb3 12 h4 Sometimes 12 ..tg5 is played im­ mediately, but obviously, including the moves h2-h4 and . . .h7-h5 is in White's favour for a future attack on the kingside. 12 h5 Black previously did not prevent the h-pawn from moving. Its sac­ rifice - 1 2 ... lDc4 1 3 ..txc4 l:txc4 14 h5 ! ? - has been seen countless times, but a definitive conclusion has not yet been made. However, it most often favours White, and ..• therefore in reply to h2-h4, the symmetrical . . . h7-h5 has now gained popularity as a means of hindering the movement of the white g- and h-pawns. 1 3 ..tgS (D) B Attempting to blow up the en­ emy fortress immediately with the help of 1 3 g4 is unsuccessful, and the bishop move to g5 looks more solid than ideas such as 13 ..th6, 1 3 l:tdgl and 1 3 �bl . 13 l:tc5 1 3 ... lDc4 and 1 3 . ..lDh7 are less convincing. 14 �b1 This waiting move, which has a definite idea that will become clear later, was suggested to me by Gel­ ler while I was preparing for my match in Merano. I had to wait five whole years to use the idea in a game ! 14 b5 (D) ••• ••• Karpov - Sznapik 103 w 15 g4! Until that time 1 5 l:lhe1 was played here, simply centralizing. hxg4 15 A fundamental reply, which we looked at in 1 98 1 . Some rounds earlier in the USSR-Bulgaria match, Kiril Georgiev chose another prob­ lematic continuation against me, 1 5 ... a5, which we had not had time to prepare for. The correct reaction was 1 6 gxh5 ! , but I went for the more modest 1 6 i.xf6 i.xf6 1 7 a3 and the game soon ended in a draw. 16 h5! l£lxh5 Obviously, it was worth Black giving up the exchange immedi­ ately - 16 ...l:lxc3 ! ? 17 bxc3 l£lxh5 1 8 l:lxh5 gxh5 1 9 'ifh2 lbc4 20 'ifxh5. White has a certain initia­ tive, but the whole game still lies ahead. 17 l£Jd5! Now the idea behind White's 1 4th move becomes obvious. An ••• analogous position has arisen in practice several times, but without the moves �c 1 -bl and ...b7-b5. In that case the manoeuvre 1 6 l£Jd5 (now it is made a move earlier) af­ ter 16...l:lxd5 ! 17 i.xdS 'irb6 is not dangerous for Black in view of his pressure on b2 ( 1 8 ...l£lxf3 is threat­ ened). Including the moves �b1 and ... bS seriously changes the situation - the exchange sacrifice on d5 is now insufficient, as the b­ file is covered by his own pawn, and after 17 ...l:lxd5 1 8 i.xd5 gxf3 19 lbr5 it is only one step to victory. 17 l:le8 (D) ••• � · � -� · ri@;· � �� � ��'- � �... ri@; � � ri@; ' � . - � • � B m• • l. ��-"l.J- - . ��-·· Bi.. B � B �" pm � �� u�ri@; � B<t>B : • • : w 1 8 l:lxhS! White sacrifices the exchange himself, hopelessly weakening the guard around the enemy king. 18 gxhS 19 'ifh2! (D) This precise position was the subject of our old analysis. Then ••• 104 Dubai OL (14) 1986 we established that White's small material outlay is repaid with inter­ est when the black king is attacked. Because the defence cannot man­ age without ... e7-e6, Black should move his queen from d8, and we considered it necessary here to re­ turn the exchange - 1 9 ... llxd5 20 .i.xd5 (20 exd5 l£lxf3 21 l£lxf3 gxf3 22 'ibh5 .i.f5 is not so clear) 20 . . . 1ib6. Now if 2 1 'ifxh5, then 2 l . . . e6 is possible, with a resilient defence, but Black's fortress can be assaulted by 21 l£lf5 ! .i.xf5 22 exf5 ltJxf3 23 'ii'xh5 (the pawn must be taken at precisely this moment; if 23 .i.xf3 gxf3 24 'ifxh5, then 24...'ifc5 !) 23 ...ltJxg5 24 'ifxg5, and pressure by White on the g- and h­ files threatens the most unpleasant consequences. 19 ... a5 is terrible for Black, for example 20 'ifxh5 a4 2 1 llh 1 axb3 22 'ifh7+ �f8 23 .i.h6 .i.xh6 24 'iVh8#. 19 ... llc4 Black prefers to give the rook back for the bishop rather than the knight, but it does not ease his problems. bxc4 20 .i.xc4 21 'ii'xhS f6 22 f4! liJf7 (D) He will be saved by neither 22 . . . e6 23 l£lxf6+ .i.xf6 24 fxe5 .i.xg5 25 1i'g6+ nor 22 ...fxg5 23 fxe5 dxe5 24 llh1 (24 l£lf5 ! is even stronger for White) 24 ... exd4 25 'i'h7+ �f7 (25 . . .<it>f8 26 'ii'g6 ! , threatening both 27 l:l f l+ and 27 l:lh7, gives White a decisive attack) 26 llfl + <it>e6 27 'ii'xg7 .i.c6 28 'ifxd4. 23 .i.b4! There is no sense in rushing, as Black is totally helpless. 'ifb8 23 Again, if 23 ...e6, then 24 l£lxf6+ .i.xf6 25 1i'g6+ is decisive; after ..• Karpov - A. Sokolov 105 23 ...lbh6 there is 24 f5 l:tf8 25 lbf4 .te8 26 lbg6 .txg6 27 fxg6 fol­ lowed by lbf5. c3 24 l:th1 'ifb7 25 b3 26 r5 lbeS 27 lbe6 1-0 If 27 . . ..txe6, then 28 'iWxe8+, and otherwise there is no defence against 28 .txf6 .txf6 29 'iWh7#. Game 25 Karpov - A. Sokolov Linares Ct (10) 1987 Queen's Indian 1 d4 lbf6 e6 2 c4 b6 3 lbf3 .ta6 4 g3 .tb4+ b3 5 6 .td2 .te7 7 lbc3 It is curious that the Queen's In­ dian was played in all five games where I had White in this match against Sokolov, and this position arose automatically. It must be said that the results exceeded expecta­ tions - I won three games, and the other two were drawn. However, my success was due not so much to good opening play, but to victories in the ending. 7 0-0 Other games in the match against Sokolov continued 7 . . . d5, and after 8 cxd5 lbxd5 9 .tg2 0-0 10 lbxd5 exd5 1 1 0-0 lbd7 a position arose which was well known from the first Karpov-Kasparov match. In situations like this, White's altered move order (.tg2 and lbc3) does not matter. But now, after Black has already castled kingside, the position of the knight on c3 is not that harmless for him, since White can seize the centre by means of e2-e4. d5 8 e4 9 cxd5 .txn 10 �xn exd5 l l eS lbe4 12 'iie2 (D) In game 8 I played 1 2 l:tc l , pre­ venting immediate regrouping by means of ...'iid7 and ...lbc6-d8-e6. However, this loss of tempo makes itself felt, and Black could gain a satisfactory position by playing ...c5 and ...lbc6. ••. B IM UMNI a (10) 1�1'1 Thll time I managed to find a more precise arrangement for my pieces. The position of the queen has great attacking potential con­ nected with preparing f4-f5. 18 -txd4 'W'f5 (D) on e2 besides everything else has the important virtue of preventing ... f7-f5; after 12 ...f5 1 3 exf6, the e4-knight is forced to abandon the centre, and White has the e-file and the e5-square at his disposal. 12 �xc3 'W'd7 13 -txc3 14 �g2 �6 The idea behind this somewhat fanciful knight manoeuvre is to bring it to e6 to blockade White's pawn centre. 15 l:the1 Introducing a natural plan: White will throw the f-pawn for­ ward and simultaneously control the e-file. The move 1 5 l:tac l was played here previously, allowing Black to construct a defensive for­ tress: 15 ... �d8 1 6 -tb2 �e6 17 l:thd l l:tae8. 15 �d8 16 �g1 c5 1 6 ...�e6 is still impossible: 17 f4 f5 18 exf6 l:txf6 1 9 f5 ! wins a piece. In the event of 16 ... f5 17 exf6 -txf6 1 8 f4 followed by �f3e5, White has the advantage. 17 f4 cxd4 Black cannot manage without opening the c-file, and he decides to do so immediately. 1 7 . . �c6 1 8 �f3 g6 is risky, since White 19 l:tad1 -tb4 �e6 20 :n It looks as though Black is on the point of seizing back the initia­ tive, but, as so often happens, he is a tempo short. If the rook were on c8, White would not have the pos­ sibility of 21 'ii'd3 due to 2 1 ...l:tc2+ 22 l:tf2 'ii'xd3 23 l:txd3 �xd4 24 l:txd4 :Xf2+ and 25 ...-tc5 . 21 'ii'd 3! Destroying Black's blockade on the light squares. 21 'ifxd3 l:tac8 22 l:txd3 23 �f3 23 f5? is too hasty: 23 ...�xd4 24 l:txd4 l:tc2+, when 25 l:tf2? fails to 25 ...l:txf2+ 26 �xf2 -tc5. 23 l:tc2+ :res (D) 24 :n ••• . w Karpov - A. Sokolov 107 33 lbf3 .:.cl (D) Obviously, now Black is simply placing his hopes on a counterat­ tack against the white queenside pawns. On other parts of the board he has lost the fight. 25 f5 ! The phalanx of white pawns finally begins its forward march. 25 26 27 28 29 30 lbxd4 �xfl g4 �3 �g3 lbxd4 lbfl+ .:.et �8 .:.n+ .:.et Checking from behind achieves nothing: 30....:.g1 + 3 1 �f4 .:.n + 32 lbf3 , while Black already has to bear in mind the direct threat of 3 1 lbc6. 3t �f4 After 3 1 g5, in the event of 3 I .. ..:.g1+ (or 3 l ...�e8 32 lbf3) 32 �f4 .:.n+ 33 lbf3 things are not looking good for Black, but after exchanging rooks with 3 I . . . .:.c3 32 .:.xc3 .i.xc3 33 lbc6 aS 34 �f4 �e8 his resistance could have been prolonged. 3t 32 h4 h6 �e8 w 34 a4 35 lbd4 36 h5 .:.bl i..e7 I spent a lot of time on this move. Of course, the continuation 36 g5 hxg5+ 37 hxg5 would have given me chances for victory, but it is more useful to fix and 'seal up' Black's kingside and his pawns. I am not exaggerating when I con­ firm that at this point I had already calculated a beautiful combination a dozen moves long. But at the same time I will not hide the fact that I foresaw the possibility of the pawn break. 36 37 � a6 i..cS /OB LlnaNs Ct ( 10) 1987 38 l0e2 d4 39 l0f4 �d7 40 e6+ �e8 Both now and later, taking on e6 is impossible since the g7-pawn would be lost. 41 �e4 aS Here the game was adjourned. Analysis showed that the game could be won by study-like means. 42 .Uf3! (D) B The sealed move, which seemed the strongest. The rook lies in an ambush, anticipating the black king wishing to cross the f-file. 42 .Ubl Preparing for 43 �d5, in which case after 43 ....Ugl 44 l£Jd3 .Uxg4 45 lbxc5 bxc5 46 �xc5 fxe6 47 fxe6 �e7 Black would have held the position. However, White's knight goes off in quite a different direction. .Ugl ! 43 l£Jd5! 43 ...�f8 caused more trouble in analysis. At first I examined the move 44 lbc7, but I stopped every­ thing when I found that after 44 ... .Ue 1 + 45 �d3 .Ue5 ! Black can hold the position. Only the next morning did I manage to find the correct route: 44 e7+ ! i..xe7 45 �xd4 .Ugl 46 �c4 .Uxg4+ 47 �b5 i..d8 48 .Uc3 ! .Uh4 49 �c6 ! .Uxh5 50 <i'd7 .Uxf5 and here 5 1 lbxb6! is the end. 5 1 . . . i..xb6 allows a rather funny mate: 52 .Uc8+ i..d 8 53 .Uxd8#. Therefore Black has to give back the exchange with 5 1 .. ..Uf3, and after 52 .Uxf3 i..xb6 the inevi­ table break b3-b4 is decisive. A unique case, in which one king is the main director and executor of an attack on the other! 44 <i'd3! .Uxg4 45 f6! (D) ••• B This is White's quintessential idea. He is threatening 46 l£lc7+ Karpov - A. Sokolov 109 and 47 e7+, and none of his pawns can be taken: 45 ...fxe6 46 f7+ �d7 (46 . . . �f8 47 li:Jc7 and 48 li:Jxe6+; 46 . . .�d8 47 f8'ii'+ �xf8 48 llxf8+ 'iti>d7 49 �xb6+) 47 li:Jxb6+ 'iti>c7 48 li:Jc4, threatening li:Je5 and li:Jg6. 45 . . . llg5 46 li:Jc7+ 'iti>d8 47 e7+ �xe7 48 fxe7+ 'iti>xe7 would have been more stubborn, with hopes for the kingside pawns. 45 46 li:Jxb6 ••• �d6 llgS Again 46 . . .fxe6 will not do: 47 f7+ 'iti>f8 48 li:Jd7+. Black is not helped by 46 . . . gxf6 either: 47 llxf6 ! fxe6 48 llxe6+ �e7 49 li:Jd5 leads to a winning pawn endgame for White. 47 fxg7 llxg7 The final hope for salvation lay in 47 . . .fxe6 48 li:Jc4 �e7 (if Black plays 48 ... �b4, then 49 �xa5 ! is decisive) 49 li:Jxa5 llxg7 50 li:Jc6+ 'iti>d7 5 1 li:Jxd4, although I suggest that I would have coped with the diffkulties. 48 li:Jc4 49 exf7+ so llxf7 �b4 llxf7 'iti>xf7 (D) The storm of the combination has abated, and the game has moved into its technical phase. This gives the impression that the worst is over for Black, but there is more to come. 51 li:JeS+! �6 52 li:Jc6 �el - - . ­ • • ••• . . - . - . . -� • d � �ttJ� • - - d · � ·�· . • • • • . . - . w 53 li:Jxb4 was threatened, while after 52 ...�c5 53 li:Jxa5 �g5 54 li:Jc6 �xh5 55 b4 the a-pawn will queen. 53 li:Jxd4 �b4 Black would not be saved by 53 . . .�f2 54 li:Jc6 �g5 55 �xa5 �xh5 56 �e2 ! , followed by 57 'iti>f3 and, if necessary, 58 �g2. The queenside pawns can easily move forwards. Surprisingly, the knight turns out to be stronger than the bishop with play on both flanks. 54 SS 56 57 58 li:Jc6 �e2 'iti>d3 'iti>c4 li:JxaS! �el �c3 �el ..tgs Both the most effective and the most striking line. Black cannot re­ fuse to take the piece: 58 ...�xh5 59 li:Jc6 'iti>g4 (after 59 ... �g5 60 b4 h5 White wins a decisive tempo by means of 6 1 li:Jd4, threatening 62 li:Jf3+) 60 b4 h5 6 1 aS h4 62 a6 h3 1 10 Stvllll Wch (2) 1987 (or 62 .tf2 63 lbd4) 63 a7 h2 64 aB'ii' h 1'ii 65 lbe5+. 58 .txaS 59 b4 .tdS 'itxhS 60 aS 61 '1tb5 .tgS 62 a6 .te3 63 'itc6 1-0 ... Game 26 Kasparov - Karpov Seville Wch (2) 1987 English, Four Knights 1 c4 I must confess that I was not in­ itially expecting Kasparov's con­ version to the English Opening in this match. However, this surprise at the start strangely turned into a gift for me straight away - at least, Kasparov spent almost an hour and half on his first ten moves ! It is in­ teresting that the novelty I used was prepared at the beginning of the 1 980s for the Merano match. However, instead of 7 lbg5, Korch­ noi retreated the knight to e 1 , and so it remained unused. 1 eS 2 lbc3 lbf6 lbc6 3 lbf3 .tb4 4 g3 0-0 5 .tg2 e4 6 0-0 .txc3 7 lbgS 8 bxc3 :es 9 f3 e3! ? (D) w It was precisely this move, sug­ gested in its time by Igor Zaitsev, which plunged my opponent deep into thought (until then Black had always taken automatically on f3). 10 d3! The correct reply, so White has not wasted his thinking time. Let us look at two other possibilities: a) 10 dxe3 b6 ( 10 ... h6 1 1 lbh3 d5 12 lbf4! dxc4 1 3 e4 is worse) 1 1 e4 and then: a l ) 1 l .. . .ta6 immediately is dangerous : 12 f4 h6 13 lbxf7 ! cl;xf7 14 e5 .txc4 (14 ... lbg8? 1 5 .td5+ '1tf8 1 6 .ta3+ lbge7 17 'iid3 d6 18 'iih7 ! winning) 15 exf6 gives White an advantage. a2) 1 1 ... h6 12 lt:Jh3 .ta6. The c4-pawn is doomed, and Black can be quite satisfied at the results of the opening. Kasparov - Karpov b) .i.b7 ! as in grade 10 d4 iLia5 1 1 'ifd3 b6 12 c5 and Black has the initiative, H.Olafsson-Naumkin, Bel­ 1988. 10 dS The only move, because other­ wise the e3-pawn will quickly be surrounded. 1 1 'ii'b 3! Yet another precise move; after 1 1 cxd5 iLixd5 two lines are possi­ ble: a) 1 2 iLie4 f5 1 3 c4 fxe4 14 cxd5 exf3 15 l:txf3 ilJd4 1 6 l:txe3 .tg4 17 .te4 'iff6 1 8 .tb2 'ifb6 ! 19 .txd4 'ii'xd4 20 'iic l .txe2 21 1itg2 .txd3 with a winning position for Black. b) 1 2 'ii' b3 ilJa5 ( 1 2 ...'iixg5 1 3 f4 iLixf4 1 4 l:txf4 leads to a sharp game) 1 3 'iia3 'iixg5 14 'iixa5 'ii'e5 ( l4 . . . b6 15 f4) 1 5 d4 ( 1 5 f4? iLixf4 16 'iixe5 iLixe2+ 17 1ith l l:txe5 1 8 .i.b2 .:r.b5) 1 5 ...'ifd6, and the game is roughly even: 1 6 .ta3 'ifc6 17 f4 'iixc3 1 8 'ifa4 iLif6. 11 ilJaS 12 'ii'a3 c6 13 cxdS cxdS (D) 14 f4 ilJc6 15 l:tb1 'ii'c7 After 15 ....i.g4 16 .l:tel (the most precise; after 16 l:txb7 .txe2 17 lbxf7 'ifa5 ! 1 8 'ifd6 .i.xfl 1 9 iLih6+ lith8 20 ilJf7+ White has to settle for perpetual check) 1 6... h6 ( 1 6 . . .'iid7 1 7 h3) 1 7 iLif3 'ifd7 1 8 ..• Ill · ·i.� · -·· �·· �a • . . . . � - • •• .• d - ··� ·� � . � � . . - . • �"' '/�!iJAo . % oA B "�' '�"' ' /,- !'? !1,::: � 8 Bl · 8 -�B � � � � - • .:* w d4, by directing the knight to e5, White is better. 16 .tb2 .tg4 17 c4 (D) Of course, it is very tempting to threaten the enemy king, but this gives Black the initiative. 17 iLif3 'ifd7 1 8 .tal .l:te7 ! ? 1 9 l:tfc l .i.h3 is not dangerous either - Black has sufficient counterplay. However, by continuing with the quiet 1 7 l:tfe l White would have retained good chances. 17 dxc4 gxf6 18 .txf6 19 ilJe4 litg7 1 9....:Xe4 ! ? 20 .i.xe4 f5 ! 2 1 .tf3 ilJd4 22 dxc4 i.xf3 23 exf3 e2 24 l:tfel 'ifxc4 is also interesting. 20 dxc4 Now the advantage fully trans­ fers to Black. The continuation 20 iLixf6 ! ? litxf6 21 .:r.b5 l:tad8 22 'ifc3+ lite? 23 'ifxc4 would have led to uncertain complications. 1 12 S1vlll� Wch (2) 1987 l:.ad8 20 More reliable than 20....txe2 2 1 lbxf6! . 2 1 l:.b3 (D) Apparently the decisive error. After the continuation 2 1 lbc3 li:)d4 22 li:)d5 there is still a stub­ born fight to face. Its direction de­ pends on the queen's choice of whether to go to d6 or c4: a) 22... 'ii'd6 23 'irxd6 l:.xd6 24 l:.xb7 .txe2 25 l:.e1 .i.xc4 26 lbxe3 l:.de6 27 l:.b4 lbe2+ 28 �f2 .i.d3 29 lbd5 lbxf4 30 lbxf4 :Xe1 3 1 lbxd3 l:.8e2+ 32 �f3 l:.e3+ 3 3 �f2 l:.1e2+ 34 �1 :Xg2 35 �xg2 lhd3 36 l:.a4, and as a result chances are equal. b) 22 ...'ii'xc4 23 li:)xe3 'ii'xe2 24 li:)xg4 'ii'xg4, with a tense situation on the board. ... B Here many commentators have ex­ amined the more natural 20 'ii'c 3, and in this case after the correct re­ ply 20 . . .'ird8! (20 .. .'1Ve7 2 1 l:.xb7 ! 'ii'xb7 22 lbxf6 �f8 23 lbxg4 ! and White has the upper hand) 2 1 dxc4 'ii'd4 22 :Xb7 l:.xe4 23 'ii'xd4 l:.xd4 24 .txc6 the chances are roughly even. Another dangerous manoeuvre is 20 lbd6, and Black has to play carefully: 20 . . .l:.e6 2 1 lbxc4 l:.d8 22 f5 l:.ee8 23 l:.b2 lbd4 24 l:.xb7 lbxe2+ 25 �h1 lbxg3+! 26 �g1 lbe2+ (neither 26 ...'ii'xb7 27 .txb7 lbxfl 28 �xfl .txf5 29 lbd6, nor 26 . . . lbxfl 27 l:.xc7 e2 28 'ii'xa7 e1 'iV 29 l:.xf7+ �h6 30 l:.xf6+ ! �g5 3 1 'ii'g7+ �f4 32 'ii'h6# will do) 27 �h l 'ii'g3 with perpetual check. However, in the last vari­ ation Black could take a risk: 27 ...'ii'c 8 ! ? 28 lbd6 l:.xd6 29 'ii'xd6 .txf5 with enough compensation for the exchange. Karpov - Farago 113 24 l:td3 25 hd8 �xe2 hd8 (D) w 26 l:te1 l:te8 The game is decided. Black has prepared a simple mating idea: 27 lDd6 ltlxd6 28 1i'xd6 �f3 29 :Xe8 1i'fl#. 27 1i'a5 bS 1i'd3 28 ltld2 29 ltlb3 �f3 30 �xf3 1i':d'3+ l:txe1+ 31 �g1 32 1Wxe1 lLle3 0-1 Game 27 Karpov - Farago Wijk aan Zee 1988 French, Winawer 1 2 3 4 e4 d4 lbc3 eS e6 dS �b4 ltle7 S a3 �xc3+ 6 bxc3 cS 7 1i'g4 I had prepared this variation for my match against Korchnoi at Ba­ guio. Ten years later, I finally got a chance to try it in practice. 7 1i'c7 8 1i'xg7 l:tg8 9 1i'xh7 cxd4 One of the key positions in this opening. 10 ltle2 A versatile manoeuvre: White's knight is covering the queenside and simultaneously supporting the pawn advance on the kingside. 10 ltlbc6 The pawn on e5 is invulnerable: 10 ...1i'xe5 1 1 �4 1i'f6 12 cxd4 with a big advantage to White. 1 1 f4 White's basic idea is to prepare a pawn advance on the kingside, and 1 1 �f4 runs contrary to this. Inci­ dentally, l l cxd4 is no good due to 1 1 .. .ltlxd4 ! . 11 �d7 dxc3 12 1i'd3 13 1Wxc3 (D) White takes the pawn immedi­ ately, and with it opens up both the c-file and the entire queenside, so he has to be extremely careful. 1 3 lDd4, 1 3 ltlg3, 1 3 �e3, and 13 l:tb1 have all been played several times, but with material equality Black ..• 1 14 W(Jl Cltlll bt 1988 can easily count on a successful outcome. There is no difference in value between taking with the queen or the knight on c3, and in any case I have tried both routes. If the queen takes on c3, the black c6knight turns out to be pinned, while the queen herself does not feel too settled (the threat is the advance ...d5-d4 ); if the knight takes it then the black knight on c6 is free to manoeuvre. B ti::lfS 13 14 l:tb1! This move has been given an ex­ clamation mark not because it is so cool or particularly strong, but due to other considerations: this game is unique in the number of short moves made by the white rooks out of the 27 moves remaining in the game, 1 5 are made by the rooks! 14 l:tc8 15 ..td2 b6 In his game against Tal in the same tournament, Farago used the successful novelty 1 5 ... a6 ! and equalized. I still think that all White's resources have not yet been exhausted here. 16 g3 (D) The exchange 16 ti::lg3 ti::lxg3 1 7 hxg3 ti::le7 1 8 "iixc7 l:txc7 19 ..td3 l:txg3 20 l:th8+ l:tg8 2 1 l:txg8+ ti::lx g8 22 ..tb4 led to a better end­ game for White in Martic-Dras­ kovic, corr. 1 987, but I prefer to maintain the tension on the board. Incidentally, only now does this game acquire independent signifi­ cance. Of course, I would have liked to play g2-g4 in one go, but if 16 l:tgl , then 1 6 ...1fd8 ! with the threat of ...ti)cd4 and ...1fh4+. ••• ••• 16 17 .d3 18 l:tg1 .b7 ti::lce7 l:tc4! (D) Karpov - Farago 1 15 • • •• • • ....... . . . . -·- ··� . ��· · -- � � � �-- � �}, . " � � � -· � rff// � -� �\Wr u �- "' · � mltj• u - . � • � � ,Wit : � · � i.. " � - w Black's rook is ideally placed on c4: it is working both along White's fourth rank and the c-file, and furthermore it is covering the fl -a6 diagonal from the possibility of transferring the queen to a6 or the bishop to a4 (or b5). 19 g4 ll'lh4 ,.c8 20 l:.g3 ll'lhg6 21 l:.b2 22 l:f3! (D) Overprotection of the f4-pawn. The white rook on g3 was beauti­ fully fulfilling its defensive func­ tion on the third rank and the g-file, but now it is time to get down to ac­ tive operations, for which it is nec­ essary to regroup. 22 ll'lc6 Black cannot repeat the position with 22 ... ll'lh4 because of 23 l:.f2 ! , and then 2 3. . .l:.xg4 i s impossible in view of 24 'ilfh3 ! , when White is better. .:hs 23 m ••• B 24 h3 ii'd8 25 'ii'g3! Following Black's example of rearranging his pieces on the c-file, I also decided to swap round my queen and rook. 'ilfe7 25 ii'cS 26 .:o 27 'itd1 White is planning to begin ma­ noeuvring the knight, therefore it is best to remove the king from the possibility of a check from c4 be­ forehand, simultaneously defend­ ing the pawn on c2. Incidentally, transferring the king to dl now is better grounded than it was on move 10. l:.a4 (D) 27 In a complex position Black is losing the thread of the game. It was necessary to play 27 ... d4 28 ll'lc l l:.c3 29 .i.xc3 dxc3 30 l:.b5 'ilfd4+ 3 1 l:d3 'ilfxf4 32 'ilfxf4 ll'lxf4 33 l:.xc3 ll'ld5, and Black holds on, ••• 1 16 Brussels World Cup 1988 although 3 1 i.d3 is stronger for White. w 28 .Uc3! Giving Black no respite. 28 lbc 1 lba5 29 l:c3 � 30 l:b4 is weaker. Now Black should return the rook to c4, as retreating the queen leads to an almost forced loss. 'ile7 28 ... 29 .!i)ct The white pieces have unex­ pectedly developed stormy activity exactly where Black usually domi­ nates. White is threatening 30 i.b5, and if29... lba.S, then 30 .Ub4 is rea­ sonable. d4 29 Somewhat spoiling White's in­ tentions, but having torn itself loose, the central pawn will need constant defence. .!i)a5 30 .Ucb3 31 .Ub4! .!i)b7 i.xa4 32 .Uxa4 ••• 33 .Ub4 34 .!i)b3 Matters are coming to a head; 34...i.xb3 is impossible because of 35 i.b5. 34 d3 35 'ili'xd3 'ilt'xd3 36 i.xd3 i.xb3 37 .Uxb3 .!i)cS Not 37 . ..l:xh3 because of 38 i.b5+. </;e7 38 .Uc3 39 i.n .Ud8 1-0 40 .Uc4! Game 28 Karpov - Timman Brussels World Cup 1 988 Queen's Gambit Accepted d5 1 d4 dxc4 2 c4 3 e4 The classical continuation is 3 .!i)f3, 4 e3 and 5 i.xc4, immediately winning back the pawn. However, sometimes I like playing in pure gambit style, acting in accordance with the name of the opening! .!i)f6 3 4 e5 .!i)d5 .!i)b6 5 i.xc4 6 i.d3 It is still not known which re­ treat is better for the bishop, d3 or b3. Before we continue I would like to pay a little attention to the Karpov second of these: 6 .i.b3 lDc6 7 .i.e3 .i.f5 and now: a) 8 lbe2 e6 9 0-0 lDa5 I 0 .i.a4+! c6 1 1 .i.c2 .i.g6 12 lbbc3 lbac4 1 3 'iVc 1 .i.e7 1 4 .i.xg6 hxg6 1 5 lbe4 l:h4 1 6 lD2g3 'iVd5 1 7 'iVc2 0-0-0 1 8 .i.g5 .i.xg5 1 9 lbxg5 l:d7 20 lDf3 l:tf4 21 l:ad1 g5 22 1fc1 ! gave White an appreciable advantage in Karpov-Speelman, Brussels 1 988 and the game ended in my favour. b) 8 lDc3 e6 9 lbge2 .i.e7 10 a3 0-0 1 1 0-0 lba5 I 2 .i.c2 .i.xc2 1 3 'it'xc2 lDac4 14 l:adi lDxe3 1 5 fxe3 and I quickly managed to seize the initiative in Karpov-Short, Linares Ct (9) 1 992. 6 ... lbc6 7 lDe2 (D) White can also play 7 .i.e3 : a) 7 . .. lbb4 looks good: 8 .i.e4 f5 ! 9 exf6 exf6 1 0 lbc3 f5 1 1 .i.f3 lD4d5 1 2 .i.d2 .i.e6 1 3 lDge2 'iVd7 1 4 0-0 0-0-0 1 5 l:e 1 ( 1 5 a4 ! ? de­ served attention) 1 5 ...l:tg8 ! , and Black had solved all his opening problems in Karpov-Short, Linares Ct (3) 1 992. b) 7 ... .i.e6 8 lbc3 'iVd7 9 lDf3 0-0-0 10 h3 lbb4 1 I .i.e2 f5 12 0-0 h6 1 3 a3 lD4d5 14 lDe1 lDxc3 1 5 bxc3 .i.c4 1 6 lbd3 e6 I 7 a4 g5 I 8 1fc2 'iVc6 1 9 :tfc I gave White the better chances in the game Karpov­ Ivanchuk, Reggio Emilia I99112. 7 .i.g4!? 8 .i.e3 - Timman 1 17 B After 8 f3 .i.h5 and 9 ... .i.g6 (White must avoid continuing 9 lDf4 owing to 9 ...'iVxd4 10 lbxh5 'iVxe5+) Black has a comfortable game (the position resembles the Alekhine Defence, Four Pawns At­ tack). 8 ... .i.xe2 9 .i.xe2 'iVd7 Black does not use up a tempo playing ...e7 -e6, preferring instead to increase the pressure on the d4pawn. In a situation like this one immediately has to resort to tactics. 0-0-0 10 lbc3 11 a4! ? With the intention of distracting White's attention from the d4square. a6!? 1 1 ... If I l .. .lbxd4, then 1 2 a5 lbxe2 ( 1 2 ... lba8 is no better: I 3 a6 b6 I4 .i.f3, and the black knight is trapped in the corner of the board) I 18 Brussels World Cup 1988 1 3 axb6 �xc3 14 iic2 cxb6 1 5 'it'xc3+ with a very strong attack for White: 1 5 ... �b8 16 e6 iixe6 17 0-0, etc. 12 aS �dS 13 i.f3 �db4 (D) 1 3 ... e6? loses immediately in view of 14 �xd5 exd5 1 5 i.g4. Exchanging on c3 or e3 favours White, as he gains a strong pawn centre. However, 1 3 . . .�cb4 looks stronger. w 14 e6!? The prelude to all further play with an unusual correlation of forces. After the quiet 14 0-0, tak­ ing the central pawn is risky: 14 ...�xd4 (Black's best chance is probably 14 ... e6, with a complex game) 1 5 i.xd4 'it'xd4, and now alongside 1 6 Wb3 e6 17 llfd 1 iixe5 1 8 i.xb7+ �xb7 1 9 llxd8 i.d6 20 llxd6 iixd6 21 �a2 c5 22 �xb4 cxb4 23 'it'f3+, which gives White a minimal advantage, he could immediately take the bull by the horns with 1 6 i.xb7+ ! �xb7 17 Wf3+, and then both 17 ...c6 1 8 llfd I �d3 19 lla4 Wd7 20 � e6 2 1 �d6+ followed by 22 llxd3, and 17 ...�b8 1 8 llfd1 �3 1 9 lla4 'it'd7 20 lla3 ! are unfortunate for Black. 14 'ili'xe6 After 14 .. .fxe6 1 5 �a4 Black does not become disentangled: 15 ... �xd4 16 i.xb7+!. 'ili'eS 15 dS There is another interesting at­ tempt, 1 5 ...Wg6. Then the continu­ ation 1 6 i.h5 'ifxg2 17 'it'g4+! looks tempting, but if instead Black replies 1 6 ...�c2+ ! 17 �fl �xe3+ 18 fxe3 Wf6+ 19 �g1 e6, the complications are not at all in White's favour. Probably here as well everything would have been reduced to a queen sacrifice for rook and knight, but in a slightly different way: 1 6 0-0 e6 17 dxc6 llxd 1 1 8 cxb7+ �b8 19 llfxd l i.d6. White's initiative compen­ sates him for his small material loss, although the chances can be considered roughly equal. 16 0-0 e6 (D) 17 dxc6! The logical continuation of the plan White began on move 1 1 . 17 llxd1 �b8 18 cxb7+ .•• Karpov - Timman ll9 24 ... lDe5 loses: 25 l:tdxc7 lDxf3+ 26 gxf3 Wxa5 27 lDc5 Wxc7 28 lDxa6+ �xb7 29 l:txc7+! �xa6 30 l:txg7 and White has a technically won ending. w 19 l:tfxd1 i.cS 19 . . . i.d6 is more stubborn, pre­ venting the white rook from pene­ trating the seventh rank. 20 i.xcS 'iWxcS fS 21 l:td7 lDc6 22 l:tad 1 White was threatening l:td8+, and in the event of 22 . . . c6 23 h4 ! (but not 23 l:txg7 lDd3 ! 24 l:txd3 'iWe5 ! ) White has an obvious posi­ tional advantage. 23 lDa4 23 lDe2 !? is also good: 23 ...'ir'b5 24 l:tcl lDe5 (the most obstinate continuation is 24 ... lDa7 ! , though here as well Black experiences se­ rious problems) 25 l:tdxc7 lDxf3+ 26 gxf3 Wxa5 (Black cannot con­ tinue 26 . . . 'iWxe2 due to 27 l:tc8+ �xb7 28 l:t l c7#) 27 l:tc8+ �xb7 28 l:txh8 'ir'd2 29 �fl is insufficient for Black. 23 24 l:tcl w 'iWxaS 25 .:Xc6 26 l:txe6 �a7 27 g3 (D) 27 h4 ! is even stronger, com­ pletely depriving Black of counter­ play on the kingside. 120 Btlfort World Cup 1988 27 g5! 28 ltxh7! ltb8 29 hJ g4 fxg4 30 hxg4 31 .i.g2 'it'a1+ 'it'xb2 32 ..th2 'it'a2 33 lthh6 c5 34 ltef6 34...ltxb7 would have lost: 35 .i.d5 ! 'it'a5 (35 ...'it'xd5 36 ltxa6+ ..tb8 37 lth8+) 36 .i.xb7 ..txb7 37 ltxa6 'ifxa6 38 ltxa6 ..txa6 39 f4!. 35 ltf4 (D) B 'it'd2? ! 35 35 . . .c4 was necessary, although it is true that after 36 .i.d5 White has taken three black pawns for his own on b7, and finally an endgame would arise with a rook, bishop and pawns on f2 and g3, against the queen, which is good for White. 36 .i.n ! Lb7 ..tb8 37 ltxa6+ ..tc7 38 ltf8+ ••• 39 .i.g2 Co-ordination between the rooks and the bishop creates decisive threats. 39 'it'd7 c4 40 lth8 1-0 41 .i.e4 Game 29 Karpov - Kasparov Belfort World Cup 1 988 Griinfeld Defence 1 d4 l£lf6 2 c4 g6 d5 3 lbcJ 4 cxd5 lbxd5 lbxc3 5 e4 .i.g7 6 bxcJ 7 .i.c4 The plan of 7 lbf3 c5 8 ltb1 0-0 9 .i.e2 has enjoyed great popularity in recent years. Another plan of development for White involves supporting his mighty pawn centre with an early .i.e3, followed by bringing the light-squared bishop out to e2. c5 7 lbc6 8 lbe2 0-0 9 .i.e3 .i.g4 10 0-0 It is well known that in the event of 10 . .. cxd4 1 1 cxd4 lba5 1 2 .i.d3 .i.e6 1 3 d5 White can sacrifice the exchange, gaining a dangerous at­ tack. Karpov - Kasparov 121 1 1 f3 lLlaS (D) Now in reply to 12 i.d3 there is the possibility of 12 ...cxd4 13 cxd4 i.e6, when sacrificing the exchange with 14 d5 is not so effective, as in a lot of variations Black has a queen check on b6. However, in this case White is not at all forced to give up the rook for the bishop. w 12 i.xf7+ I specially prepared this cap­ ture on f7 for the Seville match. Immediately after it had ended Kasparov remarked that the vari­ ation is hopeless for White. But if you judge from what happened in the five games with it in Seville, you cannot say that it is all that easy for Black to solve the prob­ lems of the opening. Moreover the proposed post-match duel in gen­ eral ended miserably for him. Only later was it established that in the event of i.xf7+ and the exchange of light-squared bishops, Black gains a favourable game. l:txf7 12 13 fxg4 l:txfi+ 14 Wxfl The point of White's idea is not to take the pawn, as his activity compensates Black for his small loss, but to block in the enemy g7bishop with the help of the pawn chain c3, d4, e5, g5 and h4. 'iVd6 14 ... This queen thrust occurred in four out of five games in Seville, and it has also been used in many other encounters. The other con­ tinuations, 14 ...cxd4, 14 ...'ili'd7 and 14 .. .'ii'e8 , are not really justified. 15 eS 'iVdS l:td8 (D) 16 i.f2 w The threat of a capture on e5 forces the white queen to vacate d 1 . In this game I moved her to a4, while later against Timman I 122 B1(fo11 World Cup 1988 played 17 Wc2, a game which also proved to be very engaging. 17 . . . Wc4 1 8 'ifb2 .i.h6 19 h4, and now: a) 19 ...'iff7 20 �g1 (20 g5 ! ltX4 2 1 e6 ! is correct; thanks to this zwischenzug, White diverts the queen, and does not allow Black to double on the f-file: 2 1 . . .'iff5 22 �g3 Wxe6 23 Wxb7 llf8 24 lle 1 ! �e3+ 25 �g1 l:txf2 26 gxh6 cxd4 27 cxd4 llf8 28 'ifb2 �4 29 llxe6 1 -0 Vyzhmanavin-Emst, Stock­ holm 1 99 1 ) 20 ...l:tf8 2 1 �g3 �c4 22 We2 Wxf2+ ! 23 'ifxf2 .i.e3 24 'ifxe3 �xe3 and a roughly even endgame arose in Karpov-Kaspar­ ov, Amsterdam 1988. b) 19 ... llf8 20 g5 ! (a universal move which guarantees White an advantage in this case) 20...'ifd3 2 1 'ii'b 1 ! 'ife3 2 2 1We 1 .i.g7 2 3 �g1 We4 24 �g3 ! (White returns the pawn but seizes all the critical squares) 24...'irxh4 25 �4 llxf2 (a desperate sacrifice, similar to what happened in the Belfort game I am commentating on) 26 �xf2 cxd4 27 lld 1 d3 28 'ire3 ! �c6 29 �xd3 'ira4 30 'irf3 'ifa5 3 1 e6 �d8 32 �4 .i.e5 33 llJd5 'ifc5+ 34 �h1 1 -0 Karpov-Timman, Rotterdam 1 989. 17 'ifa4 (D) Now after . . .b6, the queen will (as we see) retreat to c2. The ques­ tion is, why does White choose to provoke such a useful pawn move as . . . b6? The fact is that after 1 7 'ifa4 b6 1 8 'ifc2 in some variations, for example in reply to 1 8 ...'ifc4, White switches the queen to e4 and, by attacking the knight, wins an important tempo. Thus, moving the queen out to a4 has its advan­ tages. But I will not undertake to decide definitively where the queen is more comfortable, al­ though I have tried both possibili­ ties in practice. • • ••• � � -·� � - -· - • • ••• - N �- u � � � � u � � � •• D ·�· � � � � � � � · ·tb· � �p� a • •�• B 17 ... b6 White gets a better game after 17 ...�6 1 8 'irb3 c4 19 'ifxb7. If Black prefers 17 . . .�c4, then after 1 8 �f4 1Wf7 1 9 g3 �d2+ 20 �g2, White has the advantage in the event of 20. . .lbe4 (Gutman) 2 1 'ifc2 �xf2 22 'irxf2, but Black has the strong reply 20 ...g5 ! as demon­ strated by Gutman. He further sug­ gested that White should play 1 8 Karpov - Kasparov 123 g5 ! himself, followed by 18 . . .:fs 1 9 �g1 Wf7 20 ..tg3 �3 2 1 ..tf4 l'bd5 22 :n ; Black's initiative has been arrested, and his bishop is un­ der lock and key. After the continuation 17 . . .:fs 1 8 �g1 1lf7 19 ..th4 ! l'bc4 20 'ifb3 the bishop again does not manage to break free: 20 . . . ..th6 2 1 g5 ! .i.xg5 22 .i.xg5 11f2+ 23 �h1 Wxe2 24 h3 and White is apprecia­ bly superior. 18 Wc2 The queen has nothing left to do on a4. 18 :rs Now the black pieces fall into a clamp. 1 8 . . .:cs is more precise, with counterplay on the c-file. 19 �g1 11c4 (D) Black's bishop still cannot break out to freedom: 19 .....th6 20 h4 'iff7 2 1 l'Dg3, alternatively 1 9 . . . l'Dc4 20 h4. ••• 20 Wd.2! White continues to play to limit the mobility of the bishop. In the event of 20 'ife4 l'Dc6 ! ? the threat is to take on e5, and 2 1 Wxc6 Wxe2 is natural, but not a cause for de­ light. 20 We6 20 . . .11f7 is no good: after 2 1 l'Dg3 everything turns out happily for White - the knight moves to e4, and the queen to e2. Black is not saved from difficulties by 20.....th6 2 1 11xh6 Wxe2 either, in view of 22 We3 'iVxg4 23 dxc5 bxc5 24 Wxc5. 21 h3 l'Dc4 22 1lg5! (D) An important moment. Besides l'Df4, now ..th4 is also threatened when the time comes. ••• B w h6 22 22 . . ...tf6 does not work due to 23 exf6 exf6 24 l'Df4. 124 Be/fort World Cup 1988 23 'ifcl 'ii'f7 In order to gain counterplay B lack should have chosen 23 . . .b5 with the idea, for instance, of 24 tbf4 iif7 25 tbd3 (after 25 -'.g3 White would preserve his advan­ tage) 25 ... b4 !?. On the other hand 23 . . . 'ii'd5 does not achieve Black's aim because of 24 'ii'c2! , and after 23 ... h5 the queen returns to g5. g5 24 �g3 This move was condemned by the commentators, but alas, the recommended 24 ...iid5 25 tbf4 'ii'e4 26 tbe6 would also have led to a difficult position for Black. Here are three key variations: a) 26 ...l:tc8 27 'ii'b l ! 'i!Ve3+ 28 -'.f2 Wxc3 29 'i!Vxg6 'i!Vxal + 30 'it>h2 with inevitable mate. b) 26 ... liJe3 27 'ii'd2 cxd4 28 cxd4 l:tc8 29 l:tel l:tc2 30 l:txe3 'ii'c6 3 1 d5 and Black's position is bad. He could try driving the queen to e2 by playing 27 ... ltJc4 28 'ii'e 1 lLle3 29 'ii'e2 cxd4, but 30 lLlxf8 -'.xf8 3 1 'ii'f3 ! 'i!fd3 32 cxd4 'ii'xd4 33 l:te1 brightens up White's situ­ ation completely: 33 ...liJc2+ 34 -'.t2 ltJxe1 does not work because of the intervening 35 Wb3+. c) 26 ...cxd4 27 lLlxf8 liJe3 28 'ii'd2 dxc3 29 'ii'e 2 -'.xf8 30 'ii'f3 with a big advantage. 'it'd5 25 'it'c2 b5 26 -'.f2 .:tf7 27 liJg3 Forced, as if 27 ... b4, then 28 liJf5 and after 28 . . .l:tf7, White has 29 e6 'ii'xe6 30 l:te1 Wd7 3 1 cxb4. b4 28 l:te1 wrs 29 'ii'g6 29 . .. bxc3 loses immediately: 30 liJf5 Wf8 31 e6 l:txf5 32 gxf5 liJd6 33 dxc5. 30 liJe4 (D) B l:txf2 30 This exchange sacrifice hardly improves the situation, and White only needs elementary accuracy. 31 Wxf2 bxc3 �g8 32 'ii'f5+ 33 'ii'c8+ �h7 iif7+ 34 iixc5 c2 35 �g1 -'.f8 36 liJg3 ..ti>g8 37 liJf5 38 .:tcl 1-0 Black's pieces on the kingside could not break free. This was de­ clared the best game in the next ••. Karpov - M. Gurevich I 25 edition of Jnformator, and as it was played in the French town of Bel­ fort, someone as a joke called the variation with 1 7 iVa4 the 'Belfort Variation' . Game 30 Karpov M.Gurevich USSR Ch (Moscow) /988 Queen's Indian - 1 c4 lbf6 e6 2 lbf3 3 d4 b6 4 g3 J.a6 J.b4+ 5 b3 J.e7 6 J.d2 In the game Karpov-Short, Am­ sterdam 1 988, after 6 . ..J.xd2+ 7 'ifxd2 0-0 8 J.g2 c6 9 0-0 d5 10 'ifb4 lbe4 1 1 l:tc l lbd7 1 2 'ir'a3 J.b7 1 3 cxd5 exd5 1 4 lbc3 f5 1 5 e 3 the dark-squared strategy led White to victory; after 1 5 ...'ii'f6 16 l:tc2 a5 17 :td 1 l:tac8 1 8 lbe2 g5 1 9 lbc 1 g4 20 li:Jh4 m y opponent had got nowhere. 7 J.g2 c6 The gradual preparation of ... d5 has recently supplanted immediate action in the centre by means of 7 ... d5 or 7 ...J.b7 8 li:Jc3 d5. These move-orders were seen a few times in my first match against Kasparov. 8 J.c3 dS 9 li:JeS In game 1 8 of the above-men­ tioned match, at a point when my opponent was not particularly try­ ing for an active game, he played 9 li:Jbd2, and after 9 ... li:Jbd7 10 0-0 0-0 1 1 l:tel c5 1 2 e4 dxe4 1 3 li:Jxe4 J.b7 ! the game was totally equal. 9 li:Jfd7 10 li:Jxd7 li:Jxd7 0-0 1 1 li:Jd2 l:tc8 12 0-0 13 e4 (D) B 13 bS After 1 3 . . . dxe4 14 J.xe4 b5 1 5 1Wc2 h6 1 6 l:tfd l bxc4 1 7 bxc4 White had the initiative in Fta�nik­ Adorjan, Szirak 1986. Obviously, 1 3 ...c5 is also insufficient: 14 exd5 exd5 1 5 dxc5 dxc4 1 6 c6! cxb3 1 7 l:tel b2 1 8 J.xb2 li:Jc5 1 9 J.a3 ! gave White a clear advantage in Gheorghiu-Csema, West Berlin 1 986. bxc4 14 :tel 126 USSR Ch (Moscow) 1988 In the sixth game of my first match against Kasparov, I captured with the other pawn, 14 ... dxc4, and after 15 bxc4 lbb6 ( 1 5 ...bxc4 would have led to a position from the cur­ rent game against Gurevich) 1 6 cxb5 cxb5 17 l:l.c l .i.a3 1 8 l:l.c2 lba4 1 9 .i.a1 l:l.xc2 20 •xc2 •as 2 1 'iWd 1 White gained an apprecia­ ble advantage thanks to his danger­ ous passed d-pawn. 15 bxc4 dxc4 After 1 5 ....i.xc4 16 lbxc4 dxc4 1 7 a4 'iWb6 1 8 .i.fl ..,a6 1 9 a5 c5 20 d5 l:l.fd8 2 1 e5 White had more than enough compensation for the pawn in Dol.matov-Ehlvest, USSR Ch (Minsk) 1987. 16 ••4! ? Until now this encounter had followed game 2 1 of my return match against Kasparov. Then I continued 1 6 'ifc2, and after 1 6... ..,c7 1 7 lbfl e5 ! 1 8 lbe3 exd4 1 9 .i.xd4 .i.c5 chances were even. However, White's resources in this variation have not yet been ex­ hausted. .i.b5 16 .i.a3 17 ..,c2 18 lbb1 (D) White could have maintained a minimal advantage after 1 8 lbxc4 .i.xc4 1 9 ..,a4 c5 20 'iha3 cxd4 2 1 .i.xd4. 18 .i.d6 19 a4 .i.a6 B 20 lbd2 e5?! 20 ... c5 ! is more precise: 21 d5 lbes with a double-edged game. ..,e7 21 l:l.ad1 22 .i.n •e6 .i.xeS 23 dxeS 23 ... lbxe5 24 .i.xe5 .i.xe5 25 .i.xc4 is no better. 24 f4 i.c7 lbb6 (D) 25 e5 After 25 ... .i.b6+ 26 �h1 lbcs the whole game would still lie ahead, but now White gains a seri­ ous advantage. 26 f5! ...h6 lbd5 27 lbe4! 28 .i.d2 .i.b6+ 29 �h1 •h5 30 .i.e2 ...h3 31 lbgS ...h6 32 .i.xc4 Taking on f7 would have led White to his goal more quickly: 32 lbxf7 ! 'ifh3 33 lbd6, etc. 32 ... .i.xc4 Karpov - M.Gurevich 127 43 ... 44 l:xe6 45 .:eel 33 34 35 36 w 'ibc4 •e4 g4 w •hs .i.fl ..h4 ..d8 (D) B 37 .:n .i.b6 38 l:del l:e8 39 l:e2 •c7 40 .:et �7 41 •c4 c5 42 �g2 lbb4 43 e6! White makes the decisive break. fxe6 �h8 ..d7 (D) w 46 aS! .i.xa5 46 ... .i.c7 also loses; 47 .i.xb4 cxb4 48 ..xc7 !. lbd3 4 7 •xc5! lbxel + 48 •xas .:ed8 (D) 49 l:xel w 50 �g3! 51 ..a3 �7 l:d7 128 USSR Ch (Moscow) 1988 'tWbS 52 �f4 'iWc6 53 l:teS l:tb7 54 'ife3 'iWa4 ss l:te6 'ifbS 56 l:te4 'iVc6 57 �eS 1Wd6+ 58 �d4 l:tf8?! 59 �h3 An oversight caused by time trouble in a difficult position. 1-0 60 �cS Game 3 1 Karpov - Yusupov USSR Ch (Moscow) 1988 QGD, Exchange Variation 1 c4 e6 dS 2 lLlc3 3 d4 �e7 lLlf6 4 lLlf3 exdS 5 cxdS c6 6 �gS 7 'ii'c2 Instead 7 e3 (and then 'iVc2 and �d3 in one order or another) is most frequently played, or first 7 'ifc2 and then 8 e3. In any case, at this point the e-pawn is not moved more than one square. g6 7 A year later at Rotterdam, Yusu­ pov chose 7 ... lLla6 against me, but again did not solve his opening problems, and the game ended in my favour. 8 e4 ••• Transferring the game from a closed opening to an open type of game. Incidentally, the move e2-e4 in this situation was given new life by Timman in his match against Short, Belgrade 1 987. 8 lLlxe4 Short's preference was 8 ... dxe4 9 �xf6 �xf6 10 1Wxe4+ c;li1f8 1 1 �c4 �g7 1 2 0-0 l:te8 13 'iff4 �e6 1 4 �xe6 l:txe6 1 5 l:tfe1 l:txel + 1 6 l:txe1 lLld7 17 lbe4 �e7 1 8 h4 'iVb8 19 lLle5 f6 20 lLlg5 ! lLlf8 21 hS gxhS 22 'iff5 'ii'c8 23 lLld7 ! 'i'xd7 24 'ii'xd7 lLlxd7 25 l:txe7+ �g6 26 lLlf3. White had the better end­ game, and eventually came out on top. Taking on e4 with the knight did not seem to have been tried be­ fore, but the novelty did not take me unawares. �xe7 (D) 9 �xe7! Strangely enough, the alterna­ tive 9 ... 'it'xe7 loses straight away: 10 lLlxdS ! cxd5 (if lO ...'i'e6, then 1 1 lLlc7+) 1 1 'iVxc8+ 1i'd8 12 �bS+ �e7 1 3 'i'xb7+ with a wipe-out. 10 lLlxe4 dxe4 1 1 'ifxe4+ �e6 'iWaS+ 12 �c4 1 2 . ..l:te8 is probably more pre­ cise, because then 1 3 lLlgS is im­ possible on account of 1 3 ...'ii'a5+ and 14... 1t'xg5. 13 � ! Yusupov very likely underesti­ mated this retreat by the king. After ••• Karpov - Yusupov 129 w 1 3 lLld2 lLld7 14 0-0-0 .:l.ae8 Black would have overcome his opening problems. 'iWfS 13 lLld7 14 'it'e3 1 4 ...�f6 loses: 1 5 d5 ..ixd5 1 6 ..id3 'irg4 17 We5#. After 1 4...�f8, White has the good continuation 1 5 ..ixe6 'ilfxe6 1 6 'irh6+ �g8 17 g3 lLld7 1 8 �g2. 15 .:let .:l.ae8 (D) ••• 16 dS! ! Thanks to this mighty break in the centre, White gains an impor­ tant square on d4 and freedom on the dark diagonals for his queen, in return for the sacrificed pawn. I spent almost an hour thinking about it, but the time is worth it for a move like that! 16 ... cxdS 17 ..ibS! The key to my idea. Without this attacking move White would find himself at a dead end. If 1 7 lLld4, then 1 7 . . .'iWe5 ! is strong, and 1 8 1t'a3+ leads to an equal position af­ ter 1 8 ...'ird6, while if 1 8 Wxe5 lLlxeS 1 9 ..ib5, then the simple 1 9 . . .lLld7 gives Black an excellent game. Of course, before throwing forward the d-pawn, I had to fore­ see this fine bishop manoeuvre to b5. 17 ... a6 After 17 ...�f8 there is the possibility of 1 8 'ilfc3, followed by lLld4xe6+ and ..ixd7. Black should have thought about returning the pawn by means of ... d5-d4, and transferring his queen to c5. 18 ..Wa3+ �d8 Right through the game the black king does not find a safe ha­ ven, and now 1 8 ...�f6 loses to 19 ..ixd7 ..ixd7 20 'iWc3+. 19 'ii'aS+ But not 1 9 lLld4 because of 1 9... 'ii'f4. JJO USSif Ch (Moscow) 1988 19 �e7 The path to the queenside is fraught with danger, e.g. 1 9 ... �c8 20 l:.c l + �b8 2 1 'ii'c7+ �as 22 li)d4 'ii'f6 23 .txa6 ! l:.b8 24 'it'a5 'it'd8 25 l:.c7 winning. 20 'ii'b4+ �6 If 20 ...�d8, then 2 1 li)d4 'iVf6 22 .txa6 bxa6 looks reasonable. Then after 23 l:.c l ! (23 li)c6+ �c7 24 l:.c 1 l:.a8 25 lLla5+ �d8 26 'ii'b7 �e7 27 tLlc6+ �d6 28 'ir'b4+ �c7 only leads to a repetition of the position) 23 ...l:.eg8 24 l:.c6 there is no defence against the threat of 25 l:.xa6 and 26 l:.a8+: a) 24...'it'e7 25 'iia5+. b) 24 ... 'ii'e5 25 l:.xe6! . 2 1 'ii'd4+ A pure geometric pattern. The white queen is moving like a pen­ dulum. 2 1 .txd7 does not work in view of 21 ...'ir'd3+. 21 �e7 'ii' bS 22 .td3! 22 ...'ir'f6 cuts off the path of the king's retreat after 23 'iVb4+ [edi­ tor's note: for example 23 .. . �d8 24 'ii'xb7 'ifxf3 !? 25 'ir'a8+ (25 gxf3?? .th3+ 26 �g1 l:.xe 1 + 27 .tfl l:.xfl#) 25 ... �e7 26 'ifxe8+ l:.xe8 27 gxf3]. 23 h4! (D) It is almost as if White is con­ structing a study on the theme of domination. 23 �d8 ••• .•• •.. B 24 tLlgS Threatening 25 g4 ..Wh6 26 l:.xe6 and 27 lLlxf7+. 24 l:.hf8 ..Wh6 25 .te2! 26 .tf3 l:.e7 Retreating the king does not help: 26 ...�c7 27 ..Wf4+ and 28 tLlxe6; 26 ... �c8 27 l:.c 1 + �d8 28 .txd5 .txd5 29 ..Wxd5. lLlf6 27 'ii'b4 28 'ii'd6+ 28 'ii'f4! would have been imme­ diately decisive, as it attacks the knight and simultaneously threat­ ens 29 tLlxe6+ and 29 Wfb8+. l:.d7 28 29 'ii'f4 tLlg8 (D) If 29 ... 'ir'g7, then 30 l:.xe6!. �c8 3 0 .tg4! Taking on g4 leads to an elegant mate: 30 ....txg4 3 1 'ifb8#. 31 .txe6 fxe6 32 l:.cl+ �d8 33 tLlxe6+ �e7 ••• Karpov - Malaniuk 131 3 4 5 6 7 34 35 36 37 38 39 40 41 42 43 w 'ii'xf8+ .!Dxf8 .Uh3 h5 h6+ llf3+ .Ue1+ .Uf6+ g4 .Ue8 1-0 'ii'xf8 �xf8 .!De7 �g7 �6 �e6 �d6 �c7 .!Dc6 d4 Game 32 Karpov - Malaniuk USSR Ch (Moscow) 1988 Dutch Defence Malaniuk is one of the outstanding experts in this opening, and there­ fore I prepared particularly thor­ oughly for our game in this event. As a result I managed to come up with an interesting novelty. 1 d4 f5 2 g3 .!Df6 i.g2 c4 .!Df3 0-0 .!Dc3 g6 i.g7 d6 0-0 'ife8 (D) w One of the most topical posi­ tions in the Leningrad Variation of the Dutch Defence. Black is in­ tending ...e5, creating a flexible pawn centre. e8 is the most suitable square for the queen, as from there she is influencing the centre, and might come in useful on the king­ side. 8 b3! The appearance of the bishop on a3 could essentially hinder Black's plans. Other lines are less danger­ ous for him. 8 ... .!Da6 An immediate march by the e­ pawn, 8 ... e5, has not yet been fully investigated, but it is evidently not without danger for Black. 132 USSR Ch (Moscow) 1988 9 .i.a3! White's idea consists of actively positioning his pieces, with his bishop on a3, his queen on d3, and rooks on d 1 and e 1 , and then mov­ ing the pawn to e4. In this game I managed to carry out this plan in ideal fashion. Black is forced to set aside his standard advance ... e5 in­ definitely. 9 c6 10 'iVd3 It is interesting that in this na­ tional championship Malaniuk chose the Dutch Defence seven times ! Two rounds before our game Gavrikov had played the more timid 10 llc 1 against him. After 10 . . .h6 1 1 e3 �e6?! 1 2 We2 'i!fd7 1 3 lbd2 lbc7 14 'ji'd3 llab8 15 �b2 �h8 1 6 d5 cxd5 17 lbxd5 lbcxd5 1 8 cxd5 �f7 Black built a fortress which was not easy to penetrate, and in the end he even won. How­ ever, this time matters did not turn out so successfully for him. 10 ... �d7?! This bishop move has very un­ pleasant consequences. 10 ...llb8 deserves attention, as a series of games has testified. The fate of the whole variation possibly depends on the assessment of this position. 11 llfe1 l:.d8 1 l . . .d5 is more stubborn, al­ though after 1 2 lbe5 White main­ tains the initiative. �h8 12 llad1 13 e4 Having finished mobilizing his forces, White generates activity in the centre of the board. 13 fxe4 14 lbxe4 �fS (D) .•. The extremely unpleasant 2 1 Wa5 was threatened. 21 lbxe6 �xe6 �d7 (D) 22 l:.de1 He should have brought his bishop back to c8. There now fol­ lows a decisive exchange sacrifice, which, incidentally, has been ma­ turing for a long time. �xe7 23 l:.xe7! Karpov - Hjartarson 133 dxe5 29 l:l.xe5 30 'iVxeS rM7 31 d6 j_f5 32 c5 Black is in total zugzwang. h5 32 33 g4 hxg4 34 hxg4 j_d3 If 34 ...j_xg4, then 35 'iVf6+ 'it>e8 36 'ii'xg6+ and 37 'ifxg4 is decisive. 1-0 35 j_d5+! An effective final blow ended the game; taking the bishop is im­ possible in view of 35 . . . cxd5 36 'iVxd5+ 'it>e8 37 'it'e6+ mating. Game 33 Karpov - Hjartarson Tilburg 1988 QGD, Slav Defence w In this tournament there was a rare occurrence, at least in my experi­ ence - I played three straight games in the same variation, and moreover it was not in a classical variation, but a very sharp one, in which the balance of forces on the board is disturbed in the first ten moves. My results were quite fa­ vourable, 211213 . Detailed analysis of these rather complex duels only took place much later. d5 1 d4 c6 2 c4 lt:lf6 3 lt:lf3 /34 Tilburg 1988 dxc4 4 tt'lc3 5 a4 i..f5 6 lDe5 This is the so-called Central Variation (White is preparing f3 and e4). 6 e6 This move has recently com­ pletely replaced two other theoreti­ cal continuations, 6 ... lDa6, and 6... lDbd7. 7 f3 Logical; after 7 e3 or 7 i..g5 it is easier for Black to equalize. 7 i..b4 8 e4 Leading to immense complica­ tions, as Black is forced to sacrifice a piece. As a rule after the cautious 8 tt'lxc4 or 8 i..g 5, positions arise with even chances. i..xe4 8 ... tt'lxe4 9 fxe4 Black has three pawns for the piece, so there is approximate material equality on the board. 'ii'xd4 10 i..d2 10 .. .'ii'h4+ does not work: 1 1 g3 lDxg3 12 hxg3 'ifxh1 13 'ii'g4 with a strong attack for White. 1 1 tt'lxe4 'i!Vxe4+ 1 1 ...i..xd2+ is no good: 12 lDxd2! 'ifxe5+ 1 3 i..e2 b5 14 axb5 cxb5 1 5 l:la5 a6 1 6 tt'lxc4 'ii'c7 17 lDd6+ <1iie7 1 8 0-0 gave White an irresist­ ible attack in Khalifman-Ehlvest, USSR Ch (Moscow) 1 988. ••• ••• 12 'ii'e2 i..xd2+ Here in the same tournament Timman introduced the novelty 1 2 . . .'ifh4+ against me, but the in­ tervening check did not achieve its aim: 1 3 g3 i..xd2+ 1 4 'itxd2 'fle7 1 5 'ii'e 3 ! with a patent advantage. 13 'itxd2 t!fd5+ In the ending after 1 3 ... 'flxe2+ 14 i..xe2 the black pawns do not compensate for the missing piece. It is another matter if the queen ex­ change takes place on d5, because then Black strengthens his pawn chain. lDa6 14 'itc2 15 tt'lxc4 (D) • • ••• ,� . � � �'�' � � �. , . , . . . . .. . . �·ttJ· • • • • • • ft ��u Rc;t>� �-� u � � a • •i.• : B A peculiar, though standard pat­ tern from the Slav. Black has quite a wide choice here, but recently castling one side or the other has been seen most frequently. Theory is sceptical about 1 5 ...l:r.d8, 1 5 ...b5 and 1 5 ...lDb4+; 1 5 ...'iff5+ deprives Karpov - Hjartarson 135 Black of the possibility of swap­ ping queens on d5, while exchang­ ing on other squares leads to an unpleasant endgame. 15 0-0-0 With castling long, as opposed to short, the rook immediately lands on d8. 16 'ii'e5 White is aiming at the g7 -square and is prepared to swap queens, just not on d5. 16 f6 Defending the pawn and forcing the queen to clarify her intentions. Neither 1 6 ... lbb4+ nor 1 6 . ..:he8 has vindicated itself in practice. 17 'ii'eJ! c5 Black has also tried 17 ...'ii'f5+, 1 7 . . . lbc5 and 1 7 .. .'�b8, but with­ out any particular success. 18 �b3! In my game against HUhner at Tilburg I preferred 1 8 .te2, and af­ ter the continuation 1 8 ...lbb4+ 1 9 �b3 lbc6 2 0 �c3 lbd4 21 .tf3 lbxf3 22 gxf3 'ii'd4+ 23 'ii'xd4 :xd4 24 b4 :ds 25 bxc5 :xc5 a level endgame arose. This unex­ pected raid by the king looks rather strange, but it saves White an im­ portant tempo by immediately freeing the c-file for the rook. 18 lbb4 lbc6 19 :et lbd4 20 �a3 21 lba5! •.• More accurate than 21 :g1 e5 22 .td3 b5 ! 23 axb5 lbxb5+ 24 �a4, when the game is roughly even. 21 eS (D) 2 l . ..�b8 is no good because of 22 1i'g3+. ••• •.• w b6 22 'ii'c3! 23 lbbJ 'W'xb3+ 23 . ..'W'e6 is more obstinate: 24 .tc4 'ii'e7 25 lbxd4 :xd4 26 :hd1 :hd8, and Black holds on. lbxb3 24 'W'xb3 25 �b3 :d4 26 h4 Black has material equality and a flexible pawn chain, but I quickly managed to disturb its harmony. 26 :hd8 �c7 27 .tc4 28 h5 :g4 (D) Instead of pursuing the g-pawn, Black ought to generate some ac­ tivity on the queenside: 28 ... a6! 29 .txa6 (Black was threatening ... b5) J 36 Seanle Ct (3) 1989 29 . . ..:tb4+ 30 �c3 (in the event of 30 �a3 White could even lose 30 . . . .:ta8 !) 30 ....:txa4 with chances for a draw. 32 33 34 35 36 37 38 39 40 41 42 43 .Uhxf6 li6f4 .Uf7+ lixg7 lif4 �c2 �c3 �d2 .Uf7+ �e3 �xa6 e4 h5 lid4 .Ud7 lixg7 lig3+ lig2+ .Ug3+ lig4 �d6 a6 1-0 Game 34 Hjartarson - Karpov Seattle Ct (3) 1989 Spanish, Zaitsev 29 h6! By stealing along the edge of the board, this pawn undermines the black kingside. 29 lbg2 Attempting to preserve a dy­ namic pawn chain by means of 29 . . . g6 is unsuccessful on account of the weakness on h7: 30 .:thd 1 ! .Ugd4 (30 ....:txd 1 3 1 .Uxd 1 .:th4 32 �b5 .Uxh6 33 l:td7+ �b8 34 �a6 with decisive threats) 3 1 lixd4 lixd4 32 �g8 ! lid? 33 �c2 and Black will be put in zugzwang. l:txg7 30 hxg7 .Ud6 31 lien 32 lih6 Black's fortress has been de­ stroyed, and this sums up the result of the opening. ••• e5 1 e4 llJc6 2 lLlf3 a6 3 �b5 lLlf6 4 �a4 �e7 5 0-0 b5 6 :et d6 7 �b3 0-0 8 c3 lieS 9 h3 �b7 10 d4 �f8 1 1 lLlbd2 12 a3 As well as 1 2 a4, this restrained pawn advance is also seen frequently. White guards b4 from invasion by the black knight, and then prepares to develop the initiative on the queenside. h6 12 ... Hjartarson - Karpov 137 liJb8 13 ..tc2 14 b4 liJbd7 15 ..tb2 aS! ? (D) A move I had prepared specifi­ cally for this match; 1 5 ...g6 is a more widely used continuation. ... . jilli ... • - ... d · � ... �· .t. � · � ­ � - � . - � - � ­ . � . · � � ·d - -� • �-/� � � fj · d � ·� � 1":\ . fj u . .t.z.J. �- � � --'<% � -�� 0A ­ � ... /� R •• �. m � . � � '. . � � : ·· · � 7• • 1'1 � w c6 16 ..td3 17 liJb3 axb4 18 cxb4 1 8 axb4 liJb6 1 9 liJa5 "ilc7 would have led to equality. exd4 18 19 liJfxd4 (D) Now Black seizes the initiative. Things are also fine for him after 1 9 ..txd4 c5 20 ..txf6 liJxf6 2 1 ..txb5 l:be4. The correct continu­ ation is 1 9 liJbxd4 c5 20 bxc5 dxc5 2 1 liJxb5 ltJxe4 22 liJe5 liJxe5 23 ..txe4 with a roughly even game. c5! 19 A very promising pawn sacri­ fice, with which Black takes aim at the enemy centre. ••. ••• B 20 bxc5 20 lbxb5 immediately is much more precise: 20 ...cxb4 21 axb4 l:lxal 22 ..txal d5 23 exd5 (23 e5 ..txb4 leads to an approximately equal position) 23 ...l:lxel+ 24 1Wxel liJxd5 25 'lie4 liJ7f6 26 ..txf6 liJxf6 27 "ilxb7 "ilxd3 with a sharp game, in which Black has enough compensation for the pawn. 20 dxc5 21 liJxb5 ltJxe4! 22 'lic2?! This is a serious error. White had to play 22 "ilf3, when the following variations are possible: a) 22 ...liJd6? fails to 23 l:lxe8 'tWxe8 24 liJxd6 ..txd6 25 "ilxb7 l:lb8 26 '1id5. b) 22 ...'lib6 23 ..txe4 ..txe4 24 l:lxe4 'lixb5 25 a4 is unclear. c) 22...l:lb8 should be answered by 23 ..tc4, with a complicated po­ sition. Instead after 23 ..txe4?! ..txe4 24 l:lxe4 l:lxe4 25 'lixe4 138 Slcelleftea World Cup 1989 .:txb5 26 'ili'd5 'ilt'b8 27 'i't'xd7 .:txb3 28 �cl .:tb1 29 .:.Xbl 'i't'xb1 30 'ilt'd2 c4 3 1 �h2 1t'b3 Black is slightly better. d) 22 ... �df6 !? 23 �xf6 �xf6 24 .:txe8 �xf3 25 .:txd8 .:txd8 re­ sults in an equal position. 22 �df6 23 �c3 (D) White must certainly avoid the continuation 23 .:tad 1 'ilt'b6 24 tDc3 �xf2! . ••• Black threatens the deadly ...tZ::lf3+ or ...tZ::lxh3+. 27 .:te3 �d6 28 h4 tZ::le6 29 li)d1? White would fare no better with 29 .:txe6? fxe6 30 'it'g6 'it'e8 . His best chance was 29 li)b5 !? �f4 30 .:th3 li)e4 when Black would still have some obstacles to overcome. After his actual choice I was able to launch a winning attack. tZ::lg4 29 30 .:txe6 �h2+! 31 �h1 1t'xe6 32 f3 ile1 ! ! 0-1 White's position collapses: 33 fxg4 'it'xh4 wins on the spot, while 33 'ii'c4 gives Black a choice be­ tween 33 ...�f4 and 33 ....:ta4. Game 35 Karpov P. Nikolic Skelleftea World Cup 1989 Bogo-Indian Defence - B �g5! 23 24 �bS More resistance was possible by 24 tZ::ld 2, although after 24 ...'ilt'b6 Black threatens 25 ...'ilt'c6. llxe1+ 24 ikc7 25 .:txe1 26 �n Now, however, 26 tZ::ld2 can be met by 26 ...1t'f4!. Still, 26 .:te3 was worth considering. 26 'ii'c6! ••• ••• tZ::lf6 1 d4 2 c4 e6 3 tZ::lr3 �b4+ 4 �d2 A more popular continuation than 4 li)bd2. ile7 4 Black has several alternatives, including exchanging bishops on d2. ••• Karpov 0-0 s g3 i.xd2+ 6 .tg2 7 Wxd2 d6 e5 8 ltlc3 l:te8 9 0-0 .tg4 10 e4 .txf3 11 dS 12 .txf3 cS 13 l:tael a6 ltlbd7 14 b3 15 .tg2 l:.ab8 16 a4 (D) By freezing play on the queen­ side, I am decisively moving the battle's centre of gravity to the kingside. Now if 1 6 ... a5, then 1 7 ltlb5 b6 1 8 ltla7 and ltlc6, followed by f4. 'ird8 16 Preventing further movement by the white pawn. Otherwise after a5 and ltla4, I would have advanced the pawn to b4 and put pressure on the weak b7-pawn. - P. Nikolic 139 17 'ird1 First and foremost in order to control the h5- and g4-squares. After the immediate 17 f4 Black could have chosen the exchange 1 7 . ..exf4 in order to gain counter­ play by means of ...ltlg4 (or ..ltlhs) and ...'ifh4. 17 l:te7 Was 18 l:te3 19 l:tfe1 White has reinforced the e4pawn, and it remains for him to transfer the bishop to h3 and ex­ change it for the d7-knight, which would give him a significant spa­ tial advantage. 19 ... �h8 l:tg8 20 .th3 Nikolic prevents the advance 21 f4, after which the game would now continue 2 l ...exf4 22 gxf4 g5. In that case White would not have enough time to rush the pawn through to e5, and Black would thus gain a comfortable post for his knight on that square. Wd8 21 �h1 ltlhS 22 'ird2 g6 23 ltle2 ltlg7 24 aS White needs time to regroup his forces on the queenside, but Black has apparently prepared for activ­ ity on the kingside. However... 25 .txd7 Wxd7 (D) l:tge8 26 f4! . 140 Skelleftea World Cup 1989 . - . - · -� -- . . .... . . . . •• • ••• �'� �8� • d > !'! ·8·8· • � � -� . A . Ml • pm o.z.J �U � � B B � do• " . !'! 31 lbxe4 ikf5 (D) 3 l . . .'iWg4 32 c5 .:txe4 33 .:txe4 'ii'f3+ 34 �gl 'iWg4+ 35 �f2 'ii'h4+ 36 <ite2, etc., is worse, but 3 l ...lbh5 ! is interesting, and in the event of 32 c5 c;l;>gS 33 cxd6 .:txe4 34 .:txe4 lLlg3+! 35 hxg3 'iWh3+ the compli­ cations end with a perpetual check. • • a •� w 27 b4! cxb4 In the event of 27 ...exf4 28 lbxf4 the essential superiority of the white knight over his opposite number becomes clear: by landing on d3 it supports the attack in the centre. f5 28 'ii'xb4 Black is prepared to exchange the e- and f-pawns, then move the knight over to f5, achieving a posi­ tion of dynamic equality. There­ fore I have to play very carefully, but also quite decisively. 29 lbc3 The immediate 29 c5 ! deserved attention: 29 ... dxc5 30 'ii'xc5 exf4 3 1 lbxf4 fxe4 32 l:Xe4 l:be4 33 .:txe4 .:tc8 (33 ....:txe4? 34 'ii'f8#) 34 .:te7 'ii'g4 35 'ii'f2 , and White's chances are obviously better. De­ laying the break c4-c5 for three moves complicates the situation. 29 exf4 fxe4 30 gxf4 w It looks as though everything is in order for Black now. 32 lbxd6 does not work: 32 .. .l:be3 33 l:lxe3 ifxf4 threatens mate on fl , which would cost White his knight if he plays 34 .:txe8+ lLlxe8. After 32 lLlg3 Black should again continue 32 ....:txe3 33 .:txe3 'ili'xf4 34 .:txe8+ lLlxe8 and White has nothing better than a draw: 35 'iWxb7 'tie l + 36 c;li>g2 1Wxc4. But not all White's resources have been exhausted... 32 c5! The queen joins in to defend the knight, and White makes a passed Karpov - P. Nikolic 141 pawn. Events on the board take on quite a stormy character. dxcS 32 33 'it'c4 'ifxf4 :Z.eS? (D) 34 d6 Black is still under the impres­ sion that his doubled rooks are strong. 34 ... :Z.d7 would have been more forceful, preventing the pawn from advancing. After 35 'ii'c3 'ii'e5 Black holds on, while 35 'ii'xc5 is not dangerous for him either: 35 . . .'ii'e5 36 'ii'xe5 :Z.xe5 37 :Z.c l lbe6 38 :Z.c8+ �g7 39 :Z.e8 �f7 40 liJg5+ [editor's note: White may try 40 :Z.xe6 �xe6 4 1 liJc5+ �xd6 42 :Z.xe5 :Z.c7 43 liJd3] 40 ...:Z.xg5 4 1 :Z.3xe6 :Z.d5 42 :Z.8e7+ :Z.xe7 43 :Z.xe7+ �f6 44 :Z.xh7 :Z.xd6 45 :Z.xb7 with a draw. Allowing White to create a small study. After 36 ...h6 the posi­ tion is still unclear, for example 37 :Z.f3 'ii'g4 38 :Z.f7 ...e6 39 :Z.efl liJfS ! . 'it'h4 37 :Z.f3 38 :Z.f7 Not, of course, 38 •xe5? due to 38 ......xel + 39 �g2 ,..e2+. 38 Ilg8 39 'it'cl 39 :Z.xg7 :Z.xg7 40 d81i'+ •xd8 41 'ii'xe5 is also good, and after 4 l . ..'�i'e7 42 liJxc5 •xe5 43 :Z.xe5 White achieves a technically win­ ning endgame. But I managed to find a more elegant line. 39 gS 39 . ..liJf5 does not work because of 40 :Z.xf5, therefore Black is forced to advance yet another pawn from his king's shelter. :Z.dS 40 liJg3 41 :Z.f6 :Z.d8 (D) The game ends with immediate mate after 4 1 . .. �h7 42 ...bl+. 42 'iib 1 ! An elegant finale. �g8 42 It looks as if Black has extri­ cated himself, and with his next move he will seize the daring run­ away on d7, but, as it perishes, the pawn has the last word. 1-0 43 :Z.e8+! Let us see the end of the study: 43 ...:Z.xe8 44 dxe8...+ liJxe8 45 ••• ••• •.• w 35 d7! :Z.d8 If 35 ...:Z.f8 36 'ii'c3 ! :Z.d5, then 37 liJxc5 is decisive. hS?! 36 'iVcJ! 142 London Ct (8) 1989 Game 36 Karpov - Yusupov London Ct (8) 1989 QGD, Lasker This decisive duel was very strik­ ing. li)f6 1 d4 2 c4 e6 d5 3 lf)fJ i.e7 4 li)c3 0-0 5 i.g5 h6 6 e3 li)e4 7 i.h4 Not waiting for the white bishop to take on f6. This move constitutes the Lasker Defence, where Black in many respects has no preten­ sions, but relies on creating an im­ penetrable fortress. In games 4 and 6 Yusupov had already turned to this defence, and in fact I had not managed to glean the least advan­ tage out of the opening. 8 i.xe7 'iVxe7 9 llcl In game 4 I had continued 9 it'c2, and in game 6, 9 cxd5, and in both cases Black quickly equal­ ized. This time I was better pre­ pared for the game. 9 c6 10 i.d3 li)xc3 dxc4 1 1 llxc3 12 i.xc4 Recapturing on c4 with the rook also deserves attention. 12 li)d7 eS 13 0-0 Neither 1 3 ... b6 nor l 3 ...lld8 is very promising for Black. It must be said that from the Lasker De­ fence, the game has transposed to lines of the less weighty Capa­ blanca System, and moreover the position of the pawn on h6 (instead of h7) is favourable for White, as it gives him an additional object for attack. 14 i.b3 (D) This waiting move with the bishop was prepared specially for this game. Here White would nor­ mally play 14 h3, 14 jfc2, 14 'iib l , or 1 4 dxe5 li)xe5 1 5 li)xe5 it'xe5 1 6 f4. exd4 14 lDf6 15 exd4 'iVd6 16 lle1 .•• Karpov - Yusupov 143 • •.t.• -·· B mu · m - · ··••• • • . . � . - m . � . . R� ­ � � U"'l.J. d£• 8� • . • B 8_rJ .\lllr .. � � - �� . -� B lDdS 17 ltleS! 18 l:.g3 White does not hide his inten­ tions. .i.fS 18 This move is a novelty. Theory gives the continuation 1 8 ....i.e6 19 'iid2 �h8 20 l:.e4 and 21 l:.h4, which is somewhat preferable for White. 19 'tWhS .i.h7 The paradoxical move 19 ......e6!?, suggested by Taimanov, is interest­ ing. 20 'tWg4! gS (D) After 20 . . . g6 there is the very strong reply 21 h4. 20 . ..'it'f6 does not work because to 21 ltld7. After 20 ....i.g6, the following variation is possible: 2 1 ltlxg6 'tWxg6 22 'iixg6 fxg6 23 l:.xg6 l:.ae8 24 l:.xe8 l:.xe8 25 �1 l:.e4 26 l:.d6 l:.xd4 27 l:.d8+ with a big advantage to White. (6 21 h4 ••• w 2 1 ... f5 is no good because of 22 '6'f4 ! (or 22 'ii' h5 g4 23 ltlxg4); in the event of 21 ... ltlf6 White has the strong 22 'itf3, or 22 '6'f4! ltlh5 23 'ii'f3 ltlxg3 24 ltlxf7 'iixd4 25 l:.e7, etc. 22 hxgS! 22 Wh5 is insufficient: 22... fxe5 23 hxg5 .i.g6 24 gxh6 �h7 and the black king is safely covered. hxgS 22 In the event of 22 ... fxg5 the re­ ply 23 f4 simply becomes more valid. 23 f4 After the continuation 23 l:.h3 fxe5 24 l:.xh7, White would have won beautifully in the event of 24...�xh7: 25 .i.c2+ �g7 (25 ...�h6 26 'iWh3+ �g7 27 'iWh7+ �f6 28 dxe5+ Wxe5 29 'iWh6+) 26 'fi'xg5+ cltf7 27 dxe5 We6 (27 ...'6'b4 28 e6+ �e8 29 .i.g6+; 27 .. .'ile7 28 'iWg6#; 27 ....:tae8 28 'iWh5+) 28 .i.f5 . However, Black can hold on ••• · 144 London Ct (8) 1989 thanks to 24 ... 'iff6. Besides mobi­ lizing the f-pawn, White would have gained a dangerous initiative after either 23 'iih5 or 23 ltlf3. 23 .:tae8 (D) The critical moment. This rook manoeuvre forces Black's demise. He should have sheltered the king in the corner with 23 ... �h8, al­ though in this case as well I could have reached a winning ending with four rooks with 24 fxg5 fxe5 25 g6 'iixg6 26 'ii'xg6 i.xg6 27 .:txg6 exd4 28 .:te4 .:tf7 29 i.xd5 cxd5 30 .:th4+ .:th7 3 1 l:xd4 l:d8 32 .:tg5 .:thd7 33 a3 a6 34 a4 b5 35 axb5 axb5 36 b4 - Black is in zug­ zwang. ..• queen sacrifice - 25 gxf6+! ! i.xg4 26 .:Xg4+, and then: a) 26 ... �h8 27 ltlf7+ .:txf7 (if 27 ...�h7, 28 .:tg7#) 28 .:Xe8+ .:tf8 (28 ... �h7 29 .tc2+ �h6 30 .:th8+ and 3 1 .:txh7#) 29 f7 ! ltlf6 (or 29 . . .�h7 30 i.c2+ �h6 3 1 .:tg6+ 'iixg6 32 i.xg6, etc.) 30 .:txf8+ 'ili'xf8 3 1 .:tg8+ ltlxg8 32 fxg8'ii+ 'ifxg8 33 i.xg8 �xg8 34 �f2 and it's all over. b) 26 . .. �h7 27 i.c2+ �h6 (al­ ternatively 27 ...�h8 28 .:th4+ �g8 29 .th7+ �h8 30 ltlg6#) 28 .:tg6+ �h7 (28 . . . �h5 29 i.d 1 + �h4 30 l:e4+ mates) 29 f7 and White wins. i.xg6 25 g6! 26 dxeS! More accurate than 26 .:txe5 �h7 27 'iWh5+ �g7 (27 ... i.xh5 28 .:txh5+ 'ii'h6 29 i.c2+ �h8 30 l:xh6#) 28 .:txg6+ 'ifxg6 29 l:g5 .:te 1 + 30 �h2 .:te6 with an unclear game. 'ili'e6 26 Things are no better for Black after the continuation 26 ...'iic5+ 27 �h2! (but not 27 �h1 .:tf6 !), or 26 . .. �g7 27 exd6 .:txe 1 + 28 �h2 .:th8+ 29 .:th3. cxdS 27 i.xdS 28 'iWxg6+ 'iixg6 29 .:txg6+ �h7 (D) If 29 ...�t7, then 30 .:td6 .:td8 3 1 .:r.n + �e7 32 .:txf8 .:txf8 3 3 .:txd5 winning. .:cs 30 .:td6 ••• Karpov - Timman 145 . . . - ... � . • • • • •• . . •:• . · · �'� . • • • • • • • • 80 • .8• � . �� � . . . � w After 30....:d8 White has the decisive 3 1 .:d 1 . l:.c2 31 .:e3 �g6 32 .:d7+ 33 .:xb7 .:e8 The rook ending is hopeless for Black, but resigning the game was tantamount to resigning the match too, and so Yusupov played on, hoping for a miracle. If 33 ...1:.ff2, then 34 .:g3+ �h6 35 .:bs ..th7 36 .:b4 is decisive. d4 34 a3 .:xe5 35 .:d3 .:g5 (D) 36 .:xd4 If 36 ....:ee2, then 37 .:g4+ �f5 38 .:gg7. 37 .:d6+ �h5 �g4 38 .:h7+ 39 .:d4+ �5 �g6 40 .:d5+ 41 .:g7+ �xg7 �f6 42 .:Xg5+ 43 .:b5 a6 44 .:b6+ �e7 Game 37 Karpov - Timman Kuala Lumpur Ct (4) 1990 Griinfeld Defence 1 2 3 4 s 6 7 8 9 10 d4 c4 lLlf3 g3 j.gl cxd5 lLlc3 lLleS 0-0 f4 lLlf6 g6 j.g7 c6 d5 cxd5 0-0 e6 lLlfd7 lLlc6 146 Kuala Lumpur Ct (4) 1990 1 1 .i..e3 f6 ltlb6 12 ltld3 'ile7 13 b3 14 a4 .i..d7 .:tfd8 15 .i..c t 16 e3 .i..e8 'ilf7 17 .i..a 3 18 :et .i..rs 19 .ixrs Wxrs (D) As a result of complex manoeu­ vres, an advantage has crystallized for White on the queenside. I . _ ... . . . -·· . •• �· �·� � �·• ·· m •• • ••• � • A � � " �� � � Q �- U � � r.... � rt� � �� ��"l.JU �� � i. " � � m m � - ' �\lllr � � � : � � �- � � " V "� V w 'ile7 20 g4 .:tac8 21 'ild2 l:tc7 22 ltle2 ltlc8 23 l:tc5 24 f5 g5 White is very slightly better af­ ter the continuation 24.. . ltld6 25 lLldf4 gxf5 26 gxf5 ltlxf5 27 ltlxd5 exd5 28 l:txf5. 25 ltlg3 25 fxe6 'i!he6 26 ltlc3 ltl8e7 27 e4 is not so clear. After the text move the threat of the knight appearing on f5 is more dangerous for Black: 25 ... .if7 26 fxe6 'ilxe6 27 lLlfS. To defend against this, Timman sacrifices a pawn. 25 e5 (D) ••. ··- -*-•• • �. • ·-�· - . .• � V V � � . -�� M·�� r... . � � � •�- r/� • � • z.w�"l.JU � •�. wa. �- � � � i. � � -� - • -:= � w 26 Wet Obviously, the continuation 26 .i..xd5+ �h8 27 l:tfcl exd4 28 e4 ! is stronger, returning the pawn; the inevitable break e4-e5 would have guaranteed White a serious advantage. Killing two birds with one stone (attempting to control the c- and f-files simultaneously) does not work, and Black, by liqui­ dating the danger in the centre, achieves a playable game, which for a long time will be typified by manoeuvring. However, thanks to such tardy action the game quickly moved into an ending filled with fascinating events. Karpov b6 26 e4 27 l:.c2 liJd6 28 liJf2 l:.dc8 29 'ii'd2 aS 30 l:Ucl 31 �n If it did not deprive the white queen of mobility, this move would have been very perilous for Black. Attempts to rebuild with 3 1 'ii'c3 are repulsed by means of 3 1 . .. �d7 and 32 ...l2Je8. 31 li)b4 (D) 3 l . . .li)xd4 !? suggested itself, but after 32 exd4 e3 33 'ii'd3 exf2+ 34 �xf2 l:.xc2+ 35 l:.xc2 l2Je4+ 36 �f3 a draw is not far away (swap­ ping the good c6-knight for the bad one on f2 is not favourable for Black). Timman's position in the match obliged him to maintain the pressure on the board in the hope of seizing the initiative from me. ••. 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 so - Timman 147 liJdl l:.xc6 l:.xc6 l2Jc3 �f2 �el �dl �el l2Ja2 'ifxa2 �f2 'W'b2 �e2 �el �d2 'ii'c l �el � l:.c6 l:.xc6 'ifxc6 �f8 �e7 �f8 'ii'c8 �g7 l2Jxa2 'ii'c7 �f8 �e7 �d8 �c8 �b7 'W'e7 �d7 l2Je8 (D) • • •• • ···i.- • . � . . � - . � · � -·· -88. Ri·8· �� u .8. � d d � u � • d •.li.m d � � u . . � - . . . . w 51 'W'hl ! I did not want to agree to a draw, so I decided to bring about some confusing complications, by leav­ ing the queenside to its fate, while actively occupying the kingside. 148 Kuala Lumpur Ct (4) 1990 Calculating variations to the very end was obviously out of the ques­ tion. 'Wib4 51 'ii'xb3 52 h4! 53 hxg5 fxg5 54 'Wxh7 'ii'xa4 55 'Wie7 'ii'c6 56 'ii'xg5 a4 57 'ii'e7 'ii'd6 lbxd6 58 'Wixd6 i.b5?! 59 i.dl It was barely worth obstructing the path of the pawn. After 59 ...b5 the game would most likely have ended in a draw. a3 60 lbe2 61 lbcl (D) • • • • ••• • • - - . . � �� ' •-*-• · · � ­ . - " u·· -�· � . " . - � ;Q/ � . �� . � � . � . � • � . � " � i. • B Here Timman sealed a move he had been thinking about for almost half an hour. Black's activity on the queenside puzzled many commen­ tators, and the following day there were even headlines in the local papers such as 'Will Karpov Hold Out?' . In fact, my preliminary analysis showed that only White has chances to win, and his pawns are more dangerous than his oppo­ nent's. 61 <j;c7 A good sealed move, but, of course, we had also analysed oth­ ers. In the end it was clear that 6 1 . .. i.d7, 6 l . ..i.e8 and 6 l .. .'.tc6 would have left Black with less chances to save himself. lbc4 62 'iti>g3 62...<iti>d7? loses: 63 <ifi>f4 i.c4 64 g5 a2 65 lbxa2 i.xa2 66 g6 ri;e7 (66 ...lbe8 67 i.a4+) 67 g7 <j;f7 68 f6 i.c4 69 i.g4 <itg8 70 i.e6+ lbf7 7 1 <j;f5 i.b5 72 ..ti>g6 i.e8 73 i.xd5 b5 74 i.a2 b4 75 i.b3 . 63 i.e2 But not 63 <j;f4? lbb2 ! with the threat of ...lbd3+. i.e8! 63 This is the strongest defence. 63 ...lbxe3 is insufficient: 64 i.xb5 lbc2 65 g5 lbxd4 66 f6 <j;d6 67 g6 <j;e6 68 g7 ..ti>f7 69 i.e8+ <j;g8 70 ..ti>f4 winning. 64 ..ti>f4 (D) Here is a striking variation after 64 g5, ending in a problem-like mate: 64 ...lbxe3 65 <itf4 lbc2 66 <j;e5 lbb4 67 <ifi>f6 <iti>d6 68 g6 lbc6 69 i.b5 e3 70 g7 l1Je7 7 1 i.xe8 ! e2 72 lbxe2 a2 73 lbc3 ! lbg8+ 74 <j;f7 lbh6+ 75 <ifi>f8 a1 'iW 76 lbb5#. ..• ••• Karpov It would not have been too bad ending like this, but unfortunately, Black has a defence: 64 ...�d6 65 .i.xc4 dxc4 66 ltJa2 .i.a4 67 ltJc3 .i.b3 68 tiJb5+ �e7 69 ltJxa3 c3 70 �f4 b5 ! 7 1 �xe4 b4 72 �d3 bxa3 73 �xc3 .i.d5 drawing. tiJb2 64 ••• - Timman 149 67 ltJal b5 b4 (D) 68 .i.dl Bringing the king slightly closer to the pawns does not work: 68 .. .<iPd6? 69 .i.b3 tiJc l 70 cJ;h6 b4 71 f6 �e6 72 Wg7 liJxb3 73 liJxb3 .i.a4 74 ltJc5+! cJ;d6 75 f7 al 'iW 76 f8'iW+ cJ;c6 77 'il'c8+ cJ;d6 78 'il'd8+ Wc6 79 'i!i'd7+ 'it>b6 80 'ii'b7+ cJ;a5 8 1 'iWa6#. . .... . . � · � � . �� � • • • • � � · �·��� /'·� � � � � � � ... ��· � � � • • •o • .... . . . � .JL. • " -� B 65 �g5 After 65 ltJa2, the long variation 65 . . . �d6 66 �g5 �e7 67 �h6 �f6 68 tiJb4 .i.f7 69 .i.fl ! tiJd I 70 g5+ �e7 7 1 �g7 ltJxe3 72 f6+ �e6 73 .i.h3+ tiJf5+ 74 .i.xf5+ �xf5 75 rJifxf7 e3 76 'iii>g7 ! e2 77 ltJc2 a2 78 f7 e l 'ii' 79 liJxel al 'i!l 80 f8._,+ �xg5 8 1 'iWf2 promises White good prospects, but after the correct 65 ... b5 ! nothing appropri­ ate can be found. tiJd3 65 It is too late for 65 ...ltJc4 66 �f6 ltJxe3 67 cJ;e7. 66 tiJb3!? a2 ••• w tiJcl 69 .i.b3 After 69 ...'it>d6 70 cJ;f6 Black's king is kept away and (he white g­ pawn is irrepressible. 'iii>d6 70 .i.xd5 .i.b5 71 .i.c4 There is also the interesting pos­ sibility of 7 l ...b3 72 .i.xb3 liJxb3 73 liJxb3 'it>d5 74 tiJal 'it>c4 75 cJ;f6 'it>c3 76 cJ;e7 cJ;b2 77 'it>xe8 'it>xal 78 f6 cJ;b2 79 f7 al 'iW 80 f8'iW 'ifa4+ 8 1 'it>f7 'i!i'b3+. This position arose in our analysis, when we considered that 82 'it>f6 Wxe3 83 150 Brussels Ct (4) 1991 �e5 �c2 84 'iff4 �d3 85 d5 would have given chances for victory, but in our post-mortem, Timman demonstrated the more striking continuation 82 d5 ! 'ii'xe3 83 'ifb4+ �c2 84 g5 'iff3+ 85 �g7 e3 86 'ifc4+ �b2 87 'ifb5+ �c 1 88 d6 e2 89 'ifc5+ �b2 90 'ffb6+ �c 1 9 1 d7 el'if 92 d8'ff 'ffec3+ 93 'ifdf6, etc. �e7 72 i.g8 73 �h6 (D) In the event of 73 f6+ �f8 74 .i.e6 b3 (74 ... .i.a4 leads to the same position as in the game, but 74 . . . .i.d7 ! will save Black, and if 75 i.xd7? then 75 ...b3 76 �f5 b2 77 g5 lbe2! 78 g6 lbg3+ 79 �g5 lbh5 ! 80 �xh5 bxal'ff 8 1 g7+ � 82 .i.e8+ �xf6 and Black unex­ pectedly takes the initiative) 75 .i.xb3 lbxb3 76 lbxb3 i.c4 77 lba1 �f7 78 �f5 White has the upper hand. 73 Only now does Timman make the decisive error. 73 ...b3! 74 .i.xb3 lbxb3 75 lbxb3 .i.c4 76 lbal �f6 ! would have led to a draw. 74 .i.e6 i.d7 75 g5 b3 76 g6 1-0 If 76 ...i.xe6, then 77 fxe6 b2 78 g7+ �g8 (78 ...�e7 79 g811f bxal'ff 80 'it'f7+ �d6 8 1 'ifd7#) 79 e7 � 80 g8'it'+ �xe7 8 1 'it'g5+ �f7 82 'ifd5+ and after two more checks White takes the b2-pawn. 000 Game 38 Karpov - Anand Brussels Ct (4) 1991 QGD, Semi-Slav 1 d4 d5 c6 2 c4 3 lbf3 lbf6 4 lbc3 e6 5 e3 lbbd7 6 'ifc2 After 6 .i.d3 dxc4 7 i.xc4 b5 8 i.d3 the classical variation of the Meran arises. Moving the queen forward to c2 is a standard Anti­ Meran possibility. 6 .i.d6 7 .i.e2 Besides this move, White has a choice between 7 i.d3, 7 g4 and 7 b3. 0-0 7 000 ••• Karpov - Anand 151 8 0-0 Black's counterplay is linked with the advance . . .c5 or . . .e5 (or both), usually after a preliminary exchange on c4. All my games as White in this match against Anand developed in this fashion. 8 9 .ixc4 dxc4 'fle7 (D) .•• In the second game of this match after 9 ... a6 10 l:td1 'fle7 1 1 h3 b5 I did not place my bishop too suc­ cessfully: 12 .id3 c5 1 3 lDe4 c4 14 lDxd6 'flxd6 15 .ie2 .ib7 1 6 .id2 l:tfc8, and Black solved his open­ ing problems. 'ifc5+ 16 �h2 .id7 ! was being played, and Black had a favourable game) 1 3 . . . lDe5 14 .ib3 .id7 15 .ie3 lDg6 and everything was in order for Black. My dispute with Anand contin­ ued to the 8th, and decisive, game of the match. On that occasion I changed my move order and played 10 a3. Then in the event of 10 ...c5 1 1 dxc5 .ixc5 Black has to bear in mind 1 2 b4 .id6 1 3 lDb5, but Anand instead advanced his e­ pawn - 1 0...e5. Thus, in this opening variation White can play a3 or h3, while Black can play ...c5, or ...e5, and any of these moves could essentially in­ fluence the future course of the bat­ tle. Modem opening theory is built on precisely such nuances. I am convinced that assessments here are still to be redefined more than once. 1 1 e4! 12 l:tdl ! After 12 ... exd4 13 lDxd4 �5 14 .ifl White's pressure in the centre is quite tangible. w 10 h3 e5 b5 a6 In game 6 Anand improved his play by means of 10 ...c5 1 1 dxc5 .ixc5 1 2 e4 .id6 1 3 lDd4 (incred­ ibly, at that very moment on the same stage in the quarter-final match between Korchnoi-Timman, 1 3 lDb5 lDe5 14 �e5 .ixe5 15 f4 13 .tn 14 d5 c5 c4?! A serious inaccuracy; the mod­ est 14 ...l:tb8 was correct, and then in the event of 15 a4 b4 1 6 lDb1 b3 17 'iie2 l:tb4 Black would have gained counterplay. 15 a4 l:tb8 152 Brussels Ct (4) 1991 Now 1 5 ...b4 does not work be­ cause of 16 lbe2 .!Dc5 17 .!Dd2 aS 1 8 lL!g3 and the advanced black pawns become objects for attack. axb5 16 axb5 17 .:l.a5! (D) _... . -·· � � ��ma � l. � � � - l. ­ �· � � - � - . .·� a••�- • ··-�- � � � � � � · - ltj� �� . ��d �-� \lllr •· • � ,. !;,' • � .: d ·.i.� � d 8 I had this position on the board during my preparation for the game. The attack by the black queenside pawns, which is inci­ dentally forced, seems dangerous, but in fact these foot soldiers will soon be stopped. b4 17 18 .!Da4 1 8 lL!b5 •d8 ! 1 9 .!Dxd6 'ii'xa5 20 .!Dxc4 'ii'c7 !? is not so clear. 'ii'd8 18 ... 19 .:l.a7 b3!? After 1 9 . . . .!Db6 20 i.e3 b3 2 1 'ii'b l it i s more difficult for Black to create counterplay, and he does not have full compensation in the event of 1 9 ... c3 20 bxc3 b3 2 1 'ii'e 2 lL!c5 ••• 22 .!Dxc5 i.xc5 23 .:tal b2 24 i.xb2 .!Dxe4 25 'ii'xe4 .:l.xb2 26 .:l.d2 either. 20 'ii'e2 The c4-pawn for the time being is invulnerable: 20 'ii'xc4 .:l.b4 2 1 �c6 lL!b8. 20 .!DeS 21 .!Dxc5 i.xc5 22 .:tal c3! Postponing this pawn's march until a later date is unsuccessful, viz. 22 .. .'ii'd6 23 i.d2 i.a6 24 i.c3, and White's superiority is be­ yond doubt. 23 .!Dxe5! Anand had probably only reck­ oned on 23 bxc3, and then 23 ...b2! 24 i.xb2 .:l.xb2 (24....!Dxe4 25 1i'xe4 .:l.xb2 26 'ii'x e5) 25 'ii'xb2 .!Dxe4 26 .:l.d2 .!Dxd2 27 'ii'xd2 e4 28 lbel e3 ! would have led to a po­ sition with chances for both sides. After this bold knight move the en­ emy pawn reaches White's second rank, but fortunately its career comes to an end there. 23 c2 24 .:l.d3 (D) I chose this continuation, al­ though I also examined 24 .:l.el .:l.e8 25 .!Dc6 'ii'b6 26 'ii'f3 which has roughly the same value. 1t'e8 24 Now 24 . .. .:1.e8 loses: 25 .!Dc6 'ii'b6 26 e5 .!Dxd5 27 .:l.xd5 'ii'xc6 28 \i'c4. ••• ••• Karpov - Short 153 -.t.- -·� � �··· - - - - -� - ·8� - - 8 - �· • - ·-:- - 8 B • -'iVD 8. � '� ' !:'-' !:'-' .- !i! � a m -i.= + if? ' !i! B l:tb6 25 li:)c6 li:)xe4 26 .i.e3 li:)xc5 27 .i.xcS 'iVd7 28 l:te3 l:txc6 29 'iVc4 29 ...'iVd6 30 l:tc3 li:)d7 3 1 l:txb3 is no better. 'iVd1 30 dxc6 'iVd6 31 l:te1 32 'iVcJ! White has a clear-cut plan to win a pawn: the queen makes way for the bishop on c4, after which the threat of l:ta5 arises. 32 'iVd5 32 ... .tf5 33 'it'e5! 'iVxe5 34l:txe5, etc., does not help. 33 c7! 33 l:ta5 would have been too hasty in view of3 3 ... li)e4! 34 l:txd5 li:)xc3 35 l:tc5 li:)a4 ! 36 l:tb5 li:)xb2 37 l:txb3 li:)dl and Black unexpect­ edly wins. 33 .i.b7 li:)e4 34 l:taS ••. 35 36 37 38 39 40 41 42 43 lbd5 l:td3 lbb3 .i.c4! l:ta3 l:txcl l:tcJ lbcl .i.n li:)xcJ li:)a2 .i.c8 g6 cHi' li:)xcl l:te8 l:te7 10 - Game 39 Karpov Short Linares Ct (7) 1992 QGD, Tartakower - 1 d4 d5 e6 2 c4 li:)f6 3 ll)cJ .i.e7 4 .i.g5 0-0 5 e3 h6 6 ll)f3 b6 7 .i.h4 The Tartakower Variation of the Queen's Gambit, also known to Russian players as the Bondarev­ sky-Makogonov Variation, is very widely used in modem practice, and moreover many grandmasters willingly play it as both Black and White. When the h-pawn attacks the bishop, you always have some hesitation: should you exchange on f6, or wait until ... b6 has been played, and temporarily retreat the bishop to h4? In my encounters with Kasparov, this bishop di­ lemma arose 24 times (equivalent 154 Linares Ct (7) 1992 to a whole World Championship match !), and most frequently the bishop avoided immediate ex­ change. 8 .i.e2 .i.b7 .i.xf6 9 .i.xf6 exdS (D) 10 cxdS At first glance the position seems fairly simple, but this vari­ ation is quite poisonous, and de­ mands careful play by Black, as we shall see. Curiously enough, this position arose eight times in my matches against Kasparov. In spite of the stormy nature of our battles, these disputes strangely al­ ways ended in peace. B lack is already preparing to carry out ... c5, and the natural reac­ tion is b4, and after an exchange on c5, l:[bl . w 1 1 b4 cS By playing I l ...c6 Black could have avoided hanging pawns, but this looks rather passive. After 1 2 0-0 consider the following exam­ ples: a) 12 .. .'ifd6 1 3 'ifb3 lLld7 I4 :rei .:r.ad8 I 5 .:r.ab1 .:.res I6 .i.d3 .i.a8 (Speelman-Short, London Ct ( I ) I 988) I7 h3 and then e4 gives White a small but stable advantage. b) I 2 . .. .:r.es 1 3 'iib 3 and then (D): B b i ) 1 3 ...'ifd6 I 4 a4 lLld7 1 5 a5 l:tad8 I 6 axb6 axb6 17 l:ta7 'iib 8 IS .:r.a2 b5 I9 lLle1 .i.e7 20 lLld3 .i.d6 2I g3 lLlb6 22 .i.f3 .i.c8 23 l:tfal .i.f5 24 lLlc5 lLlc4 25 lLle2 gave White a secure plus in P.Nik­ olic-Short, Manila IZ 1 990. b2) Black tried to improve with 1 3 . . . a5 in Karpov-Short, Amster­ dam 1 99 1 . The game continued I4 a3 lLld7 1 5 b5 c5 1 6 lLlxd5 .i.xd4 ( 1 6 ... .i.xd5 I7 'ifxd5 lLle5 1 8 'iib3 ltlxf3+ I9 .i.xf3 .:r.cs 20 .i.c6 is no good). After 17 .:r.ad 1 lLle5 ! 1 8 Karpov - Short 155 lt:ixe5 -'.xd5 1 9 lt:ic4 'ifg5 20 g3 'iff5 2 1 1lfe1 'ife4 22 f3 'ifxe3+ 23 'ifxe3 llxe3 24 lt:ixe3 -'.xe3+ 25 �fl -'.d4 26 llxd4 cxd4 Black managed to stand firm, but 17 exd4 llxe2 1 8 1lfe 1 ! ? would have given Black certain difficulties. What­ ever, at our next meeting in the Candidates match, Short pushed his c-pawn two squares. bxcS 12 bxcS 'ii'aS 13 llb1 As a whole series of games against Kasparov showed, 1 3 .. .-'.c6 is more passive. cxd4 14 'ifd2 -'.xd4 15 lt:ixd4 16 exd4 (D) Not, of course, 16 'iVxd4 lt:ic6 1 7 'iVd2 d 4 1 8 exd4 -'.a6 an d Black i s already better [editor's note: but in The Queen 's Gambit for the At­ tacking Player 1 9 -'.xa6 'ifxa6 20 lt:ie2 llfe8 2 1 llb2 is analysed to an ending with marginal winning . chances for White]. -'.a6 16 1 6 . . . -'.c6, as seen in the 40th game of my first match against Kasparov, is weaker. I was playing B lack, and I only managed to hold the position with difficulty. 17 lt:ibS 'ii'd8 18 0-0 Now the black knight has two possibilities - it can develop to d7 or c6. After 1 8 . . . lt:id7 a rapid rout ••• B ended the game Vaganian-Geller, New York 1 990: 1 9 1lfc 1 lt:if6 20 f3 ! lle8 2 1 a4 1le7 22 -'.d3 -'.c8? 23 'iVf4 ! a6 24 lt:ic7 g5 25 ...xf6 lle1 + 26 1lxe1 'ifxf6 27 �hl 'ilfxd4 28 1lbd1 ! , etc. 18 lt:ic6 19 .:.Cd1 Soon after our Candidates match the Olympiad took place (Manila 1 992), and Short again chose this variation against Azmaiparashvili. The Georgian team captain obvi­ ously supposed that after our duel Short would have prepared some sort of improvement, so he played something different: 1 9 a4. I have also played this pawn move, if slightly later, but here White de­ cided there was no time to lose. Af­ ter 1 9 . ..'iff6 20 llfd1 1lfd8 2 1 1lb3 llac8 22 h3 ...g5 23 'ilfxg5 hxg5 24 llg3 ! f6 25 -'.g4 1lb8 26 1lc3 Black had a difficult endgame. 'iVf6 19 ••• ••• 156 Linares Ct (7) 1992 20 �n (DJ A modest novelty, which I had prepared specially for the match. It is more precise than 20 �f3, as in the game Yusupov-Beliavsky, Lin­ ares 1 988, where after 20 ... .:tab8 2 1 a4 .:tfd8 22 'iic 3 �c8 23 .:tbc 1 a6 24 'iixc6 'iixc6 25 .:txc6 axb5 26 axb5 the players agreed a draw. w 26 .:te3! lbe6 26 .. ..:tc2? loses directly: 27 .:te8+ 'iii>h7 28 'iixc2 1fxc2 29 �d3+. 27 .:te5 'ii'g4 28 �e2! 'ii'g6 29 g3 .:tbc8 (D) B 20 .:tabS 21 a4 :res 22 .:tb3! �xb5 lbd8 (D) 23 axb5 24 'ifa2! Now the white rook on the third rank prepares to move to one of three important squares, a3, e3 or f3. 24 .:tc7 If 24 ....:tb7, then 25 .:te3, threat­ ening 26 'ifxd5 and 26 .:te8+. The weak d5-pawn is causing Black a lot of trouble. 'ji'fS 25 .:ta3! ••• . �. . . • • • • ••• • ·41\·�·�•·a • • � -� IJ · � · � . � � • • • B � �� � 'ii' IJ · � �� /� u u � · � · � .:. · � � � � � � � w 30 �h5! The d5-pawn is going nowhere, and so there is no point hurrying to take it: 30 .:txd5? .:tc2; 30 jfxd5 .:td8 3 1 'ji'e4 'ji'xe4 32 .:txe4 .:tcd7. 1i'f6 30 ••• Karpov - Kamsky 157 g6 l:tc3 l:ta3 (D) 31 'ii'xd5 32 .te2 33 h4 38 . .. 'ilt'xd4 39 l:txd4 l:taa2, though it is true that here the precise 40 �f l ! would have maintained an appreciable advantage for White. �xg8 39 l:tg8+ 40 'ii'xf6 l:txe2 lbe4 41 :at l:txa2 42 l:txa2 l:txf2+ 43 'ii'd4 l:te2 44 �gl 1-0 45 d6 Game 40 Karpov - Kamsky Moscow Alekhine mem 1992 Grtinfeld Defence w 34 35 36 37 38 'ii'e4 �g2 d5 'ilt'd4 l:te8 (D) l:tcc3 �g7 lbc5 l:tc2 • . •:• • � · � � � .·� • mu � � . ."" ... - · � - � - •:n � � . � -� � � - � � - � � u . � B .I Bi.B�· . •:• . B l:taa2? 38 Black could have shown more dogged resistance by continuing ••• This game, played at the Moscow super-tournament celebrating the l OOth anniversary of Alekhine's birth, went, as they say, wonder­ fully well. A bitter struggle was conducted over the whole board, first on one flank, then the other, and in the end the black pieces be­ came totally disorientated. lbf6 1 d4 2 c4 g6 .tg7 3 lbfJ c6 4 g3 First White (by playing 4 g3 in­ stead of 4 lbc3) and then Black (by 4 ... c6 instead of 4 ...d5) avoids the sharpest continuations in this opening, giving the game a posi­ tional feel. d5 5 .tg2 1 58 Moscow Alekhine mem 1992 6 cxdS cxdS 0-0 7 ltlc3 e6 8 ltleS 9 0-0 9 .i.g5 has been played several times, but it is not compulsory to fix the position of the dark-squared bishop at this early stage. 9 ltlfd7 10 f4 (D) B Supporting the knight with the pawn is more logical than retreat­ ing it to f3. 10 ltlc6 1 0 ...f6 is also played quite fre­ quently, but practice has shown that it has more disadvantages than advantages. 1 1 ..t.e3 ltlb6 Here as well 1 1 . . .f6 is possible (see game 37). 12 .i.f2 .i.d7 In Karpov-Kasparov, Seville Wch (3) 1 987, Kasparov preferred •.• 1 2 .. . ltle7, so that after 1 3 e4 he could exchange pawns and seize the d5-square. Although Black managed to equalize, it was felt that his development was defi­ nitely awkward. 13 e4 ltle7 With the idea of placing the bishop on c6 and strengthening the centre. 14 ltlxd7 'ifxd7 15 eS (D) B l:.ac8 15 1 5 ...l:.fc8 was Kasparov's pref­ erence in Karpov-Kasparov, Se­ ville Weh ( 1 ) 1987. This game went 1 6 l:.cl .i.f8, when White could have sacrificed a pawn with 17 g4 l:.c4 18 f5 !?, achieving a serious in­ itiative. Instead the game contin­ ued 17 .i.f3 l:.c7 1 8 b3 l:.ac8 1 9 'ilfd2 ltlc6 20 'ilfb2 a6 2 1 .i.e2 'ilfe7 22 ltlb1 l£Jb4 23 ltlc3 ltlc6, ending soon in a draw by repetition. .•• Karpov 16 l:lcl a6 (D) Now Black has a ' hanging' knight on b6, which will soon put him on the spot. 1 6. ..l:lc7 would have been safer, and if 17 'ifb3, then 17 . . .ltlc4 1 8 l'Llb5 l:lc6 1 9 ltlxa7 l:la6 20 l'Llb5 l:lb6 with coun­ terplay. Generally, while Black's position on the queenside looks quite reliable, the less said about the kingside the better. - Kamsky 159 The first in a series of fine moves by the queen, which charac­ terized this game. In the variation 20 l'Lle2 l:lxc 1 2 1 l:lxcl l:lxc l + 22 tbxc 1 ltlc6 23 lDd3 White main­ tains his spatial advantage, but the position has lost its internal energy, and his chances on the kingside are not all that great. • � .•• . - •• - · · -·- · · · ·- ••••• B BiD B � u �8· • • u . • 8 !!U " mu · - • · 8• • M � .JL �U R 8 8 � • . -:� *! B lbc6 20 �a3 21 fS! All the black pieces have trans­ ferred to the queen side, but for the time being this is not dangerous, because the white pieces on the other half of the board are not yet co-ordinated. 2 1 .. .exf5 22 gxf5 'ifxf5 is unsuccessful in view of 23 �h3 'ifh5 24 �xc8 �h6 25 'ifh3 �xcl 26 'ifxh5 gxh5 27 �xb7 �b2 28 �xc6. 2 l ...�b4 would not have led Black to his aim either, due to 22 �hl intending 23 �el . l'Llb4 22 l:lcd1 ••• 17 b3 Emphasizing the awkward position of the knight on b6. ltc7 17 1 8 'ifd2 :Cc8 19 g4! The starting gun is fired for White's attack. The constrained nature of the black pieces makes counterplay significantly more dif­ ficult. �f8 19 20 '1Ve3! (D) 160 Moscow Alekhine mem 1992 Now the black pieces are getting under each other's feet. The knight is blockading his own bishop, a fact of which I instantly made use. However, 22 ...lbxd4 23 W'xd4 is no better, since the b6-knight is hanging (23 ....tc5 24 Wf4 .txf2+ 25 1fxf2). 23 'ii'h6 (D) Stronger than 23 lbbl lbc2! 24 W'h6 ..tf8. 'iWe8 23 In the new situation 23 . ..lbc2 does not guarantee equality, as af­ ter 24 lbe2 the black rooks have no entry square for their invasion. 23 .. .'ii'e7 is no better: 24 lbbl .tb2 25 ..th4 'ii'f8 26 'ii'd2 .:.c2 27 1Wel , and the black pieces are soundly tied up on the queenside, despite their superficial activity. 23 ...lbd3?! does not work either, in view of 24 lbxd5 ! lbxd5 25 .:.xd3 .tf8 26 'ii'g 5 h6 27 fxe6 ! 'ii'xe6 28 .txd5, ••• when White simply has an extra pawn. 24 lbbl! Now the bishop has to desert the important a3-f8 diagonal. 24 .tb2 25 'ii'd 2! The white queen's shuttling is causing Black more than a little bother. lbc2 (D) 25 The startling 25 ... a5 ?! can be re­ futed by means of 26 a3 (the trap is 26 ti'xb2 l:r.c2 27 'ii'a3 l:r.xa2) 26 . ..l:r.c2 27 W'e l ii'bS (27 ...lbc6 28 l:r.d2 !) 28 axb4 .:.e2 29 it'xe2 "it'xe2 30 bxa5 (or 30 .:.d2 'ii'xg4 3 1 .:.xb2) 30...lbd7 3 1 .:.d2 . •.. •.. ........, • • - • • a"i ·- ····­ • ··8�· � �� · �J � � u �� � -�· . � ··· fli.B attJa : a : = w Black has been demonstrating the maximum activity for the last five moves. However, the threaten- , ing-looking tandem of the bishop on b2 and knight on c2 has to be defended, and only reduces the Karpov - Kamsky 161 activity of his own rooks. Having siopped his opponent's aggression, White begins to develop his attack. 26 �hl ! Freeing g l for the bishop, which in its turn uncovers the fl -rook. "ile7 26 lbd7 27 i.gl 28 ·.Uf3 With every move White's attack is gaining energy, while Black is still at a dead end. 28 'ifb4 29 'ifh6! Black has no time to take the d4pawn due to the threat of 30 .Uh3 lbf8 3 1 f6. 29 'iff4 would have been slipshod: 29 ...i.xd4! 30 i.xd4 lbxd4 3 1 .:Xd4 .:c 1 + 32 .:.n .:Xfl + 33 i.xfl 'ifel 34 lbd2 lbxe5 with sufficient counterplay. 29 'iff8 30 'ifgS (D) In the event of 30 'ifh4 the cap­ ture 30 ... lbxd4 does not work due to 3 1 i.xd4 i.xd4 (or 3 1 .. ..l:lc l 32 i.xb2 ! .l:lxdl + 33 .l:lfl ) 32 .Uxd4 .Uc 1 + 33 .Ufl .Uxfl + 34 i.xfl .l:lc 1 35 lbd2 and White maintains a material advantage. However, both 30 ... 'ifd8 and 30 ...'ifg7 are quite reliable for Black. Now after either of these moves White should con­ tinue 3 1 'ifd2, and in comparison with the position on move 28, the situation has altered appreciably in White's favour. ••• ••• ••• The black queen cannot yet re­ turn to f8 in view of the reply fxe6. 31 b6 There is no other way of freeing the 'hostages' on b2 and c2. Black is preparing 3 1 . ..a5 in order to se­ cure the b4-square for his knight to retreat. aS 32 .Udn 33 h4 lbb4 34 a3 Obviously not 34 'ifxb2 owing to 34 ... .Uc2. After the text the rook invasion on c2 is not dangerous for White. .Ucl 34 35 'iff4 lbc6 36 .th3! Threatening 37 fxe6 fxe6 38 g5 lbf8 39 'ifxf8+ ! ! 'ifxf8 40 i.xe6+ �g7 4 1 .l:lf7+! , winning. lbd8 36 37 i.e3! ..• 1 62 Moscow Alekhine mem 1992 The c l -square is under control, therefore the bishop scrutinises the h6-square. 37 b5 38 ft3fl! (D) Having appreciably strength­ ened his position on the kingside, White suddenly changes tactics. Paradoxically, because of the awk­ ward position of his pieces, espe­ cially the b2-bishop, Black, after exchanging rooks, will perish pre­ cisely where he once dominated. In a few moves the white queen be­ comes mistress of the position on the c-file. ..• li)fS 47 'ife7 ! 'ifd7 48 .i.h6 ! - an amusing finale! li)xe5 43 'ili'c2 Despair. 43 . ..'ili'f8 44 'ifc7 'ife8 45 f6 is also bad, and if 43 ... li)b8, then 44 li)d2 ltJdc6 45 lbf3 is suffi­ cient, as White is effectively play­ ing with an extra piece. But now, having accepted the sacrifice, I managed to find a beautiful forced win. 44 dxeS 'iWxeS 45 'ili'c8! 'ili'e4+ 45 ...'ibe3 allows mate in three. 46 i.g2 'ili'xbl + 47 �h2 i.b2 48 'ili'xd8+ �g7 (D) · · - ••• • • · mit •l • -· ··· � • ••••• � � � �z . �-� � -·· � �� �u � �tffi � "' � - u o � • m •.t � �- � --� � r• � .l[). • .:. •� " � B b4 38 39 axb4 axb4 40 llxc2 llxc2 41 :n ftxfl 42 'ili'xfl .i.a3 Alas, after 42 ....i.c3 there is the immediately decisive 43 f6 'iff8 44 'ili'c2 i.el 45 'ifc7 'ili'e8 46 i.fl w i.xf6 49 f6+! �h6 50 i.h6+! 51 'ili'xf6 'ifcl �h5 52 g5+ 53 �gJ! Strangely, 53 �h3? would have let victory slip from my grasp: Kamsky - Karpov 163 53 ...1Wf5+ 54 1Wxf5 gxf5 55 .i.f3+ �g6. 53 'ii'c7+ 54 �h3 1-0 For this victory in the Alekhine memorial, I was awarded a prize for 'playing in Alekhine's style' . Strictly speaking, I do not feel this was not a very precise decision. In fact, in the work of the fourth World Chess Champion there were many examples of the battle being transferred from the queenside to the kingside. But here we were ac­ tually dealing with a rarer theme, as on the contrary, I managed to transfer the battle from the king­ side to the queenside with decisive effect! Then this beautiful victory also soon won the best game prize in In­ formator. ..• Game 4 1 Kamsky - Karpov Dortmund 1993 Caro-Kann Defence 1 2 3 4 5 6 6 .i.c4 tinuation. 6 e4 c6 d5 d4 ltJd2 dxe4 ltJd7 ltJxe4 ltJg5 ltJgf6 .i.d3 is another popular con­ ... e6 7 ltJ1f3 .i.d6 (D) One of the key positions in the modern treatment of the Caro­ Kann. Besides 7 ....i.d6 Black may choose between 7 ... 'fic7, 7 ... i.e7 and 7 ...h6. w 8 1We2 More active than 8 0-0, which Kasparov once played against me (Amsterdam 1 988); then after the continuation 8 ...h6 9 lbe4 ltJxe4 10 .i.xe4, by playing 10...ltJf6 !? Black could have equalized. In fact I chose another plan: 10 ...0-0 1 1 c3 e5 1 2 .i.c2 l:le8 1 3 lle1 exd4 1 4 l:lxe8+ ..,xe8 1 5 'fixd4 'fie7 16 .tf4 .txf4 17 'fixf4, and White gained an advantage. 8 h6 9 ltJe4 The standard sacrifice on e6 is not dangerous for Black in this situation: 9 ltJxe6 fxe6 10 .i.g6+ �e7 1 1 0-0 lDf8 1 2 .td3 .td7 1 3 ... 164 Dortmund 1993 lbe5 •e8 14 f4 �d8 1 5 c4 �c7 1 6 �d2 .:.cs 1 7 c5 �e7 1 8 f5 �b8 1 9 .:.ael �d8, and he had a clear ad­ vantage in the game Amason-0st­ enstad, Torshavn 1987. tDxe4 9 10 1i'xe4 In the event of 1 0 �xe4 0-0 Black quickly plays ...c5 or ...e5 with a good game. 10 lDf6 1 1 ... h4 (D) •.• B 1 1 'ii'e2 is played more fre­ quently, but in fact on h4 the queen looks quite threatening. 1 1 . .. 0-0 will not do because of 1 2 �xh6 ! gxh6 1 3 'ii'xh6, and Black cannot defend himself from the threat of 14 lLlg5 . After l l ...'iia5+ 1 2 �d2 'ifh5 1 3 'ifxh5 lDxh5 1 4 lDe5 0-0 1 5 0-0-0 he has a prospectless posi­ tion. 1 1 . . .lDd5 is recommended by theory, but after 1 2 'ifxd8+ �xd8 1 3 c3 White preserves a small but stable positional advantage. What should Black do? 11 �e7!? I thought of this paradoxical ma­ noeuvre, keeping the king in the centre, in 1 988, when I was prepar­ ing for the above-mentioned game against Kasparov in Amsterdam. I was afraid of the white queen being transferred to the king side, but ex­ changing queens struck me as be­ ing rather insipid. I sat for some time over this position before I found the correct solution. So this important novelty had to bide its time for five whole years ! With this odd king move Black unexpectedly brings harmony to his forces. Now the threat of ...g5g4 has been created, winnilfg a piece, and if White is to count on keeping his opening advantage, he will have to act decisively. 12 lDeS The most ambitious solution. Bearing in mind that the enemy king is stuck in the centre, White decides to sacrifice a pawn. Of course, he will hardly be happy af­ ter 1 2 �f4 �b4+ 1 3 �d2 (else 1 3 ... g5) 1 3 ...�xd2+ 14 'itxd2. 12 �xeS 'ila5+ 13 dxeS 14 cJ (D) 1 4 �d2 'ifxe5+ 1 5 �e3 is also possible, and it is very dangerous ••• Kamsky - Karpov 165 for B lack to take on b2, altho1:1gh the position of the c-pawn is hardly better for White than in the game itself. 20 f4 (in order to close the c l-h6 diagonal from checks) 20...gxf4 2 1 i.d4 seems dangerous, for exam­ ple (D): .. . . . . � ··� � - •.t. . � � � - � �·- mu � · � - - � � � � � - � � � � Wl1 ­ � � i.� · - u � � ��" � � -� � � � u � � � � . � � ��:, . w 14 15 16 17 18 19 i.e3 0-0-0 'ii'a4 libel 'ii'a3 ,\& 'ii'xe5+ b6 g5 cS ..td7 l:.hd8 (D) B a) 21 ...ifd5 loses on account of 22 l:.e5 Wxg2 23 l:.xc5! bxc5 24 ifxc5+ �e8 25 ..txf6. b) 2 l ...'it'c7 is also in White's favour: 22 i.eS ifc6 23 i.xf6+ (a rare combination arises after 23 ..txf4 ! ? 'ilt'a4?? 24 ..td6+! �e8 25 l:.xe6+! fxe6 26 i.g6#) 23 ...�xf6 24 i.e4 Wa4. c) 2 1 . . .'it'g5 ! 22 l:.e5 ifh4 ! res­ cues Black, and 23 g3 does not lead White to his goal in view of 23 . .. fxg3 24 hxg3 ifxg3 25 l:.xcS bxcS 26 ifxcS+ �e8 27 i.xf6 iff4+. 20 'ii'c7 The black queen abandons the centre of the board, and restrains all approaches to the king. 21 ..td4 ••• 20 g3 166 Dortmund 1993 Now after 2 1 f4 there is the strong retort 2 1 . .. ttlg4 22 .td4 gxf4 23 h3 ttle3. .te8! 21 ... 22 �b1 l:td5 22 . . . .tc6 ! 23 .te5 'ifd7 24 .tc2 'ifb7 is perhaps even better. .:.Sd8 23 f4 24 .tc2 24 .te5 1i'c6 ! 25 c4 l:t5d7 26 'ifc3 ttlg4 27 fxg5 hxg5 28 l:tfl l:td4 ! favours Black. l:t5d6 24�xf6 25 .txf6+ hxg5 26 fxg5+ l:txd6 27 l:txd6 28 c4 The only move, as otherwise the white queen turns out to be ex­ cluded from the game. �e7 28 29 ii'e3 f6 30 h4 (D) Here in mutual time trouble, Kamsky offered me a draw, but I turned it down, as Black's chances are evidently better: White has been deprived of the advantage of the bishop pair, and he is still a pawn down. gxh4 30 'ifd7 31 gxh4 e5?! 32 W'h6 In time trouble I decided to de­ fend the f6-pawn, and missed a chance to do something much more effective: 32 ...l:td2 ! 33 l:tfl ..• have been stronger. 34 W'h7+ �d8 34 .. ..tf7 35 .tg6 l:tdl + would have led to a draw (35 ...'ifxc4 36 b3 ii'e6 37 l:tgl ). 35 h6 l:td2 36 W'f5 'ifxf5 37 .txf5 .td7? (D) Karpov - Kramnik 1 67 38 i.g6? A mistake; Kamsky failed to make use of my blunder in time trouble. After 38 �c 1 ! Black would even have to fight for a draw: 38 ...l:td4 (the rook cannot abandon the d-file because 39 l:td1 would win the bishop) 39 i.xd7 ! l:txc4+ (after either capture on d7 there follows 40 l:td1 , and the h-pawn will inevitably promote) 40 �bl l:th4 (40 ... �xd7? 4 1 l:th l ) 4 1 l:td l l:txh6. l:th2 38 �e7 39 h7 i.e6 40 i.d3 fS 41 l:tg1 �f6 42 l:tg7+ e4 (D) 43 l:txa7 l:txa2 49 l:txb6 0-1 If you exclude Black's mistake in time trouble, the game was quite logical, and its result was fully in order. Game 42 Karpov - Kramnik Linares 1994 QGD, Semi-Slav 1 d4 dS 2 c4 c6 3 �f3 �f6 4 �c3 e6 �bd7 5 e3 6 i.d3 Around this time I was avoiding the Meran System, preferring the Anti-Meran 6 'ii'c2 (see, for exam­ ple, game 38), but this did not mean that I had excluded it from my opening repertoire for ever. 6 dxc4 7 i.xc4 bS 8 i.d3 After the move 8 i.e2 it would be more difficult for White to count on an advantage, as the game Kar­ pov-Kasparov, Moscow Web (29) 1 984/5, showed: 8 ... i.b7 (8 ...b4 9 �a4 i.e7 10 0-0 i.b7 1 1 a3 a5 1 2 'ii'c2 0-0 1 3 l:td 1 c5 ! 1 4 �xc5 lill.c5 15 dxc5 'ii'c7 1 6 axb4 axb4 17 l:txa8 l:txa8 1 8 i.d2 i.e4 ! 1 9 'ifc4 'ifxc5 20 jfxc5 i.xc5 gave Black 1 68 Linares 1994 an equal position in Polugaevsky­ M.Gurevich, Antwerp 1 993) 9 a3 b4 1 0 lLla4 bxa3 1 1 bxa3 !l...e7 1 2 0-0 0-0 1 3 !l...b2 c5 112-112. 8 a6 After 8...!1...b7 9 a3 b4 10 lLle4 a5 1 1 lLlxf6+ lt:Jxf6 1 2 e4 !l...e7 1 3 'ii'e2 c5 14 !l...b 5+ �8 15 dxc5 !l...xc5 1 6 !l...d 3 h 6 17 0-0 g5?! 1 8 !l...e 3 !l...xe3 1 9 'iixe3 rlig7 20 li:Jd2 lDd7 2 1 lt:Jc4 'iie7 2 2 'ifd4+ e 5 2 3 'ifd6 'iWxd6 24 lDxd6 !l...c6 25 !l...c4 I gained a big advantage in Karpov­ Shirov, Linares 1994. c5 9 e4 c4 10 d5 My theoretical duel with the young chess star soon continued in Monaco, albeit in a rapid-play. Af­ ter the continuation 1 0 . . . 'ilfc7 1 1 0-0 !l...b7 1 2 dxe6 fxe6 1 3 !l...c 2, Kramnik could have returned to the game we are examining with 1 3 . .c4, but of course it ended sadly for him, so he continued instead with 1 3 ...!1...e7 !? ( 1 3 ...!1...d6 14 lDg5 gives White a big advantage) 14 lt:Jg5 W'c6 1 5 'ii'f3 h6 16 'ii'h 3. Here 1 6...lLlf8 !? would have led to a rather complicated game, but in­ stead Black preferred a very dubi­ ous exchange sacrifice: 1 6 . .. hxg5? 17 ...xh8+ rlif7 1 8 ...h3 g4 19 'ilfh4 lt:Je5 20 f3 gxf3 2 1 gxf3. Black has no compensation for his material loss, and later I could have taken the upper hand by force more than ... once, but I was caught out by the time control. 1 1 dxe6 fxe6 After 1 1 .. .cxd3 1 2 exd7+ 'ilfxd7 1 3 0-0 !l...b7 14 .:.et , it is well­ known that Black faces an uphill struggle for a draw. 12 !l...c2 !l...b7 'ilfc7 13 0-0 14 lDg5 lLlcS (D) Black played 1 4 ...'ii'c6? ! in the game Gligoric-Ljubojevic, Linares 1 99 1 , but the stormy complications of 1 5 'irf3 ! !l...c5 1 6 'irh3 rlie7 17 lDf3 b4 1 8 lLle2 lDxe4 19 lDed4 'iib6 20 !l...xe4 !l...xe4 2 1 'ifg4 !l...xd4 22 'irxe4 !l...f6 23 'ii'xc4 were clearly not in his favour. Moving the knight to c5 looks reasonable, but... . w 15 e5! A clever break in the centre. There are other possibilities, 1 5 'irf3, 1 5 'ire2, and 1 5 f4, but as I Karpov - Kramnik 169 have the two bishops, I want to open up the position ! Generally, the idea of sacrificing a pawn with e4-e5 is not new, but I am not aware of it being seen in precisely this position before. 'ii'xeS 15 The pawn must be taken, as after 1 5 . . .'ifc6 1 6 f3 Black's battery on the long diagonal is liquidated, and things are looking bad for him. 'ii'd6 16 :et 17 'i!kxd6! This turn of events is completely unexpected; White is a pawn down, but he nevertheless voluntarily ex­ changes queens. Rather paradoxi­ cal ! .txd6 (D) 17 ... ••. w 18 .te3! Quietly finishing his develop­ ment, as the e6-pawn is going no­ where. 0-0 18 ... After 1 8 ... �d3, White should play the simple 19 .txd3 cxd3 20 .:r.ad l , and Black loses both pawns in the centre. 1 8 . . b4 is no better: 1 9 .txc5 .txc5 20 �a4, and either the knight or rook can take on e6. Finally, 18 ...�g4 is no good be­ cause of 19 .txc5 .txc5 20 .:r.xe6+ 'ifild7 2 1 .:r.dl+ r/;c7 22 �. etc. 19 .:r.ad1 .te7 Black has to lose a tempo - after 19 ...:ads 20 .:r.xd6 .:r.xd6 2 1 .txc5 .:r.d2 22 :c l Black does not have sufficient compensation for the material he has given up because of the threat of 23 .te3. .txcS 20 .txcS :res 21 �xe6 22 h3! (D) Not so much to open an escape hatch as to advance the g-pawn. The raid 22 �xc5 .:r.xc5 23 .:r.e7 .:r.bs allows Black to hold the posi­ tion. . 1 70 Linares 1994 The critical moment. The black pieces are poorly co-ordinated, and the question is whether he can im­ prove their interaction. i.f8?! 22 22... b4 loses: 23 lba4 i.a7 (oth­ erwise 24 lbb6) 24 lbxg7 'it>xg7 25 l:te7+. If 22 ... i.b4, then 23 a3, forcing Black to give up his bishop, as after 23 ...i.a5 there is the cap­ ture 24 lbxg7 'it>xg7 25 l:te7+. A later game, Topalov-Lautier, Dos Hermanas 1994, featured 22...l:tab8, and after 23 g4 i.f3 24 l:td2 b4 25 lba4 i.a7 26 g5 lbd5 27 l:te5 l:tb5 Black eventually managed to stand firm. However, the impression re­ mained that he was on the edge of the abyss for a long time. h6 23 g4 Or 23 ... .tf3 24 l:td4 ! b4 25 lba4 lbd5 26 l:te5 ! , etc. If 23 ... g6, then 24 lbxf8 l:txf8 25 g5 followed by 26 l:td7 is very unpleasant. i.f3 24 f4 25 l:td2 (D) 25 l:td4 looks more energetic, and if 25 ...i.c6, then 26 g5 hxg5 27 fxg5 lbd7 28 l:th4 and the black king falls into a mating net. How­ ever, by continuing 25 ...l:te8 Black could threaten 26 ... l:r.xe6 27 l:r.xe6 i.c5 or 26...i.c5 immediately. i.c6 25 hxg5 26 g5 lbd7 27 fxg5 28 lbxf8 •.• • • • • •• • • • • • i. R . RltJ• � ••• • • ••• 08& • ""� � � • � .t. • -8 • • 8 "i.� u � • • . . � � � � . . B 28 .l:.f2, with the idea of i.g6f7+ and l:tf4, is tempting, but af­ ter 28 ... .tc5 (28 ...lbc5? 29 lhf8+) 29 lbxc5 lbxc5 30 l:te5 White has only a minimal advantage: 30...lbd3 3 1 i.xd3 cxd3. �xf8 28 ... In the event of 28 ...l:r.xf8 the rook manoeuvre to d6 becomes more valid. 29 l:td6 29 lbd5 also looks reasonable, aiming to exchange off Black's ac­ tive bishop. b4 29 ... 30 lbe4 .i.e8?! (D) 30 ...i.d7 was necessary: 3 1 h4 (3 1 lbg3 l:te8 !) 3 1 ...l:tc6 and Black can still breathe. 31 lbg3! I was placing great hopes on this move - the appearance of the knight on f5 or h5 is extremely un­ pleasant. lidS 31 ... Karpov - Kramnik 1 71 Evidently 37 a4 is more logical, rounding up the black a-pawn later. l:td3 37 ... 38 �gl l:txb3 39 h4 (D) w Now 3 l . . ..i.d7 does not work: 32 l:te7 .i.xh3 33 ll)hS, etc. l:txd6 32 ll:)fS .i.g6 33 ll)xd6 34 .i.xg6 ll)xg6 35 ll)xc4 l:td8 (D) 35 ...l:tf8 36 l:te4 ll)f4 37 h4 �h7 38 ll)eS is suicide. - . .•• � � � � . � � - -­ � � •• . • • • D .ttJ. • • . . . -� �B • • • � %z � � � � � � � w 36 l:te4 b3 If 36 ...l:td3, then 37 �g2 and the h-pawn marches forward. 37 axb3 B ll)f8? 39 Of course, the defence 39 ... �f8 40 h5 ll)e7 4 1 l:te5 �e8 would have been more stubborn, although after 42 l:ta5 (or 42 l:tc5) White, with his extra pawn, must emerge victorious . 40 l:te8! 1-0 It is becoming clear that the black king cannot break out to free­ dom; if 40 . .t�f7. then 4 1 ll)d6+ ! and the rook is defended. While Kramnik was grieving over the fact that moving his king from g8 would not free his pieces from their paralysis, his flag fell. It remains only to say that while my game against Topalov from the same tournament won the best ••• . 1 72 Linares 1994 game prize in lnformator, this game in the very same edition won the prize for the most important theoretical game. Game@ Karpov - Beliavsky Linares 1994 Catalan Opening We have come to the end of the Li­ nares tournament, and with it the end of this book. Victory over Be­ liavsky in the final round would guarantee me a lead of 2 points over the rest of the field ! But this game is not only remarkable for the context of the competition, as in it I also managed to use a piece of opening preparation which had been awaiting its hour of glory for about twenty years ! So, let us return to the chess board for the last time. ltlf6 1 d4 d5 2 ltlfJ e6 3 c4 i..e7 4 g3 0-0 5 i..g2 dxc4 6 0-0 a6 - 7 ..,cl 8 a4 So far this is all very well known. For example, I once had the position after move 7 against the very same opponent, although on that occasion we were playing with opposite colours! After 8 'iVxc4 b5 9 'ifc2 1&..b7 10 i..d 2 lDc6 1 1 e3 ltlb4 12 i..xb4 1&..xb4 1 3 a3 i..d6 14 ltlbd2 .:c8 15 b4 a5 a com­ plex position with chances for both sides arose in Beliavsky-Karpov, Brussels 1 988. i..d7 8 9 Wxc4 i..c6 (D) w 10 i..g5 i..d5 Matters developed differently in Kasparov-Karpov, Moscow Wch (22) 1 984/5 : 10 ... a5 1 1 ltlc3 ltla6 1 2 .:act 'ilfd6 1 3 ltle5 i..x g2 1 4 'iPxg2 c 6 1 5 i..xf6 gxf6 1 6 ltlf3, and again White has· preserved only a small advantage out of the opening. c5 1 1 'ii'd3 12 ltlc3 cxd4 'ii'xd5 (D) 13 ltlxd5 It is necessary to capture on d5 with the knight. In the old game Htibner-Karpov, Tilburg 1 979, the Karpov - Beliavsky 1 73 w exchanges 1 3 ... lLlxd5 14 i.xe7 'flxe7 15 lLlxd4 lLlc6 16 lLlxc6 bxc6 left White only a symbolic advan­ tage. Taking on d5 with the queen is a new move ... 1 4 h4! ! . . . new for Beliavsky, but not for the author of these pages ! I had faced this position on the board in 1 974, as I was preparing for my match against Korchnoi in Mos­ cow - the match which was soon to give me the title of 1 2th World Champion ! It goes without saying that it is very rare for an opening novelty to be used two decades af­ ter it was prepared. Furthermore, I must admit that this novelty is shat­ tering: Black's position collapses like a house of cards. Capturing on d4 with the queen is known to lead to a better end­ game for Black, and taking with the knight is impossible, since the g5-bishop loses its support. Now, with the bishop defended by the h­ pawn, the recapture in the centre can be made by the knight. Besides this, the light-squared bishop ac­ quires incredible strength, which is extremely important. This is a very rare example of an advance by a rook's pawn practically deciding the outcome of the game from the very start. lLlbd7 14 ... 14 ... lLlc6 suggests itself, but White has at his disposal the fol­ lowing cunning variation: 15 i.xf6 i.xf6 16 lLlg5 !, and again the pawn on the edge of the board helps, this time by defending the knight: 16 . ..iif5 17 i.e4 it'a5 18 i.xc6 i.xg5 19 i.xb7 .:.a7 20 i.e4. The pawn has been won back, and White's positional advantage is ex­ tremely tangible. 15 lLlxd4 'ii'd6 (D) w 1 74 Linares 1994 If 15 ...Wa5, then 1 6 lt:Jb3 ! . 1 6 l:lfd1 lt:JcS 17 ...c4 This is stronger than 17 lt:Jf5? ! lt:Jxd3 1 8 lt:Jxd6 lt:Jxb2 19 l:ld2 .i.xd6 20 l:lxd6 lt:Jc4 or 1 7 'i'c2 ...b6 1 8 a5 'ii'b4, when White's advantage is not that large. 17 l:lfd8 (D) ... etc. w 18 b4! lt:Jxa4 Beliavsky is confused, and this is another consequence of White's 1 4th move. He should have played 1 8 . . . lt:Jce4 1 9 .i.xf6 lt:Jxf6 (but not 1 9 . . ..i.xf6? 20 .i.xe4 .i.xd4 2 1 e3 .i.xal 22 l:lxd6 l:lxd6 23 .i.xb7 with a big advantage) 20 .i.xb7 l:lab8 2 1 'i'xa6 'ilt'xb4 22 lt:Jb5 . White would have taken the upper hand, but the game would have continued. 19 ...b3! (D) Now it's all over: Black loses a piece. 1-0 20 e3 This victory in the final round was my ninth at Linares; it was even more remarkable that I won six straight games in the first six rounds ! With four draws (I man­ aged to avoid being beaten) I reached an inconceivable score, 1 11 1 3 . And, incidentally, the sec­ ond and third placed players, Kas­ parov and Shirov, were 2112 points behind me. Without false modesty I can say that in the whole of chess history, tournaments where on the one hand, all the stars of the chess world are gathered together, and on the other, the winner has demon­ strated such notable superiority over the remaining contestants, can be counted on your fingers. Karpov - Beliavsky 1 75 Kasparov once called the tradi­ tional battle at Linares the World Championship Tournament. It is a pity that this title was not made official, because it would have been a wonderful crowning addi­ tion to my collection of champion­ ship titles !


Comments

Copyright © 2024 UPDOCS Inc.