Winning Endgame Strategy
Alexander Beliavsky, Adrian Mikhalchishin
B.T. Batsford Ltd, London First published in 2000 ©Alexander Beliavsky, Adrian Mikhalchishin 2000
ISBN 0 7134 8446 2
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 permission of the publisher.
Printed in Great Britain by Creative Print and Design (Wales), Ebbw Vale forthe publishers, B.T. Batsford Ltd, 9 Blenheim Court, Brewery Road, London N7 9NT A member of the ChKalisGroup plc
A BA'J:SFORD CHESS BOOK Contents
Page
Introduction 5
I. Pawn Endings 7
2. Knight Endings 34
3. Rook Endings 56
4. Shouldering: the struggle of the kings 109
5. Complex Endings 112
6. Defence in the Ending 131
7. The isolated pawn in the Ending 136
8. Rook and bishop against rook and knight 141
9. Rook and two pawns against rook and knight 150
10. Two minor pieces against a rook 153
11. Rook against knight and pawn 172
12. Queen Endings 180
Solutions to Exercises 191 Introduction
For the authors the most import themes, such as queen, knight and ant thing is-will readers study their complex endings, were not dealt book? It was very pleasant for us with in the previous book. However when Winning Endgame Technique in some cases we have devoted a (in fact the first part of the present littl e more attention to methods of book) was deeply studied by grand play in definite types of position, masters Boris Gelfand and Ognjen rather than concrete cases, which is Cvitan. Then grandmasters Suat a fault, on the whole, of all books on Atalik and Alex Yermolinsky also the endgame. For example the studied the book for a month Encyclopaedia of Chess Endings (looking for mistakes!) and found gives replies only to some ques much of interest for themselves. tions. We have tried to find more Particular thanks to grandmaster examples in which play conforms Alexei Kuzmin, trainer of the Qatar with general principles so that team, who discovered many import readers can begin to apply these ant and interesting corrections in methods in concrete practical situ pawn endings. We still cannot guar ations. Of course, correlation of antee there are no mistakes-but methods of play in typical positions then again, generally speaking, any and techniques is not always uni one who considers themselves free form, but the authors present their from error makes more mistakes vision of practical endgame prob than others. lems and are a long way from ex The present book is a direct con hausting this theme. tinuation of Winning Endgame It remains to thank our friend, Technique and to some extent the master Oleg Stetsko, for help with second part in the sense that many the selection of practical examples. 1 Pawn Endings
Guru Grigoriev was right! Shirov-Timman Wijkaan Zee, 1996 Readers might know that the greatest connoisseur of pawn studies was the Soviet master Nikolai Grigoriev, who achieved the im possible in his complicated pawn studies. In fact he once said: "Grandmasters do not like pawn endings because they simply don't understand them". The statement sounds paradoxical but the thought is topical even today. It seems that pawn endings are simple but the number of mistakes grandmasters have made over the years has in no way diminished. Since the publica But White does not win after tion of our firstbook quite a few in 1 ... �d6 2 h4 �xc6 3 f5! 'iitd6 4 f6 , teresting endgames have been when there are two ways to draw: played and quite a few instructive mistakes made. (a) 4 ...�d7 5 �f3 �e8 6 'i¥te4 The most characteristic and sur rtic7 7 �d5 �e8! 8 rtic6 (8 rtie6 prising mistakes are ... rtif8!) 8 ...�d8, and he does not win the c7 pawn; Premature resignation (b) 4 ...c6 5 rtif3 �d7 6 rtie4 of a game @d8!, and the king holds the Yes, even today, this happens at squares corresponding to a5, c5, grandmaster level. e5-d7; a4, b4, c4, d4, e4-d8, e8; a6-e6. In the followingposition Timman resigned and both players thought Another surprising resignation this to be in order, and appropriate followed in a game played two comments appeared in ChessBase... years later ... 8 Pawn Endings
Maciej a-Grabarczyk Krumpachnik-Polak Poland, 1998 Slovenia, 1985
Although aft er 1. ..bxa 5+ 2 �xa5 Here Black resigned, reckoning c5 3 �b5 �d6 it is a simple draw. that he would lose both of his doubled pawns, but after1...�d7 2 It is also possible to resign be �xe5 �e7 3 g6 �f8! 4 �xe6 �g7 cause of an incorrect calculation of 5 �f5 �h6! 6 �f6 it all ends in a an arising pawn ending. well known stalemate.
Svidler-Lobron N atapov-Schuravlov Erevan, 1996 Moscow, 1994
Here Black resigned, reckoning Black resigned, forgetting that on the variation 1...�g2 2 @c5 there was a last chance of saving the hl='ir' 3 .!:l.xhl �xhl 4 �d5 �g2 5 game by stalemate aft er 1. •. �c6 2 'itie5, but he did not calculate to the axb5+ �xb5 3 'it>d5 �a4! end. 5 ...'itit3 6 �xf5 �e3 and Black captures the c3 pawn. Pawn Endings 9
Yudasin-Osnos After an arduous defence and Leningrad, 1987 transposition to a pawn ending White had the impression that an in teresting chance had cropped up, and he took it. 1 'it>fS?? But he did not reckon on the simple l...b4! 2 axb4+ Losing simply is 2 a4 'it>d4 3'it>g6 'it>c3 and the black pawn is faster than the white one. 2 ...'it>xb4 3 'it>g6 'it>xb34 'it>xh6 a5 5 'it>xgS a4 6 'it>h6 a3 7 g5 a2 8 g6 al='i' 9 g7 11Yf6+ and he had to After playing 1 'it>f2 Yudasin of resign. fered a draw, saying that this well Correct was the more logical l known position is in the books and 'it>e5 a5 2 'it>e4 a4 3 bxa4 bxa4 4 is drawn! His experienced oppo 'it>e5 'it>c4 5 'it>e4 'it>b3 6 'it>d3 'it>xa3 nent, and international master, was 7 'it>c3 'it>a2 8 'it>c2 a3 and here the shocked by his own 'ignorance' and extra move 9 h3! makes a draw. accepted it there and then. However Another way to the draw is 5 'it>d3 it ought to be well-known to every 'it>d5 6 h3!. one that after l ... 'it>e4 2 'it>e2 f4 3 @f2 f3 4 @fl the triangulation e5-f5-e4 is winning. It just shows A lack of understanding of basic don 't believe everybody!! principles of play in pawn endings Erroneous play in pawn endings Markovic-I van ovic Another type of very common Vrnjacka Banja, 1998 mistake-even strong grandmasters have many technical shortcomings.
Hector-Speelman Roskilde, 1998
Because of his weakened structure on the queen's flank (doubled pawns) White has the inferior posi tion, and the right way now was 10 Pawn Endings
I ... hxg4 2 'it>xg4 'it>e5 3 Wg5 d5 ! 4 Schandorff- Speelman cxd5 b5! 5 axb5 a4 6 'it>g6 a3 7 d6! Roskilde, 1998 'it>xd6 8 @xg7 a2 etc. The only chance of a draw would be the diffi cult 4 d4+! @xd4 5 'it>g6 @es 6 cxd5 @f4 (or 6 ... b5 7 d6! cxd6 8 axb5 a4 9 b6=) 7 c4 We4 8 @xg7 @xf5 9 @fl We5 10 @e7, and a draw. But Black played 1...g5+? 2 fx g6 h4?? It was still a draw after2 ...hxg4 3 'it>xg4 @xg6. 3 @n wxg6 4 @g2 @f6 5 Wh3 @g5 6 d4 c6 7 c3 And Black resigned since after 7 ...d5 winning is 8 c5 b5 9 axb5 a4 1 b4?? 'it1f6 2'it>h4 g6! and White 10 bxc6, and in the resulting queen resigned since after 3 'it>g3 g5 4 f4 ending White has too many pawns. exf3 5 'it1xf3 'it>e5 6 @fl We4 7 @e2 f5 ! 8 gxf5 'it>xf5 thanks to the dis tant passed pawn Black wins easily. Ehlvest-Shirov Correct was I @g3 ! b4 (or I . ..a5 2 Vienna, 1996 a4 bxa4 3 bxa4 g6 4 Wh3 'it>f6 5 @h4 @e6 6 'it1g5 'it>e5 7'it>h6 Wf6 8 g5+! @f5 9 Wg7 with a draw) 2 @h3 f5 3 gxf5 @xf5 4 @h4! g5+ 5 'it>g3 @f6 6 'it>h3 @g7 7 'it>g3 '.t>h6 8 @h3 @h5 9 'it>g3 a6 I 0 'it>h3 g4+ 11 'it>g3 @g5 12 'it>g2 @h4 13 @h2 g3+ 14 fxg3+ @g4 15 'it>g2 as 16 �fl @h3 17 g4!, and a draw. Black lacks a single tempo move with the a-pawn.
Plaskett-Rowson Scotland, 1998 A simple position. White only needs to meet ...'it>d5 with @e3 and there is no problem. However Ehlvest played carelessly. 1 @f4?? and after 1 ... 'it1e6 2 We3 @d5 3 'it>d3 f4 ! 4 gxf4 h4 5 @e3 h3 6 gxh3 gxh3 7 @n @xd4 8 'it1g3 @e4 he had to resign.
In the following position we have another example of an incorrect pawn move. Pawn Endings 11
Black evaluated the position poor Ostenstad-Kuzmin ly and sacrificeda pawn. Biel, 1990 1. .. g4?? 2 fxg4 lt>g5 3 'itif3 e5 4 c4 e4+ 5 lt>xe4 lt>xg4 6 b4 Wxh5 7 �f5! lt>h4 8 c5 a6 9 a4 h5 10 b5 axb5 11 axb5 lt>g3 12 c6! bxc6 13 b6!, and Black resigned. After the correct 'passive' defence l...\t>g7! 2 c4 lt>h6 3 b4 lt>xh5 4 c5 lt>g6 5 'it>e5 h5 6 b5 Wfl 7 'itid6 g4 he would easily achieve a draw. But it looked all so simple-going directly forthe h5 pawn.
Greenfeld-Golod Only one winning move is left: Is rael, 1998 1...lt>e3! 2 lt>c4 @d2 3 lt>b5 'it>c3 4 Wxb6 lt>b4, with a typical finish,but there followed 1...lt>d5?? and after 2 'it>d3! draw.
Ignorance of typical methods of struggle. Typical breakthroughs.
Adams-Lutz Wijk aan Zee, 1995
Correct play was the simple l...f5 ! 2 Wxa7 (after 2 gxf5 exf5 3 lt>xa7, simply 3 ... g4 4 a4 f4 5 a5 g3) 2 ...e5 3 a4 e4 4 lt>b6 fxg4 (4 ...f4 !?-+) 5 a5 g3! 6 fxg3 e3 7 a6 e2 8 a7 el='l'W 9 a8=it' �4+ 10 @c7 'l'Wd6+! with a transposition to a single, but easily winning, pawn ending. In the game Black decided to 'press' the white king. 1...lt>d6? 2 @xa7 'itic6 but here 1 h5?? followed 3 '1t>b8!! f5 4 a4 fxg4 (or Correct was 1 bxa4 bxa4 2 g4 ! g6 4 ... f4 5 a5 lt>b5 6 @b7! 'itxa5 7 3 gxf5 gxf5 4 lt>d3! h5 5 c3 ! 'it>c5 6 'itic6!) 5 a5 lt>b5 6 'it>b7! (Reti's c4 e4+ 7 fxe4 fxe4 8 lt>xe4 @xc4 9 typical manoeuvre) 6 ...'it>xa5 7 'it>c6 'it>e3 'it>b3 10 lt>d3 lt>xa3 11 'it>c3 Wb4 8 lt>d6 lt>c4 9 @xe6 @d3 with with a draw. a draw. l...b4! White resigned. J 2 Pawn Endings
Overlooking stalemating defences 1 h4? It is terribly dangerous to move Hellers-Eingorn the 'wrong' pawn-necessary was 1 Debrecen, 1992 @c3 'it>d5 2 'it>d3 h5 3 b4! axb4 (3 ... a4 4 e4+ @d6 5 h4 e5 6 'it>e3 exf4+ 7 'it>xf4 'it>e6 8 'it>e3 g5 9 f4=) 4 axb4 e5 5 e4+ @e6 6 'it>e3 exf4+ 7 'it>xf4 g5+ (if 7 ...h4, then 8 'it>e3 D. f4) 8 'it>e3We5 (8... f5 9 f4 g4 10 e5) 9 h4! 'it>e6 10 f4 with a draw. 1...'it>d5 2 b4? Another incorrect advance-it was necessary to play 2 e4+ @d6 3 @e3 ! (3 'it>d4? e5+ 4 fxe5 fxe5+ 5 @d3 'it>e6 6 Wc3 h5 D. g5 -+) 3 ...e5 4 f5 ! gxf5 5 exf5 'it>d5 6 'it>d3 a4 7 bxa4 bxa4 8 h5 h6 9 'it>c3 e4 IO Here easily winning is l ...l:!xa4 2 fxe4+ �xe4 11 'it>b4, and, though i.xf6 (2 c6 fxe5 3 c7 l!d4+! 4 'it>c5 Black has an extra pawn in the �di) 2 ...�a2 3 c6 l:!xg24 c7 l:!c2. queen ending, White has chances of However Hellers saw that in the a draw. pawn ending he could force the win 2 ...axb4 3 axb4 of a pawn and decided that this was sufficient. 1. .. J:Ixc5+? 2 'it>xc5 fxe5 3 'it>d5 Wg6 4 'it>xe5 'it>g5 5 'it>e4 h5 6 We5 Wh4! 7 'it>xf4 Drawn. The Black king is stalemated!
Erroneous pawn advances
Indeed, this is a great mistake to make in pawn endings.
Karpov-Kasparov Las Palmas, 1996 3 ... h6? Now comes a mistake from the other side-after 3 ... e5! 4 e4+ We6 5 'it>e3 �d6! (5 ... 'it>f7? 6 f5 !=) 6 f5 (if 6 fxe5+, then 6 ...fxe5 7 'it>t2 �e6 8 'it>g2 'it>f6 9 'it>g3 h6, and zug zwang) 6 ...gxf5 7 exf5 'it>d5 8 'it>d3 h5 9 'it>e3 'it>c4 10 'it>e4 @xb4 11 'it>d5 'it>c3 12 'it>e6b4 13 'it>xf6 b3 14 'it>g7 b2 15 f6 b 1 ='iV 16 f7 'i/Vb7 17 'it>g8'iVxf3, winning. 4 e4+ 'it>d6 5 We3 e5 6 fxe5+ fxe5 7 'it>f2�e6 8 'it>g2! Pawn Endings 13
Weak was 8 'it>g3 c;f;>f6 9 'it>g4 h5+ 6 ... �d5?? IO c;f;>g3 g5, and Black wins, while Correct was to force e2-e3 by now on 8 ... �f6 follows 9 c;f;>g3 h5 6 ...'it>e4! 7 e3 �d5 8 �e2 i;£;>c5! 9 10 f4, and a draw. 'it>d2'it>d6 !, so that on 10 'it>d3 @d5
Matlak-Tseshkovsky Lubniewice, 1995
...it will be White's move-after Black moves his position is lost, since he is forced to allow the White king to e5, and ...i;£;>c5 loses aftere4. 1 f4 ?? 7 'it>e3 'it>c4 8 i;t>d2 'it>d4 9 e3+ Any other move wins-simplest 'it>c4 10 i;£;>c2!, was 1 c;f;>M 'it>f4 2 �h5 'it>e3 3 �g6 and Black resigned since White 'it>xe2 4f4 'it>e3 5 f5 . achieves the above-mentioned posi 1...'it>e4 2 c;f;>g4 c;f;>d4! tion with Black to move. White had reckoned only on 2 ...'it>e3 3 �f5 with a win, but now Drasko-VratonjiC on 3 c;f;>f5 follows 3 ... 'it>e3!, and after Ulcinj, 1997 White moves this position is drawn! 3 'it>h5 'it>e4! Again Black will not 'buy' 3 ... i;t>e3? because of 4 f5 winning forWhite. 4 i;£;>g4 i;t>d4 s wo rs 6 'it>f2
1... f5 ! A colossal defensive resource, though it was possible to hold the position even by simple defence, e.g. l...'it>e6 2 i;£;>e4 (the assessment is not changed by 2 f4 f5 3 g5 h5=) 14 Pawn Endings
2 . . . f5+ (the only move) 3 gxf5+ A pawn ending which is quite 'it>d6 4 f6 'it>c5 5 'it>e5 a5 6 a4 difficult to evaluate. To gain victory 'it>c6!=; or 2 c5 bxc5+ (losing is Black needs to solve the problem on 2 . . . b5 3 f4 a5 4 h4, 5 h5, 6 g5! + -) 3 the queenside, where his opponent 'it>xc5 f5 4 gxf5+ 'it>xf5 5 �b5 'it>f4 has the chance to create a passed 6 'it>a6 'it>f3 7 'it>xa7 'it>xf2=. pawn. This is achieved by I...'it>d7! 2 gxf5 f6 ! 3 a4? 2 h4 'it>c73 'it>f3 'it>c84 'it>e4 'it>b75 White loses the thread of the 'it>f3 a6 -+. However in the game gam.e, starting to play on his weak followed flank. He should reconcile himself 1 ... g5? 2 @f3 'it'd5 3 c6 'it>d6 4 to a drawn result after 3 h4 'it>c6 4 �e4?? h5 'it>d6 5 'it>d3 'it>c5 6 'it>c3 a6 7 a4 The decisive mistake. The signifi a5 8 f3=. cance of a tempo in pawn endings is 3 ...'it>c6 4 f4 ?? fa r higher than in any other. White A very serious mistake. Without should immediately exploit his op need White himself gives up the op ponent's mistake and himself break portunity of exploiting the tempo up Black's pawn chain. This is done and hands back the move to his op by 4 h4! gxh4 5 'it>g4 a6 6 bxa6 ponent in a position of mutual zug 'it>xc6 7 'it>xh4 'it'b6 8 'iitg4 'it>xa6 9 zwang. It was still not too late to h4 + -. lead the game to a draw by 4 h4 h5 4 ...a6 5 bxa6 'it>xc6 6 'it>f3 'itb6 7 5 'it>d3 'it>c5 6 �c3 a6 6 f3 a5 7 f4 . h4 gxh4 8 �g4 'it'xa6 9 �xh4 'it>b6 4 .•. 'it>d6 5 'it>d3 'it>c5 6 'it>c3 h5 7 10 'it'g4 'it>c6 11 h4 'it'd6 White 'it'b3 h4- + 8 'itc3 h3 9 �b3 a6 10 resigned. a5 The assessment of the position is As a matter of fact, in their first not changed by 10 �c3 a5 - +. book, the authors made several inac 10 ... bxa5 11 'it'a4 'it>xc4 12 'it>xa5 curacies, which were corrected by 'it'd4 13 'it>xa6 'it'e4 14 'it>b5 'it'xf5 grandmaster Alexei Kuzmin. White resigned.
Sulipa-Gricak Kupreichik-Mikhalchishin Lvov, 1995 Lvov, 1988 Pawn Endings 15
1... @dS In their first book the authors placed a question mark against this move, pointing out a 'direct' path to a draw: l ...\t>f 6 2 <;¥;>f2 <;¥;>g6 3 @f3 h5 4 gxh5+ 'it>xh5 5 'it>e4 <;¥;>g4 6 f5 �g5 7 <;¥;>xd4 <;¥;>xf5 8 'it>c5 'it>e5 9 �b6 i;t>d6 etc. However they did not reckon on one finesse to which A.Kuzmin drew attention: 3 f5+! (3 <;¥;>e2 h5 4 f5 + 'it>h6=) 3 ... \t>g5? (3 i;t>f6, returning to the basic vari ation) 4 'it>e2 h5 5 f6 ! 'it>xf6 6 gxh5 In this queen ending White has \t>g5 7 i;t>d3 'it>xh5 8 'it>xd4 \t>g5 9 some chances of a win but upon c;t>c5 i;t>f6 10 i;t>b6 'it>e7 11 <;¥;>xa6 correct defence it should probably <;¥;>d7 12 <;¥;>b7+-. Thus Black's at be drawn. For example: 13 g5 'fl'el+ tempt to simplify the position at 14 'it>c4 'it>g4, and it is not easy for once is mistaken. White to improve his position. 2 'it>f2 <;¥;>e4 4 ••.hS? A dubious move. Better, missed Better is 4 ...a5 ! (6. 5 'it>d3 'it>d5 6 by the authors, is 2 ...a5 ! 3 <;¥;>f3 (or 3 f6 'iite6) 5 @f3 i;t>d5 6\t>f4 'it>c4 7 f6 'it>e2 'it>e4 4 f5 'it>e5)3 ...\t>c4 4 'it>e2 d3 8 fl d2 9 f8='fl' dl=ii=. 'it>c5 (but not 4 ...'it>d5? 5 'itid3 'it>c5 6 S gxhS 'it>xfS 6 �d3 'it>gS 7 'it>xd4 g5 hxg5 7 fxg5 'it>d5 8g6+-) 5 i;t>d3 �xhS 8 \ticsBlack resigned. (5 'it>d2 'it>d6 6 'it>c2 @e6 7 c;t>b3 Now it is clear that best was an i;t>d5=) 5 ... i;t>d5 6 f5 'it>e5 7 <;¥;>c4 immediate l ... a5! 2 'it>f2 'it>d5 3 @f3 'it>e4 8 f6 d3 9 'it>c3 <;¥;>e3 lO fl d2 11 <;¥;>c4 4 'it>e2 'it>c3 5 'it>dl <;¥;>c4 6 'it>d2 f8='ii'dl ='fl' 12 'fl'xh6+ <;¥;>f3, and in 'itc5 7 'it>c2 'it>c4 8 f5 i;t>d5 with a the resulting queen ending Black draw, since on 1...'it>d5 could follow achieves a draw without trouble. 2 a5! (the same as on l...�f6 2 'it>f2! 3 rs \ties 4 'it>e2? 'iitg6 3 f5 + i;t>f6!-4 a5!). Much stronger is 4 a5, but the authors mistakenly assumed that Vaganian-Portisch this was easily winning, giving the Ti/burg, 1992 variation 4 ...'it>d5 (4 ...h5? 5 gxh5 \t>xf5 6 <;¥;>e2 'it>g5 7 i;t>d3 l:>. @xd4-c5-b6xa6-b7+-) 5 'it>e2 'it>e5(?) 6 '>t>d3 �d5 7 f6 'it>e6 8 Wxd4 'itxf6 9 @c5 @g5 10 'it>b6 �xg4 11 <;¥;>xa6 h5 12 i;t>b6h4 13 a6 �tc. Far more tenacious in the opin ion of A.Kuzmin was 5 ...'it>d6 6 �d3 (6 �d2 'iitd5 7 �c2 i;t>d6 8 �b3 i>d5=) 6 ...�e5 7 'it>c4 \t>e4 8 f6 d3 9 <;¥;>c3 'it>e3 10 fl d2 11 f8='ii' dl='ii 12 'ti'xh6+ �f3. 16 Pawn Endings
A.Kuzmin also drew attention to 'it>c5 4 a3 h5 5 a4 a5 6 e5 'it>d5 7 this pawn ending. We present his 'it>d2 (of course White is not obliged more accurate analysis. to give up the pawn by the move 7 f5 ?) 7 ...Wd4 8 Wc2, sailing into a "The authors assessed this ending drawn harbour." in White's favour, giving 1 Wd2 It remains only to see how the Wc5(?-A.K.) 2 'it>c3 g5 (2 ...Wb5 3 game ended (comments by the g3 g6 4 h3 h5 5 g4 h4 6 g5 'it>c5 7 authors of the book). a3 'it>b5 8 Wd4 i;t>a4 9 Wxc4 Wxa3 1 h4? i;t>c5? 10 f5 a5 11 fxg6 fxg6 12 e5+-) 3 After l...h5! White is in no posi g3 (?-A.K.) 3 ...g4 4 a3 h5 5 a4 a5 6 tion to create a passed pawn: 2 �d2 e5 'it>d5 7 f5 ! Wxe5 8 Wxc4 Wxf5 9 Wc5 3 'it>c3 'it>b5 4 a3 Wc5 5 a4 a5 6 Wb5 +-. But the variation leaves a g3 g6 7 e5 Wd5 8 'it>d2 'it>d4 9 �c2 strange impression. Firstly let's look c3 10 'it>dl i;t>d3-+. at the finalposition. 2 h5 f6 ? Making it easy for his opponent to create a passed pawn: 2 ...Wd4 3 e5 'it>c3 4 f5 'it>b2 5e6 fx e6 6 fx e6 c3 7 e7 c2 8 e8='ii cl= 'ii+ . 3 We3 a5? Once again after 3 ... c3 4 Wd3 c2 5 Wxc2 �d4 6 i;t>b3 'it>xe4 7 'it>a4 'it>xf4 8 'it>a5 'it>g3 9 Wxa6 f5 10 a4 'it>g2 Black has the better queen ending. 4 a3 a4 5 e5 ! fxe5 6 fxe5 �d5 7 e6 �xe6 8 'it>d4 Black resigned.
After 9 ...'it>e6 (9 ...'it>g6 !?) 10 Poor technique in transposing 'it>xa5 f5 it is not Black, but White to a pawn ending who needs to think about saving the game: 11 Wb6 (11 Wb4 f4 12 a5 It is surprising but true that mod fxg3 13 hxg3 'it>d6- +; 11 �b5 f4 12 ern grandmasters experience great gxf4 h4 13 a5 g3 14 hxg3 h3 !-+) difficulties in transposing to a pawn l l...f4 12 a5 fx g3 (12... f3!? ) 13 ending. There are two factors-a hxg3 h4 14 a6 hxg3 15 a7 g2 16 poor knowledge of pawn endings a8='ilk gl='fl'+ with a theoretically and inability to switch from the fac won queen ending. tors of evaluation and play of com Secondly, after 2 ...g5 White can plicated endings to the factors and also win directly by 3 fxg5 hxg5 4 methods typical forpawn endings. a3 f6 5 a4 a5 6 h3 �d6 7 'it>xc4 �e5 This is a very great problem even 8 'it>b5 �xe4 9 �xa5. for high-level grandmasters and Thirdly, on 1 �d2 correct is an here the authors simply must also immediate l...g5 D. 2 g3 g4 3 �c3 name themselves. Pawn Endings 17
Beliavsky-Sveshnikov Black should correctly transfer to No vi Sad, 1979 a pawn ending and this is achieved by l ... .l:!a4! 2 a3, and the position of the pawn on a3 gives Black the possibility of more quickly creating a passed pawn. But in the game there followed ... 1... .l:!xd4+ 2 .!id3 .l:!xd3+ 3 �xd3 f6 4 i;¥(.c4 a6 S fS ! It is necessary to clear a path for the king, otherwise the march of his h2 pawn will lead to nothing. s ... 'it>d6 6 fx e6 �xe6 7 'it>d4 @rs 8 'it>dS 'it>gS 9 'ii>e6! fS 10 e3 'it>g4 11 h3+! <;¥;>gs 12 h4+! i;¥i>g6 13 hS+ <;¥;>gs 14 h6 �g6 1s h7 'it>xh7 16 How is it possible to lose here? 'itixfS 'itig7 17 'itie6 Black resigned. very simply-by transposing to a pawn ending. 1 'ii'e2?? 1 @£3, with a draw. Adams-Lautier 1. •. i;¥i>c3! Taking on e2 indeed leads Ti/burg, 1996 to a draw. 2 i;¥(fl 11Vxe2+ 3 i;¥i>xe2 'it>c2, and he had to resign since the king goes over to the white pawn on g3 while White's during this time goes to g6 after which Black wins by the well-known zugzwang .... 'it>g4!.
Kramnik-Lautier Belgrade, 1995
White has an extra pawn and can win the position as he pleases. As he pleases? In the game followed ... 1 h4? l:!.e6+!, and the pawn end ing with an extra pawn was drawn. 2 .Iles @f6 3 .l:!xe6+ �xe6 4 'it>d4 'it>d6 S i;¥i>c3 i;¥i>c7 6 b4 cxb4+ 7 �xb4 �c6, and a draw. Correct was 1 c5! or 1 .!lb5. 18 Pawn Endings
011-Benj amin 2 'llxe6 lies3 lies? Ne w York, 1995 Both players "drift" commented A.Alekhine. After 3 'it>e2 .l:Ixc6 4 lixc6 bxc6 5 b4! 'it>e7 6 'it>d3 'it>d6 7 'it>d4 White has a typically winning position. 3 ..•bxe6 ? But here Black does not exploit his chance-better was 3 ... lixc6 4 .l:Ixc6 (after 4 lixd5 lic2 and Black's active rook compensates forthe loss of a pawn) 4 ...bxc6 5 b4 'it>e7 6 We2 c,t>d6 7 'it>d3 c5 8 bxc5+ 'it>xc5 9 @c3 a5 reaching an equal position. 4 'it>e2 i;t>e7 S 'it>d3 'it>d6 6 :!.as Neither side can lose the rook lia8 7 'it>d4 fS 8 b4 lib89 a3 lia8 ending. But White decides to try to win the pawn ending. 1 lid4? .l:!xd4 2 exd4 c,t>g6 3 'it>f3 @rs 4 'it>e3 i;t>e6 s 'it>e4 fS+ 6 'it>f4 gS+ 7 'it>e3 @dS 8 f3 'it>e4 9 b3+ 'it>dS! 10 h3 hS, and he had to resign since on 11 c,t>d3 follows 1 l...g4.
However, we should not think that classical players from the past handled analogous situations better.
Flohr-Vidmar 10 e4! No ttingham, 1936 A seemingly illogical move but Black has only one weakness on a6 and White exchanges his weak e3 pawn, activates his king and rook along the fifth rank and then sets about creating weaknesses for the opponent on the kingside. 10 .•• fxe4 11 fxe4 dxe4 12 @xe4 �a7 13 i;t>f4 h6 14 h4! @e6 lS 'it>g4 .l:ta8 16 hS gS Or 16... gxh5+ 17 'it>xh5. l:!g8 18 g4+-. 17 g3 .l:Ia7 18'it>f3 Now the king transfers to the 1...'lle6? other flank. After the natural 1...c,t>e7 Black 18 ... .l:IaS 19 'it>e4 l:!.a720 .l:!eS+! has every chance of holding the Here Black has a choice: to allow slightly inferiorisolated pawn. the rook to e8 or the king to f5 . Pawn Endings 19
20 ... @d6 21 .l:te8 c5 22 .l:Id8+ Again typical-only this wins.
We also certainly come across transitions which are backed up by splendid calculation.
Cruz-Seirawan Moscow, 1994
1 ...tt::l c3! 2 i.xc3 dxc3 3 a4 c;t>b4 4 a5 Wxa5 5 Wxc3 'it>b5 6 'it>d3 Wb4! Black meets the white king's roundabout route with one of his own! 7 e3 1 ....l:Ia5 +! Or 7 rt;e4 rt;c3 8 Wxe5 rt;d2, The best solution, after l ....l:th5 2 catching up with the white pawns. 'it>b4 J::rxh2 3 c5 and White has 7 ...rt;b3 8 exf4 exf4 9 'it>e4 'it>c2 strong compensation for the pawn. 10 '\t>f5rt;d3 11 'it>xg5We3 12 'it>h4 2 rt;b3 .l:txa2 3 'it>xa2 'it>d6 4 'it>b3 Wxf3 13 g5 'it>e2Dr awn. rt;cs 5 c;t>c3 e5 At first sight it seems that White Kuzmin-Petrosian has the advantage because of his Moscow, 1979 passed pawn, but the most important factor in this ending is the paralysed white pawn chain on the kingside. 6 rt;d3 f5 7 'it>c3 e4 8 'it>b3 h6! 9 'ii>c3 h5 10 'it>b3 f4 ! Now follows a typical, well calculated breakthrough. 11 gxf4 e3! 12 fxe3 h4 13 f5 �d6! It is this which had to be foreseen -bad would have been 13... g3 ? 14 hxg3 hxg3 15 f6 rt;d6 16 c5+. 14 'it>b4 \ties 15 c5
The transfer to a pawn ending is capture on a3 the king on c3 can do forced, since after 1 ...bxc5 2 'it>d3 nothing since Black will have a and 3 \¥tc4Black stands badly. tempo after ...a5 ; h3 a4; h4 @a2; 1... .!lxc5 2 J:Ixc5 bxc5 3 'it'd3 e5 4 \¥tc2 a3!) 6 ... a5 7 \¥tc4 f4 8 @d4 f3 9 \¥tc4 f5 5 \¥txc5 h5 6 b4 axb4 7 \¥te3 <;¥txc5 10 @xf3 \¥tc4. @xb4 f4 8 a5 e4 9 \¥tc3 e3 10 c;¥(d3 ! Black resigned. Ivanchuk-Kasimdzhanov Elista, 1998 Ap. analogous idea was not taken into account by Black when trans posing to a pawn ending in the following game ...
Finkel-Mikhalchishin Belgrade, 1998
Here 1 l:td7 is quite simply win ning. But Ivanchuk was reckoning on the pawn ending. 1 l:txe6 'iVxe6 2 'iVxe6 fx e6 3 J:ixf8+ <;¥txf8 4 <;¥tg2 and all of a sudden 4 ...a5 !! Winning easily is 1...\¥te5! 2 l:l'.c3 The only move-bad was 4 ...cJite7 .!lc7 and3 ...<;¥td5, but Blackwants to 5 c;¥(f3 @d7 6 @f4 <;¥tc67 <;¥tg5cJitd5 take the bull by the horns at once by 8 c;¥(f6 a5 because of 9 b3 ! with a 1 ...@d5 ?? win forWhite. White is frightened by the pawn 5 @f3@f7 6 cJite4 ending and after . . . Nothing is gained by 6 cJitf4: 2 c;¥(f4 l:tc7 6 ...a4! 7 g4 hxg4 8 \¥txg4 \¥tg8!!, ...he could quietly resign. creating the distant opposition. And Meanwhile after 2 :id 1 + \¥tc6 3 if 6 b3, then 6 ...g5 7 \¥te3 @g6 8 l:txd7 <;¥txd7 4'it'd3! 'it>c6 (4 ... @e6 5 'it>e4 'it>h6 9 'it>d4 \¥tg6 l 0 'it>c5 c;¥(f5 'it>d4 changes practically nothing) 5 11 hxg5 <;¥txg5,with a draw. <;¥td4 a6 (after 5 ...f4 6 'it>e4 <;¥txc5 7 6 ... g5 7 c;¥(d4 <;¥tg6! @xf4 'it>b4 8\¥te5 \¥ta39 @f6 'it>xa2 We must mention the erroneous 10 cJitg7 'it>b1 11 h4 ! a5 12 h5 an ness of the exchange on h4, e.g. endgame with an extra pawn for 7 ... gxh4? 8 gxh4 \¥tg6 9 \¥te4 a4 10 White is reached. Therefore Black \¥tf4 @h6 11 <;¥te3 @g6 12 cJite4!, ri ghtly plays 7 ...@d5 8 \¥te3 'it>e5 9 triangulating, and Black is forced to h4! a6 10 a3, with a draw) 6 a3! move away his king to h6, which (bad is 6 'it>c4 f4 !, and the white gives White the opportunity of king does not get to g7, while aftera entering via f4 . Pawn Endings 21
8 @cs gxh4 9 gxh4 c,t>fs 10 'it>b5 1 a4? wxe5 11 'it>xa5 c,t>d4 and he had to In principle, a serious mistake, agr ee a draw. though White's plan is understan dable-he places a pawn on a5 and, Pawn endings in the creative work by sacrificing a pawn on the king of Robert Fischer side, breaks through with his king to the pawn on a6. However it is not Each of the gr eat champions has, possible to win this position. besides his own style, also his own E.Mednis in his book How to beat methods of play in the various Bobby Fischer assessed the position stages of the chess game. as a draw and did not criticise the Thus, upon his 'taste' depends the move in the game. Nevertheless ar ising various structures and ac White has a path to victory and it cording to his 'taste' his tr eatment consists of the move 1 g4!. In reply of them. The transfer to a pawn end Black has the following contin ing is one of the instruments for uations: realisation of an advantage or a a) l...Wd6 2 f5 gxf5 3 c,t>xf5 c4 4 method of defence. In Fischer's case bxc4 bxc4 5 'it>e4 c3 6 c,t>d3 @e5 7 this ending is met more frequently c,t>xc3 c,t>f4 8 'it>b4 'it>xg4 9@a5 Wf5 than any other champion, and the 10 'it>xa6 c,t>e6 11 a4, and White is authors were interested in how the victorious; great Fischer played them at these b) L..c,t>d6 2 f5 g5 3 a4! c,t>c6 moments. The examples show quite (3... b4 4 a5 'it>c6 5 f6+-) 4 axb5+ a broad spectrum of quality of play. axb5 5 'it>e5+-; c) l ...a5 2 a4 b4 3 'it>d3 ! (Why Fischer-Letelier not 3 f5 + gxf5 4 gxf5+ 'it>d6 5 f6 ? Mar de! Plata, 1959 Because of 6 ...c4 ! 7 bxc4 �e6!, and forthe time being White must think about saving himself) 3... 'it>d5
White clearly has the better king but the asymmetrical structure gives Black the possibility to reply to the creation of a passed pawn on the 4 g5 ! 'it>e6 5 'it>c4 c,t>f5 6 'it>xc5 king's flank with the creation of a 'it>xf4 7 'it>b5 'it>xg5 8 'it>xa5 '\t>f4 9 passed pawn of his own on the op 'it>xb4 g5 10 a5, and White reaches a posite side. Fischer continued to queen ending with a b-pawn-and play for a win. every chance of a win. This 22 Pawn Endings
variation was found by A.Bely Here is another example. avsky. But we return to the game: 1...'it>d6 2 f5?? Gligoric-Fischer Mednis passed over this move in Candidates (t ), Belgrade 1959 silence. Fischer, apparently, noticed that as a result of the correct 2 g3 'it>e63 a5 'it>d6 4 g4 'it>e6 5 f5 + gxf5 6 gxf5 'itf6 7 c,t>d5 c4 8 bxc4 bxc4 9 'it>xc4 c,t>xf5 the Black king hurries back to c8. 2 ••.gxf5+ 3 'it>xf5
Here, leading to a draw is 1.. . .l:!h5! 2 l:!c5 l:!xc5, and whichever way White retakes there follows 3 ...'it>c8 with a draw. But Fischer mistakenly played... 1... l:!h8? 2 c,t>xb5? Now Gligoric met mistake with 3 ..• 'it>d5? mistake, wrongly transferring to a (!)Mednis. Why not 3 ...c4 !, and pawn ending. Winning was 2 �c7+! White must resign at once. 'it>d6 3 �c6+ 'it>d7 4 'it>xb5 J:ib8+ 5 4 g4 'it>d4 5 g5? J:ib6 J:ih8 6 Il:b7+ 'itic8 7 'it>a6 nh6+ 5 axb5 leads at once to a draw. 8 'it>a7. 5 ...c4 ! 6 bxc4 b4 7 c5?? 2... llb8+ 3 'it>a4 J::ra8+ 4 'it>b3 After 7 g6 White reaches a slight nc8! 5 nxc8 'it>xc8 6 'it>c4 'it>b8! ly inferior queen ending but with Gligoric did not reckon in his chances of a draw. Now however previous calculations that White it's all simple-7 ... b3, and it is time could not maintain the opposition, forWhite to resign. therefore it's a draw. The pawn ending, apparently so simple, is difficult in that on the transferto it the players must calcu With time Fischer managed to late the changes that take place in correct his shortcomings and his the game by comparison with other transfer to the endgame became aspects of the ending. immaculate. Pawn Endings 23
We also come across ignorance in Lombardy-Fischer transferring to a pawn ending by USA (ch) , 1960/61 Fischer's rivals. Fischer-Bisguier USA (ch), 1959
Returning the exchange to win a pawn was also a recipe of Capablan ca himself. Black should suffer a little in the 1...lixc3+! 2 bxc3 .l:IxeS+ 3 'it>d2 rook ending by 1....l:Ig5 2 l:id4 b5 3 �xel 4 c;t>xel 'it>dS S 'it>d2 'it>c4 6 hS 'it>e2 'it>c5 4 @f3 .l:!g8 5 'it>f4 �f8+ b6 7 'it>c2 gS! 8 h6 f4 9 g4 as 10 with the idea of breaking through on bxaS bxaS 11 'it>b2 a4 12 Wa3 the second rank with the rook. But @xc3 13 'it>xa4 'it>d4 14 'it>b4 'it>e3 Bisguier decides to transfer to a and White resigned . pawn ending in which it requires in credibly accurate play to achieve a draw. Fischer-Larsen Candidates (m) Denver, 1971 1. ..Ii.dS? ! 2 'it>e2.l:Ixd2+ No help is 2 ...b5 3 �xd5 Wxd5 4 'i¥te3 a5 5 g5 b4 6 g6 We6 7 'it>d4 + -. 3 'it>xd2@dS 4 'i¥te3 @es S 'it>O! Fischer 'waits' for the weakening of Black's pawn structure-there is no win by 5 g5 'it>f5 6 'it>d4 'it>xg5 7 'it>xc4 Wf6 etc. S ... aS If 5 ...'it>f6 6 'it>f4 'it>g6 7 @e5 @g5 White goes not for the c4 pawn but for the pawn on b7-8 'it>d6! 'it>xg4 9 'it>c7 @f3 10 'it>xb7 'it>e2 11 'it>xa7 @d3 12 'i¥tb6 'it>c2 13 'i¥tc5, and Here White won easily after wm s. 1 a4 Wf8 2c3! .lii. .txc3 3 Wxc3 6 'it>e3 a4 c;t>e7 4 'it>d4 'it>d6 S aS f6 6 a6 'i¥tc6 On 6 ...b5 winning is 7 g5 b4 8 g6 7 a7 'it>b7 8 'it>dS h4 9 'i¥te6 and 'it>f6 9 'i¥td4. Black resigned. 24 Pawn Endings
7 g5 @f5 8 'it>d4 'it>xg5 9 'it>xc4 Offat last! @f4 10 Wb4 'it>e3 11 'it>xa4 c,t>d2 20... 'it>c2 20 ... 'it>d3 21 'it>b5 +-. 21 b4 Black resigned. .� � � � Rossolimo-Fischer ,• • • USA (ch), 1962/63 • • • w8· 8· 8· q '�-� �-�..- ·· .�DL��.� �·� �. �. � � mL,�m ... �m m 12 'it>b3 The position appears very simple, but in fact it is just the opposite as testified by the mistakes made in past analysis by authors even in the Encyclopaedia of Ch ess Endings. Fischer avoided 12 c4 because of Black clearly has the more active 12... Wc2 13 b4 'it>c3 14 c,t>b5 'it>d4 king but the White g5 pawn paral 15 c5 'it>c3 16 'it>a5 'it>c4 17 'it>a4 yses Black's flank and it is imposs Wd4 18 'it>b3 'it>d5 19 'itc3 b6 ! with ible to win. a draw. 1...a4 2 'it>e2 12 ... 'it>d3 13 c4 Wd2 14 Wa4 The Encyclopaedia recommends Also no help is 14 c5 'it>d3 15 2 b3! a3 3 h3 b4 4 h4, and now 'it>b4 'it>d4 16 @b5 c,t>d3 17 'it>a5 Black must play 4 ...d5 5 exd5 'it>xd5 'it>c4!. 6 c,t>d3 'it>c5 7 'it>e4 Wd6 with a 14 ... 'it>c2 15 'it>a3! draw. Again 15 b4 draw, as pointed out 2 ...a3 !? above. The last try. 15.. . 'it>d3 16 'it>b3 b6? 3 bxa3 Wc3 4 a4! bxa4 5 'it>e3 It is incomprehensible why Bis Wb2 6 @d2 'it>xa2 7 Wc2 a3 8 h3 guier did not give Fischer the And a draw. chance 'to show' the win after 16... 'it>d2. Some conclusions 17 c,t>b4c,t>c2 18 'it>a3 Clearly not 18 Wb5? because of In the earliest stage of his career 18... Wb3 !, but in the Encyclopaedia Robert Fischer had technical prob of Ch ess Endings is given the vari lems in playing pawn endings, and ation 18 b3? 'it>b2 19 'it>a4 'it>c3 20 particularly when transposing to 'it>a3 'it>d3 21 'it>b2 Wd4 22 'it>c2 them. But by the beginning of the b5!=. 60s he had already managed to 18... 'it>d3 19 Wb3 'it>d2 eradicate these shortcomings and no 19... 'it>d4 20 'it>b4 wins. longer had technical problems with 20 'it>a4! pawn endgames. Pawn Endings 25
Exercises: Pawn Endings
1 3
How can White win the endgame? What plan must White adopt to realise his advantage?
2 4
How do you assess the position What is correct l .. jxg4 or J...hxg4? with White to move? 26 Pawn Endings
5 8
How does Black save the endgame? What iscorrect: J...@e5 or 'iPg5?
6 9
Assess the outcome of playing What is correct: J ...g4 or l . . . rtJf6? the typical 1 g4?
7 10
Should White resign? What is the simplest way for White to win? Pawn Endings 2 7
11 14
How does White win? Can White hold the position?
12 15
Can Black save himself? White to move and win.
13 16
Whatis right J . . . @e6 or J ... �c6 ? Point out the right move fo r White. 28 Pawn Endings
17 20
White wins. How does White make a draw?
18 21
What is correct: l . . . e4, 1 ...g4 What is correct: or l ... @g8? l ...@g7 or l . . . rJ;h5?
19 22
What is the right movefo r Black? What is correct: l . . . a5 or l ...axb5? Pawn Endings 29
23 25
Find the right continuation How should White continue? for Black.
24 26
What iscorrect: 1 c4 or 1 b4? What is correct: l ... @xb4 or l ...rJilc4 ? 30 Pawn Endings
Exercises: Transposition to a Pawn Ending
1 3
Can Black transpose to Demonstrate a plan to realise a pawn ending? th e extra pawn.
2 4
Can White transpose to How do you assess th e endgame a pawn ending? aft er 1 c5? Pawn Endings 31
5 8
In whose fa vour is the pawn ending Wh ich should Black prefe r,
aft er I ... b5 ? the rook or the pawn ending?
6 9
Is it worth White avoiding the Can Black transpose to transposition to a pawn ending? a pawn ending? 7 10
Can I 0,d3 be played? In what way can White transpose to a pawn ending? 32 Pawn Endings
11 14
Is it possible to transpose to Find the righ t continuation
a pawn ending by I ...g 5? fo r Black.
12 15
Find the righ t continuation How can White save the game? fo r Black
13 16
Can Black transpose Find a winning plan fo r Black. to a pawn ending? Pawn Endings 33
17 20
Can Wh ite transpose to Ho w can Black hold a draw? a pawn ending?
18 21
Find the right continuation How does Wh ite transpose to fo r Black. a winning pawn endgame?
19 22
How should Black realise Find the right continuation his extra pawn? fo r Black. 2 Knight Endings
Due to their small radius of action his king and the presence for Black knights and pawns have very defi of a weak backward pawn on b6. nite features. M.Botvinnik described these features well by the very com Fedotov-Arkhipov prehensive and deep statement: Moscow, 1978 "Knight endings are pawn endings". Indeed the method of playing knight endings is very close to those with pawns. Here also a space advantage, active position of the king, pawn structure are all significant. An important role is played by the presence of outside passed pawns. Often the zugzwang motive is ex ploited in the game. However, apart from the general principles char acteristic for playing pawn endings, play in knight endings is compli cated by the possibility of pursuing However the fact that there are a knight with an enemy king. Apart only a few pawns on the board com from this, knight endings with plicates White's task. limited pawn material have specific An insufficiently deep penetration features associated with the into the features of the position possibility for the weaker side of leads to Black's defeat, practically sacrificing the knight to liquidate without a struggle. However, the opponent's pawns and thereby analysis shows that Black has very achieve a draw. This feature creates considerable defensive resources, significant difficulties in the and to achieve a win White has to realisation of a positional and demonstrate very accurate play material advantage. throughout. Let us look at a few examples Let us look at the possible course where there is a compact pawn of events. The first part of White's mass. plan consists of the creation of a Upon the central placement of passed pawn and exploitation of the pawns the pieces can have quite a weakness of the b6 pawn. broad range of activity. In the first 1 f5 exf5 2 liJxf5 ctJd43 ctJd6 position White has a positional ad White carefully controls the b5 vantage due to the active position of square, preventing the move Knight Endings 35
...b 6-b5, e.g. 3 tfJg7? @c6 4 e6 Black's possibilities. Now his best tllxe6 5 tfJxe6 b5, and the white counterchance lies in the formation pawns are exchanged. Also no good of a passed c-pawn, since giving up is 3 lllxd4 cxd4 4 rJ;fJ d3 5 e6+ the b6 pawn without a murmur leads rJ;;c7 6 e7 d2 7 e8='ili' dl=WV, with a quickly to defeat. And so ... draw. Black, for his part, is obliged 6 ... b5 ! 7 cxb5 Clld4 8 Cllc3 to control the e6 square. Now, for The realisation of the extra pawn example, 3 ...�c6 with the idea requires accuracy. Black's hopes are ...b6 -b5, does not work because of 4 linked to the win of the b-pawn and e6! rJi;xd6 5 e7, and the pawn cannot the subsequent sacrifice of the be stopped. knight for the e-pawn. Since his 3 .•.ttJe6 4 tfJb5 tfJd4 5tfJc3 tfJe6 king is already close to the b-pawn, any advance of this pawn must be backed up by accurate calculation. � �. � The second part of White's plan lies in the win of the c-pawn while holding on to his own pawns. Let us ·�,%ft.•&,�•-�� . ....� . •, ,,,M. ��%ft. ·'· see whether Black can hold the de � �J � fence by the active advance 8 ...c4 or .��.� .� . the passive move 8 ...tfJe6. -�•. �.. ; -� �� I. 8 ... c4 . . . � � � � 6 ltJd5 And so the first part of the plan is completed: White wins the b6 pawn, since the opponent's pieces are forced to control the passed e5 pawn. Also possible is 6 tlJa4 How ever it is interesting that, with the knight on a4, tactical resources suddenly appear for Black which require precise attention from his opponent. For example: 6 ...b5 7 cxb5 tfJc7. This trappy move was On c4 the pawn supported the ma not possible with the knight on d5. noeuvres of the white knight but Now, on the natural 8 tfJxc5+ fol now Black is leftwithout a base on �ws 8 ...'it>e8 ! with a draw after9 b6 d4 for his own knight which is also "Z.ld5+. However after 8 tfJc3 White deprived of its defence. This creates can return to the channels of the a motive to carry out the threat of a ma � in var iation, looked at below. breakthrough with the king to the c4 lso possible of course is 8 b6 pawn, in combination with thr eats to "Z.ld5+ 9 ctJ 'ii>fl c4 10 e6+ @c6 11 advance the passed pawn. c3! tfJxc3 12 e7 and White But the direct approach of the s hould win. But 7 ...tfJ �7 is not obli king to the c-pawn does not achieve gatory and is given to illustrate its objective. For example: 9 c;i;>g5 36 Knight Endings
Wc7 IO Wf4 Wb6 11 @e4 liJxb5 is not successful. For example: 9 with a draw. @f5 liJd4+ l 0 @e4 @e6 11 b6 liJc6 The solution lies in the creation of 12 b7 @d7 and ... @c7. Therefore a zugzwang position where Black White first endeavours to provoke simultaneously controls the e6 and the advance of the pawn to c4, aim rs squares, i.e. in this position it is ing to obtain the position in the necessary to hand over the move to previous diagram. The direct 9 liJe4 Black. This is achieved by the does not achieve its objective be endgame king manoeuvre, 'trian cause of 9 ... liJc7!, and he has to go gulation', characteristic of pawn and back, 10 liJc3, so as not to allow a indeed also of knight endgames. In draw in the variations l 0 liJxc5+ the present case, along the g6, II @e8! 11 b6 liJd5+ or 10 b6 liJd5+. and f6 squares. This method is ac It becomes clear that the manoeuvre tively exploited also later on. liJe4 is good when the king is situ 9 wn liJe6 10 @g6 liJd4 11 @f6 ated on fl and cannot immediately The objective is achieved! Now in be attacked. The solution to the the position shown in the previous position is again achieved by the diagram it is already Black to move. triangulating king manoeuvre Il...liJe6 12 @rs liJc5 13 liJd5 g6-fl-f6. liJd3 14 e6+ @d6 15 e7 Wd7 16 9'iti'g6 We7 We4 After 9 . ..liJd8, preventing the The c4 pawn is doomed. move to fl, the king breaks through Instead of lO... liJd4 there is no to the c-pawn: 10 @rs @e7 11 @e4 salvation in 10... @e7, on which We6 12 liJd5 followed by liJf4+ and possible are both 11 @rs liJc5 12 'it>d5. tlld5+ @If 13 b6 @e8 14 e6, and 11 The continuation 9 ... c4 l 0 @rs! liJd5+ We8 (1 l...Wd8 12 @fl liJc7 leads to a position looked at under 13 liJc3@d7 14 b6) 12 @f6 liJd4 13 the analysis of the first variation. b6. The continuation 9 . . liJd4 10 @fl And so 8 ... c4 loses. But what if liJe6 11 liJe4is looked at below (see the pawn is left on c5? moves 11-13). 10 liJd5+'iti'd7 II. 8 ... liJe6 On 10... We8 winning is 11 b6 liJd812 'iitf6 'iii'd7 13 e6+. 11 @t7liJd4 Losing at once is l 1 ...liJd8 12 @f6 liJe6 13 liJb6+. 12 liJc3 liJe6 13 liJe4liJd8+ The position arising after 13... c4 14 liJc3liJd4 15 @f6 is examined in the commentary to the continuation 8 ....c4. Also losing is 13... liJd4 be cause of 14 liJxc5+ @c7 15 liJe6+. 14 Wf6 c4 Black already cannot hold on to the pawn on c5: since in the event of Also here the immediate break 14.. . liJe6 winning is 15 b6 liJd8 16 through with the king to the c-pawn liJxc5+ @c6 17@e7. Knight Endings 37
15 lLic3 lLie6 16 'it>f5 , and White lLixg7 However he does not notice wins by continuing as in the analy this possibility. sis to Variation I. 6 ...tt:::l f4 ? 7 'it>e5lLi g6+ 8 'it>f5gxf6 The difficulties also encountered 9 gxf6 by the defending side with a flank pawn fonnation are well demon strated by the following example.
Kremenetsky-Razuvaev Moscow, 1981
How can he fight against the passed f6 pawn? Black cannot allow its further advance to fl since, with out support from the king, the knight will not be able to cope with it, e.g. 9 . ..tt:::lf8 10 fl 'it>h5 11 lLic5 'it>h6 12 'it>f6 ll'ih7+ 13 'it>e7 Wg7 14 White has a material and posi lLid7. The transfer of the knight to tional advantage, but his chances are this key point ends the struggle, based on the fonnation of a passed since the f8 square is indefensible. pawn which is inevitably linked to This is a typical knight manoeuvre exchanges and gives Black defens with a pawn one square from queen ive resources. ing. And so the knight is forced to 1 'it>e3 tt:Jd5+ 2 i;t>d4 lLif4 3 g5+ move to a passive position on h8. In Wh5 contrast to the variation given after More natural looks 3 ...fxg5 4 Black's 3rd move, Black's pieces hxg5 'it>h5 5 i;t>e5 'it>g4, creating de now have no room for manoeuvre. fensive possibilities after6 g6 lLid3+ 9 ...lLih8 10 tLld6 'it>h5 11 i;t>e6 7 i;t>e6 tt:Jf4+, and also upon 6 f6 �g6 12 'it>e7 gxf6 7 gxf6 lLig6+ 8 We6 tLlf4+ 9 'it>d6'it>f5 10 fl lbg6. 4 lLixf6+ 'iitxh4 After 4... gxf 6 5 gxf6 the white pawns are very dangerous. 5 lLie4 tLlh3 f66 ? On 6 g6 'it>h5 7 'it>e5 Whitemain tains the advantage, but now Black can momentarily exploit his own drawing resource, linked to a knight sacrifice to liquidate the pawns: 6 ...lLixg5 ! 7 fxg7 lLie6+ and 8 ... 38 Kn ight Endings
The black king has two squares at (b) 17 CiJe7�h6 18 CiJc6 @g5 19 its disposal: h5 and h7. Black played CiJe5 @f4 20 CiJfl CiJg6 21 ttJd6 12... 'it>hS ttJh8! 22 �e7 CiJg6+ 23 @fl CiJe5+ Why did he reject 12 ... 'it>h7? Let 24 �g7 'it>g4! (it is important to us look at a possible continuation of leave the file where the promotion the struggle. 12... �h7 13 CiJfl of the pawn will take place!) 25 CiJg6+ 14 �e8 �_g 8 15 CiJh6+ 'it>h7 CiJc4CiJc6 26 'it>g8 CiJd8. 16 CiJg4 CiJh8 17 tlle5 @g8 18 'it>e7, 17 CiJd4! and there arises a well-known posi The knight heads forthe key e5 tion from Cheron, presented in square! Y.Averbakh's monograph Compre 17 ..• @h6 18 CiJf3 tt:::lg6 hensive Chess Endings (2nd edition, On 18... 'it>h7 19 CiJe5 @g8 20 @e7 1980). This position is won in the arises the above-mentioned Cheron following way: 18... �h7 19 @f8 position. 'it>h6 20 'it>g8 'it>g5 21 �g7 �f5 22 19 'it>d7 'Llh8 20 CiJe5 'it>g5 21 CiJd7 CiJg6 23 fl 'it>g5 24 CiJe5 CiJf4 �e6 25 'it>g8 CiJe626 CiJD+and 27 CiJd4. Instead of 14 ...�g8, there is no saving himself either by 14... 'Llf 4, or 14... CiJM. For example: 14 ... CiJf4 14 CiJe5 CiJd5 16 fl CiJf6+ 17 'it>e7 CiJg8 18 'it>e6 'it>g7 19 CiJd7; or 14 ... CiJh4 15 CiJg5+'it>h6 16 fl CiJg6 17 CiJf3 'it>g7 18 CiJe5 CiJf8 19 'it>e7 ttJh7 20 CiJd7 'it>h8 21 CiJf8 (21 CiJf6 ttJf8!) 21...tLlg5 22CiJg6+. After 12... �h5 ! the king at the necessary moment can attack the pawn fromthe rear. This method of defence is presented in the same 21... 'it>h5? book by Y.Averbakh in examples Black cannot stand the tension 470 and 471. For example: and makes a mistake. With the only (a) 13 CiJfl CiJg6+ 14 @d6 (also move 21...'it>f4! he holds the posi insufficient is 14 'it>e8 because of tion. A.K.remenetsky intended the 14 ...'it>g4 15 CiJd6 CiJe5!=) 14 ... 'it>g4 continuation 22 ttJfl CiJg6 23 CiJd6 15 CiJe5+@f5 16 f7 �f6=; 'Llh8 (on 23 ...tt:Je5 winning is 24 (b) 13 'it>d7 'it>g5 14 �e6 @g6. 'it>d5) 24 'it>e7 tt:::lg6+ 25 @fl tt:::le5+ In the game followed: 26 'it>g7 �g4 27 ttJc4, so as on 13 'it>e6 'it>g6 14 'it>e7 'it>h5 15 27 ... CiJc6! to offer a draw, while on We6 'it>g6 16 tt:irs'it>gs the natural-looking 27 ...CiJD? there Great accuracy is required of is the beautiful win: 28 'it>h6! CiJg5 Black. Simpler is 16... 'it>h7, since 29 @g6 @f4 30 CiJd2! 'it>g431 lt.Je4 White does not manage to reach CiJD 32 ti:Jf2+�h4 33 CiJd3 CiJg5 34 Cheron's position. For example: CiJe5(zugzwang !) and then 35 fl. (a) 17 tllh4 'it>_g8 18 We7 CiJfl 19 22 'it>e7Black resigned. ti:Jf3 CiJh6 20 t'zJe5 CiJf5, defends On 22 ...�h4 winning is 23 @f8 successfully; 'it>g5 24 �g7 �f5 25 CiJd7. Knight Endings 3 9
The following example shows the possibilities of the weaker side, when he has compensation in the formof actively placed pieces.
Gufeld-Grigorian Daugavpils, 1979
For examEle: (a) 6 h3 8h6 7 ctJd4 'it>d6 8 @g3 �d5! 9 lt>h4 �e4 IO @g5 ct:Jf7+ 11 @f6 ctJd6 12 h4 ctJe8+ 13 @fl @xe3! 14 ctJe6ctJc7 !=; (b) 6 e4+ @c5 7 exf5 ctJxh2+ 8 �g3 ctJfl+ 9 lt>g4 �c6! 10 f6 @d7 11 ctJc7 ctJe3+ 12 �g5 ct:Jc4 and White has an extra pawn but ctJd6=. Black has sufficientcounterplay due Probably the maximum that White to the active positions of his pieces. can extract from the position is to Now White should secure the elas organise a passed pawn by 6 h3 ticity of his pawn chain by 1 h3, ctJh67 ctJd4 @d6 8 e4. retaining the possibility of the ex The attempt to obtain two con change gxf5 for an appropriate mo nected passed pawns, undertaken in ment. However, there followed: the game, is met surprisingly by an 1 gxf5 gxf5 2 ctJf3 ctJf6 energetic black counterattack, It is important to maintain the linked to the activity of the king. possibility of play in the centre and 6 ctJd4 ct:Jxh2+ 7 @g3 liJfl+ 8 on the king's flank. Passive is @f2@e4!! 2 ...ctJd6 3 @d3 h6 4 ctJd4 with a An apotheosis of the activity of wm. the king! On 9 ctJxf5 lLixe3! the 3 ctJd4 white pawns are liquidated. On 3 @d3 possible is 3 ...ctJd7 4 9 'itixfl @xe3 10 lLie6 �f3 11 ctJd4 ctJc5+. Wgl @g3 Draw! 3 ...@e4 4 ctJb5 @d5 The black king confidently ties Further activity by 4 ...ctJd5?? ends the knight down to the f4 pawn and in a sudden mate by 5 ctJd6! does not allow its own white col 5 �f3 ctJg4! league out of the cage. On 12 �hl The activity of the opponent's follows 12... h6! 13 @gl h5 and then pieces does not permit White to h4-h3-h2, stalemating the king. strengthen his position without ex Now let us look at an example of changing pawns. But this allows playing knight endings with the Black drawing chances linked to the presence of pawns on opposite possibility of sacrificing the knight flanks. In this case a decisive role is at an appropriate moment. often played by the organisation and 40 Knigh t Endings
energetic exploitation of a distant 6 ...�fS ? 7 lL\aS tLldS 8 b7 tLlb4+ passed pawn. 9 �e2! tLla6 10 'it>f3 And so White blockades the pawn Vladimirov-Novopashin and with a fine knight manoeuvre Vo lgodonsk, 1981 wins it. 10 ... tLlbS 11 h4 @es 12 tLlb3 tLlc6 13 tLlcS@fS 14 tLld3! An important finesse! If 14 tLld7 Black holds on by 14 ...tLld4+ 15 �£2 tLlc6 16 b8==\W lLlxb8 17 lLlxb8 'itig4. 14 ... tLld4 lS �f2 lLlc6 16 lLlb4! tLlb8 17 @f3 @es 18 lL\d3+ �d6 19 'it>xf4 �c6 20 @gs 'it>xb7 21 'itih6 tLlc6 22 'it>xh7
White's advantage lies in the possibility of quickly creating a passed pawn while retaining a good coordination of pieces which allows him to support the advance of this pawn and at the same time to con trol Black's activity on the kingside. However the peculiarities of the position are such that great accuracy is required from both sides for the fulfilmentof their plans. Now White This ending has theoretical sig can quickly organise the manoeuvre nificance. Will the black pieces 1 ©d4 tLldS 2 @cs tLlxf4 3 b5, but succeed in neutralising the h-pawn? he prefersproph ylaxis. 22 ... tLle7 23 hS 'itic624 'itig7�dS 1 tLle3gS ?! 2S @f6 tLlg8+ 26 @f7 tLlh6+ 27 A hasty decision, facilitating 'itig7lL\fS+ White's task. After l...h6 2 h4 g5! On 27 ...tLlg4 winning is 28 lL\£2!. Black's defence is far easier. 28 �f6 2 fxgS f4 3 tLlc4 @fS 4 bS axbS S axbS �xgS 6 b6 Let us pay attention to the differ ences in the possibilities for the two sides. The passed b-pawn distracts the black knight whereas both white pieces can attack the passed pawn. Nevertheless Black can put up stub b m resistance by breaking through � . with his king to the g4 square- 6 ...'itig4 L However he chooses another route-march for his king. Knight Endings 41
An instructive position! The black Now every move by Black loses: knight can control theh-pawn from (a) 29 ...lbe7+ 30 'it>g5 lbf5 31 the g4 and f5 squares, therefore to lLif2+rJie5 32 lbg4+ rJie6 (32 ...rJie4 advance it White must deflect the 33 lbe3!!+-) 33 rJig6 lbh4+ (or knight from these points. This task 33... lbe7+) 34 rJig7 'llf5+ 35 rJif8! can be fulfilled by the knight from Zugzwang! On any king move, win the e3 square: 28 ...lbh6 29 lLif2! ning is 36 lbe3! Leading to the same \tid6 30 'it>g6 lbg8 31 rJig7 lbe7 32 outcome is 30 ...lbg8 31 lbf2+ rJie5 ltJg4 �e6 33 lbe3. 32 lbg4+ We6 33 Wg6 lbe7+ 34 Black chooses the best plan, en rJig7 lbf5+ 35 'it>f8! . deavouring to break through with (b) 29 ...lbh4+ 30 rJig5 lbf3+ the king to the h-pawn and at the (30 ...lbf5 after 31 lLif2 leads to same time controlling the e3 square, Variation (a) above) 31 rJif6!, and and this attempt is rewarded. the h-pawn is unstoppable. 28 ... 'it>e4! 29 'it>g5 lbd6 30 lbf2+ But to conclude White missed the rJie5 31 Wg6 lt:\f5 32 lbd3+rJie4 33 win by playing 33 lbc1 ?. After lbcl? 'it>f4 34 lbe2+ �g4 35 lbg3 Black's 32nd move there arises the lbe7+36 Wf6 lLif5Drawn. position shown in the last diagram, The king attacks the pawn and de but with White to move. With the fends the knight, observing from the tried and tested 'triangulation' king h6 square. manoeuvre 33 'it>f6! 'llh6 34 rJig7 Is White's advantage in the last lbf5+ (34 ... lt:\g4 35 lLif2!) 35 'it>g6 diagram really insufficient for a he hands the move over to his oppo win? No way! After the game nent and wins. E. Vladimirov demonstrated a clear Returning to the position shown way to realise the passed pawn. In in the above diagram, it is useful to stead of 29 'it>g5 he gave 29 'it>g6! ! bring to mind a fragment from the as leading immediately to victory. game Botvinnik-Simagin (Moscow 1955), where, by exploiting his opponent's inaccurate play, M.Bot vinnik was able to save himself in a similar ending.
Botvinnik-Simagin Moscow, 1955
A picture of a position, having an instructive character! White's pieces are clearly fulfilling their functions: the king confidently squares up to the enemy knight, while the white knight remains triumphant on the d3 square. 42 Kn ight Endings
Leading to a win here is 1.. .@f5 2 1 'Lle4 @es2 c,t>d6! tiJf3g5 ltJ 3 'Llh4+ c,t>g4 4 'Llg6 tt:Jf3 After 2 'Llxg5 'Llf2 White's task is 5 'it>b4 @g5. However there more complicated. followed: 2 .•. @f8 3 'Llxc5 'Llf2 4 'Lld7+ @f7 1 ...'Lld2? 2 'iitb4 @f5 3 @c3! 5 ltJe5+ @f6 With tempo the king draws closer And so White creates a distant to the centre of events and now if passed pawn. He combines its ad 3 ... 'it>xe5 4 'it>xd2 he succeeds in vance with fine manoeuvring of containing the pawn. pieces, creating on the way threats 3 ... 'Lle4+ 4 �d4 'Llg5 5 'Lld3 c,t>g4 to the g5 pawn. 6 'Lle5+ 'it>f5 7 'Lld3 @g4 8 'Lle5+ 6 'it>d5 'Lid1 7 c5 'Llc3+ 8 @c4 @g3 9 'Llg6! 'Lle4 9 'Lld3! 'it>e6 10 'it>d4 'Llf6 11 In contrast to the previous 'Llf2 @e7 12 c6 'it>e6 example White has this defensive On 12 ...'it>d6 winning is 13 'Lle4+. resource, since the pawn finds itself 13 @c5 @e7 14 @b6 @d6 one move further from the queening square. 9 ... 'Lle6 10 @e3 'Llf8 11 'Llxf8 h4 12 'Lle6 h3 13 'Llg5Drawn.
Let us look at some more examples of a struggle with pawns on diffe rent flanks.
Vasiukov-Timoschenko Vo lgodonsk, 1981
How can he realise the extra pawn? Insufficent is the natural 15 'it>b7 'Lle8 16 'Lle4+ @e5 17 'Llxg5 @f4 18 'Llh7 'Lld6+! (for the present the g4 pawn is untouchable because of 18... @xg4 19 'Llf6+ 'Llxf6 20 c7, and Black does not succeed in play ing ...'Lle8 since the pawn queens with check) 19 'it>c7 'Llf7 (also poss ible is 19 ...'Llb5+ 20 Wb6 'Lld6) 20 g5 @f5!=. On the g5 square the In this position White can quickly pawn becomes vulnerable. convert his positional advantage to a However E.Vasyukov finds an el material one but playing this ending egant manoeuvre and wins the g5 is complicated due to the fact that pawn while the black knight oc the pawns are situated close to one cupies the d5 square-which is another which might allow Black to rather poor for the struggle against regroup ,his forces successfully for the c-eawn. defence. With fine play White pre 15 t'llh3! 'Lld5+ 16 'it>b7 'it>e5 17 vents this possibility. 'Llxg5 @f4 18 tLlh7 Knigh t Endings 43
In this lies the main point of the The main thing is to organise a refined manoeuvre begun with the distant passed pawn and support its move 15 ctJh3! . In contrast to the advance. The white knight is in no position looked at in the previous position to struggle against the note, the pawn remains on g4 with whole of Black's position. Black to move-and he is in zug 4 �e4 tt:lc5+ 5 �f5 ctJxb3 6 ctJb7 zwang. On 18... �e4 or 18... 'it>e5 Alas, on 6 tt:lc6fo llows 6 ...ctJd4+. follows 19 g5 'it>f5 20 ctJf6 tt:le7 21 6 .•.ctJd4+ 7 'it>xg5 tt:le6+! c7 'it>xg5 22 ctJd5 ctJf5 23 'it>c6! and It is important to restrict the white 24 c8=iV. This same manoeuvre knight; now it is deprived of the wins on 18... tt:le7 -19 c7 'it>xg4 20 important c5 square. tt:lf6+ 'it>g5 21 ctJd5 On 18 ...'it>xg4 8 'it>f6 b3 9 ctJd6 b2 10 'it>xe6 decisive is 19 ctJf6+ tt:lxf6 20 c7 bl='ii' 11 \t>f6 'ii'b6 White tt:le8 21 c8='ii'+. resigned. In the game followed ... 18... tt:lc3 19 g5 �f5 20 c7 and Let us look at another interesting Black resigned. ending, demonstrating the import A very instructive ending to the ance of active defence. game. Nikolaevsky-Gufeld Kochiev-Lerner Kiev, 1951 Beltsy, 1981
After 1. .. tt:la5 2 b4 cxb4 3 axb4 In contrast to the previous tt:lxc4 White gradually loses. How example Black's positional advan ever analysis shows that he is not tage is felt less in view of the more exploiting his defensive possibi active position of the white pieces. lities. After 2 tt:lcl g6 the first im With energetic play, 1 tt:le5+ 'it>c3 2 pression is that White gets into ctJI! or 1 b4 'it>c3 2 'it>e3, White zugzwang since 3 �e5 @xg5 4 'it>d5 could create defensive chances. loses because of 4 ...'it>f 4! 5 b4 However if he deprives his pieces of 'it>e3! ! , and the pawn cannot be activity, Black achieves a decisive stopped. But nevertheless White advantage. findsa saving resource: 1 tt:la7? b4 2 tt:lc6 �c3 3 tt:la5 3 tt:le2!! tt:lxb3 4 tt:lg3+ @h4 5 ctJd7 ctJf5+! 44 Knigh t Endings
And Black is forced to reconcile A position which, despite its ap himself to a draw, 5 ...�h5 6 ttJg3+, parent simplicity, is quite insidious. since he would risk defeat afterboth Black's pawn weaknesses require 5 ...'it>h3 6 lLie7, and 5 ...gxf5 6 g6 due attention from him. The cardi ttJd4 7 'it>e5! ttJc6+ 8 @d6 ttJd8 9 nal decision in the position is the 'it>d7. exchange of knights, which requires accurate calculation. For example, With passed pawns on opposite after l...ttJe6+! 2 lLixe6 'it>xe6 3 g5 flanks, as in pawn endings, the fxg5 4 hxg5 'it>d6 5 '\t>g3 �e5 6 f4 + strongest side might win even with �f5 6 'it>f3 d4 the chances are even. out the participation of the king. An In the game, however, there interesting example is from the followed ... game ... 1... h6? 2 hS gxhS? The only chance of saving the Azmaiparashvili-Novopashin game remains with 2 . . . 'it>f7 3 hxg6+ Volgodonsk, 1981 'it>xg6 4 f3 ttJd7 5 ttJf5 h5 6 ttJe7+ @fJ, and White's advantage is not so tangible. 3 ttJfS+ 'it>e6 4 gxhS lLie4 S f3 ttJc3 6 e3 ttJe2+ 7 'it>g4 �es 8 ttJxh6 d4 9 ttJf7+ 'it>e6 10 ttJd8+ @d7 11 exd4! and White won.
Godena-Lalic Portoroz, 1998
1 as �f6 2 f4 \¥tg7 3 rs 'it>gs 4 ttJd3 ttJa6 S ttJeS ttJb4 6f6 c4 7 h6, and Black lost on time.
Van der Sterren-Douven Netherlands, 1985
After 1. ..ll'ld3 ! 2 b3 'it>dS the black king cuts offthe enemy knight from the centre. 3 lllas 'it>e4 4ttJc6es s h4 White presents his opponent with good chances. Correct was 5 b4 ! �d5 6 b5 e4 7 ttJa7! �c5 8 ttJc8 Knight Endings 45
@xb5 9 tlld6+ ..t>c5 10 tllxe4+ Wd4 Barlov-AbramoviC 11 tllg5 with a clear advantage. Yugoslavia, 1989 5 .....t>d5 A critical position has arisen. 6 tlla5 e4 7 g3 Again better was 7 tllc4 tllc 1 8 tlld2, exchanging the b- for the e-pawn. 7 ...@d4 8 tllc4 tllcl 9 tlla5 tlld3 10 @e2 tllcl+ 11 @d2 tlld3 12 @e2 And White reconciled himself to a draw.
Polnareva-Akhsharumova Moscow, 1984 1. .. tlle7+ 2 Wf6 tllc8 3 Wg5 h6+! 4 c;t>h4 If 4 Wxh6 @g4. 4 ...c;t>e4 5 tllg7 @es 6 ..t>h5 @f6 7 @xh6 tlld6 Drawn. White is not able to break the blockade surrounding him.
Vyzhmanavin-Chiburdanidze USSR (ch), 1984
At first sight it seems that a com plicated and long struggle is in pros pect, but Black, exploiting the bad position of the white king, quickly achieves victory. 1 tlle5 f5 2 tlld7 @f'l3 tllc5 We7 4 tlla6 c;t>d6 5 tllb4 e5 6 tlld3 @d5 7 tllb4+ c;t>e4 8 tlla6 c;t>d3 9 tllb4+ ..t> e2 10 .!lJc6 tllf3 11 .!lJe 7 .!lJe 1+ 12 White has the advantage, but the ..t>gl tlld3 13 tllxf5 tllxf2 14 @g2 limited amount of material left com e4 15 tlld4+ c;t>d3! and White plicates the task. White's plan is resigned. linked to a breakthrough with the king to the queen's flank, since In the next game, White has a ma there is nothing in 1 tllb6 tlla5 2 terial advantage but Black finds an @d4 because of 2 ...tll c6+!. interesting possibility to save 1 @e3 c;t>g6 2 @d2 @g7 3 Wc2 himself. 'iitg6 4 @b2 tlle7 46 Knight Endings
He cannot wait any longer, the c4 27 ... 'i¥te5 28 'i¥te3! pawn cannot be held. To exploit the The last finesse: if 28 'llxg5, then remoteness of the white king, Black 28 . .. 'llf2 ! 29 'i¥txf2 'i¥tf4. strives to simplifythe position. 28... 'it>d5 29 'llf6+ 'it>e5 30 'llh5 5 lbb6 f5 6 exf6 'it>xf6 7 'llxc4 Black resigned. 'llg6 8 'i¥tc2 lDh4 He loses the knight, without man In the pawn ending-8 ...'ll e5 9 aging to capture the g4 pawn in 'llxe5 'it>xe5-White gains victory return. by 10 'it>d34 '\t>f 11 c4 e5 12 c5 'it.?£3 13 c6 e4+ 14 'it>d4 e3 15 c7 e2 16 An instructive and complicated c8='i¥ el='fl' 17 'fl'f5+ Wg2 18 knight ending, where the advantage 'iVe4+ 'i¥xe4 19 'it>xe4. of one of the sides lies only in a 9 'it>d3 ctJf3 10 h3 'llgl 11 'lld2 slightly better structure, is encoun 'i¥te5 12 'it>e3 'llxh3 13 c4 tered in the game ... With this move White rejected Black's offer of a draw. In fact, if Timm an-Ree 13 ...'ll f4 White retains the advan Netherlands, tage by 14 tbf3+ or 14 Clie4. Also 1984 other defencesare no help. 13•.• 'll gl 14 'lle4 'llh3 15 'llc5 'it>d6 16 ctJd3 e5 17 'i¥te4 'it>e6 18 'llxe5 'llt1+19 c;t>e3! 19 'it>d4 'it>d6 20 c5+ 'it>e6 21 c6 'it>d6leads to a draw. 19 .•.'ll dl+ 20 'i¥td4 ctJt1 21 c5 ctJhl 22 c6 'i¥td6 23 c;t>e3 'llg3 24 'llc4 'it>xc6 25 tbd2 'i¥td5 26 'it>f3 ctJhl
1... tbc5 The more obvious l ... e5 looks stronger. For example: 2 'i¥te2 h6 3 tbe4 b6 4 'i¥td3 'i¥tf7 5 'it>c4 'it>e6 6 'it>b5 'llb8, and it is difficult for White to obtain real chances of a wm. 2 b4 'lla6 3 a3 'llc7 4 'it>e2 h6 His defensive problems are not solved by 4 ...'ll b5 5 a4 Cllc3+ 6 'it>d3 In this apparently arid desert a lllxa4 7 'llxe6 and Black has win can still be achieved due to the difficulty involving his knight in the poor _eositionof the black knight. future play. 27 tt:'le4 5 Clle4 'it>f8 But not 27 'llfl because of More circumspect was 5 ...b6 6 27 ... 'it>d4! 28 ctJh2 'it>d3 29 'it>g2 'lld6 a6. 'i¥te3 30 'it>xhl @f2 with a draw. 6 ctJd6b6 7 'i¥td3 Knight Endings 47
Accurate calculation was required 21 ... tl:\xh4 22 tl:\xh5 'it>c6 23 tl:\g3 for the initiative-seeking 7 tl'ic8!. rJld5 24 a4 b5 25 a5 'it>c4 26 tl:\f5 For example: 7 ...tl:\b5 8 a4 tl:\c3 9 tl:\g2 27 'it>e5 Black resigned. 'it>d3 tl:\xa4 10 tl:\xa7 tl:\b2 11 'it>d4 e5+ 12 @e4 tl:\dl 13 f3 tl:\f2+ 14 The difference between pawn and @d5 tl:\d3 15 b5 tl:\f4+ 16 'it>e4 knight endings is seen when there tl:\xg2 17 tl:\c8, and White's position are passed pawns on opposite is close to a win. flanks. For example, the game ... 7 ...a6 8 tl:\c4 tl:\d5 9 'it>d4 rJle7 10 g3 'it>d7 11 f4 'it>c6 12 'ite5 Vukovic-Eingorn Belgrade 1987
12... tl:\c7 White should continue 1 'it>g2! h4 Black chooses passive defence 2 tl:\d4 'it>e4 3tl:\e6 'it>e5 4tl:\g5, and and wrongly so. In his comments to the outcome of the struggle is still this game Timman wrote that not quite clear, since it is difficult 12... 'it>b5 13 tl:\b2 tl:\xb4 14 axb4 for Black to advance both pawns to 'it>xb4 15 'it>xe6 a5 16 @fl a4 17 the third rank. But in the game tl:\xa4 was losing. Considerably followed... stronger, however, is 16 ...'it>b3 ! , 1 tl:\d4'it>e3! 2 tl:\f5+ 'it>f2 3 lt.Jg3 afterwhich 17 'it>xg7 'it>xb2 18 f5 a4 After 3 tl:\h4 g3+ 4 'it>h1 the might lead to a queen ending with pawns are frozen but the poor posi an h-pawn, where White's chances tion of the white king is decisive---4 of winning are problematical. ... tl:\e3 5 b6 @fl 6 b7 lt.Jg4 7 b8='i¥ 13 tl:\d6'it>d7 14 f5 exf5 15 tl:\xf5 tt.Jf2 mate. With simple and convincing 3 ...h4 4 tl:\e4+'it>e3 5 tl:\d6 moves White has succeeded in in On 5 tl:\g5g3+, with the following creasing his advantage and placing interesting variations: Black in a zugzwang position. (a) 6 'it>g2 (6 'it>h3 tl:\f4+ 7 'it>xh4 15... tl:\es 16 g4 lt.Jf6 17 h3 g2) 6 ... 'it>f4 7 tl:\h3+ rJlg4 8 'it>gl 17 lt.Jxh6'it>c6 18 g5 tl:\d7 19 'it>e6 tl'if4+ (8 ...tl:\e3+ 9 'it>hl h3? 10 is more quickly decisive. tl:\xh3=) 9 'it>hl h3 10 b6 h2 11 b7 17... h5 18 g5 tl:\h7 19 h4 tl:\f8 20 tl:\h3,and mate in two moves. tl:\xg7 tl:\g6 21 @f6 (b) 6 'it>g 1 'it>f4 7 tl:\h3+ rJlg4 8 21 'it>d5 is also sufficientto win. 'it>g2 tl:\e3+ 9'it>gl 'it>xh3 10 b6 tl:\g4 48 Knight Endings
11 b7 llle5 12 b8=� Clif3+ 13 �fl Of course, according to Botvin g2+ 14 �e2 gl=� 15 �xf3 'Yi'g3+. nik, playing a knight endgame, is 5 ... g3+ 6 'i¥tgl like playing a pawn ending. White's Forced. If 6 �g2 Cllf4 7 �g 1 then plan consists of advancing the first 7 ...�f3 is decisive. pawn with help of the king, after 6 ...h3 7 CZlfS+ �f4 8 CZld4 lt>g4 9 which Black gradually lands in llle2 zugzwang. The threat was 9 ...h2 10 �h 1 1 Cllf4 CZld6 2 g4+ 'it>gS 3 Cllh3+ 'it>h3. Now, however, on this could �g6 follow 11 Cllxg3 'iii>xg3 12 b6 with a If 3 ...lt>h4, then 4 'it>f4 with the draw. In order to achieve victory irresistible threat of 5 Cllf2 and 6 g3 Black must broaden the range of ac mate. tivity for his knight. 4 �g3 lllc4 9 ... Cllb6 10 �fl lllc4 11 �gl Attempting to create counterplay Clld2 by means of an attack on the pawn. White resigned, without waiting 5 lt>h4 llle3 6 Cllf4+ lt>h6 7 gS+ for Black to promote to a queen in 'it>g7 8 'it>hSCZlfS 9 tiJdS! the variation 12 b6 Clif3+ 13 lt>fl Inferior was 9 g6 in view of g2+ 14 lt>f2gl ='Yi'+ 15 Cllxgl h2. 9 ...Cll g3+! 10 lt>g5 Clle4+ 11 \t>f5 Cllg3+, preventing White attacking The comparison between knight the coordinated forces. and pawn endings is borne out with 9 •..Cll g3+ the exploitation of zugzwang in the On 9 ...lll d6 or 9 ...Clld4 would two followingendgames. have followed 10 Clle3 ! , and Black Sajtar-Benko is forced into a worse position. Budapest, 1954 10 'it>g4 Cll fl If 1 O ...Cll e4 11 �f4 Clld6 decisive is 12 Clle7!. 11 'it>f4! The main principle of the end game is not to rush! White restricts the activity of the black knight, which, in order to get into play, has to go to d2 or h2, and then the white king finds itself in a very favourable position-safe from the black knight along the diagonal. 11 ...Clld2 12 Clle3! Still more restriction of the black If in this position we exchange knight. knights forbishops or rooks then the 12 . ..@g6 13 CZlfS Cllfl 14 Cllh4+ game would be drawn. It is interest 'it>g7 15 �g4! 'it>h7 ing that Reshevsky managed to win 15 ... Clle3+ 16 'it>h5. a queen ending with such an align 16 g6+ 'i¥th6 ment of forces against Geller, Inter 16... 'it>g7 17 �g5. zonal tournament, Sousse 1967, 17 'iitf4 ! and, with c-pawns, Mikhalchishin Zugzwang in action. Kasparov, USSR (ch) 1978. 11 ... llld2 1s @rs lt>g7 Knight Endings 49
18... 'Lic4 19 'it>f6. 29 ... 'Lic4 19 g4! In this case White wins in the fol Zugzwang in tum. It should be lowing instructive way. mentioned that White does not 30 @f6 @g8 hurry with the advance of his re 30 ...'Lid6 31 We7 and then 32 serve pawn, since this cuts off @f8. squares from the king. 31 'it>g6 'Lie5+ 32 @f5 'Lic4 33 19... 'it>g8 @f6 On a move of the knight follows The familiar 'triangulation'. 20 'it>g5, while on 19... 'it>h6-20 33 ...'it>h7 34 'Lig3! 'Lig2 'it>g721 'Lif4. A new reconstruction. Now bad is 20 'it>g5 'Lie4+21 'it>f4! 34 ...'it>g8 because of 35 'Lif5 and 36 21 'it>h6'Lif 6. 'Lie7, and also 34 ...'Lie3 because of 21... 'Lic5 22 'Lif5 35 'it>f7 Step by step White creates a very 34 ... 'Lid6 35 ctJf5 'Lie4+ strong position, and Black must 35 ... 'Lie8+ 36Wf7. continually watch the threat of a 36 @f7'Lig5 37 'it>e7! 'Lie4 king infiltrationto f6 or h6 37 ...'it>g8 38 'it>f6 'Lie4+ 39
the same idea was utilised-only What else can he do? He must White probably defended in weaker give up a pawn-though it was still fashion. possible to attempt a breakthrough 1 ...ClJb6 2 CZJc6 ClJd7 3 0,e7 ClJf6+ with the king to the d8 square, but 4 c;f;>h4 gs+ s '\¥i>g3 g6 6 CZJc6WhS 7 then, with the king on e7, the knight llJd4 llle4+ 8 Wf3 tLld6 9 c;f;>g2 llJfS from e5 goes to d3, and White is 10 CZJe6 g4 11 ClJf4+ WgS 12 CZJe6+ again in a blind alley. c;f;>h4 13 tl:lf4 CZJe7 14 CZJe2 llJdS lS 6 ...@xe7 7 @cs @f6 8 c;f;>dS llJd3 Wh2 gS 16 CZJg3 ClJe3 17 CZJe4 ClJfS 9 CZJc6 ClJel 10 'itie4 ClJg2 11 ClJd4 18 c;f;>g2 g3 19 0,f6 CZJe3 20 c;f;>gl CZJe3 12 ClJfS CZJc4 13 c;f;>d3 CZJeS+ 14 CZJg4 21 llJh7 CZJes 22 'i¥i>g2 QJd3 23 'itie2c;f;>g6 lS 'i¥i>f2 CZJc416 Wg2 c;f;>f6 @gl CZJf4 24 Whl g2+ 2S c;f;>h2 g4 17 h3 tl:leS 18 ClJd6 '\¥i>g6 19 CZJe4 26 llJf6 g3+ 27 c;f;>gl c;f;>h3 White c;f;>h6 20 'i¥i>f2 'i¥i>g6 21 'i¥i>e2 Wh6 22 resigned. CZJcs'i¥i>g6?
It is interesting that in both examples the stronger side had doubled pawns on the g-file. But what will be the case if the pawns are situated on another file? To us it seems that the weaker side will make a draw only with pawns on the edge file, since then the king of the stronger side has no exit.
Van Wely-Adams Groningen, 1997
Black must have the possibility, on CZJd3, to reply.... ClJg6, defending the f4 pawn. 23 h4? Correct was 23 CZJd3! ClJc4 24 h4, obtaining a winning position. The transposition of moves gives Black the possibility of saving himself. 23 .•• gxh4 24 CZJe6 h3! 2S CZJxf4+ @gs 26 CZJxh3+'it>h4 27 llJf2 c;f;>g3! The Black king breaks through to the white pawns and a draw is 1 ClJd4? inevitable. Correct was the transferto a pawn 28 gS CZJxf3 29 g6 ClJd4+30 c;f;>d3 ending by 1 llJd6! '\¥i>e6 2 ClJe4 tl:lxe4 CZJe6 31 We3 Drawn. 3 fxe4 �xe5 4 @f3 Wf6 5 h4 ! gxh4 6 'it>xf4, and then the king travels to An interesting ending arose in the h3. . following game where despite, the 1...ClJd7 2 e6 ClJeS 3 'i¥i>d2 'it>d6 4 approximate equality, the struggle 'i¥i>c3'i¥i>e7 S 'it>b4Wd6 6 e7 was still not over. Knight Endings 51
lvanchuk-Eingorn exploitation of this advantage USSR (ch), 1988 requires filigree technique. 1 f3 'lld7 ,_._.,, Neither now, nor later is there any � � � 0 _eossibility of l...c5 since then 2 . tbb5 a6 3 ctJd6 @e7 4 ctJb7 'lld7, and the white king goes to the centre followed by a3 and b4. •� •!m %�•.�� n•�-�� 2 �f2 f5 3 @e3 �e7 4 b4 e5 5 a4 �,��o,, .v, � .� @d6 6 @d3 tllf6 7 c5+ We6 . . .� Better is 7 ...�c7 8 Wc4 a6. ,,.,,v, � ��·� �. 8 b5! �d7. 9 �c4 Wc7 . .0 v, 0 Bad is 9 .. e4 10 @d4. � �� �,,,,, � 10 a5 a6! The threat was a6, bxc6 and ctJb5. 1 f3 h4 2 b5 a5 3 'llgl �d6 11 b6+ More active is 3 ...@f 4 provoking After 11 bxa6 follows 1 l...@b8 _elay such as 4 @d5 'llb6+ 5 @c6 12 ctJa2 ct:Jd5 and 13... Wa 7. tLla4 6 b6 ctJxb6 7 c,t>xb6 @g3 8 11...@b7 12 g3 h5? @xa5 c,t>g2 9 'lle2 @xf3 10 tllg 1 + He should not freeze his structure. �g2 11 'lle2 with a draw. Correct is 12... 'll d7. 4 �d4 tllc5 5 ctJh3 ctJe6+ 6 @c4 13 h4! ctJd714 f4 ! @c7 7 @dS @b6 8 @d6 @xb5 9 f4 Closing the way for the king. �a4 10 c,t>e7 �xa3 11 @xf7 'lld4 14... exf4 15 gxf4 tllf8 16 'lle2 Foreseeing 12 @xg6 a4 13 f5 'llg6 17 'lld4 tllxh4 ctJxf5 14 @xf5 @b2 15 g6 a3 16 g7 After 17 ...'llxf4 18 tllxf5 g6 19 a2 17 g8='fl' al ='iii', the rivals agreed 'lle7 Black is in zugzwang. to a draw. 18 'lle6! c,t>cs To defend against 'lld8. Sveshnikov-Sokolov 19 tllxg7 ct:Jg6 20 tllxh5 Black Moscow, 1991 resigned.
Torre-Portisch To luca, 1982
White has the advantage on ac count of his majority on the queen side and greater space, but the 52 Kn ight Endings
The endgame appears very diffi Ilincic-Abramovic cult for White but, with a pawn Ti vat, 1995 sacrifice, he activates his king. 1 b6! axb6 2 ttJc4 bS 3 ttJe3+ @cs 4 fS ! Weaker is 4 ttJg4 ttJe6 5 �e3 ttJc7! 6 ttJf6 ttJd5, and the pawn ending is hopeless for White. With the sacrifice of yet another pawn White creates his own passed pawn, which is a principal factor in this endgame. 4 ... gxrs s h4 f4 6 lLig4 rs 7 ttJh2 lLic6 8 hS ttJeS+ 9 �e2 b4 10 h6 b3 11 h7 lLig6 12 ttJf3b2 13 ttJd2 @d4 14 @f3 �c3 lS ttJbl+ It looks like White has some prob Now Black cannot win the knight lems converting his passed d-pawn on bl because after h7-h8, ttJxh8, to a win, but really it is very simple. the king captures both black pawns. 1. ..g6 2 @d4 'it>f73 ttJcS! He has to try and go with the king to Speculating on a transferto a win the h7 pawn. ning pawn endgame, White clears 1S... 'it>d3 16 'it>t1 'it>e4 17 ttJd2+ the way for his pawn and king. 'it>dS 18 'it>e2 'it>e6 19 �d3 �f6 20 3 ..•ttJb6 4 d6 '\t>f6 S d7 �e7 6 'it>c2 'it>g7 21 ttJf3 �xh7 22 'it>xb2 �eS hS 'it>h6 Zugzwang-after 6 ...ttJc4+ 7 �d5 ttJa5 (preventing 'it>c6) 8 g4 g5 9 '\ties the white king penetrates one flank or another. 7 'it>f4 ttJc48 'it>gS ttJeS 9 h4! And because of zugzwang White wins yet another pawn. Black resigned.
Maric-Zaitseva Tivat, 1995
It is interesting to compare this ending with the two following ones, where doubled pawns were success fullyrealised. 23 'it>c2 'it>hS 23 'it>d2 'it>g4 24 ttJd4 lLieS 26 �e2 ttJf7 27 'it>fl ttJgS 26 �t1ttJe4+ 29 'it>g2 ttJd2 And Black did not manage to real ise his material advantage. Knight Endings 53
In many cases 4:3 on one flank is Usually such endgames with a winning for the stronger side, but distant passed pawn give great win here it isn't so clear. ning chances. 1 'it>f3 f6 1 ctJeS More or less necessary-White's Black has a compact pawn struc plan was 'Lld4, g2-g4 and f4 -f5 ture and White tries to create some creating a weakness on e6. weaknesses so he can penetrate with 2'it>e3 his king, taking advantage of the Playable was 2 ctJd4. fact that Black must spend time to 2 ...'LlbS 3 g3 'it>e8 4 'it>d3 'it>d7 S win the a-pawn. 'Llb4 fxeS? 1...f6 2 'Llc4 'it>b8 3 @e2 'it>a7 4 Clearly better would be 5 ...'Lla7 6 @e3 'it>xa6 S'it>e4 tf:ic7 6 'Lle3 'it>bS 'it>e3 ctJc6 7 ctJd3 'it>e7 with the idea 7 ctJdS 8 ...g5 and transfer of the king via f7-g6to f5 . 6 fxeS 'it>c7 7 <;¥;>e3 'Lla7 8 ctJd3 'Llc6 9 h4 g6 10 'LlcS ctJd8 Otherwise after h4-h5 Black's structure would be completely blocked. 11 g4 h6 12 gS hxgS 13 hxgS 'it>c8 14 @d4 <;¥;>c7 lS 'it>d3 �b6 16 'Lld7+�c6 17 'Llf8�cs 18 'Lld7+! Rather dangerous was 18 'Llxg6 ctJf7 when the e5 pawn is quite weak. 18 ... <;¥;>bs 19 'Llf8'it>b4 20 'Llxg6! 7 ...'Lle6 Now is the time. The pawn endgame after 7 ...�c6 20... 'Llf7 21 'Llf4 'LlxgS 22 'LlxdS! 8 'Llxc7 'it>xc7 9 @d5 �d7 10 f4 Draw. <;¥;>e7 11 f5 would be too dangerous for Black. 8 ctJxf6! Alexandria-Marie White has no other way of playing Tivat, 1995 for the win. 8 ... gxf6 9 @rs 'Llcs 1 o f4 The direct ap12roach was 10 �xf6 'Lle4+ 11 <;¥;>g7 lllxf2 12 �xh7 <;¥;>c5 13 h4 <;¥;>d5 14 h5 �e5 15 �g6 (15 h6 'Llg4 � 16... 'Llxh6=) 15... 'Llg4 16 'it>g5 lllf6 17 h6 @e6 18 �g6 <;¥;>e7 19 h7 ! gave White winning chances. But better would be 14 ...'it>e4 15 'it>g6'Llg4 16 'it>g5 'Lle5 17 h6 ct:Jf7+ and l 8 ...ctJxh6=. 10 •.. 'Lld3 11 g3 'it>c6 12 'it>xf6 'it>d6 13rs Or 13 'it>g7 'it>e6 14 'it>xh7 'it>f5 15 h4! �f6 ! 16 h5 ctJf2 ! 17 h6 ct:Jg4=. 54 Kn ight Endings
13... lLieS 14 h3 After 14 @g7 h5 15 'it>h6 'it>e7 16 'it>xh5 @f6 17 g4 'Llf3 18 h4 lLie5 the draw is obvious. 14 •..lLid7+ 15 'it>gs @es 16 g4 lLif6 17 h4 Or 17 'it>h6 Wf4 18 @g7 @e5 19 @fl h5 20 g5lLie4 21 g6 @xf5 22 g7 lLif6=. 17 •..h6+! 18 @xh6 lLixg4+ 19 'it>g5 lLif6 20 @g6 lLid5Drawn.
9 ...@c6? Sermek-Hulak Slovenia, The only chance was 9 ... g5 ! 10 g4 I 995 ( 10 'it>g6 g4 11 Wg5 'it>c6 12 �xg4 'it>d6=) 10... 'it>c6 11 @g6 'it>d6 12 @xg5 'it>e7 13 'it>g6 'Llc3 14 @h7 (14 f4 lLid5 15 f5 @f8=) 14 ... @f6 15 f4 lLie4 16 'iith6 'it>flwith a probable draw. After the move in the game White's task is easy. 10 @xg6 'it>d7 11 g4 'it>e6 12 g5 lLid6 13 �h7 lLie4 14 g6 lLif6+ 15 @h8 rl;e7 16 f4 @f8 17 g7+ @f718 g4 lLig8 19g5 Black resigned.
Beliavsky-Tratar This is a very similar endgame to Bled, 1996 the previous one but here Black's pawn structure is weaker. If now 1...lLid5 2 lLidl with the idea 2 . . . 'it>xa4?3 lLic3 +-. 1...@b3 2 a5! And here White is forced to sacri fice a piece-if 2 lLidl then 2 ... 'Llxa43 @e3 lLic5 4 'it>d4 'it>b4. 2 ... @xb2 3 a6 lLib5 4 @e3 Wc3 5 'it>xe4 @b4 If the black king tries to go forthe white pawns by 5 ...@d2 then White creates second passed pawn by 6 g4 ! We2 7 f4 @f2 8 f5 +-. White's king is much more active 6 We5 Wa5 7 @f6 @xa6 8 @g7 than Black's and this gives him @b6 . chances to make progress. Or 8 . . lLic3 9 'it>xh7 lLie4 10 f4 1 @c6 We7 2 'iitc7 lLia6+ 3 @b6 lLixg3 ,1 1 Wxg6 +-. lLic5 4 'it>c6! lLib3 5 lLixa4 lLid4+ 6 9@xh7 'it>b7! lLib5 7 lLib6 lLic3? Knight Endings 55
The way to the draw wasn't easy: Black's task is to create yet 7 ...ltJx a3 8 Wc6 ltJbl 9 ltJc8+ 'iiff6 another passed pawn. And so there IO ltJxd6 ltJc3 ( 10... ltJd2 11 'it>d7 followed... 'itig5 12 'it'e6 'it'f4 13 'it'f6 ltJb3 14 1...fxg4 2 hxg4 h5 ltJb7! 'it'xe4 15 d6 ltJd4 16 d7 ltJc6 Now White has no time to winthe 17 ltJa5!+-) 11 'it'd7 'it>g5 12 'it>e6 d3 pawn. \t>f4 13 @f6 ltJa4 14 ltJb5'it'xe4 15 3 g5+ @g7! d6 lt)b6 16 ltJc3+ 'it'd4 17 ltJa4 The king must retreat, otherwise ltJd7+ 18 'it>e7ltJb8=. the white knight becomes highly ac- 8 a4 ltJxe49 a5 ltJc5+ 10 'it'c7 tive: 3 .... 'it>e6 4 ltJd4+ 'it>d5 5 ltJf3. Wrong was 10 'it>c6? e4 11 ltJc4 On 5 ... lt'c4 6 f5 gxf5 7 g6 'it'c3 lit'd8! 12 'it>d6e3! 13 ltJxe3ltJb7 +=. possible is 8 'it>f4!. If 3 ... @f5, then4 IO..• ltJa6+ tlJe5. After 10... e4 11 ltJc4 @f6 12 4 ltJd4 h4 'it>xd6 e3 (12... ltJb7+ 13 'it>c7 ltJxa5 Necessary in view of the threat of 14 d6!) 13 ltJxe3 ltJb7+ 14 'it>c7 5 ltJf3. ltJxa5 15 d6 'it'e6 16 d7 ltJb7 17 s rs h3 ltJd5. After this White himelf obtains a 11 @c8! protected passed pawn. Once again 11 'it>c6?was wrong, 6 f6+ 'it>f7 because of 11.. .. e4 12 ltJc4 ltJb4+=; If 6 ...@f8, then 7 lbf3 lbe4 8 11 Wb 7 ltJc5+=. 'it'xd3 lbxg5 9 tlJh2 'it'f7 10 'it>e3 ll... e4 @xf6 11 Wf4, and Black is obliged Or 1 L.ltJb4 12 'it'b7. to let go of the h3 pawn. 12 ltJc4 ltJb4 7 ltJtJ 'it>e6 8 'it>d2@rs Or 12... Wf 6 13 'it>d7. The king hurries to support the 13 'it>b7 ltJxd5 14 a6 'it>d7 15 a7 h-pawn, while the f-pawn will be ltJc7 16 ltJb6+ 'it'd8 17 ltJd5 ltJa8 watched by the knight. 18 'it'b8! Wd7 19 ltJf6+ Wd8 20 9 f7 ltJd7 lbxe4 d521 t'llc5 d4 22 'it>b7 d3 23 Of course not 9 ...t'll e6 in view of lbxd3'it'd7 24 lbc5+Wd6 25 lba4! 10 lbd4+. t'llc7 26 ltJb6Black resigned 10 Wxd3 Wf4 11 'it'e2 'it>g3 It seems it's all over. The knight f3 is doomed, whereas the white Zotkin-Kudrin king is too far from the g6 pawn. Moscow, 1965 But White finds a study-like idea. 12 ltJe5 ltJf8 13 tlJd3 h2 14 lbf2 'it>g2 15 ltJbl In this lies White's idea. After a few moves the players agreed a draw. The question arises whether Black could have won. Instead of 5 ... h3, played in the game, he had at his disposal the more effective move 5 . . . d2!, pointed out by I.Zaitsev. Af ter 6 f6+ 'it'f7 7 'it'e2 (7 'it>xd2 ltJb3+!) 7 ...h3 8 lbf3 ltJe4 9 'it>dl decisive is 9 ... lbf2+ 10 Wxd2 t'llg4. 3 Rook Endings
Rook and pawn against rook Emms-Riemersma Gausdal, 1993 The most classical and primitive (though not for everyone) endgame which should not, it seems, present any particular problem for players of grandmaster rank. But John Nunn wrote an interesting book about these endings on the basis of com puter analysis where he gave quite a few complicated positions. However in everyday practice it is much more simple and tragic (or more confus ing). Knowledge of precise posi tions and methods of defence here have exceptional significance. 1... .l:Ial? Correct is 1...�a3 ! 2 �g2 :a2+ 3 The edge pawns �fl a4 4 J::rh3+ 'it>c2 5 1::rh2+ �b l 6 lih3 .l:Ic2 7 c,t>e1 lib2 8 .l:Ia3 11b4 9 The most frequentcas e-rook and 'it>d1 'it>b2 10 lid3 a3 11 lid2+c,t>b 1, pawn against rook. and the a3 pawn cannot be stopped. 2 c,t>g2? People have already programmed But now White misses a draw by computers for this type of ending, 2 �g5 ! a4 3 J::rg3 �c2 4 .llg2 'it>d3 5 and so for the right method of play .l:Ig3 'it>e4 6 .l:Ig4 �f5 7 l:tb4 �e5 8 we should now tum to the Endgame Wg2 'it>d5 9 lih4 a3 10 l:!.h3! when CD ! White reaches a well-known position. Here are a couple of characteristic 2 ... a4 3 @f2 a3 4 .l:Ib5+'it>a2 examples. Rook l:!,'ndings 5 7
There was a quicker win by Tosic-Gyimesi 4 ...c;t>a4! Yugoslavia, 1998 5 @e2 .l:Ibl! 6 .l:Id5 @b2 7 l:td2+ c;t>b3 8 .!id3+ c,t>a4 9 .!id4+ l:tb4 10 lld8 a2 7 'it>d3 @b3 White resigned.
Bagirov-Kraidman Grieskirchen, 1998
l .l:Ia5? Simply driving back the king first by 1 l:tc5+ @b2 2 l:tc8 l:!.xh4 3 'it>d2 draws. 1. ..J:ixh4 2 .l:i.a8 2 l:!.a7 is also enough fora draw. 2 .•..l:Ib4 3 'it>dl? The decisive mistake. 3 .!lc8+! @b2 4 @d2! .l:Id4+(4 ...a3 5 l:tc2+!) 1... :ic5? 5 'iite3 .!ld5 6 .l:i.b8+! saves him. Correct is 1 ... l:!.g5, holding the 3 ... 'it>b2 4 �d8 a3 5 .!id2+ @bl draw as in the previous example. and White had to resign. 2 'it>d4 J:icl3 .l:If6+? Simpler is 3 a5 winning easily. Herrera-Vasquez 3 ...'it>g5 4 .l:Ib6.li(a l 4 .!la6 Cuba, 1998 4 .!lb4 led to a draw after4 ...'it>f 5, and the king rushes to c8. 4 ...@f5? Again he should go for the draw ing mechanism 4 ...l:tfl !? and check on the f- file. 5 a5 .!idl+ 6 'it>c5 l:ral 7 'it>b5 @e5 On 7 ....l:i.b l+ there is 8 @c6 l:tal 9 'it>b6 l+ .!lb 10 'it>c7 l:ta1 11 .!la8 @e5 12 a6 'it>d5 13 .!ld8+ @c5! 14 .!ld6!, winning. 8 .!ih6! @d5 9 'it>b6 .!lbl+ 10 @c7 .!lcl+ 11 @b7 l:tbl+ 12 .!lb6 .l:Ihl 1..Jlal? 13 .l:i.c6 .!ih7+ 14 .!lc7 I:l.hl 15 a6 He can achieve a draw by any l::tbl+ 16 @c8 'it>d6 17 a7 Black move except this, including even resigned. 1.. Jle7, but best of all is 1.. . .!le8!. 2 a4 'it>f6 3 @d6, Black resigned. 58 Rook Endings
Vyzhmanavin-Lerner Simferopol, 1988, where the same USSR (ch), 1984 mistake was repeated! 2 .l:Ic6! 'it>g5 3 l:!'.c8! .l:Ih7 4 .l:Ig8+ leads to a draw. It is nearly always necessary to at tack the king from behind, and not in front, because in the end the king will get the better of the rook. 2 .•.l:l'.h7! 3 .l:Iel In the above-mentioned game fol lowed 3 l:l'.h3 'it>g5 4 �e2 'it>g4 5 .!lhl h3, winning because of the poor position of the rook. 3 •..h3 4 .!lhl 'it>g5 5 'it>e3 'it>g4 6 @f2 If 6 1:1'.gl+ 'it>h4 7 l:1hl, then 1 'it>d2? 7 ...l:1a7 followed by 8 ... .!la2, 9 . . . h2 There was a simple draw by 1 and then 10 . .. 'it>h3. �b2 .l:Ig3 2�c2 c,t>d5 3 'it>d2 .!la3 4 6..• l:1f7+ 7 c,t>gl l:1a7 White 'it>e2 etc. resigned. 1...h2! 2 c,t>e2 l:.al! and White Since after 8 .!lh2 there follows resigned. not 8 ... c,t>g3? 9 .l:Ig2+!!=, but 8 . ..lial+ 9 @f2 l:tbl! with a decisive Vladimirov-Rashkovsky zugzwang. Chelyabinsk, 1 975
Novikov-Lalic Manila, 1992
1 l:1c3? Correct was 1 .llg8+ �f5 2 l:tf8+ Wg4 3 l:!.g8+ 'it>h3 4 .l:Ig5 and after the advance of the h-pawn a draw is It seems that it is not easy for achieved since the white king is cut Black to defend himself, but he off only by three files. finds his only saving resource: 1...h4 2 .!le3?? 1..• .l:Ih8! 2 'it>b7�dS 3 l:1g4 Interestingly, this mistake is typi After 3 a6 �c5 4 l:tbl l:1h7 the cal for grandmasters. A mirror im draw is inevitable. age of this position was encountered 3 .•.'it>cs 4 l:1g7 .l:Ih6! 5 a6 l:!'.b6+ 6 in the game Dvoiris-Kovalev, c,t>a7 .l:Ibl 7llb7 :!al Drawn. Rook Endings 59
Kamsky-Karpov champion does not 'go for' for the Linares, 1994 side. 1 Wf4? Correct was I @f6, threatening a check from g4. If now l...a3, then 2 .l:Ig4+ Wd5 3 .li(g5! with a drawn pawn ending. 1. .. l:!e8! 2 'it>f3 'it>d3 3.l:Ib2.!if8+ Cutting off the king looks a deadly blow forWhite. 4 'it>g3 Wc3 5 .l:!b7 .l:Ifl 6 .l:Ib8 After 6 'it>g2 .l:Ia 1 the position is lost. 6 ... .l:Ial 7 @f3.!lxa2 8 We3 .l:!h2 9 l:1c8+'iii> b2 10 .!lb8+ 'it>cl! What system of defence should he Clearly not 10. ..'it>a l? because of choose: wait for the approach of the 11 �a8 .l:Ih3+ 12 'it>d2 a3 13 Wcl black king to the a4 pawn or adopt with a draw. the more active method. It is clear 11 .l:Ic8+ 'it>bl 12 .l:Ib8+ .l:Ib2 13 that the second solution is more ef .!la8.l:Ib3+ 14 'it>d4 a3 15 'it>c4 Wb2 fective, but it requires accurate play. And in this theoretical position 1 .l:If6+! 'it>e42 .!lg6!'it>e3 3 .l:Ig4! White resigned. The main thing is to maintain con• tact with the opponent's pawns. Rook endings 3 ... a3 4 .l:Ig3+ 'it>e4 5.l:Ib3 l:!a2+6 Several pawns on one flank Wg3 Obviously it is better not to go to This type of ending is met quite the first rank. often in practice. With various posi 6 ... 'it>d4 7 �f3l:1al 8 'iii>g2 tions of the pawns on both sides in And the opponents agreed to a the majority of cases the weaker draw. side should hold a draw even in Euwe-Alekhine those instances where his pawn World Championship (m) , 1937 structure is weakened. Beliavsky-Spraggett Elista (o l), 1998
White is in a dilemma, whether to go forward or back, and the world 60 Rook Endings
1 h4? 16... 'it;>g2 17 �e2 �h3 Drawn. Correct was 1 g4 hxg4 2 hxg4 and then 3 g5, shutting in the black king, Bagirov-Berzinsh whereas 1 f3 was also a possibility Riga, 1998 with the future transfer to another type of classical position, looked at in the next section. 1. ..'it;>g7 2 1Ig5 'it>h63 g4 Now the plan with f3 promises less since it is not clear how Black's position can be breached if White, after an exchange on f3, puts his own pawn on e5. 3 ...hxg4 4 .l:!xg4 l:!e5 5 @fl 'it>h5 6 1If41Ie6 7 �e2 .l:te5 8 'it>el! White gains a tempo-on 8 'it>d2 there is 8 ...l:rf 5. 8 ... @h6!! Black demonstrates a clear and Black has calculated the pawn instructive way to realise his ending clearly. advantage. 9 @d2 .l:If5! 10 'it;>c3 1... J:id5! 2 �f2.l:Ig5 3 �f3f5 Rather better was 10 .l:Ixe4 .l:Ixf2+ An important link in the plan: 11 'it>c3, but even here there are no now White's pawn fortifications are real chances of a win. quickly smashed. 10... 1Ixf 4 11 exf4 'it;>h5 12 'it>d4 4 gxf5 exf5 5 'it>f4 J:ixh5 6 1lb7 'it;>xh4 13 'it;>xe4 1Ig5 7 .l:Ia7 'it>h7 8 J:ib7 'it>g6 9 llb6+ 'it>f'7 10 1Ia6 g6 11 'it;>e5 J:ig3 12 .:i.a7+ �e8 13 'it;>f6 h5 14 l:!.h7 Also there is no salvation in 14 e4 fxe4 15 1Ie7+'it>f8 16 lhe4 g5. 14 •.. �d8 15 :Ia7 h4 16 1Ih7 h3 17 1Ia7 h2 18 1Ih7 1Ig2 19 'it>e5 g5! 20 �d6 'it>c8White resigned.
Madsen-Hansen corr, 1974
On 13 'it>e5correct is 13... 'it;>h3 !. 13•.. �g4 14 'it>e3 Clearly not 14 'it;>e5 �f3. 14•.. �h3 !! Only this leads to a draw. 14 ...'it;>h4? is no use due to 15 f5 !. 15,'it>f3 'it;>h2! 16 'it>e3 After 16 'it>g4 'it>g2 17 �g5 there is the only move 17. . . @f3! !. Rook Endings 61
The position to all intents and pur 1 h4! poses concurs with the previous He cannot allow Black the possi one, and here 1 .!le4 l:tc1 2 'it>g4 bility of playing ... g6-g5. should be played. But White 1.. .l:!b8 chooses another system of defence. After l...e5 follows2 'iS?g2l:rc2 3 1 l:l.'.b4 J:icl 2 l:ra4 @g7 3 l:tb4 g4 J:icl 4 g5+ 'it>f5 5 l:If7+ Wg4 6 nn+ 4 'it>e3 l:!f6@xh4 7 l:Ixg6 winning. Here already no help is 4 c;t>g4 h5 2 'it>g2 l:tg83 g4 g5 4 h5 5 @h3 �el 6 l:!b5 �e4 - +. This is more accurate than 4 4 ... J::!f5 5 .!le4 hxg5+ l:!xg5 5 'it>g3 e5 6 l%h6+ with After 5 'it>e4 g5 6 hxg5 hxg5 7 some chances of victory. J:tb l Wg6 8 nhl �f2 9 c;t>e3 �a2 10 4 ... l:ta8 5 �b7 .lih8 6 @fl! g4 l:Ia5 11 @e4 l:!a4+ 12 c;t>[3 l:!f4+ Now the king can proceed to the 13 'it>g3 l:!e4 Black wins a pawn. other side. 5 ... h5! 6 We2 g5! 7 hxg5 @g6 8 6 ...l:!g8 7 'it>e2.!Ia8 8 h6 .!Ial 9 h7 @e3 J::!xg5 9 'it>f2 c;t>f5 10 l:!e3 l:!hl 10 Wd2 @es 11 c;t>c3 nb4 12 After 10 .!If4 the pawn ending is @c4 l:!h2 13 'it>c5 �hl 'it>c614 'it>f6 won. 15 l:!d7! 10 .. Jig7 11 .Ub3 Making a future path for the king. Or 11 @f3 llg4 ! 12 'it>f2 h4 13 15... .!Ih2 16 @c7 i::!.hl 17 'it>d8 gxh4 I:!'.xh4 14 @g3 l:l.'.g4+ 15 @f3 Black resigned . .l:If4+ 16 @g3 @g5 17 nel �f5 18 l:te4 'it>g6!, and the king works his Beliavsky-Hodgson way round to d5. In a previous ex Cacak, 1996 ample the king also had the possibil ity of attacking the h-pawn in this way. 11.. . .l:Ih7 12 l:!b5 h4! 13 gxh4 �xh4 White resigned, since the pawn ending after 14 l:ta7 'it>xe5 15 l:!xf7 J::i.f4 + is lost.
H uzman-Mikhalchishin Lvov, 1985
1 .l:If6 Stronger would have been 1 h4! 'it>g7 2 �c6 f5 3 f3 !, creating weak nesses in Black's pawn structure. 1. . . @g7 2 llf4 f5 3 h3 'it>g6 4 g4 hxg4 5 hxg4 fxg4 6 Wg3 e3! Black's only chance lies in the creation of doubled pawns in White's position. 7 .lixg4+@f5 8 kif4+ @g6 9 fxe3 nh3 10 @f2 'it>g5 11 'it>f3 ki.a3 62 Rook Endings
After l 1....!lb8 12 @e4 .!le8+ 13 One extreme but typical case 'it>d4 .!ld8+, adopting a frontal at occurred in a practical position. tack, secures Black a draw. 12 1If8 .!lal 13 @e4 J:Ia4+ 14 'it>d5 l:1a5+ 15 'it>d4 .l:Ia4+ 16@c5 Szab6-Keres .l:.a5+17 'it>b4J:ia 2?! Moscow, 1956 Again correct was 17... �e5 ! 18 .l:If3 .!le8 19 @c4 .!lc8+ 20 'it>d3 .!id8+ 21 @c3 l:tc8+ 22 'it>d2 .!ld8+ 23 @el .l:.a8 24 @f2 .!la4! with a draw. 18 .!lf2 J:ic2 19 e4 @g4?! Rather better, though already in sufficient,was 19 ...@g6. 20 e5 'it>g5 If 20 ...@g3, then 21 @b3 ! l:rc822 .!lf5 @g4 23 e4 with a win. 21 'it>b5? Correct is 21 J:ig2+! 'it>h6 22 @b5 .!lc8 23 e6 .!le8 24 @cs .!lxe6 25 Black has great problems and it is @d5! with a theoretically winning quite complicated to transpose into a position. drawn position with f and h-pawns. 21... @g6 22 e6 J:Ic823 J:tf3 1... h6 If 1...h5, then 2 g5+ @f5 3 .!lc6+, and there is no defence against 4 .l:If6. 2 J:ih7!.l:Id6 3 h5! Possible is 3 1Ixh6 @g7 4 g5 fol lowed by 5 f4 and h4-h5. 3 ...@g5 4 .!lg7'it>h4 5 .!lxg6 .l:.d2+ 6 @fl 'it>g3 7 .:i.f6! .l:Ia2 8 @el l:th2! 9 @dl! l:tg2 10 'it>cl.!lh2 11 @bl! The king moves further away from its opponent. 11 ... .!lg2! 12 J:ixh6 @xf3 13 l:1g6 @e4 14 h6 .l:Ih2 15 g5 @d3 16 .l:i.g7 23 •.• @g7? @c3 17 h7 @b3 18 l:rb7+ Black Missing the typical drawing resigned. mechanism 23 ...@gs 24 e7 .!le8 25 l:!.e3@f 4! 26 .!le6'it'f5. And here we see how difficultit is 24 e7 .!le8 25 .l:Ie3 'it>f7 26 Wc6 even for a class grandmaster to .!lxe7 27 l:txe7 Wxe7 28 @d5 Black defenda classic position. resigned. Rook Endings 63
Nikolic-Ftacnik Better than 14 .!lb6, played by Pu/a, 1997 Capablanca. White's plan, with the rook on the 8th, is to to play f5 -f6, and Black's defence, with the white king on f2,is to play ...h6-h5. 14 .. J!b3? Correct is 14 ....l:!e3 !, hindering f4-f5. 15 'it>f2? Correct is 15 .!id8! with the threat of f4 -f5-f6, e.g. 15... .l:Ib2+ 16 'it>g3 .!lb3+ 17 'it>h4 .l:Ie3 18 .l::.e8 ! and f4-f5. 15 ... .l:Ia3 16 .l::.d7! 16 .l:Id8 badis because of l 6 ...h5 1 I:l.c5! 17 g5 h4 18 .!ld7 'it>g6 19 !td6+ Correctly preparing for h4-h5 in 'it>g7 20 .!lh6 h3 21 f5 .lla5 with a order to take on h5 with the rook. draw. 1 ... .!lal 2 h5 .l::.gl + 3 @f3 'it>g7 4 16 ...lia2+ 17 'it>f3 .l::.a3+ 18 'i¥te4 .f:!.a5 gxh5 5 1Ixh5 .l::.al 6 l:!.d5 .l:!gl .!la4+ 19 �d4! Not yet a mistake but more This is the point of deploying the technical is 6 ....l::. a4, cutting off the rook on the d-file. After 19 @f5 white king. 'it>f8!, according to an analysis by 7 .l::.d6 'lt'h7?? Kopaev, the game is drawn. A serious mistake-correct was to 19 ... lial 20 f5 ? attack the e3 pawn by 7 .. Jlal 8 'i¥te4 Correct now was to move aside :a3 9 .l:!d3 .!la5 10 �d5 l:i.a3 11 with the rook 20 .!lb4! .l:Ig l 21 'it>f3 'it>d4 .!lb3 or 7 ....l:Iel 8 e4 .llfl 9 'it>g3 .Ila1 22 .l::.b8 �fl+ (after 22 ...lia3+ I:l.gl+ 10 'it>h3 .l:r.fl, and the rook 23 'it>g2 the king goes via g3 to h4) does not have the f6 square. 23 'it>e3 .l::.el+ (23 ....l:Ig l 24 f5 ! 8 e4! .l:Ifl+ 9 'it>g3 .!lgl+ 10 Wh3 .l::.xg4 25 f6+ 'it>h7 26 e6!) 24 Wf2 .l::.fl 11 I:!'.f6! Wg7 12 e5 .l::.e4 25 'it>f3 :!.el 26 I:!'.b7! @f8 27 .l::.b4 'it>g7 28 'it>f2.l::. al 29 I:!'.b8h5 30 g5 h4 31 'i¥tf3! h3 32 1::rb2 winning. 20... .l::. el+ 21 'i¥td5 I:!'.gl? This position is from the game Duras-Capablanca, New York 1913 (with colours reversed), and here 21...'it>f8! 22 'i¥td6.l::. al gives a draw. 22 'it>d6 lial 23 .l::.c4 l:!.a8 24 .lac7 lla6+ 25 'it>e7l:!.a4 26 e6! fxe6 Or 26... l:!.xg4 27 Wd6 winning. 27 f6+ 'i¥tg6 28 f7 J::rf4 29 f8='ii' .l::.xf8 30 'it>xf8 e5 31 .!lc4 Black resigned. White obtains a winning position which is in every text book. Even with a very lucky author and 12 ...J::r f3+13 'it>g2.l::. a3 14 .l::.d6! a passed e-pawn it's drawn! 64 Rook Endings
Novikov-Beliavsky and the plan of the stronger side Graz, 1997 consists of creating a passed pawn on the d-file which promises him great chances of success.
Bagirov-Veingold Ta llinn, 1981
1...l:!c72 .l:Ia5 We6 He had to try the plan with the ap proach of the king via h6. 3 .!le5+ Wf6 4 J::ra5 Wg7 5 h3 'it>h66 I:!'.e5 I:!'.c2+7 'it>g3 'it>h5 8 .!le7 h6 9 .lle5 J:ic3+ 10 Wh2 l:tcl ll g3! After 11 'it>g3? J::rfl 12 l:ta5 there 1 �al is 12... g5 ! 13 .!lxf5 e3 with a win. With the idea of defending the Now, however, despite the 'cut-off pawn and freeingthe king. king from the firstrank, Black does 1...'it>g6 2 .!ifl.!la2 not succeed in improving his After 2 ...f5 3 l:!a1 fxg4+ 4 hxg4 position. the e6 pawn would be too weak. 11... J:ic2+ 12 Wgl J::rc8 13 'it>g2 3
(8 ...g5 9 a6 .!:!'.a3 10 hxg5 + -) 9 a6 Material is nothing-activity (in f6 10 a7 l::Ixa7 11
Keller-Mikenas corr, 1992
1 ....:.a3+? In analogous situations where the pawn has not advanced to the fourth rank, it is correct to activate the king, but the rook gives a frontal check. 2
Black must lose because of the S 'it>dS 1Ia3 6 a6 .!lxf3? weakness of the e4 pawn. This is a decisive mistake-after 1 a 6! g6 6 ...'it>f6 7 'bc6 .l:Ixf3 8 1Ib8 .!la3 9 After l...h4 2 gxh4 :!.al+ 3 'it>e2 �b6+ 'it>f5 10 @b7 @g4 11 a7 .!lxa7 �a2+ 4 'it>e3 .l:Ia3+ 5 'it>xe4 lhh3 6 12 'it>xa7 'it>xg3 13 .!lb4 f6 ! there is f4 ! �xh4 7 .l:Ia8 .!lh l 8 l:tf8and 9 a7 an easy draw by ...g6-g5. wins. In the event of passive de 7 �b8 �a3 8 .l:Ib6 .!lxg3 9 'it>c6 fence by Black, White pushes his .l:Ia3 10 Wb7 gS 11 hxgS h4 12 a7 pawn up to a7, then the king goes to h3 13 a8='i' .l:!xa8 14 'it>xa8 h2 lS e5, the Black rook must keep watch .l:Ih6f6 16 .l:!xh2 fxgS 17 .!lt1!! from the a4 square. Then White We advise paying attention to this plays 'it>f4 and, on ... 'it>h7, '\t>f5 with move. a win. 2 �a8 'bf7 3 a7 'it>g7 4 'it>el 'it>h7 S 'it>dl 'it>g7 6 @cl Wh7 7 'it>bl I:!'.a6 8g4! Determining the pawn formation on the flank. 8 ... hxg4 9 hxg4 @g7 10 gs 'it>h7 11 'bb2 'it>g7 12 'bb3 'it>h7 13 'it>b4 r:j;g7 14 'it>bS �a2 lS Wb6 .l:Ib2 16 'bc6 .!lc2+ 17 'bd6 .!la2 18 .!lc8! lha7 19 .!lc7 and the pawn ending is easily won. Black resigned.
Lerner-Dorfman Now the white king avoids being Tashkent, 1980 shouldered away and hurries over to the opponent 's pawn, therefore Black resigned.
Konopka-Shcherbakov Pardubice, 1996
1...'it>g7? Passive play-correct is l...'it>e6 -or l....:i.a2 detaining the king on f6 . 2 as �a4+ 3 'bes .!la3 4 'it>e4 lfa4+? This position resembles the previ Simpler is 4 ...@f 6! 5 a6 'be6 giv ous one, but there are a number of ing a draw. important differences. Black has a 68 Rook Endings
doubled pawn, and the greater num 1 g4! g5 2 a4 .l:Ial 3 a5 'it>g6 4a6 ber of pawns creates more prerequi 'ith6 sites fora White victory. White also wins after 4 ...h5 5 1...f5 gxh5+ 'it>xh5 6 l:!a8 'itg6 7 a7 'itf5 8 Worthy of attention is 1 ...e4 !?, for 'itg3 with the idea 9 f3. the present maintaining the pawn on 5 J::ra8 'it>g6 6 a7 'it>g7 7 'it>g3 fl so as to defend the king against .l:Igl+8 'it>h2 l:!.al 9 'itg2! h6 checks fromthe flank. After 9 ...'it>h6 White wins by 10 2 .l:Ib8 e4 3 h3 h5 4 h4 J::rbl+ 5 J::rf8 lixa7 11 .l:Ixf6+ 'it>g7 12 lie6 c;tie2 lib2+ 6 Wfl J::rbl + 7 <;tig2 lib2 lia4 13 cJi>g3 'iitfl 14 lih6 'it>g7 15 8 b4 :!bl 9 b5 'itg6 10 b6 Wh7 11 .l:Id6lia7 16 lle6.l:Ia4 17 h4. f4 ! exf3+12 Wxf3Wg7 13 b7 10 'it>h2 'it>h7 11 lif8 Ji(xa7 12 On 13 Wf4 there is 13 ...Wf6 ! with lixf6 Wg7 13 .l:Ie6 .l:Ia4'it>g3 14 lib4 the idea . . . llfl . 15 h4 gxh4+ 16 'it>xh4 l:ta4 17 Wg3 13 ... Ji(b4 14 We2 llbl 15 Wd2 .l:Ib4 18 .l:Ie5 'it>g6 19 llf5 :gb3 16 Wc2 J::rb5 17 Wc3 l:!.b6 18 With the idea of llf5-f4, f2-f3. Wc4 :!bl 19 Wc5 l:tb3 20 'itc6 19 ... .l:IbS 20 'it>f4 .l:Ie8 21 lie5! l:tc3+ 21 'itd7? .l:Ixe5 22 'it>xe5 <;tig5 23 Wxe4 c;tixg4 Stronger is at once 21 c;tib5 l:tb3+ 24 'it>d3! Black resigned. 22 'it>c5!. 21...lid3+22 <;ties! Kozul-Nikolic After 22 c;tixe6 l:txe3+ 23 Wxf5 Sarajevo, 1993 l:tf3+=. 22 ... lib3 23 'it>e7 l:tb6 24 'it>d7 l:!b3 25 'itc6 llc3+ 26 'it>b5 J::rb3+ 27 'it>c5! :gbl If 27 ...c;tih7 winning is 28 c;tic6 llc3+ 29 Wd6 .S.d3+ 30 'it>e7 J::rb3 31 'it>f6!. 28 'iitc6! J::rcl + 29 c;tid6 :!bl 30 'it>xe6.:.b5 31 .l:Id8Black resigned.
Andersson-Hubner Ter Apel, 1997 1 a7 White sees no chance of going with his king to the queenside to help his a6 pawn. For this he has to sacrificethe f2 pawn. 1... c;tie5 Also not bad is 1...<;tig7 with the idea of2 ...f5 or 1...l:ta3. 2 f3 l:!.a3 3 'it>t1 1ial 4 'it>e2 'it>f6 5 f4 Ji(a3?! The simplest way to achieve a draw is 5 ...c;tie7, and there is no way for the king to approach. Rook Endings 69
6 @d2 .!ia2+ In the game after 7 ...g6? 8 l:th8 A move such as 6 ...We7 can al Black resigned. ways be made. 8 @d2 .!la3 9 i;t>e3 .!IaS 10 @xd3 7 @c3 .!la3 8 Wc4 llaS 9 'it>b4 :Ia4 11 Wc3 .!lal 12 'it>c4 .!Ia2 13 .U,a210 @cs J::ral 11 Wc6 'it>cSJ:ial 14 'it>c6.!Icl + lS 'it>d6 No use is 11 l:th8 .iha7 12 l:txh6+ Preparing to meet @es with the @g7 13 fxg5 .lla5. move .!la5, but now he threatens to 11.. . .!la2 12 @cs J:Ia6 13 'it>d4 tranpose to a pawn ending. lla4+ 14 @d3 .!la3+ lS We2 l:taS? lS ... J:ial 16 .!lc8! .!Ia6+ 17 .llc6 Completely losing his sense of .!lxa7 18 J:ic7+J:Ixc7 19 Wxc7 danger-why not 15... 'it>g7? 16 �h8! l:txa7 17 l:!.xh6+ 'it>g7 18 fx gS k.i.a3 19 'it>f3.ll aS 20 h4 I:i:bS 21 ltf6 Black resigned.
Y ermolinsky-Seirawan USA , 1997
Note an important fine point-the tempo g2-g3 is decisive. With the pawn on g3 it would be a draw. 19 ... @e7 20 i;t>c6'it>e6 21 @cS! Destroying the opposition. 21...'it>f6 22 @dS g6 23 hxg6 @xg6 24 'it>e6, winning. But with 1.. . .!lxa2? the pawn on g3, 24 . ..'it>hS 25 'it>xf5 Black does not suspect any danger is stalemate. associated with the the white pawn on h5. Thereforecorrect is l...f4!. Vujala-Smith 2 f4 ! d4 corr, 1993 The other chance is to hide on h7 and wait with 2 ... 'it>h7 3 @fl . But then the white king enters the game via bl, attacks the d5 pawn and as a result of zugzwang captures the d5 and f5 pawns. 3 @fld3 4 'it>el! @rs Losing is 4 ...l:txg2 5 l:l'.d7 .!la2 6 a7 @h7 7 'it>dl, and the white king captures the f5 pawn, and then goes to b8. S .!la8+! 'it>e7 6 a7 @f7 7 'it>dl lla4 70 Rook Endings
1 g4! not have enough tempi to save Fixing the opponent's weak himself. nesses. 2 J:ib4 'it>a5 1. .. h6 2 'it>g2 .l:Ia3 3 h4 i;t>f6 4 h5! At first sight 2 ...gxf3+ 3 'it>xf3 Guseinov-Beliavsky Pula, 1997 1. .. White has a weakness on c4, but No help is 3 gxf4 'itf5. this is small and Black needs to 3 ...J:ib2+ 4 'itglfx g3 5 I:!'.f6+ create another. It is necessary at least to force t. .. hS! 2 Wf4 back the black king. After 2 gxh5 .!lxh5 3 @g3 .!lc5 4 5 ... @g7 6 I:!'.xe6l:l.'.h2 ! l!c3 follows 4 ...'itd7! with the threat It is precisely this pawn which is ... @c7 and ...d5. important. 2 ... gS+! 3 'itt3 h4 7 J::rd6 J::rxh4 8 .llxdS Now there is a weakness on h3. After 8 ctig2 l:l.'.g4! 9 .!lxd5 h4 fol 4 l:rdl lows h4-h3 and g3-g2. Preventing .!le6-el and hi. 8 ....!le4! 9 e6 h4 10 .!lhS .!lxd4 11 4... l:IcS 5 :!:!'.el+ ctid7 6 �e4 I:!'.eS! e7 @f7 12 .!lh7+ @es 13 Wg2 I:!'.g4! 7 bi.d4 'itc6 Zugzwang. 14 'itgl g2. White Now the king is included in the resigned. attack on the weak c4 pawn. 8 .l:Id3 WcS 9 J::r dl f6 ! 10 J:id2 bi.e6 11 J::rc2 ctid4 12 :!:!'.cl I:!'.e3+ 13 Mednis-Gurevich \tg2 .l:r.g3+ 14 Wh2 .llc3+ 15 J::rdl + New Yo rk, 1982 �d3 16 .l:Icl :n 17 l:l.'.dl+ @cs 18 l:rcl .!lf2+ 19 'itgl I:!'.f4 White resigned. Mikhalchishin-Khmelnitsky Sibenik, 1990 In this rather untypical position there is only one possibility for Black to save the game. 1...cS!! 2 bS After 2 bxc5 'itd5 3 ctid3 l:!.c4 4 @e3 'itie5! Black achieves a draw. White can do nothing against the 2 ... .!lb43 l:l.'.b2! .!lxh4 4 b6 .!ih8 5 break ...f5 -f4. @c4 ctid6 6 @bS @d7! 1...l:!.bl! 2 @f2 It was still not too late to lose: Bad 2 @f4 .!lei 3 lle3 (3 'itf3 l:te4 6 ... �c8 7 .l:Id2+ \ties 8 b7 .l:Ib8 9 with the idea ...f4 ) 3 ....l:Id l, and .!id7 'ite6 10 'itc6 with a win for there is no defence. White. 2 ... f4 ! 3 �13 7 @a6 J:Ia8+8 ctib7.l:Ia4 ! 72 Rook Endings b7, and White wins; or 13 ...l:!'.c8 14 b7 l:!b8 15 'it>c6 with the same re sult) 13 b7 .l:Ib3 14 lte2+! (only now, when the rook stands on the b3 square!) 14 ...�f5 15 'it>c7, and White wins. Interesting methods of defence in rook endings Now if White plays 9 J::rd2+, then Black has the possibility to trans pose into a drawn pawn ending by 9 . . l;Id4. 9 �c2!! c4 If 9 .. . 'it>d6 10 .l:Id2+ :!:i.d4? 11 l;Ixd4+ cxd4 12 �c8, and White queens with check. 9 . . .l:Ia5 10 l:!.d2+ is also losing. 10 lld2+�e6 11 'it>c6 c3 In defending this typical position, exceptionally important is the pres ence of the two white pawns-the g or h pawns do not win, the others do quite simply, by advancing them to the seventh rank. Knowledge of these types of position is extremely important in the different variations when transposing to a rook ending. Chaunin-Friedman 12 l:!e2+�f5 13 b7 Moscow, 1951 If 13 'it>b5, then Black achieves a draw by the manoeuvre 13 ...l:Ia3 14 @c4 �a6! 13 ...l;Ib4 14 .l:Ih2 Nothing is changed by 14 �c7=. 14 ...J::r xb7! 15 'it>xb7 'it>e4 16 'it>c6 'it>d3Drawn. However at the end of the game �nalysis showed that in the position m the last diagram, 12 �c2! de served consideration. Now if 12 ...Ita3 (12 ... l::t c4+ 13 'it>b5'>t>d5 14 Rook Endings 73 Here White can win easily by 1 Here, an analogous winning 'it>xg3! 2 h4!, after which the f3 method is demonstrated by Shirov. pawn goes to f7. However White 1 b6! cxb6 2 .l:Ih8 Black decided that simpler was 1 hxg3?, resigned. keeping his pawns connected, but it turned out that after 1. .. g4+! 2 fxg4 Mokry-Pribyl he cannot win even with an extra Olomouc, 1977 pawn. Polugaevsky-Parma Sochi, 1965 After 1...Wh7? 2 Wc6 .l:Ia2 3 .l:Id8! .:i.xa7 4 l:l'.d7+ �xd7 5 Wxd7 g5 6 'it>e6! Blackresigned. And here in a classic game Black The method which helped White resigned since he saw 1 h6, and to win is typical of such positions. It thought there was no defence. How is possible that Black did not even ever after l...Wg6! and 2 ...Wh7 ! think about the fact that it was still Black calmly makes a draw. possible to give up even a centre pawn. On the correct defence- Shirov-Kramnik 1.. .Ir.a1 !-the operation to transpose Belgrade, 1999 into a pawn ending is not achieved since after 2 'i¥tc6 Black checks along the file until the king is forced away from the a-pawn. For example: 2 ....:i.c l+ 3 'it>d6 .:i.d l + 4 'i¥te6 .l:.al 5 l:td8 (5 :ge8? l:l'.a6+!) 5 ...na6+ 6 l:td6�xa7 7 l:Id7+nxd7 8 Wxd7 'it>f6! 9 'it>d6 g5! 10 hxg5+ 'it>xg5 11 Wd5 e3! 12 fxe3 Wg4 with a draw. The right method of defence was not found in the following game. 7 4 Rook Endings Holmov-Timoschenko Here a mechanism like 2 ... llel Pavlodar, 1982 does not work, since there follows first 3 l:.ta3+! and only then 4 l:txa2. There is also no win for Black after 2 ...l:!.c l 3 �a3+ l:tc3 4 l:!.xa2 l:.tc2+ 5 :rxc2 With White to move, Black wins a It is of no significance at all what pawn: 1 @h2 @f3 etc. With Black the move is here. It is necessary to move, after 1... The only path to victory is to cre with a draw since Black cannot go ate a passed f-pawn, but how can to the c-file with his king: 18... 'it>c7 this be done? If Black manoeuvres 19 hxg5. with his king to the d4, c4 squares, etc, then White gives check along We return to the game Holmov the file. But even here White needs Timoschenko. to take care. For example, after 3 .l:!c8+ @d4 4 .l:Id8+ @es mistaken is 5 .l:r.e8? 'it>f5 6 lia8 g5 7 J::ra5+ 'it>g6 8 hxg5 f5 !. After capturing on g5 Black man ages to create a passed f-pawn, which wins. Instead of 5 lie8? it is necessary to continue S �a8! Now the above-mentioned ma noeuvre does not work: 5 ...g5 6 hxg5 f5 7 .l:Ia6!h4 8 g6, and already Black has to think how to save himself. On 5 .l:Ia8 he will try to send the king to g7, so as then to play 10 .l:Ia3+�xe4 11 .l:Ia4+? ...g6-g5: S ... 'it>e6 6 lia6+ 'it>f7 7 Also here still possible was 11 .l:Ia7+'it>g8 8 J::ra8+ 'it>g7 �h2 and then g2-g3. On this follows a check on the ll... We3 12 l:ta3+'itid4 13 �a4+? rank-9 .l:Ia7+, and after 9 .•.'it>h6 'it>c3 14 .l:ta8 fS ! lS .l:Ia7 f4 +! 16 the manoeuvre 10 .l:Ia6 prevents the 'it>h2 'it>d4 17 .l:Ia4+ 'it>eS 18 .l:!a3 advance ...g6-g5. 'iitfS 19 .l:Ia6 �g4 20 lixg6+ Wxh4 There is nothing else for Black, 21 �a6 'it>gS 22 l;Ia8 h4 23 l:!'.g8+ besides 10 ... fS . Possible then is 11 'it>f6 24 .li(a8 h3! 2S gxh3 f3 26 .l:Ia3 .l:Ia7 gS 12 l::ta6+ 'it>g7 �es White resigned. Now Black can choose two paths, each of which leads to a favourable result for him. Let us look first at Activity is more important the direct 13 hxg5 h4 14 gxh4 f4 15 than material h5 f3+16 @f2lih l 17 h6+! 'it>h7 18 l:ta7+ 'it>g6 19 h7 .l:Ixh7 20 lia6+! It has Jong been known that in �xg5 21 @xf3 with a draw. rook endings activity is more impor The second path also leads to his tant than material. This means maxi objective: 13 lia7+ 'it>f6 14 laa6+. mum possible active deployment of Black is at the crossroads. the king and rook in coordination After 14 ...�e5 15 hxg5 h4 16 g6 with one's own passed pawns and in he might even lose. the struggle against the opponent's But 14 ... 'it>f7 15 lia7+ �e8 pawns. And even the very idea of (15 ... We6 16 hxg5! h4 17 g6) 16 the priority of material will be fatal. l:ta8+ c;t>d7 17 lia7+ 'it>d8 18 .l:Ia8! Here are a fewstriking examples. 76 Rook Endings Larsen-Browne Arbakov-Gurevich Las Pa/mas, 1982 Moscow, 1978 The black rook is hopelessly pass 1 . . . .Ucl! 2 @f3 :gc4 3 <;t>e3 e5! 4 ive and White's plan is to go with <;t>d3 .l:!a4! his king to h5, place the rook on b6 Clearly not 4 ...e4+? because of 5 and break up the black pawns with �xe4. the pawn march f4 -f5-f6. 5 <;t>e3 e4 6 .l:Ig3l:!'.a2 7 _:g4 1 '\t>g4! '\t>e6 Or 7 h4 g4 8 h5 .l:Ia3+ 9 @f2 After 1 ... g6 winning is 2 l:tb6+ �xg3 10 '&t>xg3 e3 with a win. '\t>g7 3 f5 @h7 4 Wf4 and 5 '\t>e5. 7 . . ..Ua3+ 8 <;tin l:.d3 White 2 fS+ �e5 3 .:M! g6 resigned. On 3 ...Wd6 follows4 <;t>h5. 4 fx g6 fx g6 5 �b6! '\t>d4 6 .l:Ixg6 "Active positions of the pieces in �xb7 7 !rxh6 rook endings are worth a pawn" At a necessary moment the b7 -Smyslov. Here is a classic pawn is given up in exchange for example. the win of a pawn on the opposite flank. This results in a theoretically Capablanca-Tartakower winning position. New Yo rk, 1924 7 ...ki.g7 + 8 <;t>f4 �t7 + 9 'it>g5 'it>es 10 g4 .t!'.f8 11 @h5 �f7 12 g5 'it>fS 13 .l:Ih8 and Black resigned. In the following position the white rook is finnly posted, but it is immobile and passive. Black finds a plan with a transposition to a pawn ending, exploiting the position of the rook on g4. Rook Endings 77 l @g3! l:hc3+ 2 'ith4 l:f3? Lilienthal-Smyslov Materialism in such positions is Moscow, 1941 simply ruinous. This position sparked a great debate in 1998 in the magazine 64. Master Goldin maintained that 2 ...a6, intending to create a passed pawn with maxi mum speed, gave Black drawing chances. However after 3 g6 b5 4 axb5 axb5 5 @g5 b4 6 :n+ @g8 7 Ii'.xf5 b3-master Barsky pointed out 7 ...l:lg3+! 8 'ith5! (8 @f6 .l:rg4! leads to a blind alley-8 ...b3 9 l:lxd5 b2 10 .l:rb5 @g7 11 :l:xb2 !:.xg6 12 :!c2 l:ld6 13 l:.c4! @f6 14 @g4 c6 15 @f3'it>f 5 16 l:Ic5+ 'ite6 There followed: 17 @e4, gradually driving back the l . . .@e4 ! 2 1:1'.xcS f4 !! black pieces) 8 lixd5 'itg7 (8 ...li'.c6 By sacrificing a third pawn, Black 9 @h6 J;;td6 IO l:.a5! l:ld8 11 li'.b5! creates cover against checks to his with a win) own king. 3 exf4 Or 3 @fl .l:ral + 4 ®e2 f3+ 5 @f2 l:a2+with perpetual check. 3 ... @f3 4 h3 i:lal+ with a draw four pawns down. Kozlov-Mikhalchishin Vladikavkaz, 1978 9 l:ld7+@f8 10 d5 ! (but not IO f5 b2 11 g7+ @g8 12 @g6 l:lc6+ 13 f6 bl=� with check!) 10... @e8 (after 10... b2 11 @f6 @e8 12 l:le7+ @d8 13 g7 l:tg3 14 ltel White wins eas ily) 11 l:Ih7 b2 12 l:lh8+ @e7 13 l:lb8:c2 14 l:l'.b7.l:rg2+ 15 @f5 l:lc2 16 d6! ®xd6 17 @f6, and White wins. 1 e4! 3 g6! l:l'.xf4+ 4 ®g5 �e4 5 @f6! A passed pawn in conjunction The f5 pawn protects the white with a centralised king is worth two king, but not the black one. pawns. But where does the capture I 5 ...® g8 6 l:lg7+ @h.8 7 :txc7 lie8 .l:rxb3 lead? After ...l J:l.xg3 2 ®e4 7 'itxf5 li'.e4 8 @f6 l:If4+ 9 ®e5 l:Ig4 h5 3 ®f4 l:lg4+ with a further ...g5 10 g7+, etc. White won. White is in a deplorable state. 78 Rook Endings 1.• .lhg3 2 e5 'it>g6 3 e6 l:id3+ There is a very clear path to the After 3 ...lie3 4 'it>d6h5 5 lib8 h4 draw. 6 e7 the white pawn queens. 1. .. a5! ! 2 bxa5 l:id4 3 I:!'.xb5 lia4 4 'it>e5 h5 5 �b8! 4 1::rb3 5 e7 does not work because of Or 4 l:i.b6+ 'it>e5 activating the 5 . ..@f7, but now this is a threat. king. 5 .. Jle3+ 6 'it>d6 'it>f6 7 J::rf8+ 'it>g5 4 ... l1xa5 5 'it>f4 .l:Ia4+ 6 J::rb4 8 e7 b2 9 �b8 'it>f6 10 .l:If8+! 'it>g5 lixa3 7 J::rb6+ 'it>g7 8 .l:Ie6 lib3 9 11 J::rb8 Drawn. .l:Ie3! l:!xe3! 10 'it>xe3 'it>t7 11 'it>d3 'it>e7 12 'it>d4 'it>d6 Drawn. Taimanov-Chekhov Kishinev, 1976 A classic activating of the king was seen already in this game: Owen-Morphy London, 1858 The threat is 1 e4+, and then 2 a4, 3 'it>g4.Therefore correct is 1. ..I:!'.d3! 2 �xb6 l:!a33 .l:!b2.l:!a4 ...foll owed by ...g5-g4 with a draw. 1...'it>t7! After 1.. . .l:Id3 2 lic8+ 'it>h7 3 .l:Ie8 A frequently met theme is the White has great chances of a draw. sacrifice of a pawn to activate the 2 lic7+ Wf6 3 .l:Ixb7 �d3 4 'it>f2 rook. l:!xd4 5 We3 e5 6 b6 l:i.b4 7 l:!b8 Barlov-Schiissler 'it>e7! Hanninge, 1988 Black chooses a plan to liquidate White's passed pawn. Inferior is 7 ...J::r b3+ 8 'it>d2 We6 9 'iitc2, and it is not clear how to win. 8 b7 'it>d7!9 �g8 �xb7 10 �xg7+ Wc6 11 .:i.g6+ 'it>c5 12 .l:!xh6 l:l'.b3+ 13 'it>e2 e4 After the pawn sacrifice all Black's pieces have become more active and his pawn further advanced. 14 lih8 'it>d4 15 l:!g8 1::rb2+ 16 'it>dl'it>d3 White resigned. Rook Endings 79 Schlechter-Lasker succeed because of3 ... h3 4 a5 .!lg2+ Berlin, 1910 5 Wfl f4 6 a6 f3 7 a7 h2, and it is time to resign. 3 ...h3 4 .llb8 J:ig2+5 wn J:id2 Quite possible is 5 ...l:ta2, but Black wants to use the rook to cover against checks from the side, while he simply pays no attention to the a-pawn. 6 'it>gl f4 7 .:i.g8+ �f3 8 �h8 �dl +! 9 'it>h2 'it>f2 Black wants simply to promote the f-pawn to a queen. If now 10 .l:!xh3, then 1 o ... f3 11 l::!.h8 l!d3! 13 a4 'it>e2 14 �f8l:!.e3 ! 15 a5 �e7! 16 The best defence is ... a6 f2 with a win. 1... .l:!e4! 2 .l:Ic5 'it>f6 3 .l:Ixa5 l:tc4! 10 a4 f3 11 a5 Wfl 12 a6 .!lal 13 4 l:ta6+ 'it>e5 5 l:!a5+ 'it>f6 6 .l:!a2 .!las �e5 7 J:ib2 J:ic3+ 8 'it>g2'it>f6 9 �h3 After 13 c;t>xh3 f2 14 .!lg8 .l:Ixa6 .!lc6! with a draw. 15 'it>g3 J:if6! the win is straight forward. Kramnik-Beliavsky 13 ... f2 14 a7 lia6! Groningen, 1993 Correct was the natural 1 l!b8, ac The right idea. Now on 15 �xh3 tivating the rook and not fearing 'it>gl 16 l:!g8+ 'it>hl 17 �f8l:l.'.a3+ 18 1.. . .llg3+ 2 'it>f2 l::!.xa3, since the @h4 Wg2! is reached a postion from rook ending with the f-pawn is a classic study by Lasker, where drawn. But White decides to defend Black wins by shouldering the white the pawn and at a suitable moment king to the seventh rank, while on to obtain counterplay with the help 15 .:i.b8 follows 15... l:!xa7 16 .!lb l+ of the a-pawn. But this proves to be 'it>e2 17 l!b2+ We3 18 I:!'.b8c;t>e4 19 a decisive mistake. J::!b4+ c;t>e5 .l:Ib5+20 'it>e6 21 .l:Ib6+ 1 :!.al? .l:!g3+2 Wf2 Wg4 3 l:r.bl 'it>e7 22 l:!bl l:!a3! 23 :!:!'.fl l:l.'.f3 fol Here also he had to go back; lowed by the approach of the king. counterplay with 3 a4 does not 15 'it>hl h2! 16 J:ib8 80 Rook Endings If 16 'iitxh2, then 16 ....!lh6+ 17 Bojkovic-Kakhiani Wg3 'it>gl 18 z:rf8 l:Ig6+ 19 'it>h4 Erevan, 1996 .!la6, again reaching Lasker's study. 16 ... bi.xa7 17 .l:Ibl+'it>e2 18 .l:r.b2+ We3 19 .:b3+ 'it>e4 20 .!lb4+ Wd3 21 .i:tbl .!lf7 Also good is 21 ...l:i.e7 with the idea of22 ...Ite l+. 22 bi.fl We2 23 1Ixf2+ Wxf2 White resigned. Activating pieces also means acti vating the king to coordinate with a passed pawn. Let's try to evaluate the position Smagin-Naumkin -White has an extra pawn, but Moscow, 1983 Black has a strong passed pawn on g3 plus an active rook plus the pos sibility of activating his king-it is Black who is playing for the win. 1 llg2 l:'th3! Usually the rook is placed behind, but here Black does not allow 2 'it>c2 to be played because of 2 ...l:l:h2 winning. 2 l:i.gl Wd6 3 'it>c2 i;t>e5 4 'it>d3 g2+! If 4 ...Wf 4, then 5 e5! 'it>xe5 6 'it>e3 with equality. 5 i;t>c4? The endgame with f and h-pawns Correct was 5 'it>e2 .U.g3 6 'it>f2 is drawn, but it is necessary to trans .l:lb3 7 'it>xg2 l:rxb2+ 8 Wf3 .!lb3+ 9 pose to it at once! 'it>e2 brxa3 10 .!lb1 with a draw. 1 'it>g4Wf7 2 'it>g5!!b4 3 h6 .l:tbl 5 ... z:tg3 6 'it>c5 'it>xe4 7 b4 'it>d3?! Black reluctantly decides to give Stronger is 7 . . . !!g6 with a win. up a pawn-which he should have 8 'it>b6 done earlier. But now, psychologi Or 8 b5 bi.g69 a4 b6 10 �d5 We3 cally, he is not ready fordefence. wmnmg. 4 �xa4 Wg8 5 f5 :tgl+ 8 .. .llg7 9 a4 'it>c4 10 b5 Wb4 11 Correct is 5 ...W h7. a5 .l:tg6+ 12 'it>xb7 'it>xa5 13 b6 6 .!lg4! 1Ial 7 'it>g6 .:a2 8 f6 .!lal l!xb6 14 i;t>c7 �g6 15 'it>d7 'it>b4 16 9 f7+Wf8 10 h7 l:i.hl 11 �f6 Black i;t>e7 '>t>c4 17 'it>f7 l:r.g3 White resigned. resigned. Rook Endings 81 Eliskases-Levenfish Zugzwang-bad is 19 ...�f4 20 Moscow, 1936 f6=. 20 �f7 �f4 21 �g7 'i¥tg5 White resigned. Azmaiparashvili-Kupreichik Kuibyshev, 1986 White has a pawn more but Black has an active king and a far ad vanced passed pawn. White must play 1 l:!e7! c3 2 a4 lia1 3 lic7 l:!a3 4 a5 �e5 5 a6 �d4 6 a7 �d3 7 f4 c2 8 .l:td7 with a draw according to Readers will ask-what has all an analysis by Smyslov. However in this got to do with activating in a the game he quickly played: rook ending. This is what it has to 1 lia5+?'i¥te6! do with it. White has the exchange Sacrificing yet another one. fora pawn but Black threatens after 2 l:ta6+ �d5 3 l:Ixh6 c3 4 lih8 ...lid7 to seize the initiative. So .l:Ial 5 l:!c8l:txa3 6 �h3 White decides to transpose to a rook After 6 f4 g4 7 f5 �e5 8 lic5+ ending a pawn down, but in the �d4 9 l:tc8 l:!.a1 10 f6 l:tfl 11 llc6 process activating all his remaining �d3 12 :gd6+ �c4 13 l:!c6+ 'i¥tb4!, pieces. and he has to give up the f6 pawn. 1 .l:txc5!! 'l'Wxc5 2 'fkxc5 bxc5 3 6 ...@d4 7 �g4 @fl @f6 4 �e2 �e6 5 �d3 �d5 6 Bad is 7 f4 gxf4 8 gxf4 c2 9 �g4 'i¥tc3! .l:Ic3 +-. Now 7 lidl+! is threatened fol 7 ...lia5! lowed by �c4. 'Building a bridge'-the main 6 ... c4 7 �b4! .l:tc7 8 lid4+ �e4 9 manoeuvre in rook endings. b6 .l:tc6 10 �b5 :gc8 ll l:td7! 8 f4 llc5 9 l:!d8+'i¥te3 10 l:1dl With the threat of lic7. No help is 10 l:1e8+@f2 11 .l:Ia8 11 ...lib8 12 l:1c7!�d3 13 l:rc6! c2 12 l:1al gxf4 13 'i¥txf4 cl='ii' 14 Here is the key move---defending llxcl l:txcl 14 g4 lic4+ 15 �f5 the b6 pawn and keeping in his �g3 ! 16 g5 �h4 17 g6 �h5 18 g7 sights the c4 pawn, White wants to lig4 +-. take on a4 and obtain connected 10... c2 11 l:!.cl gxf4 12 gxf4 @d2 passed pawns. 13 l:tal �cl= 14 lixcl lixcl! 15 13 ... f5 14 @xa4 c3 15 �b5 g5 16 g5 �e3 16 f5 �e4 17 g4 'ite5 18 a4 f4 17 gxf4 gxf4 18 a5 e4 19 �g6 l:!c6+19 @g7 :ga6! 'i¥tb4! 82 Rook Endings The last finesse-now on 19. . . c2 White has a distant passed pawn, follows 20 'it>b3. but his cut-offking, indeed also the 19 ... 'it>e2 20 a6 e3 21 a7 li:tf8 21 centralised black king, gives Black a fx e3 Black resigned. decisive advantage. 1 .l:Ie7+ 'it>d4 2 l:!e6? Van der Doel-Klovan Correct is 2 .l:If7!. Gelsenkirchen, 1998 2 .•. d5 3 l:!xh6 'it>e3 4 lif6 Also bad is 4 lie6+ 'it>xf3 5 l:te1 d4. 4 ...d4 5 l:!xf5 d3 6 l:re5+'it>xf3 7 �f5+ 'it>e3 8 .:i.e5+ 'it>d4 9 l:!e8 d2 10 �d8+ 'it>e3 11 .l:Ie8+ @f3 12 I:!'.f8+'it>g4 White resigned. There will be cases when it is nec essary to sacrifice all one's pawns formaximum coordination of all the pieces. Each of White's remaining pieces Pelletier-Rozentalis are clearly more active than his op Erevan, 1996 ponent's. Therefore any delay by Black would be equivalent to death. 1 ...l:!c 8!! Forcing the capture of the pawn, which allows Black, through the opening of the d-file, to invade the opponent's position. 2 .l:Ixa6 l:id8 3 I:!'.b6 l:td2 4 lixb5 �xf2 5 a4 lixg2 6 a5 f5+! 7 �xe5 f3 8 .l:Ib4 Drawn. Smyslov-Epishin Rostov, 1992 1...g4!! Inferior is l...f4 2 gxf4 gxf4 3 llxb4 e3 4 fxe3 fxe3 5 l:ta4 .:i.b2 6 b4! 'it>e4 7 l:!a8 'it>d3 8l:ra3+'it>d2 9 lial!=. 2 llxb4 f4 ! 3 l::ta4 �b2 4 gxf4 g3! 5 fxg3 e3 6 f5 'it>e4 7 f6 'it>f3 8J:fal !Ig2! 9 f7 e2+ 10 'it>el 'it>e3 White resigned. The rule of the two weaknesses the possibility of transferring an Rook Endings 83 attack from one weaknessto another Fercec-Mikhalchishin until the time comes that the No va Gorica, 1997 opponent cannot defend all his weaknesses-is an exceptionally universal one. Nikolic-Movsesian Polanica Zdroj, 1996 Black clearly has the more active king and rook, and the method of realising the advantage lies in the creation of two weaknesses and at tack on them. I... liti>e5 Less logical is l...g4, and White Even here the two weaknesses on can avoid the drawn pawn ending. 2 a6 and c6 would not mean anything .l:tf4 .l:xf4 3 exf4 @f5 4 '>te3 gxh3 5 were White not to have chances of gxh3 c5 6 a4 b6 7 b3 a5 8 'it>f3 d4 9 organising another weakness on the cxd4!? cxd4 10 h4 g6 11 'ittg3 'iti>e4 other flank. 12 'ittg4=. It is equally a draw after9 1 h5! gxh5 c4. If L.g5, then after 2 'itie4 kingthe 2 b3 breaks through to the black pawns. More principled is 2 g4 g6 3 b3 f5 2 !:lxh5.l:.i. g8 3 .l:.i.h4! 4 .Ug3 f4 5 l:.f3 c5 +. Weak is 3 �xh6? :xg4 4 ltxf6 2 ... g4! 3 ll:fl :lg3+ and the endgame is drawn, 3 ltg3? loses at once because of while the passivity of the rook is 3 ... f5 . The attempt 3 hxg4!? de· provisional upon the arrival of the serves attention, but even then white king to the defence. 3 ....l:i'.. xg4 4 g3 b5 5 b4 ll:g66 a3 lih6 3 ...'>td7 4 'itte4 '.te6 5 liti>f'3 :l'..h8 6 with the idea 7 ....Uh3 leads to a win .Uh5 Ith7 7 'it>g3 .l:i'..d7 8 !fa5! :a7 9 ning position forBlack. 'itth4 ..ttn 10 'itths 'iti>g7 11 rs 'iti>h7 3 ••.g xh3 4 gxh3 lth4 5 .:tf'3 b5 6 12 ltcS.l:i'.. c7 13 a4! 'iti>g7 '>te2? After 13... .l:i'.. b7 14 .Uxc6 .l:i'..xb4 15 A decisive mistake. Better is 6 b4 :lxf6 White wins easily. g5 7 ltg3 @f5 8 ll:f3+'iti>g6 9 l:.g3 f5 14 b5 axb5 15 axb5 .l:.i.b7 16bxc6 intending 10... g4 -+. l.c7 17 ll:cl l:.c8 18 c7 'it>t719 ltc6 6 ••. g5 7 '.tf2 8a5 .l:.i.g3 'iti>g7 20 lifi>h4 'iti>ti 21 '.tg3 Black If8 'iti>g3.Ue4 9 '.tf2b4 -+. resigned. 8 ... c5 9 'ifi>g2 84 Rook Endings If 9 a3, then 9 . ..a4 10 'it>e2 axb3 This was the last time White had 11 cxb3 �h8 with the idea 12... .l:!a8 the possibility of activating his rook: -+. 4 J::!c6! I:!'.xa2 5 h4! gxh4 6 'it>xh4 9 ... b4 IO cxb4 axb4 11 .!if3 .l:Ih8 with equality. 12 .llfl .!la8 13 J:Ial 'it>e4- + 14 'it>f2 4 ..• h5! 5 J::!c2 f5 15 'it>e2 J:ih8 16 .l:!hl .!lh4! Now there is already no saving It was still not too late to let the himself : win slip. 16... g4? 17 h4. (a) 5 h4 hxg4 6 hxg5 f5- +; 17 .!lgl .!lxh3 18 :xg5 1Ih2 19 (b) 5 gxh5+ 'it>xh5 6h4 g4 7 1If2 Wdl c,t>xe3 20 .llg3+ �d4 White f5 8 'iitf4 .!la4 9 'it>xf5 g3 -+; resigned. (c) 5 gxh5 Wxh5 6 :t:Ic2 f5 7 .l:!d2 f4 + 8 @g2 'iio>h4 intending ...l:!c3 , Typical mistakes in rook endings a7-a5-a4-a3 .l:Ic3-cl-bl-b2 - +· (d) 5 h4 hxg4 6 'it>xg4 f5+ 1 'iitg3 In rook endings there are a great g4 8 .!lf2 'it>h5 9 J:ifl J:Ia4! 10 fxg4 many typical methods of play, but J:Ixg4+ 11 @f3 'it>xh4! 12 .l:.cl 'iitg5 also, naturally, also a great many 13 a3 l:!a4 14 .l:!c3 'it>f6 15 .!lb3 a5 typical mistakes. We acquaint you -+, Levenfish. with the most typical of these. 5 .•. h4+ 6 �f2 a6 7 .l:!b2 .l:Ic3 8 'iitg2 a5 9 J:if2 .!la3 IO 'iitfl Wf'l 11 Making Passive f4 gxf4 12 .!lxf4 'it>g6+ White resigned. Ilivitsky-Taimanov USSR, 1955 Vaganian-Schlosser Germany, 1994 1 f3? He should play 1 h4!, boxing in 1... .l:!b7? the black king. An analogous mistake. He should 1. ..gS! 2 c,t>g3'iit g6 3 l:!c2? activate his forces atonce by 1 ... h5 ! Again White sticks to waiting tac 2 g4! 'it>g7 3 'it>e2 J::!e7+ 4 Wf3 tics. The correct path was 3 :i.d6+! :.c7 5 h4 h6?! 6 'iitg3 .l:!c3+ 7 f3 f6 4 h4! gxh4 5 'itxh4 .l:Ixa2 :tc7 8 'iitf4 :b7 9 h5 .l:Ib4+ IO Wg3 (5 ....l:.xf3 6 .l:r.a6=) 6 f4 I:!'.a4 7 f5 + gxh5 'iitg? 8 Itd7, maintaining equality, If 10... l:tb7 White gains the ad Levenfish. vantage by 11 hxg6 fxg6 12 f4 fol 3 •..f6 4 l:th2? lowed by Ita5-a6, f4 -f5 ±. Rook Endings 85 11 gxh5 l:!'.b7 12 'i!tg4 1.tbl 13 f4 1... ki.e52 h4 h5 3 .a'.c4l:l'.f 5 4 l:i.e4 After 13 llxa7 Black activates 'i!ta6 5 .I!e7 g5 6 hxg5 fxg5 7 ti.es himself by 13... l:i. gl+ followed by h4 8 a'.a8+ 'i!tb7 9 ti.gs 'i!tc6 10 J;.g5. l!g6+? 13 ... 1.tb7 14 ¢>f5 lte7 15 a4 'fri.c7 A mistake. He should activate the 16 'Ot>g4 l:.d7 17 lta6 Xlb7 18 a5 king by IO ®c4!. lic7 19 f5 l:.c4+ 20 ¢>g3 ltc5 21 10... @d5 11 llxb6 'lt>e5? @f4 :Ic4+22 'Ot>e5 l:i.c5+ Now Black meets mistake with Worthy of attention is 22 ...Jlh 4!? mistake. Activity of the king is the (Schlosser) 23 ltxa7 lixh5 24 a6 main thing in rook endings. Black l:lhl 25 llb7l:te l+ 26 \td6llal 27 maintains the advantage by shoul a7 'it>f6. dering the opponent's king with 23 ®e4 ltc4 24 'i!td5 Jlh4 25 f6+ l l...®d4!. Now however it's a ®h7 26 !Ixa7%Ixh5 27 'i!tc6\tg6 ? draw. Better is 27 ...ltf5 !=, Schlosser. 12 'lt>c3 'iiff4 13 Zib4+ 'lt>g3 14 2S a6 Il:a5 l:tg4+'lt>f2 'it>d2 15 Drawn. White wins also in the event of 28 ... l:thl 29 .:td7Jla l 30 @b6. 29 'iifb6 Ital 30 ltd7 Xlbl+ 31 Abra movie-Nikolic ®c7 @xf6 32 ltd5!! ltcl+33 @b6 lgalo, 1994 'ii?g6 34 a7 l:cS 35 l:ta5 h5 36 aS=W a'.xaS 37 .l;;xaS @f5 3S 'Ot>c5 ®e4 39 'it>d6@f 4 On 39 ...f5 winning is 40 l!a4 with the idea of@e5 +-. 40 !:thS! f5 41 @d5! 41 ltxh5? would be a mistake be cause of41...®e4 with a draw. 41 ...@e3 42 :es+ @t'3 43 'it>d4 f4 44 @d3 h4 45 lth8 'Ot>g3 46 'i.t;e2 ®g2 47 ltgS+ 'iifhl 4S @n Black resigned. Grunberg-Brunner 1... llf6 ? Germany, 1992 Allowing White to create counter play on the king's flank. He should restrict this by l...f5! 2 g4 hxg4 3 fxg4 fxg4 4 @g3 'i!td8 5 ®xg4 'lt>c8 6 .!Ig7b5 with advantage to Black. 2 g4 'it>d83 ®g3 ®cs 4 '11e7 b5 5 lle5 a'.b6 6 gxh5 gxh5 7 :!xh5 b 4 8 ltd5 8 l:l'.c5rather is worse. 8 ... b3 9 ltdl b2 10 llbl 'it>d7 11 @f4 ®e6 12 ®g5? White does not exploit the oppor tunity presented by his opponent. 86 Rook Endings The only chance for him was 12 Shirov-Morozevich ct;>e4! �f6 13 @d4 Wg6 14 @c3 Amsterdam, 1995 ct;>h5 15 I:!'.xb2 .l:Ixb2 16�xh4 @xb2 17 @c3 ct;>g3 18 @d4 �xf3 19 ct;>e5! with a probable draw. Now, how ever, Black has no difficulty realis ing his advantage. 12 ... l:1bS+ 13 @h6 @f6 14 hS .l:Ib3 lS @h7 l:1b8 16 f4 1::rb3 17 @gs J::rbs 1s h6 ct;>g619 rs+ @r6 20 Wh8 llb3 21 'it>g8 �b7 22 @h8 @gs White resigned. Marie-McNab Hastings, 1995 1 l:tg4? Better is 1 .l:Igl with the idea of connecting his pawns in the rook ending by f2-f4-f5. 1 .. . @e7 2 .l:Ia4? l:ta8 3 f4 Wxe6 4 Wg3 @dS S @g4 @xcS 6 WgS �bS This tempo was made possible as a consequence of 1 lig6-g4. 7 1:1'.al llg8+ 8 @f6 llf8+ 9 �eS l:te8+ 10 WdS l:td8+ 11 'it>eS l:l'.e8+ 12 @dS .l:i'.d8+13 @e4 .l:Ie8+ 14 Wf3 lth8 1s llhl @cs 16 rs @d6 17 @f4 a4 18 @gs @e7 19 f6+ @f'l20 llbl l:!xh321 lib7+ct;>rs Drawn. 1 ... aS 2 �h2? Correct was 2 f3 ! a4 3 @g3 l:ta1 4 Vaulin-Voikhovsky f5 ! ct;>xf5 (if 4 ...g5 5 l:l'.a6+ @xf5 6 Russia, 1997 lia5+ with a draw) 5 l:txf7+ We5 6 lie7+ @d4 7 lia7 a3 8 Wf4 a2 9 �a8 with equality. 2 ... �f3! 3 l:!a6+ ct;>rs 4 l:!.xaS+ @g4! S Wg2 l:l'.xf4 6 lla7 Wxh4 -+. Analysis. Technically weak play in rook endings In the following position White, of course, has a winning position but, it is not so easy as it seems at 1 e4? .l:Ia3+ 2 @g2 fxe4 3 llf4 firstsight. 'ii>eS 4 l:l'.xg4 @d4 S llg8 @e3 6 .l:Id8 ct;>e2 7 �b8 Rook Endings 87 The a-file is inaccessible. with an easy draw and therefore 7 ... l:ra2 8 .:i.bl e3 9 @g3 l:1a8 10 played �b2+ Wd3 11 l::tb3+ 'it>d2 12 J::rb2+ 1... .l:Ia2?, i;tic3 13 l:Ib7 .l:!e8White resigned. But he did not reckon on 2 h4!, Krasenkov-Iskusnik after which White has a winning Russia, 1996 endgame, whereas he could have made a draw by I....l:Ibl! 2 h4 l:1gl+ 3 'it>h3 J::rhl+ or I...1::rb8!, preparing against h3-h4. Materialism instead of activation Schmitdiel-Mikhalchishin Berne, 1994 .i � � �· .�..,�.�,,�� ..�f.�,,,,,v,� 1...'it>f5 ? After the simple l ... .l:If4 �. .�. there is an easy draw. 2 c6 @e6 3 �.!.�.�- �c2+-. .;·!•,,�•A A narrow spectrum of noticing D. • ,,; "fj,2" the opponent's threats �,,,�---. ,,, ,v,. , r, (one threat he sees, the other not) 1 J::rdl? Kozul-Mikhalchishin, He should think about the liquida Bled, 1996 tion of his opponent's activity (re stricing the mobility of the pawns) by I .l:Ia5! 'it>f8 2 c,t>g3 @e7 3 c,t>D @d7 4 'it>e3 'it>c6 5 @d4 with excel lent chances of a draw. 1...'it>f8 2 J::rxd6 as 3 'it>g3? White's only chance was to strug gle against the a-pawn by activating his c-pawn: 3 c5! @e7 4 l::td5! a4 5 c6 a3 6 c7 a2 7 J::i.dl, with a draw-clearly better is 4 ...@e6! with chances of victory. Now, however... 3•.• a4 4 'it>f35 a3 �dl 'it>e7 6 @e3 Black saw only one threat I .l:tf5+ a2 7 �al 'it>d6 8 @d4 �a4 9 f4 h5 and prepared to repulse it by 10 g4 h4 11 g5 g6 and White 1...l:rb3+ 2 @g2 .l:!b2+ 3 @fl .l:Ib3 resigned. 88 Rook Endings Mikhalchishin-Stangl A well known drawing mecha Dortmund, 1992 nism without the b2 pawn-the rook goes to the sixth rank (f6) and upon the approach of the king to b5 begins an endless checking se quence. But in the game there followed ... 1.. . .l:Ib4?2 .!lc7.l:Ia4 3 lic8+? For what reason? Why not 3 .!lc6 �xh7 4 @xb2, winning easily. 3 ...@xh7 4 .!lc6 .l:Ib4? After4 ...@g7 5 �xb2 @fl 6 �b3 .l:Ial 7 @b4 @e7 8 @b5 @d7- draw. S a7 .!la4 6 "IJ,.c7 @g6 7 �xb2 @f6 1...l:td3? 8 @b3 J::ral 9 @b4 'it>e6 10 @bS A technically incorrect attack @d6 11 .!lc6+ @dS 12 .!la6 :!:!.bl+ necessaty was 1...l:td1 + 2 @g2 lia1 13 'it>as @cs 14 .l:Ic6+! and Black 3 .l:Ia7 J:Ixa3 4 b5 a5=. could resign. 2 a4 aS 3 bxaS J:id4 4 a6 .l:Ixa4 S a7 �h6 6 @n gS Inaccurate technical execution Also no help is 6 ...lia2 7 �el, and the king goes to b 1. Ribli-Mikhalchishin 7 @e2 .l:ta3 8 @d2 @g6 9 @c2 Germany, 1993 gxh4 10 gxh4 @rs 11 @b2 .l:Ia6 12 @b3 lial 13 .!lc7! Preventing 13 ... �g4 because of 14 .!lc4+and 15 J::ra4. 13.•. l:ra6 14 @b4 and Black resigned. Ignorance of typical drawing mechanisms Ivanchuk-Lautier Horgen, 1996 1...l:!.b2? Occupying the b2 square, which is needed for the king-correct is 1...J:ic2!, and Black has no difficul ties. Now, however, he lacks a tempo. 2 hS a2 3 @g2 @c3 4 'itig3 'it>b3 S f4 J:ibl 6 fS al='iV 7 l:txal J::rxal 8 g6 @c4 8 ...hxg6 9 fx g6 �c4 10 @£2 ! +- 9 f6 hxg6 10 f7!Black resigned. Rook Endings 89 Typical Rook Endings In the endings it is difficult to find anything new-everything has already been played. But among various positions resembling one another it is still possible to find great differences and great simila rities at one and the same time. Our attention was attracted to the ending from the game 11 .!lf8! ! lhf8 12 exf8='YW+ Wxf8 Hiibner-J .Polgar 13 Wd6 with a won pawn ending. Dortmund, 1996 2 .l:.a6 Wf7 3 .U.f6+ 'it>g7 4 e6 l:idl 5 .!if7+ 'it>gs 6 We4 .!lgl 7 li:tf3 .llel+ After 7 ...'it>g7 8 e 7 lie1 9 .!le3 an easily winning pawn endgame is again reached: 9 ...l:txe3+ 10 <;fo>xe3 @fl 11 We4 We8! 12 'it>d5 'it>d7 13 e8='ii'!, etc. 8 'it>d5 <;fo>g7 9 .!lf7+<;fo>g8 10 <;fo>d6, and Black wins. This reminds me of a very similar endgame that I had a year earlier. Barle-Mikhalchishin Here Black resigned. Slo venia, 1995 1...<;fo>f8 After 1...l:!.b4 Hubner gave the following variation: 2 .l:!a6 (also winning is 2 .l:.g7 �b6 3 'it>e4.!la6 4 .!lc7 'it>d8 5.!lc5 .l:!a4+ 6 'it>d5 .l:!g4 7 <;fo>d6 .l:!xg5 8 l:ta5 +-) 2 ...Wfl 3 .!lf6+ 'it>g7 4 e6 .l:!b5 5 'it>f4 J:Ia5 (5 ...�b4+ 6 We5 .!lb5+ 7 'it>d6 lhg5 8 .!ifl+ <;fo>g8 9 .l:!a7 .!lgl 10 l:l.'.a8+ @g7 11 e7 +-) 6 e7 l:l'.a8 7 'it>e5 .l:.a5+ 8 'it>d6 .!la6+ 9 'it>c5 .l:!a5+ 10 @c6 .!la8 90 Rook Endings There followed: 'it>f6 obtaining a draw. Correct is 1 1...g4+ 2 hxg4 hxg4+ 3 �e3 �e5! .l:Ie7+ 2 �f4 l:ta7 3 I:!'.b6l:tc7 4 After3 Wf2 .l:Ib2+ 4 @fl arises a .l:tf6+ Wg7 5 e5 reaching an easily position which is very similar to the winning position, known from the previous one, with the following previous examples. variations: 4 ...�e4 5 J::rc4+ @e3 6 1.. . .!ld7! J::[.c3+ 'it>d4 7 l:ta3 .l:r.b4 (7 ...l:tc2 !?) 8 The only chance-the threat was @f2 e4 9 J::ra2 e3+ 9 �e2 �e4 10 2 e6, followed by �e5, 1::rd5 and .!lc2 J:id4, again threatening a .!ld7. transfer to a winning pawn 2 'it>e4 J:idl 3 .l:Ib7+'it>e6 4 l:tb6+ endgame. We7! 3 ...l:rb2 4 lk8 l:!.g2 5 I:!'.f8+ 'it>e6 6 To achieve a draw it is necessary 'it>e4 to give up yet another pawn. After 6 J:ie8+ @f6 7 I:!'.f8+ �e7 8 5 l:txg6 J:iel+ 6 @f4 .!ifl+ 7 @g3 .l:If5 We6 Black wins easily. l:!.gl+! 6 ...Ibg3 7 J:ie8+'\t>f6 8 J:Ixe5 In the game 7 ... .Ile1 8 J:if6 l:!e4 9 White reckoned on 8 ....!le3+? 9 I:!'.f4! was played, with a win. @xe3=, but more tenacious was 9 8 'iitf2 l:tg4!, l:rf8+ 'it>g7 9 .l:Ia8 (9 l:tfl �g6 10 And a'ccording to an analysis by Wxe5 J:If3 -+) 9 ....l:tf3 10 .!la6 l:tf6, M.Yudovich-it's a draw. and then the black king goes to g6 and g5, winning. More complicated variations with 9 .. JU3 10 l::te8 @gs 11 .l:Ih8 l:!.ti the addition of the h-pawns are met 12 @e3 g3 White resigned. in the game A similar endgame was analysed a Akopian-Almasi very long time ago. Lj ubljana, 1995 Filipov-Kopatsny USSR, 1968 There followed: 1...'iitf4 Inferior is 1... .!lh 1, since after 2 There followed: .l:Ia4 he cannot play 2 ....!lxh2? 3 1 e5? 'it>gl. If 1 .l:tb6 �a5+ 2 Wd6 l:txg5 3 e5 2 J::rc2 l:tgi 4 'it>d7 (4 .l:tb7+ 'itf8,there and After 2 g3+ hxg3 3 hxg3 'it>g4 4 is no win) 4 .. J::tdl+ 5 :d6 �el 6 e6 !te2 'it>f5 5'it>g2 J:ib3 6 �f2 J::!d3 7 Rook Endings 91 @g2 e3 the threat of .l:!d2 is on g3, which also, however, leads to unstoppable. a draw. 2 ...@r s 15 Ita5+ 'itid4 16 l'.Ia4+ l:!c4 17 On 2 ...l:ih I Almasi gave 3 g3+ l'.Ixc4+! hxg3+ 4 hxg3+ J...@dS 2 Itf6 e4+ 3 @e3 l'.Ib3+4 'it>f2l'lb2+ 5 @fl This passivity is forced, since af ter 5 @g3 h4+! 6 @h3 e3 the pawn cannot be stopped. 14... g4? 5 ...@d4 6 l:lg6litd3 Afterthe correct 14... @f5 15 liit:f2 After 6 ...'ite3 ?! 7 l:[g3+ @d2 8 White defended in the following way: 1 .!le7 �d32 g4! An attempt to create very quickly a passed pawn on the king's flank. 2 ...J::! d8 After 2 ...�d4 3 �f3 �d8 4 l:!a7! the position is analogous to the game. 3 'it>f3 �f8+ 4 'it>g2 .!lf4 After 4 ...e2 5 l:id7+ 'it>el 6 �d6! .l:If2+ 7 'it>gl �f8 8 h3 the threat to Here Black has several different take on h6 saves White. plans: 7 J::!a7! (a) 15 ....l:Ic2 16 h3 'it>d3 17 l:!a3+ Transposing the game to a well 'it>d2 (after 17... 'it>e4 18 �a5 'it>f3 19 known position with a rook attack :i.f5+ �xg3 20 l:!xh5 with a draw) from the long side. 18 :!al l:!b2 19 �gl 'ite2 20 �g2 5 ...e2 6 .!la2+'it>d1 7 .l:Ial+Draw. l:id2 21 g4 'it>d3+ 22 �fl �h2 23 .:i.a3+and a draw was agreed. The correct plan of defence for (b) 15... 'it>d3 , and now this class of position was demon (b l) 16 'it>gl? 'it>e4 17 l:!a8 :!b l+ strated in the game 18 �g2 e2 -+; (b2) 16 h3 'it>e4 17 �a8 'it>f3 18 Portisch-Pietzsch .l:If8+ 'it>xg3 19 .l:Ie8 .:i.f2+ 20 'it>e1 Madrid, 1960 'it>f3 21 l:if8'it>g2 22 �e8 �f3 23 h4, and the win is not easy; (c) 16... h4 !?, and, for example, 17 gxh4 'it>d3 18 'it>gl 'it>e4 19 h5 .!lb8 20 h6 l:!g8+ 21 @fl 'it>f3 22 .!la3! , l :-1 ! and White maintains equality. • • �� An analogous ending is ... · · ·=·· Chiburdanidze-Gal!iamova -:·,:·· �'� Groningen, 1997 a�. �. �."---- 1...:i:!.bl! 2 h4 .l:In+ 3 '\ties l:!gl 4 l:!c7+ 'it>g6 5 .!lc6+ �f7 6 'it>f4 :i:Ifl+ 7 'it>g5 h6+! Weaker is the more passive 7 ...l:!.e l 8 l:!c7+ @f89 '\t>f5 .l:Ifl+ 10 �e6 .l:If6+ 11 'ite5. 8 'it>h5 :i:!.f4 9e5 l:!e4! The black rook has gained maxi mum activity, but the position still requires accuracy. Rook Endings 9 3 10 e6+ 'it>f6! 'it>e4! 6 ltc4+ c,t>d3 followed by Not 10... 'it>e7 11 �c7+ 'it>f8 12 ...l:i.f6 and the advance of the a5 �c8+ 'it>e7 13 l:i.g8 +-. pawn. 11 g5+ i;¥i>e7! 4 'it>g2 !td7 5 .l:Ixe5 :i.d2+ 6 'it>gl Again not 1 l...'it>f5? 12 l:!c5+ 'it>h3 7 lig5? '\t>xe6 13 .l:!c7 g6 14 'it>xh6 +- or White sees a defensive idea, but ll...h��5 12 hxg5 'it>e7 13 'it>g6 +-. his execution of it is inaccurate. Af 12 .Uc7+ 'it>f8 13 .l:!cS+ @e7 14 ter 7 �e6! l:i.g2+ 8 'it>hl l:!xg3 9 lic6 l:t.�g6! Black can give up trying to After 14 l:!g8 hxg5 15 hxg5 .l:Ixe6 wm. 16 lixg7+ i;¥(f8draws. 7 •...l:Ig2 s 'it>hl :!:i.f2!! 9 'it>gll:If6! 14 ... @fS 15 l:!b6 'it>e7 16 I:!'.b7+ Now, however, it's zugzwang. i;¥(f8 17 .llf7 i;¥i>gS 1 S lie7 @f8 19 10 .l:Ia5 l:!f3 11 g4 l::!.g3+ 12 'it>hl gxh6 gxh6! i;¥i>xg4White resigned. Clearly not 19 ...'it>xe7? 20 hxg7 J:e5+21 i;¥i>g4 +-. 20 l:tf7+ @es 21 l:tf6 'it>e7 22 Bogoljubow-Rubinstein !txh6.l:r.e5+! and a draw. London, 1922 Trabattoni-Barlov La Va letta, 1979 Rubinstein is considered the 'king' of rook endings, but his play in the following ending leaves us 1... lib7! profoundlybewild ered. The main task is to drive offthe 1...g6? rook from the sixth rank and cover Correct is 1...g5 !, transposing to the king against checks along the the position in Barle-Mikhalchishin. f-file. Then follow preparations for 2 ltbS lia2? the advance of the e5 pawn. Clearly better is 2 ....l:Ia3+, gaining 2 lia6l::!.f7 3:!:!'.as some tempi. Another possibility would have 3 .l:If8'it>g7 4!tes 'it>f7 s l::!.bs lih2 been 3 I:!'.a4 lif6! 4 !l:b4 lta6 5 .l:Ic4 What is he doing? He can't take the e4! 6 l::!.c5+ 'it>g4 7 l:tg5+ 'it>h3 8 @fl pawn! .l:!f6+ 9 'it>e2.l:!f 3!, winning. 6 :cs .l:!a2 7 h4 l:!a7? S 'it>f4 @f6 3 •..'it>g4 ?! 9 .:.rs+ 'it>g7 An interesting plan, but better And a draw was agreed. Rubin looks 3 ...�f 6 4 .:i.b5 lia6 5 .l:!c5 stein's worst endgame! 94 Rook Endings The fact that matters are not quite ltal @g6 22 .tlbl .l!te5 23 @d4 'itf6 so simple is shown by the game 24 .tlel? Stronger was 24 .tlfl+. Fischer-Geller 24.. Jta5! 25 :l.xe4?! Curacao, 1962 Again stronger is 25 :.fl. 2S ... 'i.t>fS ! 26 l:e8 'itg4! 27 @e3 @g3! And in this theoretical endgame White resigned. The last game did not answer sev eral questions of defendingthe end game with the e and g·pawns. In modem chess there is one more very interesting example. Balashov-Ulibin Uzhgorod, 1988 1 g5+?! An extraordinarily crucial deci sion-it would be simpler to wait with 1 l:.c7or I llb5. 1. ..hxg5 2 hxg5+ @g6 3 !1.e7 !1.e3 4 'itf2? Not to this side--it was necessary to exchange the g5 pawn for the e6 pawn and not stick the king's head out. Therefore 4 @h2! r!e5 5 @h3 leads to a draw. 4 ... �e5 5 @13llf5+ 6 'ite3 After6 @e4 :.f7! 7 :.xe6+ 'itxg5 the win is easy. 1 llf8 6 ...eS 7 'ite4lhgS 8 !1.e8? The preliminary I @f2 does not Preferable is 8 :a7! Wh5 9 :al threaten 2 e5 because of 2 ...lla5 3 with the idea after 9 ... :g2 to play .tle4 @f5. I 0 We5 g5 11 'ite4 though 11...@g4 1 ... lta2+ 2 @13 lla3+ 3 'i.t>f2 wins. The general idea to win with l:i.a2+4 We3 Il'.a3+S @d4 lla4+! the g and e-pawns consists of giving Not 5 ...ltxg3 ?? 6 e5 +-. up one pawn and transposing to a 6 @dS J:.a5+7 'itc6.a:a6+ 8 'itb5 theoretically winning endgame. l:.a3! 9 .:.rs+ 8 ... llgl! 9 'itt3llfl + 10 'itg3 :l.f5 On 9 e5 :!'.e3 10 .li!e8 'i.t>f5 11 g4+ 11 llb8 @gs 12 lte8 'itf6 13 ltf8+ 'itxg4 12 @c5 Wf5 draws. 'ite6 14 lte8+ 'i.t>d5 lS :as l:!:f7 16 9 ... 'itg6! 10 g4 lte3 11 eS .:e4 12 'iPg4 !1.e7! 'i.t>c6 :.xg4 13 ltf6+ 'i.t>g7 14 �d7 When the rook is placed behind, lla4 then· half the business is done. A drawn ending is reached with 16 :.as+ We6 17 .tla6+ 'i.t>f7 18 the weaker side's king on the @13 lle6 19 l:.a8 e4 20 @e3 g5 21 kingside. Rook Endings 95 15 lic6 l:!a8 16 l:!.c8 �a7+ 17 .l:Ic7 easily wmmng and described in �a8+ Drawn. every book on rook endings; ( d) l...'it>d3 ! (this was the only possible continuation) 2 .:i.fl (after2 How many roads lead to Rome? l:Ie5 'it>d4 3 :b5 @c4! the rook is continually hounded) 2 ...@e2 3 :f4 We have already repeatedly said 'it>e3 4 l:ta4 (there is nothing else) that in the endgame there are usu 4 ...lixf 5 5 a6 l:if8 6 a7 .l:Ia8 7 'it>xh3 ally two paths (moves)-right and 'it>d3 (similar to variation (b ), but wrong. But at times the number of the cut-off king has proceeded one paths is surprisingly greater and it is rank further which is of decisive exceptionally difficult to find the significance) 8 @g4 @c3 9 \t>f5 right one. @b3 10 �al @b4 11 'it>e6'it>c5 Beliavsky-Azmaiparashvili Portoroz, 1997 (After l l...'it>b5? 12 'it>d6 a posi tion is reached from variation (b), while exerting control over the d6 The black king has six(!) possible square is the key to evaluating the moves, but only one(!) leads to a position) 12 'it>d7 �b6 13 l:!.b l+ draw. 'it>c5! (the point) 14 :b7 :h8!, and (a) l...@f3 (f4) loses because of 2 a theoretical draw has arisen. :al l:!a6 (he cannot draw by taking on f5) 3 f6 winning easily; Ward-Baburin (b) 1...'it>f2 2 l:!a1 .:i.xf5 3 a6 :f84 Isle of Ma n, 1997 a7 .l:Ia8 5 .lla3 ! (a very important moment to cut offthe king) 5 ...'it>e2 6 'it>h3 'it>d2 7 Wg4 �c2 8 'it>f5 'it>b2 9 :a6 @b3 10 'it>e6 @b4 11 'it>d6! (shouldering away the king and preparing to set up a mating net) l 1...'it>b5 12 :gal@b6 13 :i:Ibl+ 'it>a6 14 'it>c7! 1ixa7 15 'it>c6, and White wins; (c) l...'it>d2? 2 l:!.e5 @d3 3 a6! l:txa6 4 'it>xh3 'it>d4 5 .l:Ie6, and Black resigned, the rook ending is 96 Rook Endings White has an extra pawn but Branicki-Sefc White's cut-offking plus the strong Prague, 1955 passed e4 pawn and centralised Black king means that it is only Black who can play for the win. Baburin assessed the position as winning for Black. Let's have a look. (a) 1 .l:!b2 l:!.c7! 2 'it>b3 'it>e5 3 l:rc2 lid7! 4 'it>c3 e3 5 f6 We4 6 l:!.g2 .l:!c7+ 7 'it>b2 J:if7 8 'it>c2 lixf6 9 l:1g3 (On 9 Wdl, 9 ...'it>d3 is unpleas ant. Instead, without the h2 pawn, White saves himself by means of the stalemate 10 l:rd2+!) 9 ... !!c6+ 10 'it>dl 'it>d3 11 h4 .lla6 12 Wcl :Ial+ Here there are several ideas for 13 'it>b2 J:ihl 14 h5 .l:!xh5 15 'it>cl+ White to realise his enormous ma l:thl+ 16 'it>b2 'it>d2 White terial advantage, but only one of resigned. them leads to its objective. (b) 1 f6 'it>e5 2 l:1f2 l:tf7 3 'it>b3 e3 (a) 1 .l:!a6 .U.dl 2 g4+ 'it>f6 3 d7 4 llfl 'it>e4 5 'it>c2 e2 6 l:1gl 'it>e3, 'it>g7 4 .l:!a7 'it>f6 5 'itig3 'it>e5! and there is no apparent defence (shouldering away the white king, against l:rxf7 or 'it>f2; now after 6 d8='i!V lixd8 7 .l:!f7 l:!g8 (c) 1 l:!b2 .U.c7 2 J:ib5 e3 3 l:1b3 a well known draw is reached) 6 (on 3 f6 there is 3 ...e2 4 l:rb l l:ra7+, @f3 lid3+ 7 'it>e2 l:rd4 8 c;t>e3 .lldl 9 and an exchange of rooks) 3 ...lie7 4 l!b7 '\t>f6! (on 9 . . . .lld6 there is 10 g5 J:id3 (if 4 f6, then 4 ... e2!) 4 ...'itic5 5 '\t>f5 11 g6!) 10 'it>e4 .llel+? and lidl 'it>c4, and again there is the now after 11 'it>d5 l:!dl+ 12 'itic6 threatof 6 ...a2; 'it>e7 there is no defence against (d) 1 h4! (logical, it is necessary 13... .l:!d6 +!, capturing the d7 pawn. to urge the passed pawn on) 1...'itie5 This is how the game went. 2 h5 c;t>xf5 (after2 ... '\t>f4 3 h6 e3 4 (b) 1 g4+ 'itig6 2 d7 'it>g7 3 .l:!b7 1Ih2 l:!h7 5 f6'it>g3 6 f7 Black can 'it>g6 4 .l:!a7 'it>h6 5 g5+ 'it>g6 6 'it>g4 not win) 3 h6 l:rh7 4 l:1h2 e3 (On lid4+7 �f3 'it>g5,and a draw; 4 ... 'it>g6 there is 5 l:!.h4 e3 6 J::!e4) 5 ( c) 1 d7 ! lixd7 2 g4+ 'itie5 3 g5 Wb2 '\t>f4 6 Wc2 'itig3 7 J:ih5 e2 8 and after 4 l:!f6 a well known theor Wd2 @f2 9 l:!'.h2+ with a clear and etically winning position is uncomplicated draw. obtained. Nevertheless material is material. And now a few examples on the theme of choosing the best continu There was a wide choice for ation. The solutions are given at the White in the following game. end of the book. Rook Endings 97 1 How the "one-legged" Viktor Lvovich tested the youngsters in the endgame Viktor Korchnoi's play has al ways been characterised by the highest class and technique. Before the start of the Berne tournamnent and his match with Lucas Brunner, the veteran broke his foot and the organisers offered to postpone the match but to their greatest surprise the 'patient' had not even thought about refusing to play! From chess Wh at is correct: history it is a well-known paradoxi I Wb5; I 'it>c5 or 1 <:ild5? cal fact that grandmasters with bro ken limbs play very strongly! We 2 mention just two examples-Jan Timman, with a broken foot, won brilliantly at the super-tournament in London 1983, while Alexander Beliavsky, with a broken hand, won the board one prize at the Thessalo niki Olympiad in 1984! Inciden tally, both breakages were sustained playing football. It was rather un usual to see the active Korchnoi sit ting motionless for all his games and only at the end with difficulty moving away on crutches. But he played splendidly, gaining particular How does Black make a draw? success in a couple of rook endings. 3 Brunner-Korchnoi Berne, 1996 What is correct: l . ..f3, l ...'13.. el or l ... '§.cJ? 98 Rook Endings There followed 1 .l:Id2? e3 16 :Ie8+ e2 17 J::l.e3 follows Correct was 1 .l:Id7 f6 2 Wg 1 l:!.e2 17... f4 ! 18 gxf4 l:!.d3!) 15 ...'it>e5 16 3 h3, and it is difficult for Black to l:te8+ lt>d4 17 I:!'.d8+ lt>e3 18 l:tf8 carry out his plan. @£2! 19 l:txf5+ @gl, and White 1. ..lt>h7 2 Wg2? cannot defend against mate. Here it was still not too late to re Well, now we return to the game. turn to 2 I:!'.d7!. 7 ...Wf5 8 J:ih8 2 ... g6 3 fxg6+? If 8 .l:.a3, then 8 ...�d2! (with the The last chance was 3 f6 g5 4 @£3 threat of 9 ...J:id3+! 10 :xd3 e4+ 11 �a4 5 l:te2 g4+ 6 lt>f2 e4 7 11£2+ We3 exd3 12 'it>xd3 Wg4) 9 lt>e3 and 8 l:tf4 with some chances of l:td4 1 o lia6 .l:Ib4 11 Wf3 l:tb3+ 12 holding the game. Now a 3:2 end Wf2 e4 with the unpleasant threat of game is reached, which it seems is ...'it>f5-g4 practically winningfor Black. 8 ... e4+ 9 We3 J:ib3+ 10 'it>f2 Wg4 3 •.. 'it>xg6! 11 .l:Ig8+lt>h3! This is stronger than taking with the pawn, since the passed e5 pawn must be supported by the f-pawn. 4 lla2 h5! Threatening to create a weakness on g3 forWhite after h5-h4. 5 h4 l:tb46 �a8 !l:b2+7 'it>f3 After7 Wh3 ... Usually such an approach of the king leads to a decisive outcome. 12 �g5 There were rather more chances remaining with the preliminary 12 J:ig7, when Black can choose be tween 12 ...f5 and 12 ...:£3+ 13 We2 �xg3 14 lixfl Wxh4. is obtained practically an identical 12 . .. lif3+13 'it>e2f6 ! 14 �g6 copy of the famous game, Smyslov On 14 l:!.xh5 Korchnoi intended to Gligoric, Warsaw 1947, (only with play 14 ... lt>xg3 15 .:.h6 lt>g4 16 h5 White to move and the black pawn f5 17 :h8 lt>g5 18 h6 @g6 19 h7 on e4). There the very instructive lih3, obtaining two connected continuation was 8 lie8 .l:Ie2 9 :e7 pawns. f5 10 �e6+ 'i;g7 11 �a6 (If 11 14 .•.fS 15 lig5 lt>g2 16 'it>el l:!.f2! .l:Ie7+, then l l...'it>f6 12 �h7 Wg6 17 1ig8 13 .l:Ia7 .l:If2 - +) l 1...llf2 12 .l:Ie6 If 17 Wdl, then 17... 'it>fl ! and Wfl '13 l:txe5 �f6 14 lle8 lld2! e4-e3-e2. (zugzwang) 15 J:If8+ (after 15 .l:Ih8 17 ... e3 18 J:ig7 Rook Endings 99 Or 18 �g5 f4 19 gxf4+ 'it>f3 20 7 Wc2 c4?! :g 1 lte2+ 21 Wd 1 lla2 22 �fl We4 The question is where is the pawn 23 :g1 'it'd3- +. best placed-on c4 or on c5. 18 ... f4 ! 19 gxf4 Wf3 20 �c7 We throw in the variation 7 ... @fl For 20 .l:i:gl see above. 8 Wb3 li.a5 9 :t::!'.e3 f5 10 Wc4 'it>f6 20 ... Ira2White resigned. 11 Wd5 f4 12 gxf4 gxf4 13 l:!.c3 Wf5 14 f3 'it>g5 15 'it>c6 Wh4 16 Korchnoi-Kengis Wb6 lta8 17 Wxc5 Wg3, and White Berne Cup, 1996 should not win. 8Wd2 At first sight the endgame looks completely drawn-only the passed a-pawn is potentially stronger than s ...Wg6?! the passed c-pawn. Now Black can After 8 . ..Wfl 9 We3 'it'e6 10Wd4 calmly transfer his king to d6, but Wd6 11 g4 (11 �f3 c31=) 1 l...We6 he decides to display activity on the 12 ..tics We5 13 Wb5 l:l.'.a8 14 a4 king's flank, since White obviously ..t>d5 Black has his own counter intends to move his king over to the chances. a3 pawn. 9 'it'e3 Wf5 10 Wd4 'it'g4 11 �c5! 1...g5 2 hxg5 hxg5 3 Wf3 �a4 'it>h3?! It is always useful to cut off the It is not quite clear where to go enemy king. with the black king. Better looks 4 We2 'it>g7 l 1...lia8 12 Wxc4 f5 13 Wb3 l:i:b8+ Again it is worth trying to go to 14 �a2 lle8 15 llc2 l:i'.a8, and it is d6, though here White can prevent very difficult for White to find a this by 4 ...�f8 5 l:te3 while on winning plan. 5 ...lld4 6 l:i'.d3 l:.c4 7 Wd2 We7 8 12 Wb5 .l:ta8 13 a4! Wg2? :ic3 lla4. After 13... f5 !? 14 a5 l:!.b8+ 15 5 .l:.c3 f6(?) Wxc4 l:.b2 16 a6 l:.xf2 17 lla3l1c2+ The authors would prefer 5 ...c5!? and 18... li.c8 Black holds the draw. 6 '>t>d2 'it'f6 7 'it'c2c4 8 'it'b2 'it>e5 =, If 15 'it>c6, then 15... .i!b2 16 a6 i:Ia2! but Kengis probably feared 6 l:i'.xc5. 17 Wb7 ltxf2 18 lla3 J::[.b2+ 19 Wc6 6 Wd2 c5?! c3 20 li.al c2 21 a7 l::l'.bl. Again preferable is 6 ...Wf7 7 Wc2 14 f4 ! Wh3 15 fx g5 fx g5 16 a5 g4 (7 11e3 f5 8 Wc2 f4 =) 7 ...We7 8 17 a6 �h2 18 !Ia3 Wh3 19 'it'xc4 '>t>b3 bla6 9 a4 'it>d6. blc8+ 20 'it>b5 l:Ib8+ 21 Wc6 .l:tc8+ J 00 Rook Endings 22 �b7 l:th8 23 a7 l:th7+ 24 'it>b6 It looks like Black has defended .l'.Ih8 himself-White cannot queen since then a drawn ending arises because of the distant white king. However Korchnoi findsa path to victory. 25 'it>c6 .l'.If8 26 .l'.Ib3! 1If6+ 27 'it>b5 l:tf828 'lta5 l:ta829 'it>a6 .l'.If8 30 l:Ib8 .l'.Ifl31 a8=tli l:lal+ 32 'ltb5 lha8 33 lba8 'it>xg3 34 Wc4 Black resigned. The young grandmasters made a surprising number of mistakes as Korchnoi demonstrated by clear-cut play. Rook Endings I OJ Exercises: Rook Endings 1 3 Demonstrate the correct What is correct: I ... @f4 or I ... a4 ? plan of defence fo r Black. 2 4 What is correct: I 'f!d6 or 1 .l:.e8+ ? Evaluate the position and find the right plan of defence fo r Black. 102 Rook Endings 5 8 Find the right method of defence Find the right plan fo r White to fo r White. realise his advantage. 6 9 Whatis correct: 1 Wc6 or I c6 ? Find an accurate order of moves fo r White. 7 10 Find the right continuation fo r Black. Find the right continuation fo r Wh ite. Rook Endings 103 11 14 Wh at plan should Wh ite choose? Choose the correct continuation fo r Black. 12 15 How should Wh ite conduct Find th e right continuation the defence? fo r Wh ite. 13 16 What continuation should Whatis correct: 1 'fJ.d5 or 1 'fJ.dJ ? Wh ite choose? 104 Rook Endings 17 20 Whatis the best way fo r White Wh ite to play and win. to achieve a draw? 18 21 How does Whitewin ? What is Wh ite 's winning plan with an exactly calculated variation. 19 22 What is correct: l .. J:!:x:h4or L.'1:..gl+ ? Ho w doesBlack defend? Rook Endings 105 23 26 Wh ite to play and win. What is correct: J ... a8=Vlllor 1 @b7? 24 27 What is correct: 1 . ..@eJ or l ...@xj3? Wh ite to move. How does he win? 25 28 What is correct: J ... h3 or J...@hl ? What is the winning move: L.i:Jilg2 or J...11.al? J 06 Rook Endings 29 32 How does Black win? Can White make a draw? 30 33 How does Wh ite win: 1 h6 or 1 @g4? What must Black do to draw? 31 34 Can Black save the game? Whatgives chances of a win: a) J l:lxa4or b) 1 d7 ? Rook Endings 107 35 38 Threatened by l ...'IJ.g3, how can Where is the clear draw fo r Black? Whitemake a draw? Calculate the variations. 36 39 How can Black make a draw? How does Black make a draw? 37 40 Whiteto play and win. How did Black play? I 08 Rook Endings 41 42 How does White win thega me? Can Black make a draw? 43 Point out a winning plan fo r Black 4 Shouldering: the struggle of the kings A reciprocal attack by a piece is not usually restricted by either time or, more so, by space. However, when we find ourselves talking about kings then the situation is cardinally changed-the king can neither attack, nor, more so, capture the enemy king. Every king creates around itself some strong squares in which can enter any enemy piece, except the king, of course-this feature makes its biggest impression in pawn endings. Manukovsky-Pustovalov 1 'it>e6! 'it>c3 2 'it>d5! 'it>b43 'it>c6 Rosija, 1992 'it>a54 cJi>b7 wins. Also possible is double shouldering. Zinar 1984 .,� • �•••• � � 1...'it>f4! ... . Only so. I...'it>xf3 2 \t>f5 leads to a ... draw. . • After the text Black resigned. ,_.,• .• The mixture of shouldering and zugzwang is indeed deadly. ...• Or yet another classic study by 1 'it>e2! �g2 2 a4 'it>g3 3 'it>e3 Maizelis. 'it>g2 4 a5 'it>n 5 'it>d4wins. J l0 Shouldering However, it is possible to define L.Mitrofanov shouldering, particularly in pawn endings, as a form of zugzwang. Mikhalchishin-Azmajparashvili Tbilisi, 1980 1 rid7! Bad is 1 l:Xt7+@e 4! 2 Ae7+ 'it>d5 3 .Ud7+ 'it>e6 4 nds .Uc5+ and 5 ....Ud5. 1. . . 'it>e42 @g4! The only move-if 2 'i.tg6, then 2 . .l:Ic6+ 3 @g7 ltc7with a win. It is obvious that White is strug 2 ... :c4! 2 Axd2 @e3+ 4 @g5 gling for the draw, and the question @xd2 5 h4 'i¥te3 6 h5 Ac5+ 7 'it>g4!! is whether Black's king will succeed This move was made before, but in getting back. There followed nearer to the opponent's king it se 1 'il;xt7al ='ii' 2 Ibal :xal 3 f5! cures a draw, while the 'more ac It is dangerous to advance the tive' 7 'it>g6? loses because of other pawn. 3 b5? 'it>c4 4 b6 'it>d5 5 7 ... 'it>f4 8 h6 l:Ic6+ 9 'i¥tg7 'i.tg5 IO b7 ltbl, and White has problems. h7 l:Xc7+ 11 'i;;g8 'it>g6 12 h8=(:ZH 3 ...@c4! @f6. If 3 . ..@xb4, then 4 f6 'il;c55 'i¥tg7, and there is no defence against Sometimes even the greats make f6-t7. mistakes in the endgame. 4 'it>e6! Alekhine-Bogoljubow Here is the shouldering-king Wo rld Championship (m) 1929 against king, shoulder to shoulder. There is nothing else-if 4 f6 , then 4 ...@d5 5 @g7 'it>e66 t7 lta7 etc. 4 ...l:Iel+ If 4 ....Ua6+, then 5 'it>e5!-the king goes to the other side, but the main thing is not to allow in the enemy king. 5 'it>d6! Afl 6 @e6 ltel+ and he had to agree a draw. This idea was expressed simply brilliantly in one study. Shouldering I I I 1 . .. 'it>g4?? 2 b7 f5 3 b8=1i' .l:Ixb8 4 .l:Ixb8 f4 Forgetting about the shouldering. 5 'it>d5 f3 6 'it>d4 f2 7 �f8 'it>g3 8 t ...'it>e4 2 b7 f5 3 b8='iV! .l:Ixb8 4 'it>e3 and Black resigned. �xb8 f4 5 .l:Ie8+ 'it>d3 6 J::!f8 c,t>e3 7 @d5 f3 would draw. 5 Complex Endings Playing to exploit a pawn i.xf6! i.xg3+ 8 �xg3 gxf6 9 c6 majority on the flank WVes+ 10 @f3 rs 11 iYcs+ @f7 12 °iVb7+Black resigned. One of the most important factors in the endgame is the possibility of A classical example. creating a passed pawn-and in the majority of cases this is formed from a majority, i.e. from a pawn Botvinnik-Rabinovich Leningrad, 1934 advantage of two against one or three pawns against two. From this follows that the majority is the pre cursor of forming a passed pawn and in itself is an important posi tional factor in the endgame. The fewerthe pieces, the more important this factor, and the greater the role it plays in the position. Playing to exploit the passed pawn in the endgame is the most ex treme case of the majority. Portisch-Ribli The presence of open lines and Skelleftea, 1989 White's pawn advantage on the queen's flank determines his su periority. Among his tasks now is to create a passed pawn. 1 a3 .l:.b7 2 l:!b2 i.g6 3 b4 'it>f8 4 @f2 rJ;e7 5 rJ;e3 �d7 6 i.e2 �c7 7 b5 lib8 After 7 . . . a6 8 a4 axb5 9 axb5 White controls the a-file and m vades the opponent's camp. 8 a4 f6 8 . ..Wb6 follows 9 lid2. 9 a5 More accurate is 9 f4 ! . 9 ... e5 10 lid2 i.f7 11 f4 ! exf4+ 12 �xf4 l:!e813 i.f3i.e6 Complex Endings 113 After 13... ii.. xc4 wmmng is 14 12 @d3 .l:Ibl 13 �el l:!b6 14 lic2 ii..d3 15 l:txc5+ @b8 16 ii..c6 Wc2! lld8 17 l:tcI! with the idea of �e I. Now White needs to defend the 14 ii..c6 g5+ 15 @f3 lif8 16 ii..d5 bl square. lldS17 We3 ii.cs 14 ... �d6 15 .l:Ie4 b6 16 a3! The pawn ending is hopeless. It is better to give up the pawn on lS �a2 ii..b7 19 lid2lies+ the third rank-prophylaxis! On 19 ...ii.. c8 follows 20 l::td3 with 16 ... �g6 17 a5! ligS the sequel We3-d2-c3, h3-h4xg5, There is no saving himself by a5-a6, l:id3-h3. 1 7 ...bxa5 18 ii..xa5 l:!a6 19 ii..d2 20 Wf2 :ids 21 g4 ii..cs 22 lid3 ii..xd2 20 @xd2 l:Ixa3(!) 21 lle5 lif8 23 lie3 f5 24 gxf5 �xf5+ 25 with a decisive advantage forWhite. 'it>g3 The rest is forced. White's position is winning due to 18 a4! I:l.f8 19 l:!e2 bxa5 20 'it>d3 the threats b5-b6, a5-a6. .lldS 21 l:!e4 'it>c6+ 22 @e2 @b6 23 i.el! i.cl 24 g3 hxg3 25 fxg3 l:thS 26 h4 gxh4 27 gxh4 i.b2 2S ii..d2 Botvinnik-Kan i.d4 29 Wf3 i.b2 30 ii.gs lif8+ 31 USSR, 1955 .l:If4 ligS32 llf5 ii..d4 33 h5 @c6 34 h6 @d7 35 'it>g4 @e6 36 l:!f4 �bS 37 lie4+! @f7 38 lie7+ WgS 39 Wf5 with a win. Vukic-Pietzsch Sarajevo, 1967 . � �,�,�,�� ��,�,a�f .�.•�•0. v., �.!.,�,. ;-�� From what we have said before it . follows that White needs to ex �.�.���� change as many pieces as possible. !i•a•n9n, And he sets about realising this � plan. �� r�--��----��� 1 l:id4! f5 2 �e2 ! g5 3 l:!.xe4 lbxe44 liJd2 liJxd2 5 Wxd2! 1... lbe8 The king must head towards the Intending to transfer the king to majority. the queen's flank to neutralise the 5 ... llh6 6 e4 ii..f4 + 7 Wel fx e4 8 opponent's pawn superiority. :rxe4 c5 9 ii..c3 a4?! 2 liJd7! e6 3 Wg2 Wg7 4 �d3 Stronger is 9 ....l:Ia6. l'tJf6 5 liedl liJxd7 6 l:.xd7 h5 7 10 bxa4 l:!b6 11 We2 @d7 @n 'it>f8 8 'it>e2 @es 9 l:!xc7 l:l.'.xc7 On 11.. . .:.b I would have followed By placing his pieces in good 12 .ie5. positions, White forces his opponent 114 ComplexEndings to exchange off, thus increasing the Georgiev-Khalifman significance of the majority factor. Pardubice, 1994 10 �d4 f6 11 f4 @e7 12 b4 b6 13 c;t>d3 J::[.c8 14 @c3 e5 15 fx e5 fxe5 16 J::!d5 @e6 17 a4 a6 17.. . .!lf8! ? would create more problems for White. 18 a5 b5 19 l:!c5!! .:i.xc5 20 bxc5 c;t>d7? And here it was still not too late to escape with a draw by 20 ...b4+! 21 �xb4 @d7=. 21 cxb5 axb5 22 h4! Black resigned. lf5! Martinovic-Yudasin The d4 pawn is well blockaded Krynica, 1998 and White has a serious advantage on the king's flank. 1...�f8? Correct is 1....l:!e8 2 h4 .l:!xe2 3 @xe2 i.d6 4 h5 i.e5!, defending the black king. 2 h4 �e8 3 h5 .l:!xe2 4 c,t>xe2i.d5 Better is 4 ... i.d6 5 f6 gxf6 6 gxf6 i.e5 7 i.h6+ c,t>g8 8 .llg7! with the idea h5-h6-h7. 5 f6 gxf6 6 gxf6 @g8 7 i.h6b4 8 b3! i.e6 9 @f3 i.d5 10 @f2 axb3 11 axb3 i.xb3 12 i.g7 i.f8 13 h6 with a winning position for White. Whose majority is better White's or Black's? After I c;t>e2 Kovacevic-Tosic l2'ld6 2 @d3 e4+ problems with the Belgrade, 1998 b3 pawn appear for White. Best was I f4 l2'ld6 2 @f3 with equality. In the game, however, there followed 1 g4? He should only play on his weak flank when it is possible to create weaknesses for his opponent. 1...l2'ld62 gxf5+ gxf5 3 c5 bxc5 4 bxc5 l2'lb5 5 c6 f4 6 l2'lc2 c;t>d6 7 @e2 'it>xc6 8'it>d3 c;t>dS And Black has a winning endgame. Complex Endings 115 Against doubled pawns, the natu On 16... t'll xe5 decisive is 17 t'llxe5 ral strategy is to exploit the majority i..xe5 18 'it>a5, and if l 6 ... i..xe5, on the other flank. then 17 b5 axb5 18 cxb5 i..d4 19 1 h3! a4 2 'it>f3 l:!.d7 3 l:Ixd7 bxc6. 'it>xd74 g4 hxg4 5 hxg4 b5 6 b3! 17 i..c5+ We8 18 i..d6 f6 19 exf6 Black cannot create a passed i..xf6 20 t'llc5 e5 21 t'llxa6 e4 22 b5 pawn. i..e5 23 i..c5 Clld4 24 i..xd4! i..xd4 6 •.. axb3 7 axb3 'it>e88 t'llh4 25 'it>b3 'it>d7 26 c5 h5 27 g4 hxg4 Preparing f4 -f5. 28 hxg4 g5 29 'it>c4 e3 30 @d3 and 8 ••. t'Llf79 'it>e4 g5 Cllf310 gxf4 11 Black resigned. wxf4 'it>f8 12 t'lld2 t'llh8 13 'it>g5 If 30 ...'it>e6 winning is 31 b6!; Wf7 14 'it>h6! t'llg6 15 t'llf3 t'llf4 16 while on 30... Wc 8-3 1 c6 i..b6 32 'it>h7 c6 17 t'llg5+ 'it>e7 18 Clle4 c4 t'llb4 and 33 Clld5. 19 bxc4 Clld3 20 g5 Black resigned. The following game serves as an example of a bad majority (i.e. a Psakhis-Bonsch Trn ava, 1988 case when the majority can become an object of undermining and attack). Kan-Ke res Moscow, 1952 1 'it>fl! The main trump in White's posi tion is the pawn majority on the queen's flank. Psakhis exploits this advantage very instructively. 1...a5! 1...'Wc8 2 'it>e2 Clld7 3 'it>dl! i.f8 Trying to create a weakness. 4 'it>c2 g6 5 i.c3 .1i.g7 6 a3! .1i.f8 7 2 :i:!.xd8 .l:Ixd8 3 l:idl t'lld5 4 b4 cxb4 8 axb4 t'llb6 9 'it>b3! Cllxd5+ i..xd5 5 c4? White has in effect an extra piece: Creating a bad majority. his king actively supports the pawns 5 ... i.e4 6 :xd8 'it>xd8 7 i..dl on the queen's flank. 'it>c7 8 @fl Wd6 9 f4 e5 10 g3 i..bl 9 ...�e8 10 Cllb2 'iVd7 11 i.d4 11 a4 exf4 12 gxf4 We6 13 @f2 'bes 12 i.e3t'll e7 13 'iVa8!'iVc6 Wf5 14 'it>e3 h5! 15 'it>f3 Black is forced to exchange If 15 h4 i..e4 ! , zugzwang. queens and the remainder becomes 15 ... h4! 16 We3 'it>e6 17 .ig4+ a matter of technique. If 17 @f3 i..f5 18 'it>g2 �f6 19 14 'iVxc6 tllxc6 15 t'lld3 i..g7 16 'it>h2 i.e4! and Black has a decisive �a4! Wf8 advantage. I J 6 Complex Endings 17 ...�f 5 18 'it>f3 'it>f6 19 �xf5 1 'it>t1 'it>t7 2 g4! ctJe6 3 CLixetS 'it>xf5 20 'it>e3 f6 21 'it>f3 g5 22 fxg5 'it>xe6 4 ctJe4 d6 5 g5! fx g5 6 fxg5 23 'it>g2 g4 24 hxg4+ 'it>xg4 25 '2Jxg5+ 'it>f5 7 '2Jxh7 'it>xf4 8 ctJf6 <;tin'it>f4 White resigned. �cs 9 '2Je8 �f5 10 c3 �bl 11 a3 �a2 12 b4 c5 13 '2Jxc7 �c4 14 h4! Ljubojevic-Ivanchuk cxb4 15 axb4 i;t>e4 16 ctJe8 <;ties 17 Buenos Aires, 1995 'it>e3 �f7 18 '2Jc7 �g6 19 ctJb5 'itd5 20 CLixa7 'it>c4 21 ctJc6 Black resigned. Toothill-Heemsoth corres, 1986 White has a pawn majority on the queen's flank, therefore he can boldly improve the positions of his pieces which only increases his positional advantage. 1 c4! �xg2 2 1Ixd7 .l:!xd7 3 .l:Ixd7 l ...li f8! 'i!Vxd7 4 'iitxg2 'i/Nc6+ 5 'it>gl f5 6 With the unequivocal intention of �b6!! �e4 7 b4 e5 8 b5 f4 9 gxf4 driving back the opponent's king exf4 10 bxa6 bxa6 11 'i!Vxa6 'ilYel+ from the centre. 12 'it>g2 and White has a decisive 2 :i.d2 f5+ 3 gxf5 gxf5+ 4 'it>f3 advantage. c4! 5 a4 a6 6 �f4 .l:Id8! 7 .l:Ixd8 Hiibner-Spassky Taking into account the previous Candidates (m) , 1985 comment, more chances for White rested with 7 lic2,retaining as many pieces as possible. 7 ...�xd8 8 �e3 b5 9 axb5 axb5 10 �cl �c7 11 h3 'it>d5 12 i;t>e3 b4 13 �d2 b3! 14 f3 f4+ 15 'it>t1 ..)te5 16 .)tel �d4+ 17 'it>el ..)te3! White resigned. Mutual Majorities Each side has his majority, and the main problem is how to limit mobility fromthe opponent's side. White's extra pawn on the king's With level pawns, if one of the flank proves a decisive factor. rivals has a majority on one flank, Complex Endings 117 then for the other it is the opposite Yates-Alekhine -this is a fact. But often it will TheHa gue, 1921 happen that one majority is better than the other, then we speak of a qualitative majority. Karpov-Yusupov Dortmund, 1997 A classical position-White loses it without a struggle. 1 g3? Incorrect prophylaxis! 1 ...@f7 2 c5 '\¥i>f6 3 �c4 .tc8!! A terrible move. For Black the 1 'Llc4! bishop is more important-so it will It is necessary to keep in his sights stand firmon the c8 square. the opponent's weaknesses 4 a4? White's plan consists of an ex Here it is worth playing 4 f4 !? change of queens and advance of exf3 5 c,t>f2 l:!d2+ 6 i;£;>xf3 .!lxh2 7 b5 pawns on the king's flank. with counterplay. i...'Lld72 b3 rs 3 c,t>n 4 ...gs 5 b5 f4 6 c,t>n .!id2 7 gxf4 Also good is 3 g4 g6 (3 .. .fxg4 4 gxf4 8 @el l:!b29 .te2 hxg4 'i/Nf3 5 'ii'f5 ) 4 gxf5 gxf5 5 After 9 l:idl �g4 10 .!ld6+ 'i¥i>e7 'if'c3±. 11 1Id4 �f3 12 �d5 .l:Ibl+ 13 'it>d2 3 ...@f7 4 f3 'i¥i>e7 5 'iite2 iie6 6 e3+! 14 fx e3 .lldl+ 15 'i¥i>c3 .!lxd4 'ii'c3 'Llf6 7 'it>f2! ? 16 'i¥i>xd4 �d5 -+. The king will be needed on the 9 ...'\¥i>e5 10 c6 bxc6 11 .l:Ixc6 �e6 other side. 12 �dl .!lbl 13 .Iles+'i¥i>d4 14 .llc2 7 ...'i¥i>d7 8 g4 'i¥i>c6?! e3 15 fxe3 fxe3 16 .l:Ic6 .tg4 17 Better really is 8 ... fxg4 9 hxg4±. .lld6+ 'it>e5 18 h3 �h5! White 9 'ii'e5! resigned. Now the knight ending is But correct prophylaxis was hopeless. possible: 9 ...'iYxeS 10 'Llxe5 'i¥i>d5 11 'Llc4 (a) 1 f4 ! (intending c,t>g l-f2-e3) fxg4 12 'Llxb6+! i;£;>c6 13 'Llc4 gxf3 l:id2 2 c5 i;£;>f7 3 .tc4 .txc4 4 .llxc4 14 'Llxa5 'i¥i>d5 15 i;£;>xf3 g5 16 tt::lc4 i;£;>e65 b5 i;t>d5 6 c6 bxc6 7 bxc6=; h5 17 'Lld2! i;£;>e5 18 e4 'Lle8 19 i;£;>e3 (b) 1 f3! (to create a weakness on 'Llc7 20 'Llc4+ 'i¥i>f6 21 'i¥i>f2! 'Lla622 e4) l...e3 2 f4 ! .lld4 3 l:!e1 .l:!xf4 4 'it>g3 'Llb4 23 h4 ! 'Llc6 24 a5 'Llb4 l:!xe3 .tc4 5 .txc4.:i.xc4 6 .:.e7 with 25 'Lld2!±. a drawn ending. J J 8 Complex Endings Bronstein-Rantanen Portisch-Kramnik Tallinn, 1975 Biel, 1993 1 g4! 1.l:tc3 Advancing on the other side-but If 1 'l'Wa3 tt:Jc4 2 �4 b5 3 a4 here it is important to force back tiJb6! 4 axb5 tiJd5! Black maintains and restrict the black pieces. a small advantage due to his control 1...a6 2 g5 tt:Je8 3 a4 l:!'.a7 4 h4 over the central squares. �b7 1. .. b5 2 :i.fcl Better is 4 ...@f 8!?. Bad is 2 l:!'.a3? because of the 5 l:1d3 tactical blow 2 ...tt:Jc4 3 .!lxa6�7. Now it is important to control the 2 .•.tt:Jc4 3 'i¥e2 d-file. On 3 tiJd2 there is the effective 5 ....l:tc5 6 llel li:td7 7 .l:Iedl! 3 ...a5 !; also in the event of 3 e4 Tactics. If 7 ...e5 there is 8 ctJf5±. there is 3 ...f5 !. 7 ... g6 8 tt:Je2! 3 ... tt:Jb6! Typical. With the idea of capturing on c3. 8 •.• .l:r.xd3 9 l:1xd3 b5 10 cxb5 4 .l:Ic7 axb5 11 .!id7 @f8 12 a5 .!lc6 13 4 i.e5! ? deserves attention. .!lb7 b4 14 .l:Ib8 i.c5 15 tt:Jg3 Black 4 ... 'ii'e6 5 i.g5 resigned. And now the only chance was 5 i.e5! ? with a slight advantage for A classic example of exploitation Black. of a pawn superiority on the queen's 5 ...liJd5 6 .l:I7c5 h6 7 i.h4 b4 + 8 flank is presented by the following WVh2 position from a modem tournament. There is no saving himself by 8 'l'Wc4 l:1xc5 9 'ii'xc5 tt:Jc3 +. Here Black has a pawn superiority 8 .•.tt:Jc3 9 .l:txc8 l:!.xc8 10 'it>hl on the queen's flank, typical forthe 'it>h7!! Griinfeld Defence, and therefore Prophylaxis. 10... a5 11 a3 tt:Ja4 12 White should keep as many pieces .l:Ixc8 'i!Vxc8 13 ii'b3 'ii'cl+ 14 'i¥th2 as possible so as to maintain the '¥.Wxa3 15 '¥.Wd5 gave White tension. counter-chances. Comp lexEndings 119 11 l:Ial Not 12... exf 4 because of 13 .l:!f2. Other moves are also no help: 11 13 J:Ie2 dS 14 i.esi.c7! �3 tllxa2! or 11 a3 tlla4. An exchange of strong pieces. 11... aS 12 'iVb3 °iVxb3 13 axb3 g5 15 i.xc7.l::!. xc7 14 i.g3 a4! +. But now the weakness of the b2 and g3 pawns is felt. In complicated endings are 16 �fl J:ib7 17 c3 �bS! 18 l:!c2 usually to be found complicated 'it>c4 19 @e2 .:i.b3! 20 �e3 d4+! plans (many stages), various pieces A blow! and asymmetrical formations. 21 �xe4 dxc3 22 bxc3 :xa3 23 Therefore it is necessary to pen g4 hxg4 24 l:!g2 l:!b3! White etrate deeply into the position, tak resigned. ing care over the exchange of pieces. Mikhalchishin-Beliavsky Moscow, 1981 Balashov-Korchnoi USSR (c h), 1969 Black has a weak pawn on c4, but together with this there is strong White seems to have a very cen counterplay on the d-file. White tralised position and a solid pawn should combine the improvement of formation, but Korchnoi begins to the positions of his pieces with undermine the white structure. prophylaxis. 1...!If8! 2 tlld2 f6 3 .lift 1 tllc5! i.dS Whit� wants to exchange rooks to If 1 . . ..:i.d2, then 2 i.fl! .!lxa2 3 neutralise the pressure. tllxe6 fxe6 4 i.xc4 with material 3 ... i.d8! 4 J::rf2 Ji.c7 advantage. Increasing the pressure on the 2 f3 .!lc83 tlla4 i.e64 e4 centre-improving the position of White's cunning manoeuvre has the bishop. cut offBlack's play along the d-file 5 Cllfl i.b6! 6 :d2 fxeS 7 i.xeS and restricted Black's setup with an J::rfi 8 @e2 d6 9 i.c3 i.c4+! 10 ideal structure f3-e4. Only nowdoes Wet ii.xfl! his king make an appearance in the A surprising exchange of the centre. strong bishop, but now it is more 4 ... llld7 s Wf2 lllb6 6 tllc3! important to create a passed pawn. The knight is best placed on a 11 'it>xfl es 12 'it>g2 e4! blockading square. 120 Complex Endings 6 .•.f5 7 'it>e3 fxe4 8 fx e4 @f8 9 It is very difficultfor White to im .l:Ibl ! prove his position, therefore he re Now he has time to improve the sorts to such unnatural maneouvres. position of his rook-the threat is But 3 h4 deserved serious attention. l:tbl -b5-a5. 3 ... e5 4 �e2 'llg5! 5 �cl Clie6 9 ... �d7 10 'it>d4 rtie7 11 l:!'.b4 With each move Black improves .!:i:d8 the positions of his pieces-the If 11...Wd6, then 12 e5+ 'it>e7 13 same cannot be said of White. a4!, taking from the knight its last 6 Cllc3 rs 7 a3 �c6! 8 b4 i.d4 9 good outpost on b6. Cllb5?! 12 �xc4! i.h3+ 13 Clld5+ Clixd5 White did not think that his oppo 14 exd5 �xg2 15 We5! nent would so readily part with his Pawns are equal but the activity of bishop, therefore better was 9 i.d2. each of White's pieces is clearly 9 ...�xb5! 10 cxb5 @f6 11 �c4 greater-this is also a typical case of i.c3! domination. Preparing the d4 square for the 15 ... :gd7 16 lib8! �f3 17 .l:Ig8 g6 knight, while preventing a3-a4 or 18 d6+! lixd6 19 l:!'.g7+ @f8 20 i.cl-d2. l:tf7+ Black resigned. 12 c,t>n 'lld4 13 f3 f4 ! Fantastic technique in a compli cated endgame was demonstrated by lvanchuk in the following game. Dao Thien Hai-lvanchuk Moscow, 1994 A decision by a grandmaster of extra class-weakening the white squares rather too much, but in re turn restricting to the maximum the white king and bishop cl. 14 @f2 h5 15 g3 If White waits, then Black pro At first sight it seems incredible ceeds with ...g7-g5-g 4 and then that Black should be playing for a ...@f 6-g5-h4 with the threat of win-White has a pawn majority on ...g4xf3 and ...@h4-h3 . the queen's flank. First of all Black 15 .•. gS 16 Wg2 'llc2! improves his structure on the king's Surprisingly, after the knight flank an� thereby restricts the oppo move many threats appear-for nent's pieces. example, not possible is 17 i.b3 1. .. f6 ! 2 Clie3 'it>f73 'lldl because of 17... Cll el+- this is a Complex Endings 12 J consequence of the restricting ma 4 .l:Id5 \¥tg6 5 g4 i.f6 6 Wg2 i.g5 noeuvre ...f5 -f4 7 i.g3 i.e7 8 h4 lic4 9 lid7i.f 6 10 17 @fl i.b2! f4 One cannot leave out the tactical Black already has some trouble element-it has been well known -White activates his phalanx to the since the time of Capablanca that to utmost. achieve the maximum one must 10 ... .:i.c5 11 .l:Id6 'it>g7 12 i.f2 transpose positional factors into .l:.a5 13 lid3�b5 14 'it>g3 tactical ones. Now that the rook defends the 18 i.xb2 third rank, White improves the If 18 i.b3, then 18. .. i.xcl 19 position of his king. i.xc2 fxg3 20 hxg3 i.a3,and with 14 .. Jla5 15 i.b6 .l:!b5 16 i.e3 the presence of the passed pawn on i.b2 17 'it>f3 i.al 18 �dl i.c3 19 the h-file Black ought to win easily. .l:Icl .l:Ib320 'it>e4! 18 ...l'De3+ 19 'it>e2 l'Dxc4 20 i.cl Beginning a very important Wf5 21 @d3 l'Dd6 22 a4 g4 23 stage-activating the white king. fxg4+ 'it>xg4! 20 ....l:!b4+ 21 Wd3 i.f6 22 g5 Precisely this-the main thing is hxg5 23 hxg5 i.d824 �c6! to break through to the h2 pawn. Preventing 24 ...f6 , on which fol 24 gxf4 exf4 25 i.b2 f3 26 i.d4 lows 25 .l:Id6 i.e7 26 l:!d7!+-. 'it>h3 27 a5 'it>xh2 24 ...I:!'.b3+ 'it>e4 25 f5+? Not everyone would pay attention It is not clear why-the fact is that to such 'trifles' as the b6 pawn. White does not threaten the move 28 axb6 axb6 29 i.xb6 'it>g2 and 26 f5 because after 26 ....l:Ixe3+! 27 White resigned. 'it>xe3 i.xg5+ Black has a certain draw. 26 'it>f3i.e7 27 :gc7 Wf8 28 'it>e2 Here is a very complicated re Ji(b2+29 �d3 �b3 30 'it>d2 We8 31 alisation with pawns on just one i.c5! i.d8 32 .l:Ig7 i.a5+ 33 We2 flank. .l:Ib2+ 34 Wd3 J::!d2+ 35 Wc4 l:!e2 36 g6! .l:!e4+37 'it>d3 4.l:!xf 38 I:!'.e7+ Romanishin-Nunn Wd8 39 g7 .l:Ig4 40 .:i.tl! Black Debrecen, 1992 resigned. A finepiece of work. N ezhmetdinov-Romanovsky Moscow, 1957 1 g3 h6 2 i.f4 'it>h73 lid2 .l:tc6 Stronger is 3 ....l:Ic5 and 4 ...h5. 122 ComplexEndin gs What plan should White choose? 1 lic3! 1 g4!! Clearly not 1 a4 because of With the intention of playing l ...Cll b4, and the black knight jumps further f2-f4 and e4-e5, gaining to freedom. ground, therefore Black's reply is 1...lie6 2 llle5 g5 3 ..tg3 Wg7 4 forced. Cllf3 h6 5 h4! 1...hxg3 2 ..txg3 Cllf8 3 'it>h2 llle6 Such moves are always unpleas 4 f4 b5 5 .l:Igl g6 ant. White latches on to the weak White has threats along the g-file, ness and wants to 'occupy' it. If therefore Black is forced to weaken 5 ...f6 , then 6 h5 Wf7 7 .l:Id3 with the himself. idea Clld4, though Black can of 6 Cllh6+ <#;;g7 7 lllg4 f6 course defend himself for a long How otherwise to deny the knight time. access to e5? If 7 .....tf6, then 8 e5 5 ... g4 6 Cllel ! and 9 f5 with a strong attack. It turns out that the knight is head 8 f5 ! ing for the f4 square. Reminiscent of the play in a 6 .....td6 7 ..txd61ixd6 8 .l:Ic4 h5 9 famousLasker-Capablanca game. <#;;e2 8 ...lllg5 9 .l:Igfl c5 10 e5! ..txf5 The exchange of bishops has not On 10 ...fxe5 White plays 11 eased Black's position very much. ..txe5+ Wh7 12 lixd8! l:!xd8 13 9 ...lll b8 10 lies 'it>h6 11 llld3 fxg6+'it>xg6 14 :i.f6+ 'it>h7 15 .l:Ih6+ Clld7 12 lia5 a6 13 g3 .l:!b6 14 with decisive threats. lllb4! 11 exf6+ 'it>f8 By covering the file, White pre On l l...i.xf6 there is 12 Cllxf6 vents the activation of the black 'it>xf6 13 i.h4. rook. 12 llle5! b4 13 Clld5 i.xc2 14 .. Jie6+ 15 'it>fl 'it>g7 More stubbornwas 13 ... lllf7. If the passive 15... lll b8, then 16 14 f7 lie6 15 llld7+ Black Clld3 l:.b6 17 Cllf4 .l:Ib5 18 .l:Ixb5 resigned. axb5 19 'it>e2 llla6 20 llld5! with a winning knight endgame. Now Lputian-Tukmakov however any capture is quite good. Moscow, 1983 16 lixa6.l:!e8 If l 6 ... lle4, then 17 Clld3 llle5 18 Cllf4 ! Clif3 19 lllxh5+ 'it>f8 20 lia8+ We7 21 lie8+!! , exploiting the tactics. 17 Clld5 lllc5 18 I:!'.c6 llle6 19 lic4 lib820 a4 lib2 21 lllf4 libl+ The rook ending is hopeless, since the king simply goes to his passed pawn. 22 'it>e2 Wf6 23 Cllxh5+ We5 24 J::!xg4 And Black soon resigned. Complex Endings 123 Salov-Khalifman Nezhmetdinov-Luik Candidates (m) 1994 USSR 1950 White has a majority plus a space White has the advantage on the advantage. king's flank plus a more active de 1 h4! f6 2 h5 ployment of pieces-there followed Spoiling Black's majority. 1 b4! 2 ... a6 3 tl:ic3 tl:ie5 4 tl:id5+ c,t>dS 5 With the threat of 2 i.c5,whereas b3 b5!? l...i.xb4? is impossible because of Without this move Black would 2 .!la8+i.f8 3 i.c5. not have any real counterplay. 1...J..eS 2 i.c5@f7 3 g6+! 6 cxb5 axb5 7 b4! i.g4 S a3! Completely cramping the oppo @cs nent-not possible is 3 ... hxg6 4 After 8 ... i.xh5 9 tl:ie6+ attacking fxg6+ 'it>xg6 5 tl:ixe5+. the g7 pawn. 3 ... hxg6 4 fx g6+ @rs 5 h5 i.xc5 9 tl:ie7+! @c7 10 tl:ig6 i.xh5 Already he has to allow White to Or 10... tl:ixg6 11 'it>xg4 tl:ie5 12 create a passed pawn. @f5 +-. 6 bxc5 J::rc7 7 .!las l:!xc5S h6! 11 tl:ixe5! fxe5 12 c;t>xe5gs Here also an advantage on the After 12... i.g4 13 tl:ie6+ i.xe6 14 king's flank expresses itself in a c;t>xe6 the pawn ending is simply tactical form. lost. S ... gxh6 9 tl:ih4 l:!'.c6 10 tl:ifS .lle6 13 tl:ie6 @b6 14 'it>f6 i.e2 15 @g6 11 l:!bS J..fl 16 g3 'it>c6 17 'it>xh6 'it>d5 lS White plans to go with his king to tl:ixg5 i.e2 19 @g6 'it>c4 20 tl:ie6 h5, therefore he places his rook to cJi>c3 21 @rs 'it>b2 22 tl:ic5 'it>xa3 24 hold up the b-pawn. tl:ia6! and there is no defence 11... h5 12 'it>g3 b4 13 c,t>h4 b3 14 against the march of the g3 pawn. @xh5 b2 15 'it>h6c;t>gs 16 J::rxb2 And as Black is in zugzwang, he Exploiting a space advantage ... resigned. 124 ComplexEndings An entertaining endgame motive It seems that there is no hint of was produced in the game danger in White's position, but... 1 ..id2 ltJeS 2 'it>g3 ..ic6 3 f4 Vaganian-Smirin USSR (ch), 1988 liJd3! 4 i.e3 eS S fxeS fx eS 6 'it>h4 rJlf6 7 'it>g3 'it>e6 8 c;t>M 'it>f6 9 'it>g3 aS 10 ..igl a6 11 ..ie3 Black's two doubled pawns successfully counter White's three thanks to the excellent position of the knight on d3, while the passed e5 pawn has become dangerous, since the white king cannot get over to the e-file. 11... 'it>e6 12 rJlh4 ii.bl 13 rJlg3 'iitd7 14 ltJdl lbb4 lS ltJc3 rJlc6 16 'it>h4 ltJc2 17 i.f2e4 18 <;fo>gS e3 19 i.g3 'it>xcS 20 ltJe2 i.f3 21 ltJgl ..idl 22 h3 hS 23 'it>xg6 hxg4 24 1 c6 is striking with the intention hxg4 ..ixg4 2s 'it>gs ..idl 26 @rs of winning in the variation l . . . ..ixb4 'it>d4 27 'it>e6ltJb4 28 ii.el ltJxa2 29 2 c7 lbb6 3 lbc4. But Black plays ..ixaS ltJcl 30 b4 ..ig4 31 rJld6 e2 I...lLlb6! 2 c7 (or 2 b5 ..id6 3 lbg2 White resigned. 'it>f8 with equality) 2 .....id6! 3 c8=='i¥ lbxc8 4 ..ixc8 ..ixb4, and is in no Space plus control of an open danger. The game, however, went line. 1 'it>g2 Surprisingly for White, Black is Botvinnik-Balashov saved by a piece sacrifice. Hastings, 1967 1...ltJxcS!? 2 bxcS ..ixcS 3 ltJg4 ..id4 4 ..ic6 c;t>g7 S h4 ii.cs 6 <;fo>g3 ..id6 7 'it>f3 i.e7 8 hS gxhS 9 lLlf2 fS 10 liJd3 'it>f6 11 ltJeS i.d6 12 lLld7+rJle7 13 i.bSh4 The draw is obvious. Neverov-Dreev Lvov, 1985 1 a4 f6 2 aS White increases his space advan tage and thereby drives back the black pieces. 2 . .•..id8 3 l:!c8'it>f7 4 e4 �d7 S bS ..ie76 ..id2 ..id67 l:l.'.a8b6 8 a6! Now not only is there a weak a7 pawn but, when the opportunity Complex Endings 125 presents itself, also the manoeuvre The invasion of the king is decis .l:!a8-b8-b7. ive-this is more accurate than 8 ...i.cs 9 i.e3 @e7 10 h3 �d6 i.xe7 and ll:lc8+. 11 lle8 :c7 12 e5+! 'it>d7 12 .l:!g8! 27... ll:lg6+ 28 @hS ll:le7 29 i.xe7 There is no sense in winning a @xe7 30 'it>g6! Blackresigned. pawn at the cost of exchanging rooks. Playing to exploit a weakness. 12 ... fS 13 :xg7+ Wd8 14 l:!g8+ @d7 15 £i.h8 i.e7 16 i.d4 .l:!c2+ 17 Korchnoi-Pinter 'it>d3.l:!a2 18 .l:Ia8+Black resigned. Reggio Emilia, 1987188 Tikhomirova-Morozova Kiev, 1966 1 ll:le2! Hurrying to activate his pieces would be punished: 1 .l:!b3 ll:lc5; 1 White's plan is simple-to push 'it>e3 ll:le5. the pawn to a6 and still further i...:c8 2 l:!c3 g6 3 I:!'.b4 .l:!c7 4 hamper the opponent, while in the ll:lf4:hc8 event of an exchange on a5 the c5 The threat was 5 ll:ld5. pawn will be very weak. 5 as @d6 6 llbl ll:le5 7 h3 ctJd7 8 1 ll:lcl! ll:lf5+ 2 'it>d3 �e8 3 a4 h4 'it>d7 4 ll:lb3 ll:lh4 5 g3 ll:lf5 6 a5 Gradually occupying space. ll:ld87 a6 @c7 8 ll:lbd2 8 •..h5 9 'it>e3 ll:le5 10 l:tc2 @e7 11 The knight has done its business 'it>d4!Wd6 and now transfers to the other flank. On l l . . . �d8+ White had prepared 8 •.• @b8 9 ll:lf6! ll:lh6 10 ll:lde4 12 ll:ld5+! (12 'it>xe5 f6 mate) @c7 ll ll:le8+'it>d7 12 ll:l4f6+ i.xf6 12... 'it>e6 13 rt;e3. 13 ll:lxf6+ rt;c7 14 h3 ctJf5 16 g4 12 :d2 l:td7 13 @c3+ 'it>c7 14 Now it is time to expand on the lidb2rt;d6 15 a6 other flank. Securing a bridge-head for an in 16 ... hxg4 17 hxg4 ll:le7 18 ll:le4 vasion along the b-file. 'it>b8 19 .Itel ll:lg8 20 i.h4 @c7 21 15 ...1:.dc7 16 :b3 ll:ld7 17 l:tdl+ @e3 rt;d7 22 @f3'it>c7 23 ll:ld6 f5 rt;e7 18 ll:ld3 .l:Id8 19 l:tdbl l:tdc8 24 gxf5 gxf5 20 .l:!b7'it>d6 21 f4 ll:lb6 Somewhat better is 24 ...exf5 . A mistake in a difficult position. 25 i.g5 @d7 26 @g3 ll:le7 27 He should not lose control of the e5 @h4! point. 126 ComplexEndings 22 tl\e5 'it>e6 White activates his pieces to the The threat was 23 tbxfl. maximum and also the passed pawn, 23 t2Jc4 .l:Ixb7 24 axb7 :gb8 25 and in the process threatens to enter tba5! with the king on b5. Leading to complications was 25 8... @d7 tl\xb6? .l:Ixb7 26 f5 + gxf5 27 exf5+ He must activate at the cost of a 'it>xf5 28 l:tfl+ 'it>g4. pawn. 25 ... @d6 26 Wb4 'it>c7 27 �cl 9 i.xg5+ 'it>c6 10 i.xf4 :xr4 11 'it>d728 'it>a3 f5 29 e5 tl\d5 30 e6+ :a7 l:.f3+ 12 'it>c4 .U.f4+ 13 'it>c3 Black resigned. .:i.f3+ 14 'it>b2 'it>xc5 15 l:Ixa5+ Connected pawns in the majority Weaknesses in the endgame of cases win against a single pawn. and their exploitation 15 ...'it>d4 16 1:1'.aSe5 17 a5 .l:If6 If 17... 'it>c5, then 18 1:1'.bS!, cutting The presence of weaknesses in the offthe king. pawn structure is a great minus, and 18 :es e4 19 b4 e3 20 'it>b3 l:tfl the correct technical exploitation of Or 20 ...l:\'.f5 21 a6 l:l'.e5 22 .l:Ixe5 them is the most important factor in 'it>xe5 23 a7, and in the queen end realising an advantage. ing White wins the enemy queen. 21 I:!'.d8+ 'it>e4 22 a6 .l:Ial 23 b5 Topalov-Beliavsky e2 24 .l:Ie8+ 'it>d3 25 Wb4 'it>d2 26 Linares, 1995 'it>c5 el=11Y 27 1:1'.xel'it>xel Black has two passed pawns but 28 'it>b6! at the same time these are serious A typical means of advancing the weaknesses which need looking b-pawn to queen-the only way to after. victory. 1 c4 .i.g3 28 ...'it>d2 29 'it>a7 Black resigned. Bad is 1...g5 because of 2 .i.f6!. 2 'it>c2i.f 4 3 i.f6 g5 4 l:te7+ As in rook endings a single weak Inferior is 4 .lig8 l::tc5! with the ness will usually be defensible, to idea of ....l:Ic5 -f5 achieve victory it will often be 4 ...'it>c8 5 .l:Ig7 l::tc5 6 'it>d3 a5 7 necessary to create a second i.e7 l:\'.f5 8 c5! weakness. Complex Endings !27 Kasparov-Andersson Hiibner-Ftacnik Belgrade, 1985 Polanica Zdroj, 1995 '":•i•Z . �. ��.��- nf!--�-�e-���.�r �.�����. r,ci . ;i;� ;i;� ?;,, !Y,�f'� m.,,,, .�e"""!Y,�r� �+� • u � � .. .. ,•i•",.,.!Y,��r0{'"'!Y, All Black's pieces are forced to 1 g4! i.d5 2 i.c2 g6 3 tt::lg3 protect the b7 pawn, while White @e7?! must create a weakness in the oppo The first dubious manoeuvre nent's camp. To the question better is 3 ... i.d6 4 tt::le4 i.e7 where?-the reply is simple--0n 4 tt::le4 tt::ld8 5 g5! the g6 square! Beginning a weakening of Black's 1 h4! @n 2 h5 @g7 3 Wg2 l:!e7 4 position. .:i.b6 lif'l 5 i.d5 .l:Ifd7 6 lilb5 .l:Ie7 5 ... fxg5 6 tt::lxg5 h6 7 tt::le4 i;i{f'l 7 @g3 lied7 8 @f4 .l:!c2 9 i;i;>g5! But not 7 ...Ci:Je6? because of 8 J::rxf2 10 hxg6 hxg6 11 lixg6 @rs tt::lc3 12 i.b3 tt::lf'l 13 Wf6 f4 14 e4 l:!b2 8 tt::lc5 i.d6 9 tt::le4 i.e7 10 tt::lc3 15 e5 f3 16 e6 f2 17 il.c4! Black i.e6 11 @g2 g5?! resigned. Another weakening-better is Commentary here is superfluous 11...tt::lc6 since White's play was direct and 12 i.e4 i.d6 13 tt::le2 i.c4 14 logical. tt::lg3 We6 15 @f3! With the threat to go to h5 15 ...i.d5 16 @g4 i.xe4 17 tt::lxe4 Play in equal positions tt::lf'l 18 @h5 b6?! Stronger is 18. .. @f5 19 tt::lxd6 In equal positions first and fore tt::lxd6 20 @xh6 g4, but White right most the task is to create weak ly plays 19 tt::lg3 @f6 20 f4 . nesses in the opponent's camp. This 19 @g6 ii.rs 20 i.c3 tt::lh8+ 21 is usually achieved through the Wh7 tt:::in 22 i;i;>g8 i.e7 23 f3 h5 24 greater activity of one's pieces, Wg7 g4 25 fxg4 hxg4 26 @g6 a5 27 which, with the help of an attack, i;i;>h5 Wd5 28 tt::ld2 axb4 29 i.xb4 force a weakening of the pawn i.xb4 30 axb4 i;i;>c6 31 i;i;>xg4 Wb5 structure. 32 �f5 @xb4 33 @g6 tt::ld8 34 @f6 +-. 128 Complex Endings Endings with a pawn sacrifice Barlov-Seira wan Zagreb, 1987 In principle, a material advantage in the endgame is the main factor, but it can happen that other factors, particularly the presence of a passed pawn, are decisive. Rublevsky-Shariazdinov Elista, 1996 Though Black is a pawn down, the American grandmaster tries to exploit the active positions of his pieces and does this, as we shall see, successfully. 1...f4 2 gxf4 gxf4 3 lZ'ic4 :gd3 4 �al h5 5 h4 'itf5 6 .l:Ia8 Wg4 Black obviously does not rush to capture the white pawns, for the 1 a5! i.h42 i.d5i.a6 3 lZ'ic7! time being improving the position The threat is b4-b5. of his king. 3 ...i.d8 4 lZ'ie8! 7 �g8+ @xh4 8 J:ig6J:ixd5 9 Wf3 If 4 lZ'ixa6? bxa6 5 i.b7 i.g5! l:td4 10 .l:Igl i.c5 11 @e2 i.b4 12 with the idea 6 ...i.d2=. @f3i.c5 13 @e2 l:te4+ 14 @f3d5 4 ...@h7 5 'itg2 15 lZ'id2 �es 16 J:idl @gs 17 lZ'ibl 5 lZ'ixd6? i.e7 would be bad d4 18 lZ'id2 wrs 19 :ghl .:i.h8 20 materialism. lZ'ie4 i.b4 21 J:igl h4 5 ... g5 Slowly the black pawns crawl On 5 ...'itg6 there is 6 h4!. along to their queening squares. 6 i.f7! 22 l:1g5+ @e6 23 _:g6+ Wf7 24 Zugzwang! l:tgl d3 25 Wxf4 d2 26 'ite3 l:td8 27 6 ...i.e7 7 lZ'ic7i.d8 8 lZ'ixa6! We2 .:i.e8 28 f3 h3 29 .:bl :ds 30 Now that the g5 square 1s 'it>dl inaccessible to the bishop. If 30 lZ'if2 Black has 30... h2 31 8 ... bxa6 9 i.c4 'it>g7 10 i.a6 d5 l:1xh2 dl=�+ 32 lZ'ixdl .:i.d2+. 11 exd5 i.e7 12 i.c8 i.xb4 13 a6 30 ... �aS 31 We2 h2 32 lZJf2 i.c5 14 @f3@f6 15 d6! h5 16 We4 If 32 lixh2?, then 32 ...dl= v&'+ 33 i.d4 17 @d5 @f7 18 i.f5 @f6 19 @xdl lr.al+ and 34... lia2+. @c6! Black resigned. 32 ... J:ieS 33 lZ'ie4 .l:td8 34 lZ'if2 i.c5 35 lZ'idl i.gl- +. Co mplex Endings 129 Chekhov-Karsa 1 @f2e5 2 J:id2 CZJa4 3 ..tn lLib6 Lvov, 1983 4 ClJc2 ..tc4 5 ClJe3 ..tb3 6 g4 h4 7 g5 ..te6 S ClJg2 h3 9 ClJh4 ..tb3 10 ..txh3 llJc4 11 .!ide2 .!idS 12 f4 ..tdl13 .li(e4 llJxb2 The strategy is completed and the game transposes to tactical channels. 14 @gl .!id3 15 ..tn l:txc3 16 f5 ! ..tc2 17 f6 l:teS lS J:Ixe5 .!lxe5 19 .!lxe5 l:txa3 20 .Iles+ 'it>h7 21 .!lf8 ..tb3 22 llJf3 l:!a4 23 llJe5 l:!e4 24 lLixf'l.l:Ig4+ 25 @f2 ..txf'l26 .l:Ixf'l+ 'itgS 27 �g7+ �f8 2S .!lxg6 J:if4+ 29 'itel J:Ie4+ 30 @d2 ClJc4+ 31 1 b5! cxb5 2 a6 ..tc4 3 ..txc4 ..txc4 l:!xc4 32 l:th6 Black bxc4 4 ClJc3 ..tc5 resigned. Also bad is 4 .....tb4 5 llJd5 ..ta5 because of 6 ClJc7! and 7 a7!, Domination is a superiority in the deciding the game. functional activity of each piece. In 5 ..txc5 lLixc5 6 a7 �as 7 ClJd5 the endgame this increases many c3! S llJxc3! times over. Black returns the pawn and White must not fall for 8 tLlxb6? .l:Ixa7! 9 l:!xa7 c2 10 .!lal ClJb3 with a draw. Morozevich-Balashov S ... ClJb3 9 l:ta6 ClJd4 10 ClJd5 lLic6 No vgorod, 1997 11 .!lxb6! lLixa7 12 l:ta6! Black loses a piece, therefore he resigned. .�.�.fa� Shirov-Van Wely . . . Amsterdam, 1995 - �, . -.��-(''""���-.�) �-��:''''%� % � •. �.fJJ '/� l 1 � 0¥;a� % • � ��,.,, .. � 1 h4! lLie6 2 'i1Nf6 h5 There is no other apparent defence against ClJg4. 3 ClJd5�c2 Defending against lLie7+ and Black has achieved the ideal lLixg6. structure for the Sicilian Defence, 4 ClJe7+'it>f8 5 ClJcS! 'fl'e4+ 6 �gl but now he begins a principally in 'fl'd4 correct advance on the king's flank. If6 ...'itg8, then 7 ClJd6. 130 Complex Endings 7 �e7+ 'it>g8 8 tt:Jd6'i!Vdl + 9 Wg2 Or 15... 'it>h6 16 tt:Jf7+ 'it>g7 17 tllf4+ tt:Jg5+. The last chance. 16 'iltbs 'it>h7 17 'Yifxa7 @gs 18 10 'it>h2! ctJe2 11 'fl'xti+ 'it>h8 12 �f7+ @h8 19 'ikf6+ c;t>h7 20 @g2! �e8+ c;i;>h7 13 'fl'd7+ 'it>h8 14 'fl'c8+ Black resigned. 'it>h7 15 'iltb7+ 'it>h8 6 Defence in the Ending A complicated, important and Fercec-Cvitan exceptionally large theme. We can Zadar, 1995 touch upon it only partially. Playing for stalemate How strange it is that this method, for example, rather frequent in pawn endings with an extra pawn, is sel dom met in its purest form. Beliavsky-Gelfand Belgrade, 1997 1 . . . i.�g2? 2 11xg3!! 'ii'xg3 3 'Wh6 'it>g84 Wih8+\t>ti 5\\Ve8 Drawn. Exploitation of a lucky chance Gelfand-Lau tier Belgrade, 1997 The threat is ...@h3-h2 and then g2+, so 1 llf.3!! g2+ 2 \t>xg2 i.xf3+ 3 @xf2 Drawn. 132 Defence in the Ending Here winning are 1 l:tc1 or 1 Wd2, Difficultpractical defence but Gelfand decides to play more in the ending simply. 1 l:tc5??, Beliavsky-Mikhalchishin Now l ... b4! is winning. However USSR (ch), 1984 after... 1.. . i.c4? 2 'it>d2 Black resigned. Shirov-Lautier Belgrade, 1997 Black has two pawn weaknesses on e6 and b7 (For the present Black is coping with defending them) and a rather worse deployment of pieces. The question is-how can he improve his position? Correct is 1...g5!, commencing play on the Thinking a long way ahead, king's flank,but Black played Lautier went forthe variation 1.. . e5?! 1 l:txe4 Wxe4 2 f6 It is not usually worth moving a Reckoning on 2 ...i.e3 3 'it>h4 weakling. i.h6 (3 ...Wf5 4 'it>xh5 'it>e6 5 Wg6) 2 f3 'it>d5 3 lba4 lbe8?! 4 'it>xh5 i.f8 5 'it>g6 d5 6 h4 d4 7 Too intricate-simpler is 3 ...@d6 ! i.a5 d3 8 h5 'it>e3 9h6 with a win. 4 lbb6 tbd5 5 We4 lDf6+! with a However there followed draw. 2 ...i.xf 6!! 3 i.xf6 d5 4 Wh4 d4 5 4 lbc3+ 'it>e6 5 'it>e2 lbd6 6 �b4 'it>xh5 'it>f3 ! 6'it>g5 @d7 Or 6 b3 g3 7 hxg3 Wxg3 8 Wg5 Black wants to defend the b7 @f3 9 @f5 d3 10 i.c3 'it>e2 11 'it>e5 pawn with his king, though also not d2 12 i.xd2 'it>xd2 13 'it>d5 Wc2, bad is 6 ...�c7 reaching the pawns. 7 J:ib6 Wc7?! 6 ...d3 7 i.c3 c4! 8 @rs 'it>e3 9 Very strong was 7 .. .1H6!, and Wxg4 d2 10 i.xd2+@xd2 then 8 ... g5 and 9 . . . l:!.h6, obtaining And a drawn ending with a c counterplay. pawn is reached. 8 lbd5+'it>d7 9 b3 h5?! Defence in the Ending 133 This is only a weakness-better 1 'it>c4 h4! was the transfer of the rook via f8, Cutting down pawn material is al c8 to c6 The text just weakens the ways usefulfor the defence! g6 square. 2 'it>d3 hxg3 3 hxg3 i.d8 4 'it>e3 10 llle3 rbc7 11 ll'lc4! l:id7!! ..tg5 5 'it>f3 'it>e8 6 ..td6 rbd7 7 i.e5 After 11...ll'lxc4 12 bxc4 the rook c,t>es 8 'it>g2 'it>d7 9 'it>f3 'it>e8 10 proceeds to g6 and then to g5. 'it>e2 c,t>d7 11 'it>d3 ..td8 12 'it>c4 12 1Ib4 i.b6 13 f3 i.f2 and White cannot If 12 lllxe5, then 12... lle 7! 13 f4 improve his position. Therefore it's ll'lf7 with a draw, while the pawn a draw. ending after 12 lllxd6 J:ixd6 13 �xd6 'it>xd6 14 'it>e3 'it>c5 15 g4 b6! was also drawn. Rozen talis-C hristiansen 12 .•. e4! 13 lllxd6 exf3+ 14 'it>xf3 Groningen, 1992 l:rxd6 15 'it>e4 b6! Completely equalising by ex changing the important a5 pawn. 16 .l:!c4+ 'it>d7 17 b4 bxa5 18 bxa5 z:!'.c6! and drawn. A quite rare means of defence is without pawns-against a central pawn and bishop he succeeds in constructing an impregnable fortress. Gretarson-Magerramov lf4 Groningen, 1993 He should try 1 g4 ..tbl 2 g5, pre paring an attack on the h7 pawn. 1... 'it>d6 2 rbf2 l:!e7! Not allowing the white king into the centre. 3 i.f3 h5 4 a6!? i.g4!? An interesting decision-the ex change of bishops weakens the d5 pawn. 5 ..te4 5!..tf Seeing the idea through to the end! 6 .to ..tg4 7 ..te4 ..trs 8 axb7 llxb7 9 Ji(c6+ 'it>e7 10 i.xf5 gxfS 11 In endings with same-colour 'it>e3 l:!xb2 12 'it>d4 l:txh2 13 'it>eS bishops very often an extra passed .lle2+ 14 'it>xf5 lie3 15 lic7+ 'it>d6 pawn 'will not run'. 16 lixa7 lhg3 17 l:!xf7 'it>xd5 I 34 Defence in the Ending Gurevich-Andersson Leningrad, I 987 An interesting rook ending Black must play exceptionally accurately in order to hold the position. 1 g5 19 lih7Wd6! Finally White goes over to the The king must come closer at any attack. price. 1. .. hxg5 2 fx g5 lLld7 3 'fl'e4 20 'it>f6 l:!.h3! 'fl'b3+ 4 'it>h4 lLlf8 5 tLle7+ 'it>h8 6 But now he needs to advance his 'fl'f4 lLlh7 pawn. Leading to a quick defeat. It 21 .l:Ih6!? 'it>d7 22 .l:Ih8 llhl 23 f5 seems to us that at the very last mo h4 24 'it>t7 h325 f6 h2 26 .a'.h6 �c7 ment Black could still surprisingly 27 .l:Ih3!? save himself, in a truly fantastic White wants to transfer his rook way, by playing 6 ... 'it>h7!!. Now 7 to the second rank, and then, with 'i¥f5+ g6 8 'i'xf8 leads to an im the rook on e2, try to enter with his mediate draw in view of 8 ...'fl'g3+ king. But Black too activates his ('i/Nxh3+) 9 'it>xg3(h3) stalemate!. king. Meanwhile Black threatens after 27 ... 'it>d6! 28 1::rd3+ 'it>e5 29 l:!.d2 7 ...tLlg6 to bring the game down to a \t>f4 Drawn. drawn queen ending, therefore it is Black heads for the h2 pawn and necessary to accept the knight sacri then plays .l:Ifl . fice: 7 'fl'xf8 'fl'c4+! (but not 7 ...'ii b4+? 8 'it>h5 nor... 7 'fl'a4+? 8 During defence, mistakes due to 'it>g3 'fl'a3+ 9 'it>g4! 'ili'a4+ 10 'fl'f4 tiredness from protracted defence 'fl'dl+ 11 'iVf3, and White wins) 8 can occur. �g3 'il'd3+! (again avoiding a crafty pitfall: 8 ...'fl'c3+ 9 'il'f3 'i¥e5+ 10 In the following game, over the �g4 ilixe7 D_g6+! mating) 9 '1Wf3 course of the last 50 moves, with the (On 9 'it>g4 'iie4+ 10 �f4 'fl'xe7 same correlation of forces, White White lands his queen in a poor po had stubbornly tried to improve the siiton-11 g6+ already does not positions of his pieces and pawns, lead to mate. On 9 Wg2 follows while Black defended technically 9 ...'i/Ne2+, while 9 \t>f4? is imposs and cooly. ible because of 9 . . . 'fl'fl+) 9 ...�d6+ Defence in the Ending 135 10 \t>g4 'ib4+! (this is the point of You should always take the op Black's idea: he refrains from the portunity to spoil the opponent's immediate capture of the knight-in pawn structure. view of 11 g6--and with a correct White has two extra pawns and, it intermediate check disrupts the seems, an easily winning position. coordination of White's pieces) 11 But... h5 'iVxe7 12 'iVf5+ 'it>g8 13 'iVc8+ 1.. . h3!! 2 gxh3? �fl with a drawn queen ending. How could he not take a third 7 'iVg4 'iVc2 8 'iVf3 Black pawn? Correct was 2 g4 tLld2 3 resigned. .l:td4! .!Llxe4 4 J:Ixe4 lixc3 5 'it>g3, and by defending the f3 pawn with the rook White then wins the h3 Gelfand-Shirov Munich, 1993 pawn. 2 ....!Lid2 3 i.d3 l:1xc3 4 1Id4 �c6 5 'it>e3 .!Llb3 6 lid5 .!Lies 7 i.b5 lie6+ s Wf2 .!Llb7 9 h4 .!Lld6 10 i.d7lif6 11 i.g4 .!Lle8 12 h5 g5! The pawn must hold two white pawns. 13 lid8.!Lld6 14 'it>e3 The last chance would have been 14 lih8+ 'it>g7 15 I:!'.a8 Wh6 16 l:ta5 .!Lle8 17'it>g3 followed by h2-h4. 14 ....!Llf5+ 15 'it>e4 .!Llh4 16 'it>e5 .l:tf4! 17 :!:i.d3 'it>g7 Black has blockaded everything excellently-therefore it's a draw. 7 The isolated pawn in the Ending The isolated pawn is a weakness, 7 'i¥i>f6 'i¥i>c4 8 'it>xg6 c,t>d3 9 e4 but this weakness is felt in different Black resigned. ways. The strategy everywhere is similar-blockade by utilising strong blockading pieces or direct Yurtaev-Temirbaev attack on the isolani itself. Elista, 1998 We begin with the simplest example-a pawn ending. Ehlvest-Rausis Riga, 1995 Here the b6 point is firmly de fended and White does not succeed in penetrating via the c5 square. 1 h4 h5! 2 a4 'i¥i>e6 3 a5 c,t>d6 4 a6 'it>e6 5 e4 Sooner or later Black will run out Sooner or later it is necessary to of useful moves, and he will be try and exploit the slightly better forced to allow the white king into position of the king, but this is his camp via the e5 or c5 squares. insufficient. 1 . . . b6 2 a4 c,t>e6 3 f3 @d6 4 a5 5 ... dxe4 6 fxe4 fx e4 7 @xe4 g6 8 bxa5 5 bxa5 'i¥i>c6 6 'it>e5 'i¥i>c5 'i¥i>f4 'it>f6 9 g3 b5 10 g4 hxg4 11 After 6 ...'it>b5 7 'i¥i>xd5 Wxa5 8 'it>xg4 'it>e6 12 'i¥i>f4 'i¥i>f6 13 'i¥i>e4 'i¥i>c5 White also wins. c,t>e6 Drawn. The Isolated Pawn in the Ending 13 7 Pelling-Miles 1 i.f5! i.a6 2 f4 I/ford, 1974 It is important to exploit his ma jority on the flank. 2 ... lZJg8 3 h5 ctJf6 4 c;t>f3 i.b7 5 i.g6 @e7 Bad is 5 ...i.c8 6 g5 i.g4 7 Wg3 i.xh5 8 gxf6 i.xg6 9 fxg7 with a winning position. 6 g5 lZJe4 7 i.xe4! The knight here is undoubtedly stronger than the bishop. 7 ... dxe4+ 8 @e3 .id5 9 b4 axb4 10 axb4 .tf7 11 lZJf5 @rs 12 lZJg3! .ib3 13 @xe4 Wf7 14 c;t>es hxg5 15 fxg5 .ic2 16 g6+ c;t>g8 17 @f4 .ibl 1... a5! 1s c;t>gs .td3 19 CZJrs c;t>rs 20 lZJd4 Weaker is l...h5 2 b4! f6 3 h3 g5 .ic4 21 h6 gxh6 22 c;t>xh6 i.a2 23 4 fxg5 fxg5 5 @e3 e5 6 dxe5 @xe5 @gs c;t>g7 24 lZJf5+ and White won. 7 'itd3 Wd5 8 c;t>e3 @c4 9 c;t>e4 with a draw. Panno-Donner 2 b3 a4! 3 bxa4 bxa4 4 h3 Palma de Mallorca, 1967 After 4 g4 g5! White is badly placed, while on 4 c;t>c3 there is 4 ...�e4 5 c;t>b4 @xd4 6 �xa4 @c4! 7 c;t>a5 f6 8�b6 e5, and the queen ending is easily won since Black manages to exchange queens. 4 ... h5 5 h4 f6 6 @c3 We4 7 @c4 f5 ! 8 Wc5 We3 9 �b4 �f3 10 c;t>cs! Wxg3 11 'itd6 @xf4 12 �xe6 @g3, and Black just as easily won the queen ending. Rausis-Farag6 Germany, 1996 How is it possible to win this position forWhite? 1 .ib5 lZJe4 2 f3 lZJec5 3 @f2 lbe6 4 i.c3ctJc7 5 i.a4 The problem is that while for Black there is no active plan-for White there is. 5 ...We6 6 i.b4ctJe8 7 �e2 ctJd68 .ib3 There appears the first threat of a3-a4. 8 .••f5 9 h4! Now already arises a real threat to exchange the dark-squared bishop, 138 Th e Isolated Pawn in the Ending after which all Black's weaknesses, Two parts of the plan have been riveted on white squares, will be completed, White goes over to the exposed. third. 9 ... t'Lle5 10 i.c5 t'Lld7 7 ....l:Id8 8 �c3+ rt;b7 9 g4! :i.d6 After 10 ... t'Lldc4 11 i.a4 t'Lld6 12 10 a3 a5 11 h4 i.f7 12 f5 ! l:id8 13 i.xd6! 'it>xd6 13 i.e8 rt;e7 14 f4 ! :c2! t'Llf7 15 i.xf7 White wins in the A prophylactic move, eliminating pawn ending. the threat of l:id6-c6. 11 i.d4 t'Llb5? 13... .l:Id7 14 g5! hxg5 15 hxg5 Better in fact was 1 l...t'Llf6 con fxg5 16 l:l'.g2 :i.d6 17 .l:!xg5 g6 18 ceding the d5 pawn. fxg6 lixg6 19 i.xd5+ Black 12 i.b2 t'Llc5 13 i.c2 rt;d6 14 g4! resigned. hxg4 Also no help is 14... 'it>e6 15 gxh5 Nikolic-Portisch gxh5, since he cannot hold the two Ter Apel, 1996 weaknesses on f5 and h5. 15 fxg4 'it>e6 16 h5! t'Llc717 hxg6 Black resigned. Szabo-Penrose Bath, 1973 1.l:!c2! A very important prophylactic move-the a2 and :f2 pawns are the main target for Black's counterplay and therefore ought to be defended to the maximum. White plan consists of three parts: 1...t'Lle4 2 t'Lle5 (a) transferof the king to d4; In these kinds of positions block (b) transferof the bishop to f3; ading the isolated pawn is not so im (c) transfer of the rook to the portant-the main thing is to attack eighth rank and an advance of the it, as Bent Larsen said about such pawns on king's flank. situations! 1 l:Ia4 a6 2 :b4 l:1d7 3 '>t>c3 rt;e7 2 ...lib7 3 rt;n rt;f84 rt;e2 rt;e7 5 Or 3 ...d4+ 4 .i:txd4 .:i.c7+ g4 t'Llf6 6 h3 g6 7 f3 h5 (4 ... l:Ixd4 5 rt;xd4 i.xa2 6 i.c4+! As the defender is supposed to do, with a winning pawn ending) 5 rt;b4 Black tries to exchange pawns. i.xa2 6 b3 lie1 7 rt;a3i.b 1 8 rt;b2, 8 g5 t'Lle8 9 rt;d3 t'Llc7 10 I:!'.c5 d4 winning. Black has to sacrifice a pawn, 4 rt;d4 rt;d8 5 i.e2 rt;c7 6 i.f3 b6 since 10... rt;d6 11 .l:Ic6+ 'it>xe5 12 7 J:!b3 f4 + leads to mate. Th e Is olated Pawn in the Ending 139 11 exd4 .!Lle6 12 J::ra5 .!Llxg5 13 16... i.c7 17 .!Llc6 f6 18 f4 .!Lixg6+ Wf6 14 .!Llf4 and Black with With the threat of lt>d4, h3, g4 a clear conscience can resign. etc. 18 •.. lt>f7 19 lt>d4 lt>e6 20 h3 .l:Ih7 Magerramov-Kohlmeyer 21 .!Llc3! Bad Wo rishofe n, 1993 Play over the whole board! 21...l:!.d7 22 g4 hxg4 23 hxg4 Ji..d6?! More stubborn is 23 ... J::rh7 24 .!Llxd5 J:id7, in spite of 25 f5 +! gxf5 26 gxf5+ lt>xf5 27 .!Lle7+ lt>e6 28 J::rc7 l:!.xc7 29 .!Lixc7+ We7 30 .!Llxd5+ Wd6 31 .!Llf6!±, Mager ramov. 24 .!Lla4 i.c7 25 J:ihl! l:!'.g7 26 .!Llc3 .l:Id7 27 J:ih6�g7 If 27... i.xc6 28 bxc6 J:id6 29 l:!.xg6 White also wins. 28 .!Llxd5 �d7 29 f5 +! An instructive example of the Black resigned in view of struggle of two knights against two 29 ...gxf5 30 l:!'.xf6 mate. bishops, with rooks on the board, this being a rare case of the com"' A case of an isolated pawn with plete superiority of the knights. bishops was given in the book 1 .!Lla4 Winning Endgame Technique. As Nimzowitsch said, weaknesses Bishop against knight is also a very do not tell unless they are attacked! well known endgame, and the side 1...�d6 2 �acl z:!'.c8 3 J:Ixc8+ with the isolated pawn has to main Also 3 f3 ! ? deserves attention tain the equilibrium. -Magerramov. Very interesting also is the case of 3 ... i.xc8 4 .!Libs li:td8 5 .!Lld4! rook endings with an isolated pawn Worse is 5 .!Llxb6?! ii..g4 6 l:!'.xd5 for one of the sides . .l:Ib8with strong counterplay. s ••• .l:!d6 6 f3 i.d77 .!Llc3 h5 8 Wf2 Bareev-Farag6 lt>f8 9lt>e2 Ji..c8 Rome, 1990 Upon 9 ...@e7 10 Wd3 ii..e8 11 :Icl nothing good awaits the black king in the centre. 10 b4! axb4 11 axb4 i.a6 12 b5 Fixing a second weakness on b6. 12... i.b7 13 l:.al i.f6 14 lt>d3 l:.d7 The only move. If 15... lt>e7? 16 .l:ta7 l:!.d7 17 l:!.xb7! l:txb7 18 .!Lixd5+ White wins. 15 .!Lla4Ji.. d8 16 .l:Icl Successfully combiniag threats along the a and c-files. 140 Th e Isolated Pawn in the Ending 1... h5? 8 ... 'it>d6 9 b5 �e6 10 g3 g6 11 He should immediately take .l:Ie8+! �d6 12 .l:If8 'itie6 13 !!c8 measures on the queen's flank by �d6 14 .l:Ic6+ �e7 15 g4 @f7 16 l ... a5, though after 2 g4 White still gxh5 gxh5 17 l:tc8 maintains the advantage. Now it will be very difficult to 2 b4 'it>e6 3 'it>d4 f6 4 h4 'it>f5 5 t3 hold on at once to the three wea 'it>g6 knesses on a7, d5 and h5. If 5 ...g5 6 hxg5 fxg5 7 1If8 'iitg6 8 17.•. �e6 18 z:!'.e8+ 'iitd6 We5 there is no defence against 9 If 18 ... Wf7, then 19 l:tb8 and 20 �e6. .l:Ib7. 6 a4! �f7 7 a5 �e6 8 a6! 19 .l:Ih8 l!i.c7 20 lid8+ �e6 21 The weakness on a7 is now lixd5 f5 22 e4 and Black can permanent. resign. 8 Rook and bishop against rook and knight In general the line-up of rook plus White begins a plan associated bishop is rather stronger than rook with an attack on the a5 pawn or the and knight (to a large extent this de forcingof the advance ...a5-a4. pends on the structure), but a few 6 ...c,t>e7 7 l:!.al! @d6 grandmasters, for example Evgeny After 7 ...�d7 8 a4 l!tb4 9 lllxa5 Sveshnikov, think the reverse. Rook l:!.xa4 10 !!xa4 �xa4 11 �b4 and and knight is stronger when there then 12 �c5 White wins the d5 are fixed weaknesses or, for pawn. example, when there are doubled 8 a4 �b7 9 lllxa5 �c7+ 10 @d2 pawns. �d7 11 lllb3 �c4 12 a5 �c6 We look at a few cases of the ad No help is 12. ..�a4 13 a6! vantage of rook and knight, since winning. with this line-up the methods of 13 a6 .llb4 14 llla5! �a8 15 'itic3 play are more complicated. �b8 16 lllb3 J:ie8 17 @d3 @c7 18 lllc5 @b6 19 !!bl+�a7 20 c,t>d2! Sakaev-Novikov Preparing an exchange of rooks Moscow, 1998 -technically the simplest way to win. 20... �c6 21 .li(b3 l:re7 22 l:!e3 with an easily winning position. Mikhailov-Volchok corr, 1995 1 �d2 b4 2 llla4! There is no sense in White open ing the a-file-but here it is possible to open the b-file. 2 ...bxa3 3 bxa3 :h7 4 'it>c3 c,t>f8 5 lllc5 l:!.b5 6 lllb3! J 42 Rook and bishop against rook and knight Black has a majority on the king's Milov-Pelletier flank,but the knight has to carry out Biel, 1997 an unpleasant defensive function and, in addition, the pawns are ham pered by their white counterparts. 1 g5! J:id5 2 .!lgl Not only are White's chances not 2 l'llh3! ..tcs 3 lllg5 Wf6 4 Clio better, but Black simply has a clear We6 5 I:!'.b5! advantage. For the time being activating the 1 ....l:Ie4! rook. A surprising exchange of rooks. 5 ... i.d7 6 I:!'.b2 Wf6 7 l'lle5 ..ta4 8 2 .l:Ixe4 lllxe4 3 i.c2 l'llc3 4 g4 lib8 @g7 9 @d2 @f6 10 �c3 @g7 �d6 11 @b2 It was also possible at once to fix Also possible was 11 d4±. the opponent's queenside pawn 11 ... 'it>h7 12 @a3 i.c6 13 @b3 structure by 4 ...l'll b5!? 5 a4 l'lld6. �g7 14 a4 i.b7 15 a5 ..tcs 16 @c3 5 �el @c5 6 @d2 l'llb5 7 a4 l'llc7 �f6 17 .l:Ib2 g5 18 lih2! gxf4 19 s @c3 l'lld5+ 9 Wb2 @b4 10 g5 b5 gxf4 e6 20 d4 cxd4+ 21 �xd4 @e7 11 axb5 cxb5 12 i.e4 l'llc3 13 i.c6 22 c5! i.a6 23 .l:Ih8 i.fl 24 lih7+ White wants to obtain counter �d8 25 llhl i.g2 26 .llgl i.d5 27 chances by doing away with the 1Ig8+ �e7 28 a6 @f6 29 1Ig6+@e7 black f7-pawn. 30 .l:Ih6! 13 ...l'll dl+ 14 @a2 t'llxf2 15 i.e8 Zugzwang! l'lld3 16 i.xf7 llle5! 30 ...i.g2 31 lllg6 @es 32 l'llh4! Now the knight dominates the i.d533 l'llxf5 Black resigned. bishop, while the king transfers to the other flank. 17 i.e8@c5 18 @b2 b4! 19 Wc2 Genba-Irzhanov Wd5 20 h5 gxh5 21 i.xhS �e4 22 Russia, I 998 @d2 g6 23 i.dl @f5 24 @e2 'lt>xg5 25 @f2 'it>h4 26 �g2 l'lld3 27 i.e2 lllc5 28 i.dl �g5! 29 �g3 'it>f6 30 @g4 �e5 31 i.c2 l'llxb3! 32 i.xb3 @e4 White resigned. Beliavsky-Neverov Koszalin, 1998 Sometimes chessplayers them selves do not know how to exploit the possibilities of their own pieces. 1 i.xa6? Driving the bishop to a passive position. He should play 1 a4! 1Ia8 2 b5, creating a_passed pawn. 1...li:ta82 b5 t'lld7 3 1Ia4? 1 ... f5 ? Better really was to leave the Better was l...i.f5 2 g4 i.d7 3 g5 pawn and activate the bishop by 3 i.f5 4 llle4 i.xe4 5 �xe4 f5 + with i.b7. chances of holding the rook ending. 3 ...lll c5 4 .l:Ia3@f6 144 Rook and bishop against rook and knight Now the black king enters the Kamsky-Cvitan game with decisive effect. Palma de Mallorca, 1989 S 'it>g2 ri;;e7 6 'it>f.3 d4! 7 'it>e2 ri;;d6 8 Wd2 'it>c7 9 h4 'it>b6 10 ri;;dl Iif8 11 ri;;e2 .l'.td8! 12 lilt3 d3+ 13 'it>dlttla4! Nobody wants the bishop on a6! 14 ri;;d2 ttlb2lS a4 ttlc4 16 'it>dl d2 17 h!b3 hS 18 aS+ 'it>xaS White resigned. With a struggle on one flank the advantage of knight over bishop in conjunction with some other pieces can be very great. 1 ll:lfS+'i;; g6 2 .l.Ig8+ �f6 3 ll:le3 ! j_cs 4 ll:ldS+ 'it>e6 s liXe8+ 'it>d6 6 Lastin-Alferenko .l'.td8+ 'it>c6 7l:tc8+'it>bS?! Moscow, 1998 Obviously better was 7 ... Wd6, though even here after8 h!c7 Black. has serious problems. 8 .l.Ic7! After this move everything becomes clear. 8 ... 'i;;c4 9 l!xti l:ia2 10 .l'.tf.3!ri;; d4 11 ttlf6 h4 12 gxh4 j_e7 13 hS l:t'.a6 14 ll:lgS! j_gs lS .i:tg3 j_f4 16 l:t'.g6 Black resigned. Gulko-Sveshnikov Volgodonsk , 1983 1 Wf.3 llc7 2 l:bS -*.f6 3 ttlcs j_e7 4 ttld3 j_f6 S l:b4! Guaranteeing the centralisation of his king. S ....l.Id7 6 We4 .UdS 7 l:b7+'it>f8? Slightly better is 7 ...'i;; g8. 8 f4! exf4 9 ttlxf4 Iles+ 10 Wf.3 :rs 11 e4 l!aS 12 ttlxg6+ 'it>g8 13 ll:lf4! Black resigned. Rook and bishop against rook and knight 145 Activity of the pieces always Preventing both ... f7-f5, and also comes before material advantage. ...'Li e8-d6- f5 1 .l:Ib7 .l:!xa2 2 .l:Ixb6 'it>g7 3 :g_b7+ 2 ... 'it>d83 b4 .l:Ie7+4 'it>d3 axb4 S 'it>h6 4 i::i.e7 :g_xb4 lt>c7 6 l:!.bl 'Lif6 7 aS ctJd7 8 White tries to limit the activity of �al! the black knight. With the threat of 9 a6 bxa6 10 · 4 ..• .l:!b2 S l1e3 libl 6 'it>g2 .l:Ib2 7 :g_xa6 attacking the c6 pawn. 'it>h3 'Lids 8 :d3 'Lif6! 8 ...'Lib8 9 h4! Again trying to control the e4 The knight has gone over to the square. opposite flank. There he can attack 9 .l:Ie3 ctJdS 10 .l:If3 'Lif6 11 g4 on the other side. hxg4+ 12 i.xg4 gS! 9 ...'Lia6 10 .l:Ibl 'Lib8 11 i.f3 The transfer to the rook ending is 'Lid7 12 .l:Ial l:te8 13 llfl lie7 14 lost: 12... 'Lixg4? 13 'it>xg4 �xh2 14 i.g2 'Lib8 IS i.e4 'Lia6 16 .l:Ibl J::[.h3+. 'Lib8 17 i.rs 'Lid7 18 l1al g6 19 13 fx gS 'it>xgS 14 i.e6 'Lie4! i.xd7! .l:!xd7 20 �fl 'it>b8 21 .l:If6! Finally the knight achieves its ob 'it>a7 22 hS! jective-the pawn is not enough but Black's weaknesses start to the activity of the scattered pieces is crystallise. sufficientto achieve a draw. 22... 'it>a6 23 gS! IS lifS+ 'it>g6 16 l:!.bS 'Lid2 17 b4 The passed pawn comes first! 'Lin 18 i.rs+ c,t>f6 19 i.d3 lixh2+ 23... hxgS 24 h6 'it>xaS 2S h7 .l:Id8 20 lt>g4 .l:If2 21 .l:IcS 'Lie3+ 22 'it>g3 26 .l:Ixf7bS 27 cxbS 'it>xbS .l:!b7+28 :d2 23 i.e4 .l:Id4 Drawn. 'it>a6 29 �g7 .l:Ih8 30 'it>e4 and White has a winning ending. However you cannot always man age to achieve such coordination Balashov-Vaganian with a rook and knight. Very often a USSR (ch), 1972 rook and bishop compliment each other splendidly. Karpov-Hort Tilburg, 1979 1 'Lic2! White plans to reconstruct with his king on d4 and knight on e3. Now Black should go over to counterplay by I...lib5 2 'Lib4 i.e6 1 .l:Ibl!'it>e7 2 g4! 3 c,t>d4f6 with a tenable position. 146 Rook and bishop against rook and knight 1 ... i.e6? 2 'it>e4! i.fS+ 3 'it'd4 18 .. J:tcs 19 .l:1b5 :es 20 'it>xc4 i.xc2? rl.e4+ 21 @d5 l:txh4 21 .tixaS 'it>g7 Vaganian makes the mistaken 23 b4 .l:1hl 24 bS h4 25 c4 h3 26 decision to play a rook ending, .l:1a2 h2 27 'it>cS @f728 rl.e2Black though even after 3 ...i.e6 4 lLie3 resigned. �c5 5 f5 i.xf5 6 lLixc4 l:!.c6 7 lLie3 i.e6 8 c4 things are not easy for him. Tal-Balashov 4 rl.xc2 �cs S �e2+ 'it>f66 l:Ie8! USSR, 1974 Now it would be a mistake to transfer to a pawn ending: 6 !f.e5? :i:lxe5 7 fxe5+ 'it>f5 8 'it>xc4 'it>g4! 9 'it>b5 'it>xg3 10 'it>xb6 @xh4 11 'it>xb7 'it>g3 12 c4 h4, and Black is quicker. 6 ... �bS 7 :!::i.e2 :cs 8 a4 g6 9 Zle8 �c7 1 a4! Depriving Black of the possibility of ... a5-a4. 1. .. b6 2 lLid2! Now White improves the position of the knight, whereas Black's �ol lowing activation assumes an unreal character. Now White needs to finda plan to 2 ...i.eS 3 f3 l:.f8 play for a win, and Balashov dis Or 3 ...l!d4 4 'it>e2 and 5 lLie4, and covers a splendid transfer of the if Black does not exchange rooks, rook to b5, where the rook attacks then 6 !Ihl and 7 h4. the b6 pawn and does not allow the 4 lLie4'it>c7 S llhl! king to penetrate to the fifth rank. It is necessary to open the h-file, 10 lieSltc6 11 :bs! 'it>e6 12 'it>e4 whereas the d-file has no 'it>f6 13 fS gS 14 J:idS gxh4 lS gxh4 significance. 'it>e7 16 WeS! 5 ... J..f4+ 6 'it>e2 'it>d7 7 h4 gxh4 8 Provoking ...f7-f6, so taking this rl.xh4 rtle7 9 l2Jf2! square away from the opponent's Now the best place forthe knight king. will be d3. 16 ...f6+ 17 Wd4 �cs 18 :bs l!c6 9 ...l'l.d8 10 lLid3 i.g5 11 .l:lhl 16 a5! bxaS 17 J:Ixb7+ c,t>f8 18 e5!? 'it>dS! A vain attempt to obtain activity All White's pieces are active, and with a pawn sacrifice. Black's passive. 12 :.dl'it>e6 13 l2Jf2! Rook and bishop against rook and knight 147 White again wants to exchange pair of rooks, which is in White's rooks, but Black is forced to avoid favour. this exchange. 3 b3 c,t>g7 4 l::ta6 l:tbc8?! 13 ... .l:Id4 14 CZJe4i.f 4 15 ClJc3! He should defend by 4 ...l:1b7 5 White's main objective is the b6 CZJc5 .:c7, though after 6 b4 (but not pawn. 6 .l:Ixe6? in view of 6 ...i.b2) and 15... i.g5 16 CZJd5 i.d8 17 I:l.d3 Black has no counterplay. c,t>f7 18 lbc3! 5 .l:Ic5! It is necessary to drive the black The best way to exchange after rook away from the important d4 which the white knight appears on square. c5. 18... i.e7? 5 ... l:!xc5 6 CZJxc5 i.b8 7 b4 Better is 18... i.g5 !, which was the Also here an exchange on e6 was only chance. not possible because of 7 ....l:Ic8. 19 ClJb5 e4 20 l:te3! l:td7 21 .l:Ixe4 7 ...@f6 8 l:l'.a5! i.f622 f4 i.g7 23 f5 'i¥tf6 24 I:l.e6+ The rook not only cuts off the Black resigned. black king along the fifth rank, but also improves its own position: its The following ending serves as objective-the b7 square. yet another example where the 8 ...l:l'.d5 knight is stronger than a bishop Hoping for 9 .l:Ib5 i.d6 10 lbe4 when there are rooks on the board. c,t>e5 11 ClJxd6 'it>xd6 12 l:rb7 a5 with probable equality. 9 a4 Pelletier-Arencibia Now already the above-mentioned Ub eda, 1998 variation is threatened, because at the end White has b4-b5. 9 ...@e7 On the more tenacious 9 ... @g7 White intended to play 10 lbb3! .lld7 (the exchange of heavy pieces leads to a quick win for White: 1 O . . . llxa5 11 CZJxa5 a6 12 CZJc6ii.c7 13 b5 axb5 14 a5! etc) 11 lZJd4 i.d6 12 lbc6 'JJ.c7 13 l:l'.a6with the future march of the king to a4 and b4-b5 and the win of the a7 pawn. On the other side of the board Black does not get anything. In order to maintain the balance 10 1Ib5 i.d6 11 l:tb7+ 'it>e8 12 Black must activate his rooks along CZJe4?! the d-file and then subsequently at Probably, sufficient for the win is tack White's king's flank. 12 lba6 l:tg5 13 g4 h5 14 gxh5 1 h3 i.e5 2 'i¥te2 llab8 l:rxh5 15 l:!xa7, but simplest was 12 Already Black does not succeed ClJd3a5 13 b5. in organising counterplay along the 12... ii.e7 13 l:!xa7 �e5 d-file because of the exchange of a Or 13... f5 14 lbc3. 148 Rook and bishop against rook and knight 14 J::!a8+ Wd7 15 l:l:a7+ @es For the present beginning an at If the Black king stepped to d8, tack on the black pawns. then 16 tLlf2 .Jtxb4 17 l!xf7would 9 . .. aS 10 tLlc6 Wg7 11 tlixa5 lia8 decide. 12 l:l:bS i.f6 13 :t;.i.a2 i.c3 14 tllc6 16 @d3 i.xb4? l:!.xc4 15 aS It would be better to play the rook The a-pawn, supported by the ending after I 6 ...f5 17 tLlc5 i.xc5 knight c6, can only be stopped at the 18 bxc5 :xc5 19 lixh7 g5 (19... Ila5 cost of material loss. 20 lih4), though 20 .l:Ia7 was suffi 15 .•..l:Ixa6 16 l:i.c2licxc6 17 dxc6 cient forvic tory because of the very l:!.xa5 18 :b7 nal+ 19 Wg2 .Jta5 bad black king. 20 lla7! 17 tllf6+ Wd8 18 Itd7+ Wc8 19 And Black resigned, since there .l:!xtil:[f S is no defence against the manoeuvre 19 ...h5 20 tLld7!? lid5 21 Wc4 .l:tc2-b2-b7. il..a5 22 tllc5 also loses. 20 g4! lif3 21 We2 �xh3 22 The knight is particularly strong l:.xh7 .l:!:xh7 in attacking various weaknesses, After 22 ...Itg3 23 @f2 the black whereas a bishop is incapable of de rook is in a 'hole'. fending light and darksquares. 23 tt:Jxh7 @d7 24 Wd3 We7 25 tLlgS �d6 26 tllf3 and Black Van Laatum-Mikhalchishin resigned. Eeklo, 1991 Adams-Almasi Dortmund, 1998 .�.�.!. � �. �. � •.•• • •• • • �.�.�1�. �.��-r�·�z• 0. •�. �.�lf ...� .�.�-..�....�� �... -0. i...tt:Jrs There was another, and perhaps more unpleasant plan for White: 1 �dS! 'tiixdS 2 exdS 1...f5 !? followed by 2 ...@f6, 3 ...f4 White is obliged to transfer to an and 4 ...@f5 . ending where for him the knight c6 2 l:.g2 controls all the important squares After 2 i.d5 l!xg3 3 .i.e4 �g4 4 around itself, while even the dark .Jtxf5 gxf5 it is not easy to defend squared bishop, controlling the long the endgame. diagonal, will feel redundant. 2 •..tll e3 2 ... g6 3 l:l.'.abl i.g7 4 c4 i.c3 5 Interesting was 2 ...tll d4 with the .l:tecl .Jtf6 6 @fl �e4 7 g3 i.gs 8 same plan ...f7-f 5-f4 . .l::!.c2 �ae8 9 tllb8! 3 na2 f6 !? Rook and bishop against rook and knight 149 A new idea: ... g6-g5-g4 and 7 i.c8 'lld6 8 i.d7 l:td3 9 i.e6 ...'ll e3-fl. 'lle4 10 J:ig2g5 4 i.e6 f5 There was also another plan: After 4 ...g5 White brings the 10 ... �D and l l...'llf2. bishop into the defence by 5 hxg5 11 hxg5+ 'i¥txg5 12 i.f7 �g4 13 hxg5 6 i.h3. i.e6l:te3 5 d4 �f6 6 d5? Avoiding the trap 13... 'll xg3? 14 This pawn should be 'discarded' i.xf5+!. by playing 6 i.c8 and 7 i.b7, trans 14 i.c8 �c3 15 i.e6 �d3 16 ferring the bishop to the long i.f7? diagonal. It was necessary to give up the d5 6 ... 'llc4! pawn, if only by the move 16 i.c8 Now the knight shows itself in all followedby 17 i.a6. its splendour. 16... 'll g5 ! White resigned. 9 Rook and two pawns against rook and knight (many pawns) It would seem that having more 3 bxa3 l:rxd44 �xd4 bxa3 5 'it>c3 pieces should confer an automatic a2! 6�b2 'ittc4 advantage, but the knight has such Aftersacrificing a pawn, the black limited mobility in a battle with king rushes over to the opponent's pawns that quite frequently nothing pawns, it seems without success ... can be done with it. 7 lt:Jd6+'it>d5 8 lt:Jf5 'itte4 9 lt:Jxg7 'iilf4 10 h3 'lt>g3 11 lt:Jf5+'iil xh3 12 lt:Jxh6'it>h4! Mikhalchishin-Sveshnikov But this is the point-and typical USSR (ch), 1978 forthe struggle of the knight against pawns-there is no defence against Cllg5, thereforea draw. Smagin-Mikhalchishin Moscow, 1989 1:d2 The decision to exchange rooks was not easy. Other possibilities were I lt:Jd2 or the roundabout way I l:.f2! with the idea of2 l:rf8. l ...�b5 2 l:rd4? 1...lt:Jb4?! After 2 l:rxd5 cxd5 3 lt:Jd2 a3 4 Clearly Black wants to attack the �d4 l.ta4 5 bxa3 bxa3 6 lt:JbI! a2 7 pawns, but stronger seems L.h6, ".lJc3+ �b3 8 lt:Jxa2 the pawn ending though it is not usually recom is won. mended to play on the side where 2 ... a3! the opponent is stronger. Bad is 2 ... I:.xd4? 3 'i!ixd4 a3 4 2 a3 lt:Jd33 g5 I:.e6 4 .l:.d2 lt:Jc5 lt:Jd6+! 'lt>a4 5 bxa3 bxa36 �c3 After 4 ... .l:.e3+ 5 'lt>g4 lt:Je5+ 6 with a win. 'ittf4 .t'td3 7 .Uxd3 lt:Jxd3+ 8 'it>e4 Rook and two pawns against rook and knight 151 ctJxb2 9 'it>d4 the white king Or 6 'Lid4 gxh3+ 7 'it>xh3 'it>f4. threatens to break through to the 6 .•.J:ixb3 7 l:txd5 'it>xg4 8 ctJxh6+ pawns on the queen's flank. 'it>f4 Drawn. 5 'it>g4 ctJxb3 6 �d7+ 'it>g8 After 6 ...l:l'.e7 7 :i.d6 there is equality on the board. Korchnoi-Ljubojevic 7 h5, and a draw was agreed, Bath, 1973 since the threats l:td8, h5-h6 give White sufficientplay. Mikhalchishin-Savchenko Postojna, 1991 1 .•. .:i.bl? After the correct l...Wf8 2 'it>f3 :!.bl Black is in no way worse, whereas now problems start to ap pear with the fl pawn. 2 'Lid3 'it>g8 3 .l:!b7 'it>g7 4 ctJe5 Black has two plans of play-the 'it>f6 5 ctJxf7 l:rcl 6 ctJxh6 �c2+7 best one was the plan to advance the 'it>O b2 8 'it>e4! l:l'.e2+ 9 'it>d4 l:l'.h2 passed pawn. 1...l:1dl !? 2 ctJc5 (2 10 g5+ We6 11 J:rb6+ �e7 12 'it>c3 l:ta2 d4 ! 3 .:i.xa6 .:i.d2+ =) 2 ...d4 3 �f8 13 l:1xb2 l:i'.xh4 14 .l:!b7 J:if4 15 ctJxa6?! d3 4 J:If6 h5 5 :i.d6 l:l'.d2+ �d3 Black resigned. with distinct counterplay. But Black decides to exchange pawns on the queen's flank, which is clearly Hertneck-Narciso weaker. Berlin, 1998 1. .. a�?! 2 'Lig7?! But White decides to play for mating threats. whereas after the correct 2 �d2! Black would en counter serious difficulties. 2 .•.� bl 3 h3+ If now 3 �d2, then after3 ...'it>f 4 4 ctJe6+ 'it>e3 5 l:l'.xd5 �b2+ 6 'it>g3 l:.xb3 7 �xa5 We4+ 8 �g4 l:l'.b2 9 h3 �g2+ 10 'it>h5 �g3 Black elimin ates White's last pawn. 3 ... 'it>h4 4 I:i:d2 g4! 5 ctJf5+�g5 6 hxg4 J 52 Rook and two pawns against rook and knight Here Black is three clear pawns 14 ltJxb7 g3 15 'it>e3 g2 16 �f2 ahead, but after I lLie5! with the 'it>e4 17 'it>xg2 'it>d3 Drawn. idea of �f4-e4 and then .l:Ic2 seri ous problems arise forBla ck, since he cannot create a passed pawn. But Romanishin-Markowski in the game there followed... Polanica Zdroj, 199 3 1 'it>f4? @f7! 2 'it>es \¥tg6 3 .!lxd4?! A primitive decision, more inter esting is 3 'it>e6! ? or 3 lLih2! ? . 3 ... .l:Ixd4 4 'it>xd4 cJi>g5 5 'it>e5 'it>h4 6 lLif2 �g3 7 ltJe4+ Wxh3 8 ltJg5+ @g4 9 t'llxh7 'it>f.3 10 c;t>d4 'it>f4 11 tllf8 g5 12 lLie6+ @rs 13 tiJd8 1 e5! dxe5? Also not winning is 1 ... .!ld4 2 exd6 l::!.xd5 3 h4 ! gxh4 4 .l:Ie4 l:i.d4 5 d7! with a draw. But necessary was l....!le4! 2 exd6 .l:Ixe6 3 dxe6 tbxd6 4 Wg2 ltJf5! 5 �f3 tiJh6! (against h2-h4, so as to have g5-g4) 6 'it>e4 �f6 7 �d5 lLif5 8 h3 ltJg7 with an easy win. 2 d6 tiJd4 5 i::i.xe5 'it>f6 4 d7 ltJc6 13 ...g4 ! 4 �e8 .l:Id4 5 .l:Ic8! By sacrificing yet another pawn, This is the idea-to take out the Black diverts the opponent's king. knight-now it's a draw. 10 Two minor pieces against a rook Endings with an uneven align of defensive resources the defender ment of forces are some of the most should strive for an exchange of complicated in chess-and also pawns, bearing in mind the possibil comparatively little investigated. ity of transferring to a theoretically Among them can be included end drawn ending, exploiting the stan ings where a rook struggles against dard methods of: (a) from the rook's two minor pieces. side-sacrificing the exchange; (b) In the middlegame, particularly in from the minor pieces' side-s a positions bearing a closed character, crificing a piece. two minor pieces will be stronger Let's look at some possibilities in than a rook even where there is a positions where the rook is playing deficit in pawns. However, in end against passed pawns. The rook's ings where the rook has room for specific characteristics depend on action, the situation often depends lengthening lines of communication. on the arrangement of the pawns. The broader they are, the more Thus, when there is a distant passed chances the long-range rook has in pawn on the board, the rook can its struggle against the less mobile prove to be stronger than two minor pieces, and it is even possible pieces. However with a balanced to think about playing for a win. pawn formation the situation may You see, whereas the minor pieces be reversed. are riveted to the blockade of a When playing such endings, it is passed pawn, the rook can be uti very important for the side having lised also on another flank to organ the minor pieces to put right their ise new objects of attack. coordination, and then the field of To counteract the rook, one of the activity of the minor pieces may be postulates of play for the side hav sufficient to struggle successfully ing the minor pieces is to put right against a rook. their interaction while controlling According to the generally held the opponent's passed pawn. But the opinion, in standard situations with various plans of activity of the two pieces one can count on having forces on the board produce play of an advantage. But rooks are not a complicated character, and the without their own chances too, par exploitation of one's resources is ticularly when there is a passed not always successful even where pawn on the board. However, first highly rated chessplayers are and foremost, as we go over to a concerned. study of positions of this type, we Naturally, the outcome of the will tum our attention to the possibi struggle in great part depends on the lities of the defending side. In terms arrangement of the pawns: the more 154 Two minor pieces against a rook compact they are, the shorter the Wd3 (after 8 'it>f3 Black forces a communications of the rook, and draw: 8 ...l:txc2! 9 lllxc2 Wb2) naturally the narrower their 8 ...c;tib2, and Black prevents White possibilities. from putting right the coordination of his pieces. Beliavsky-Miles 2 tlldl + 'it>a33 'itie2 l:1hl Th essaloniki (o l), 1984 There is no saving the pawn: 3 ...�g l 4 llle3 c3 5 @f2!i:g5 6 f4. 4 ..txg6 .l:Ih2+ 5 c;tie3 'itib46Wf4 c3 7 llle3 White has won a pawn without losing control of the opponent's passed pawn. 7 ..•'it>b3 8 ..td3 'it>b2 9 c;tie4 It is useful to pay attention to the interaction of the white pieces now the time for White's passed pawn ha,s come. 9 ... 'it>cl 10f4 l:!.h8 11 .tbs .l:Ib8 After 1 l...c2 12 ..ta4 Black loses Despite the far advanced passed a pawn and cannot derive any bene pawn, supported by all Black's fit from the temporarily uncoordi- forces, the chances of White's nated white pieces: 12 ...c;tid2 13 pieces are higher. They firmly lllxc2 .l:Ia8 14 ..tc6! or 12... l:ta8 13 blockade the pawn, while the quite i.b3! l:ta3 14 ..txc2Wd2? 15 lllc4+. narrow range of activity between 12 ..td3 'it>d2 the flanks allows White to fight for Or 12 ... l:te8+ 13 'it>d4 'it>d2 14 the creation of a passed pawn of his lllc4+. own. 13 lllc4+ c;tidl 14 @f3 lih8 15 1 ..tc2�fl ?! ..te4 Wel 16 llla3 Wdl 17 f5 If 1...g5? 2 tlldl+ 'itia3 3 'it>c3 the And White, maintaining the passive rook is unable to prevent the blockade of the opponent's passed win of the c4 pawn: 3 ...lic l 4 llle3 pawn, realised his own one. and 5 lllxc4. Therefore Black activates his Morovic-Yusupov rook, attacking a pawn. But this is Tunis, 1985 inaccurate: on the f-file the rook is in the sphere of activity of the king. Therefore stronger is 1...�gl, de fending the g6 pawn and riveting the white pieces to the passed c pawn. Here, after 2 ll:ld1 + c;tia3 no good is 3 'it>c3 :!fl or 3 'it>e3 l:!'.g2 4 ..te4 Wb3. But also on the active 3 f4 l:!'.g3! the rook succeeds in stand ing up to the pieces. For example: 4 tlle3 lif3 5 llld5 c3+ 6 c;tie2 .l:Ih3 7 etJe3 (7 ..txg6? l:th2+) 7 . ..l:!h2+ 8 Tw o minor pieces against a rook 155 In this example the sphere of ac the king penetrates to the weakened tivity of the rook is quite broad, and queen's flank. this means the more possibilities it 10 ctJf2+ 'it>d5 11 b4 g3 12 hxg3 has. In fact, despite material equal hxg3 13 ctJd3 ity, the advantage is on the side of Also after 3 ctJg4 'it>e4 14 t'Lie3 the rook. It is necessary only to axb4 15 i.xb4 l:!h6 the passed pawn create a distant passed pawn, while must decide the game. not forgetting to prevent it being 13 axb4 blockaded by the minor pieces. This It was also possible to win the ex is achieved by the advance of the change: 13... 'it>e4 14 bxa5 g2 15 a6 pawn pair after the preparatory gl='ilY 16 i.xgl .l:Ic2+ 17 'it'd! l ...l:!f 5. 'it>xd3 18 a7 .l:Ixa2 19 'it>cl b4 20 But in the game there followed.... 'it>b1 b3 21 i.f2 Wc4 22 i.e3 'it>c3 1...g4?, (A.Yusupov), but Black prefers to Breaking the coordination of the retain the passed g-pawn. pawn pair, which allows White to 14 i.e3 l:!.c2+ 15 'it>dl l:tc3 16 blockade it with the move 2 i.el!, 'it>e2 'it>e4 17 t'Lic5+ 'it>f5 18 ctJd3 without upsetting the coordination 'it>e4 19 t'Lic5+ 'it>d5 20 ctJd3 lic2+ of his own pieces. However he is 21 'it>dl .l:Ih2! deflected froma direct attack on the Now the g-pawn cannot be d4 pawn. stopped. 2 i.h6?! 'it>e6! 22 i.gl l:thl 23 ctJf4+ 'it>e4 24 No expense spared for creating a ctJe2 Wf3 White resigned. passed pawn. 3 ct:Jc5+? In the context of defensive possi It was still not too late to returnto bilities and technical realisation of a the blockading plan-3 i.f4 h5 4 distant passed pawn, we present a i.g3 ! 'it>f5 5 i.h4. The bishop con classical example fromthe praxis of trols the el-h4 diagonal, while the Yuri Razuvaev. king has nowhere to go along the light squares, since on any . . . 'it>f3 Razuvaev-Kirov there is ctJel+.Moreover, also poss Sofia, 1981 ible is the energetic 3 i.f2!?, taking aim at the d4 pawn. 3 ...Wf5 4 ctJd3 lic7 5 'it>d2 .li(c6 6 i.g7? Now already the creation of the passed pawn cannot be prevented, whereas 6 i.f4! 'it>e4 7 i.g3 'it>f3 8 lt'iel+ would have set up a solid blockade. 6 ... h5! 7 i.xd4'it>e4 8 i.c5 8 ...l:td6 was threatened. 8 ... h4 9 'it>e2 a5 More energetic was 9 . . . g3 10 hxg3 hxg3 11 b4 l:i.c8, imposing an With such a specific arrangement original zugzwang on White: 12 a3 of pawns on the king's flank, the 1:tc6 13 ctJel �f6 14 Wd2 Wd5, and bishop itself can resist the rook J 56 Two minor pieces against a rook when there are no pawns on the 7 ltb7 ii..e7 8 @f4 r;.td69 'lt>e4 @e6. queen's flank. It goes without say Now the king cannot break through ing that here the king of the weaker to the c4 square without loss: l 0 side should be close to the pawns. <;t>d3e5+ lt::J 11 'itc3xg4 lt::l 12 hxg4 Therefore to achieve a draw it is hxg4 13 'it>c4 (or 13 a5 tl'ie5 14 a6 enough for Black to sacrifice the lt::lc6) 13... lt::J e5+ 14 'it>b5 i.d8. knight for the passed a-pawn. He Upon the attempt to proceed to c4 should also avoid any unnecessary via the b3 square it is necessary to weakening of the position. White reckon with the threat of ... lt::Ja5. should take this into account and or The rook's intended penetration to ganise offensive action. Black has the 8th rank by l 0 lic7 r;.td6 11 succeeded in constructing a defence: lic8?i.d8 ends in its capture. his minor pieces are well coordi Also nothing is offered by l 0 l.t.b5 nated, preventing the advance of the i.d8. passed pawn and preparing a break 2 'it?g2'it>f6?. on the king's flank. In addition the As before, Black senses no danger bishop does not allow in the rook and breaks the coordination of his since on .l:ib7 he can cover up by pieces since now the position of the i.e7. Here, however, Black need bishop is limited by the blockading lessly weakens his king's flank with function of the knight. Better was the move 2 ... r;.td6 3 lib7 i.c7 with the idea 1. .. h6? ...lt::J c6-a5. ...and creates objects for attack, 3 .l:c5 tl'ia5 4 @f'3 @e6 5 'it>e4 thus increasing the attacking possi i.b6 6 lic8@f6 7 h4 'lt>e6 bilities of the rook. Now, as well as the h6 pawn, it is possible to add the invasion point f5, which can be created by undermining the g6 pawn with h2-h4-h5. After L.'it?d6 2 h3 (more accurate than 2 'itg2, on which Black quickly organises an attack on the passed pawn by 2 ...lt::J e5 3 h3 lt::Jd? and tl'ic5) Black has two possible plans of defence: (A) Active-2 ... tl'ia5 (preparing the attack a4 after transferring the knight to c5) 3 'it?g2 'it?c6 4 �e5 lt::lb7 5 �e8 'it>c56 �h8 'it>b47 �xh7 8 h5 lt::lc5 8 h4 r;.txa49 h5 gxh5 l 0 gxh5 It is important to fix the weak i.g5, and Black, after giving up his pawn. Too hasty is 8 l:Ih8 lt::lb7 9 bishop for the h-pawn, reaches a �xh6 tl'ic5 10 @f3 @f7 11 h5 gxh5 theoretically drawn ending 'rook 12 l:txb6 hxg4+ 13 @xg4 tl'ixa4 against knight'. with a drawn ending. Now, how (B) Passive-2 ...'it?e6 3 <;t>g2 @d6 ever, in the event of 8 ...gxh5, 9 %lh8 4 @g3 (White presently controls the is unpleasant. 5th rank in order to prevent the 8 ... 'it>t79 'it>d5 gxh5 10 gxh5 @f6 move ... h7-h5 exchanging a pair of 11 �b8 i.e3 12 'lte4! i.d2 13 �b2 pawns) 4 ...'it>e6 5 h4 'it'd6 6 h5 'it>e6 i.el 14 l::te2 i.b4 15 �g2 @f7 Two minor pieces against a rook 157 The weakness of the h6 pawn is While no additional objects of at self-evident. Combining an attack tack have been created on the king's on it and the knight, White breaks flank, the advance of the pawns to the coordination of the opponent's vulnerable squares is premature. pieces. On 15... i.fS follows 16 �g6 More logical was the exchange of @fl 17 �a6, and the a-pawn starts the f- pawn after3 ...g6, which forces to move. the white king to remain far away 16 .li(g6 i.d2 17 lld6! i.gS 18 from his pieces and increases the J::[.dS! scope of the rook. An ideal square for the rook. 4 i.c3 f6 18 .•. tt:Jc6 19 as @e6 20 a6 i.e3 Also here, stronger is 4 ...g6. Trying to blockade the a-pawn on 5 i.b4 the finalfrontier, but the mechanism The pawn structure on the king's of breaking the coordination of the flankis stabilised, and Black intends pieces again comes into operation. to attack White's pawn weaknesses. 21 .l:IbS i.gl 22.li(bl i.f223 �b2 We look at Black's possibilities on i.cS 24 .l:Ib7 tt:Ja7 25 �h7 the basis of comments made by And White wins the h-pawn and G.Kasparov in his 1987 book Two the game. A very fineending! Matches. 5 ... �f7?! And here is an example where Here the king does nothing, since pieces blockade the passed pawn there are no prospects for it in the from more distant 'remote-control' centre. Much more accurate was squares. 5 ...�h7. In this case the method of defence applied by Karpov in the Karpov-Kasparov game would place White under Wo rld Championship (m) , 1985 threat of defeat. For example: 6 i.c3 k!b8 7 i.b4? �b5! 8 g4 l:1b8.A zug zwang position has been reached and White is forced to allow the rook into his camp: 9 @d3 l:i.d8+ 10 �c2 hxg4 11 hxg4 �d4 12 i.xa3 �a4 13 �b3 lixg4 14 i.cll:i.g3+ 15 lLlc3 lif316 �c2 �xf5 17 @d3. A simple evaluation of such a position is not possible, but in practice Black would have excellent chances. But after 7 lLlb4! White holds on: 7 ....l:Ib5 8 g4 .l:Ib8 9 @d3 ! lia8 10 lLla2 l::ta4 11 i.b4 �h6 12 i.d2, and Here the bishop will not allow the the fortress cannot be breached. rook to get behind his lines via the 6 tt:Jc3.li(b8 7 lLla2 �bS 8 g4 �b8 queen's flank (due to the covering At first sightthere is no basic dif �b4 and i.c3), whereas the squares ference from the above-mentioned of invasion along the central files zugzwang position, but... are covered by the king. There 9 �d3 l:i.d8+ 10 @c4 �dl 11 followed: i.xa3 :i:Ial 12 'it>b3 �bl l...!tb82 i.b4 �d83 @e2 a3?! Not 12 ...h4? 13 lLlcl. 158 Tw o minor pieces against a rook 13 gxh5 lhh3+ 14 'Llc3 allows only 'in exchange' for a This is where the position of the counterattack behind enemy lines. black king tells: after 14... J:Ixh5 All this is not easy. Let us look at 'Lle4 the f5 pawn is inviolable. the possibilities of the two sides. 14... l:!.B based on G.Kasparov's analysis. But now he cannot, without loss, 1 l:!g3ctJd5+ 2 @f2'iit f5 cope with the h-pawn. The king moves closer to the h2 15 .tel! .l:!xf5 16 h6 g6 pawn. Right up to the draw there re 3 li:tf3+@g4 4 .l:!g3+ @h4 5 @f3 mains the sacrifice of a piece for White strives to activate his king, two pawns. On 16... g5 this is since he cannot hold on by the pass achieved by 17 'Lle4�g6 18 h7 .l:!f3 ive 5 l:!b3 CLlf4 with the threat (18 ...�xh7 19 'Llxf6+) 19 @c4 J:ih3 ... 'Lld3+. Now 5 ...'Llxb4 is not poss 20 'Llxf6. ible because of 6 �g4+. 17 'Lle4l:!h5 18 ..tb2 5 .•. 'Llf6?! And the f6 pawn is lost, since on Stronger is 5 .....td7 !, keeping l 8 ...f5 follows 19 h7. Drawn. under threat b6, while on 6 We4 the knight manages to attack the h2 If, together with the rook, there pawn: 6 ... 'Llf6+! 7 'iite5 'Llg4+ 8 are no passed pawns then roles are �d6 ..tb5 9 llg2 'Lle3,and the black reversed. It remains only for the king draws closer to the h2 pawn. rook to counteract the opponent's 6 'itig2 'Llh5 7 l:te3 ctJf4+ 8 �gl attempts to create his own passed 'itig4 pawn. But none the less the defen Now, with the white king driven der's resources are quite substantial. back, Black puts right the interac tion of his king and pieces. Short-Kasparov 9 l:1g3+'iit f5 10 J:if3 g4 Belgrade, 1989 Not letting out the king, since in the event of 10 ...@e4 11 @£2 'Lld3+ 12 @g3 'Llxb4 13 .l:!f8 White again activates his forces. 11 l:!.e3 'Lld5! Again insufficient is l l ... 'Lld3 be cause of the activating of the rook-12 �e7! (12 @g2? @f4+ 13 lle7 ..tc6 in Black's favour) 12... 'Llxb2 13 nm @g6 14 .l:!f8 'Lld3 15 @g2 'Llxb4 16 l:tf4. 12 .U.b3 @f4 13 @n Only the b3 and g3 squares re main accessible to the rook, but now In this position Black's chances he should encircle the king. How can only be linked to the creation of ever, on 13 @g2? 'Lle3 14 @£2 a passed g-pawn. Therefore he en 'Lldl+ 15 'iitel .Jla4 16 �a3 'Llxb2 deavours to win the h2 pawn. How 17 lla2 'Lld3+ 18 @d2 .tbs White ever to do this it is necessary to loses a yawn without compensation. break through the line of demarca 13 •.. tiJf6 14 l:!a3 'Lle4+ 15 �g2 tion-the 3rd rank-which the rook .Jle2 Two minor pieces against a rook 159 The manoeuvre 15... 'Lid2 16 .!lc3 'it>c2 'Lixh2 33 .!lxh2 g3 and Black 'Lic4 17 .li(c1! 'Lixb2 18 .:i.c8 allows won. the rook to escape to freedom. 16 .l:!'.b3 ..tfl+ 17 c,t>gl i.c4 18 One more example from the l:ta3 'Lig5 19 .llc3 i.d5 20 .l:!'.a3 classics. i.e4?! The a6 pawn is indirectly de Reti-Bogoljubow fended: 21 l:txa6?? ctJh3+ 22 @fl Bad Kissingen, 1928 i.d3+, but this move, losing time, also allows the rook to escape to freedom. Stronger was 20 ... ltJD+ 21 'it>g2 'Lies22 @£2..tc4. 21 @m ..tb7 1... b6 For the side having the rook it is very important to create a distant passed pawn quickly. 2 axb6 l:Ixb63 c,t>e2 22 l:td3?! A loss of tempo. He should quick White misses his chance to acti ly put right the coordination of his vate the rook by 22 l:ta5!, threaten pieces, combining an attack on the ing to exchange the b-pawn with e5 pawn with detention of the a tempo-22... 'Lie 4+ 23 @gl ctJd6 24 pawn. For example: 3 ctJc4 l!i.b5 l:l'.c5, and, it seems, maintaining the (3 ....!lb4 4 'Lixe5 .!lxe4 5 'Lic6 in equilibrium. tending 'Lia5 and ..td2) 4 i.d2 .l:Ic5 22 ... 'Lih3+23 'it>el? 5 'Lia5'it>f? 6 'it>e2'it>e6 7 'i¥td3 and 8 After moving the king further 'Lic4. away f,rom the h2 pawn the game 3 ... l:l'.b4 4 'it>f3 'it>f7 5 .lth4 .llbl 6 already cannot be saved since now 'Lic4 'it>e6 7 ..tg3 Black cuts it off from this pawn. This threat is a sham, since it has Necessary was 23 @fl ..tc624 .!lc3. a tactical refutation. 23 ... i.c6! 24 .!id4+'iitf3 25 .:i.d3+ It is useful to stop the pawn as We4 26 .!lg3 'iitf4 27 l:ta3 soon as possible. For this purpose More tenacious is 27 l:.c3 ..tb5 28 the blockading move 7 J..d8 is l:l'.c8, though after28 ...'it>D 29 l:tf8+ appropriate. 'i¥tg2 30 .!lg8 'Lif2! White also loses 7 ... .!lcl! 8 'Lias the h2 pawn. 8 'Lixe5 is no good in view of 27•.. ..tb5 28 l:l'.c3 'Lig5 29 .!lc8 8 ...1Ic3+ 9 c,t>f4 g5+ 10 'it>xg5 \t>e3! 30 .!lh8 ctJf3 31 'it>dl i.e2+32 lhg3+. J60 Two minor pieces against a rook 8 ... l!al 9 lbc4 l:f.a4 10 ll'ie3 as 11 Black has two distant passed pawns .tel �a3 12 'itie2 a4 13 ll'ic2 l:f.b3 and the opponent's pieces are in no 14 i.b4 state to contain them without ma At last White manages to put right terial loss. the coordination of his pieces and 31 'it?b4 a3! 32 lbxa3 :xa3 33 blockade the passed pawn. Trying to i.xh4 :l'.e3 34 'it>c4:.xe4+ and soon obtain additional objects of attack, White resigned. Black organises pawn pressure on the king 's flank, supported by the Let's look at yet another example, active rook. illustrating the strength of the rook 14 ... h5 15 i.f8 gS 16 h3 l!c3 17 in conjunction with an outside 'it?d2:b3 18 'it?e2 l:tb2 19 'itid2l!bl passed pawn. 20 'it?e2 :.c1 21 'itid2 �fl 22 i.cs l:i.f4 23 'it?d3 l:f.f7 24 i.e3 l:i.d7+2S Miles-Van der Sterren, @e2 l:f.g7 Wijk-aan-Zee, 1984 26 g4? Material advantage is on White's This superficially active move, side. He has a passed pawn, har pursuing the objective of blockade, monious deployment of pieces and turns out to be the decisive mistake. can reckon on playing for a win. White creates in his position a weak However the strength of the passed h3 pawn, which the rook can attack, pawn in conjunction with an active exploiting the fact that White's rook is such that Black's counter minor pieces are tied down to the resources prove sufficient to obtain threatened advance of the a-pawn. full equality. For a start he ties the After 26 lba3 g4 27 hxg4 :xg4 28 minor pieces down with the threat to 'it?f3 .Ug8 29 i.c5 White defends advance the a-pawn. himself successfully. l...a3 2 @d3 l:f.a5 3 i.a2 'it>g7 4 26... h4! 27 .Itel? �c3 @f6 5 ttJf3l:f.bS The only move was 27 i.d2!so as The rook transfers to an active to meet 27 ...l:i.c7 with 28 'itid3, and position. Weaker is 5 ...gS?! -6 e5+ 27... .l:Ib7 with 28 i.b4. @f5 7 ll'id4+ 'it?g6 8 e6 @f6 9 @b4 27 ...:.c7 28 'itid3 :b7 29 .txg5 lira8 10 lbc6, with advantage to :t.!b3+30 @c4 :xh3 White. And so the main idea of attacking 6 eS+ 'it>e7 7 lbd4 l:::tb2 8 i.dS with the rook is seen in a clear light: lln 9 ll'ic2 !te2 10 ll'ib4 Two minor pieces against a rook 161 White hopes to realise his material Analysis shows that in fact he al advantage and therefore jectsre the ready has to worry about maintain draw after 10 l'Llxa3 1:!.xe5. ing equality. For example, in 10 ...'it>d7 11 Wd4 gs Black's favour is 15 Wd5 l:.te3!? We are acquainted, from the (sufficient for a draw is 15... .i::i. cl, previous examples, with the method but Black is striving for more) 16 of exploiting a 'quality advantage' i.b5+ 'it>e7 17 l'Llc6+ 'it>f8 ! 18 i.c4 (in the words of A.Nimzowitsch). :Ie2 19 l'Llb4 l:tb2 20 l'Lla2? [20 12 h3 hS 13 g4?! 'it>c5! t:>. 20 ...l:i.h2? (20 ... 'it>e7!) 21 As shown by the endings previ 'it>d6 lixh3 22 l'Lld5 Ith1 23 e6 l:te1 ously looked at-this is a risky 24 e7 J::rxe7 25 l'Llxe7 h3 26 l'Llf5+-] weakening of the position. However 20 ...l:th2 etc or 15 e6 'it>e7! (but not White plays for the win and does 15... 'it>d6?-16 l'Llc2 J:ih 1 17 l'Llxa3 not want to reconcile himself to the .l:Ixh3 18 l'Llb5+'it>e7 19 l'Lla7 and 20 'drawing zugzwang': 13 l'Lla2 lld2+ l'Llc8+, and White has the advan 14 c;t>e4 l:l.'.e2+ 15 'it>f5l:i.d2 etc. tage) 16 l'Llc2 �h 1 17 l'Llxa3 l:.xh3 13 ... h4 14 i.c4 18 l'Llb5 l:tg3 etc. It seems he should reconcile him self to a draw by playing 15 l'Llc2! Ith1 16 l'Llxa3 l:txh3 17 l'Llc2! l:tg3 18 l'Lle3=. lS ... �dl 16 'it>e4 J:id2 17 i.c4 llb2 18 l'Lla2 .l:.h2 19 @rs �xh3 20 'it>xgSJ::r hl An inaccuracy. Far stronger is 20 ...J:ih2 with the threat of 21...J::rxa2 and 22 ...h3. White cannot at the same time hold the two distant passed pawns (there you are, don't you recall the move 13 g4?!). For 14... :!:!.el ! example: 21 'it>f5 J:if2+ 22 'it>g6 Black should not let the king go l:.txa2 23 i.xa2 'it>e7, and one of the on to c5, since his own king cannot pawns queens; or 21 Wf4 l:tf2+ 22 hold back the onslaught of the white 'it>e3 lic2 23 i.d5 h3 24 l'Llb4 lib2 pieces. For example: 14 ...l:tb2 15 25 l'Lld3 a2 (or) 25 . . . h2), with a de 'itc5 l:th2 16 e6+ We8 (16... 'it>e7 17 cisive advantage. l'Llc6+'it>e8 18 i.d3) 17 l'Lld5 l:!.c2 18 21 @f6 h3 22 l'Llc3 !:I.el 23 i.e6 'it>d6 Ji(xc4 19 l'Llf6+, and White 'it>d8 24 i.ds h2 2s gs no+ 26 wins; or 14 ...l:.th2 15 Wc5 l:txh3 16 'it>e6 .:!.cl 27 l'Lla2 � hl= 28 i.xhl e6 :Ie8(16 ... We7 17 l'Lld5+) 17 l'Lld5 .l:.xhl 29 g6 .l:.gl? with the threat of 18 i.b5+ and a Having achieved a winning posi mating attack. tion, Black makes a mistake---and lS i.d3?! victory escapes him. After29 ...'it>e8! Continuing to play for the win, White's passed pawns are neutra White suddenly breaks the coor lised. For example: 30 @f6 l:tfl+ 31 dination of his pieces, and the initi 'it>e6 'it>f8 32 'it>d6 J::rf2 33 e6 J::rd2+! ative passes to Black. Could White etc. or 30 g7 �h6+ 31 'it>d5 Wfl 32 have improved his position? Wc4 l:th3,with a winningending. 162 Two minor pieces against a rook 30 c;t.n l:tfl+ 31 We6 J:igl 32 Wf7 pawn, while 10 .id8 loses after J:tfl+ 33 'it>e6 l:tgl Drawn. 10... �c5) 8 ...fxg5 9 fxg5 (also poss ible is 9 b3 with the threat of 10 The opportunity for the side hav ctJb4, but risky is 9 f5 ?! because of ing the rook to create a distant 9 . . . J:ih2 10 ctJa l c;t>fl followed by passed pawn quite often appears as the creation of a passed h-pawn) a leit-motif of the struggle. 9 ...�xg5 10 c;t>f4 and then b2-b3. By reducing the number of pawns, White can count on holding the Novikov-Kaidanov position. Vilnius, 1984 6 ...Wf7 7 We3 Why not to e4? 7 ... 'it>e6 8 g4 ri;;dS 9 ri;;d3 J:Ic5 10 b3 Allowing a tactical solution to the position. As seen from the previous examples, the exchange sacrifice does no,t achieve its objective after 10 b4, but in this case the break through with the king decides- 10... I:!'.c4 11 .id2J:id4+ etc. 10 ... .l:txc3+! 11 Wxc3 'it>e4 Black has reached a technically won endgame, since White's king is 1...l:!.c4 riveted to the a2 pawn and the Black immediately exploits the knight will not succeed in contain break in coordination of the oppo ing the pawn break on the king's nent's pieces. In view of the threat flank. We present the main vari 2 ...l:txc3 White is forced to allow ation: 12 g5 @xf4 13 gxf6 gxf6 14 the creation of a passed pawn. @b2 ri;;e4 ! 15 ctJc2 f5 16 ctJa3 f4 17 2 ctJc6a3 3 ctJd4 ctJc4 'it>d3 18 ctJe5+ 'it>e2 19 'it>xa2 The piece sacrifice3 bxa3 l:txc3 4 f3 20 ctJxf3 'it>xf3 21 'ita3 h5, and a4 does not work because of 4 ...a6! White is lacking one tempo to save 3 ...l:!.c5! 4 @f3 himself. With the pawn placed on The knight cannot move: 4 ctJc2 b4 it would be drawn. axb2, or 4 ctJb3 I:!'.xb5. The game ended in the following 4 ...a2 5 ctJc2l:!.xb5 6 ctJal? way: He cannot hope to hold the posi 12 f5 'it>f4 13 'itb2 'i;;xg4 14 ctJc2 tion with passive defence. Stronger ri;;h3 15 ctJd4 'it>xh2 16 ctJe6 'it>g3 17 is 6 'it>e4 with the idea of transfer ltJxg7 'it>g4 18 ctJe8 ri;;xf5 19 ctJd6 ring the king to c4. For example: 'it>f4 20 ltJe8 'it>g5 21 ctJd6 h5 22 6 ....l:th5 (on 6 ...'i;;fl possible is the ctJe4+ ri;;f4 ! 23 tt:'ixf6 h4 24 ctJd5+ prophylactic 7 g4 followed by 8 'itf3 and White resigned. 'it>d4) 7 h4 g5 8 hxg5 (dangerous is 8 .i.xf6 because of 8 ...gxh4 9 gxh4 Let's look at a more complicated Wfl 10 �d4 .l:txh4, with the cre example of the creation and realisa ation of a second distant passed tion of an outside passed pawn. Two minor pieces against a rook J 63 Beliavsky-Dolmatov l'tJf4 .l:Ic3 5 �e2 .l:Ixb3 6 i.c4 lib2 7 Minsk, 1979 �d3 b3; or 4 l'tJf2 l:tcl 5 i.e4.l:Ia l 6 ctJd3 l:!.xa2 7 ctJxb4 (White should also lose in the event of 7 h4 �f2!) 7 . . . .l:Ixh2 8 ..id3 f5 9 'i¥te3 (on 9 .txf5 winning are both 9 ...�b2, and 9 ...'it>f6) ... 9 'it>f 6 10 @f4 l:l'.f2+ 11 @e3 .l:r.b2 12 i.c4 'i¥tg5 13 i.xf7 'i¥tg4 14 i.e6 'i¥txg3, and each of the passed pawns will cost a piece. 4 ...J::r cl 5 ctJf4 !!'.al 6 ctJd5l:l'.xa2 7 l'tJxb4Itxh2 8 i.e2.l:Ih3 9 @f2 Not allowing the rook to assist with the advance of the a-pawn, which would have been the case White has a material advantage, after 9 @f3 .l:Ihl 10 i.c4 l:ta1 but the presence of a weak pawn on (otherwise White can blockade the a2 and the lack of cooperation be a3 pawn by 11 ctJa2 and 12 b4) 11 tween his pieces allows the oppo ctJc2J::r a2 12 ctJb4lib2 etc. However nent to count on creating a distant Black finds a new resource. passed pawn. The position bears a 9 ... f5 ! concrete character and in the event In the event of 9 ...�hl White suc of a premature fixing of the weak cessfully solves the problem of re pawns, similar to the previous stricting the rook by 10 i.fl .l:Ih2+ examples, White will manage to put 11 ..ig2 l:l.'.h8 12 ctJa2 l:tc8 13 i.e4. right the coordination of his pieces Now however the threat of the break and prevent the invasion of the ...f5 -f4 is irresistible. On 10 ii.fl rook. For example: l...a3 2 ctJf4 follows 10 .. .l:th2+ 11 i.g2 J::l.h8, and l:l'.c8 3ctJe2. the rook breaks through on the However Black breaks through to queen's flank. the a2 pawn in a tactical way. 10 ctJa2f4 11 gxf4 l:.xb3 12 ..ic4 1. .. c4!! 2 i.xc4 1::rh3! 13 i.fl After2 bxc4 lib8 Black organises Black is preparing a breakthrough the advance of the passed b-pawn: 3 with his king, whereas White cannot ctJf2 b3 4 a3 (or 4 axb3 a3 5 .id1 force the rook from the 3rd rank lixb3 etc) 4 ...b2 5 ii.bl lib3 6 �e2 without material loss. For example: 1::rc3 7 ctJe4 l:l'.xc4 8 ctJd2 lie1. 13 'i¥tg2l:re3 14 'iitf2 J::l.e4 etc. 2 ...J::l. c8 13 ...lih2+ 14 i.g2 @f6 15 @g3 In the event of 3 ctJf2?there is the lih5 16 i.fl lies 17 i.d3 J::rd5 18 threat of an exchange sacrifice i.a6 l:!d2 19 i.c4 l:!.d4 White together with a pawn break- resigned. 3 ...l:!xc 4! 4 bxc4 b3 5 axb3 a3. 3 ..id3 a3 ! 4 'it>e3 An analogous idea of an ultra The most tenacious continuation. sharp breakthrough was exploited The attempt to place the knight at by Krarnnik against Ivanchuk. once loses quickly. For example: 4 164 Two minor pieces against a rook Ivanchuk-Kramnik Tavadian-Tseshkovsky Linares, 1998 Irkutsk, 1983 1. . . c4!! 2 bxc4 To realise his material advantage On 2 i.xc4 follows 2 ....li(d2+ and White should adjust the coordina 3 ...l:!.xc2 and 4 ....l:Ixa2, and Black tion of his pieces in such a way that has a passed pawn. However it is the threat to advance the f-pawn impossible to avoid the loss of a hampers the activity of the rook and pawn. king and wins the pawns on the 2 ...l!b8 3 c5 .l:!b2 4 c6 @e7 5 queen's flank. Together with this it lLixg6+ c;t;>d6 6 lLie5 .l:Ixa2 7 ttJc4+ is important to finnly blockade @c7 these pawns. After 7 ...'3;xc6 8 i.xf5 there is no 1 ttJc6 .l:Id3+2 c;t;>f2 play 5 ....l:Ie7? because of 6 i.d2 Superficially the impression is lt>g3 7 CZ'lf5+ or 5 ....!le8 6 'it>f2 .!lb8? that Black's advantage is of a 7 i.d2 'it>e5 8 CZ'lc6+) 6 'it>f2, and an decisive charcater. However elegant 'dance' of the king along the analysis shows that, even though the 'triangulation' e2-fl -f2, typical for d5 and g7 pawns are doomed, his pawn endings, culminating in the defensive resources are in no way win of the b-pawn. exhausted. 3 ... .!ld7 4 �e3 'it>f5 5 f4 ? 1 J:ig2 �g8 2 �g5? It was still not too late to return to In such endings it is very import the above-mentioned plan. However ant to activate the rook. This is White's careless advance of the f achieved by 2 l:Ig5 i.e4 3 f5 CZ'lxd5 pawn removes his control of the e4 (if 3 ...f6 4 �g6 CZ'lxd5 5 'it>g4 the square, allowing the rook to develop chances are already on White's side) great activity with support from the 4 f6 ! CZ'lxf6 5 .l:Ixb5 'it>xg7 6 l:Ib6!, king. In effect the material advan with the idea of 6 ...d5 7 lt>g5. The tage is now devalued. position is simplified and White ob 5 ...�h7 6 i.e5 lld7 7 'it>f3 lid2! 8 tains definite counterplay, asso 'it>e3 .!ld7 9 CZ'la2 :i:Id3+! 10 lt>e2 ciated with the possibility of lib3 creating a distant passed pawn on The white king is cut offalong the the queen's flank. 1st-3rd ranks, and when the oppor 2 ... i.e43 .!lg3 CZ'lxd54 'it>h6?. tunity presents itself an advance of White again misses the opportun the b-pawn is threatened. The ity to activate the rook, though in a chances of the two sides are even. less favourable light. Correct was 4 11 'it>d2lid3 12 'it>c2 �f3 13 CZ'lb4 f5 'it>xg7 5 �g4 tLlf6! (5 ...i.f3 looks 'it>e4! 14 CZ'la6 lt>d5 15 CZ'lc7+ �c6 tempting on account of 6 .!id4 f6+ 7 16 CZ'le6 �d5 17 CZ'lg5 li.g3 18 i.c3 'it>h4 CZ'le7, winning the f5 pawn, .!lg2+ 19 @cl !!f2 20 i.d2 b4! 21 however, by playing 6 l!tg3!, White i.xb4 lixf4 and the game soon is forced to repeat the position: ended in a draw. 6 ...i.e4 7 .!lg4, since it is dangerous for the bishop to leave the bl-h7 di Here is a good illustration of the agonal because of the threat f5-f6) 6 failureto exploit one's chances. l:.g l i.c6 7 h4! (it is useful to hold the check to the king in reserve: Miles-Adorjan after 7 \t>f4+ �h6 it is not easy for Wijk-aan-Zee, 1984 White to hold the position, for example: 8 .!idl CZ'ld5 9 'it>g3 'it>g5, or 8 l:tel �h5 9 .!lgl CZ'ld5, with advantage to Black; nevertheless possibilities of resistance were of fered by 8 h4 ! 'it>h5 9 �g3 !, but not 9 .!lg7? i.d5-+) 7 ... CZ'le4+ 8 �f4 �h6 9 .!lg8, and with an active rook White can successfully defend himself. 4 ... tt:Je7 5 !!c3 CZ'lf5+ 6 'it>g5 'it>xg7 7 l:tc8 CZ'ld4 8 l:te8 CZ'le6+ 9 'it>g4 i.c6 10 .!lb8 CZ'ld4 11 f5 ? 166 Two minor pieces against a rook Now also the f-pawn is lost. We 1. ..@f8 2 tLld2 �e7 3 tLlb3 .!lb44 mention that on 11 l:.td8 Black re tLl3c5 groups his pieces by 1 l...i.f3+ 12 And so the coordination of the 'it>g3 i.e4 followedby 13... tLlf 5, and knights is put right. They defend the d-pawn quickly advances. one another and therefore are 11... 'it>f6 12 h4 tlixfS 13 hS unassailable. Belated activity by the passed 4 ... fS 5 'it>gl .!ibS 6 'it>f2 'it>f6 7 pawn. tiJd7+ 'it>e6 13 ... tLle3+ 14 �h4 dS lS .l:td8 d4 His hopes of driving away the 16 .l:td6+ 'it>eS 17 h6 tiJfS+ 18 'it>h3 knight from the a6 pawn are unreal. tLlxh6 19 lixh6 d3 20 .l:IhS+ rs Stronger is 7 ...�g5, trying to keep White resigned. back the king and create a passed f- pawn. Let's look at an example where 8 tLlb6! the minor pieces have to hold back an onslaught by an active rook when there is relative material equality. Keres-Szab6 Candidates (t) , Zurich, 1953 The knights have regrouped and again are unassailable. 8 ... �eS 9 'it>g3 l:tb3+ Worth considering is 9 ...f4 +, but even in this case the king is not ob liged to step back. For example: 10 �g4 f5+ 11 'it>g5 lib3 12 tLld7+ The initiative lays with Black: his (weaker is 12 tLlc4 because of rook is very active, whereas the 12... �d5 13 tLlbd6 f3! 14 gxf3 knights are far away from each .!lxf3, and the knights are "hob other and, what is no less important, bling", while Black is able to create they have no support points. Besides a passed pawn) 12.. .�d5 13 tLldc5 this White needs to reckon with the f3 14 gxf3 l:txf3 15 tLlxa6, with a possibility of a transfer to a pawn probable draw. ending. White's main task is quickly 10 �h4 l:!.c3 11 tLlcS! to put right the coordination of his A tactical solution to the problems knights. Let us proceed with of defence-typical for such end P.Keres' logical and beautiful fili ings. White exchanges pieces and gree manoeuvre which solved the transposes the game to a drawn problem of the defence. pawn ending. Two minor pieces against a rook 167 11.. . .l:IxcS 12 tL'ld7+ 'i¥i>d6 13 A speedy advance of the distant liJxcS 'it>xcS 14 'it>gS @bS lS c,t>h6 passed pawn should lead to a win And the game was drawn after a forBlack. few moves. 8 @es h4 9 .i.e4+@f7? Carelessness, after which White With the pawns placed on one saves the game with a surprising flank the task of the defending side tactical blow. The natural continu still remains quite complicated. ation appears to be 9 ...'it>h6, and the h-pawn swiftly advances. For example: 10 'it>f6 h3 11 .i.b4.lle2 I2 Miles-Kindermann Bath, 1983 .i.g6 .l:Ie6!, or IO c,t>fs .l:!g2 I I .i.b4 h3 I 2 'i¥i>f6 �e2-in both cases with an easy win. Black has the advantage, since he has the possibility of creating a passed pawn. But the small number 10 .i.xh4!! of pawns gives White a defensive Now it all comes down to a theor resource-at an appropriate moment etically drawn ending. to sacrifice his bishop and obtain a 10.•• gxh4 theoretically drawn position. The situation is not changed by 1. .. 'it>e62 .i.c3 10... .l:Ixh4 I I '\¥i>f5 .l:Ih6 I2 'i¥i>xg5 The best practical chance! In the .:.f6 I3 Jl..f5 and I4 'it>xf4, with a event of2 Jl..c7 l:i.c2+3 "it>b6 ter(af 3 draw. 'i¥i>b5 f4 4 .i.d8 nxh2 White loses a 11 @xf4 I:!'.g2 12 .i.f5 pawn, since he cannot play 5 .i.xg5 White would have replied with because of 5 ...l:.h5) Black plays not this move also after I I...J::rhl. There 3 ... f4-4 Jl..d8 J::rxh2 5 .i.xg5 with followed: possibilities of defence, but first of 12 ... "it>f6 13 Jl..h3 l:.gl 14 .i.g4 all 3 ...h6, with the idea 4 ...f4 5 h4 .:.bl lS .i.h3 :i:Ib4+ 16 'it>e3@es 17 gxh4 6 Jlxf4 h3-+. .i.g4 .:.b3+ 18 'it>f2 '\¥i>f4 19 'it>g2 2 ... I:!'.xh2 3 JldS+ 'it>e7 4 'i¥i>d4 .l:!b2+ 20 'i¥i>h3, and Black soon con @f6 S 'it>e3+ '\¥i>g6 6 .i.el f4 + 7 'it>d4 vinced himself of the futility of h5 playing fora win. 168 Two minor pieces against a rook When there are pawns on one 14 'it>f3 .l:Ib4 15 i.d5 l:ra4 16 i.f7 flank for both sides, the rook can .l:Ib4 17 'it>e21::rb7 18 i.c4J::r d7 successfullycontend with the minor The king breaks through in the pieces. centre, and Black changes the rook's horizontal resistance to a Dan-Pytel vertical one. Austria, 1978 19 'it>f3lk7 20 i.d5.l:Ie7! 21 i.c5 .l:Ie8 22 i.e4 .l:Ic8 23 i.d4 l:Ie8 24 'it>e3 Black has prepared against the breakthrough g4-g5. For example: 24 g5 fxg5 25 fxg5 lid8 ! 26 'it>e3 'it>g8, defending against 27 gxh6. Now in the event of 27 gxh6 gxh6 or 27 g6 Black is ready to sacrifice the exchange for the dark-squared bishop, reaching a theoretically drawn position. 24 ...l:re7 25 i.c5 .!id7 26 i.f5 .l:Id5! Material advantage is on White's In preventing the attempted side, but the inevitability of pawn breakthrough of the king to the d5 exchanges when pawns are ad square, Black once agains includes vanced, gives the defender the the rook for defence along the rank. possibility of exploiting the ex 27 i.d4 lia5 28 �e4 lib5 29 i.e6 change sacrifice motive to reach a White can attempt to break theoretically drawn position. through with the king, by playing 29 1... f6 ! i.d7. Here are some sample vari The more active l...f5 creates too ations: (a) 29 ...l:!a5 ?! 30 i.b6 J::ra6 much space for the bishop's 31 i.c5 lia5 32 Wd5 'it>h7 33 Wd6 activities. Wh8 34 i.d4t; (b) 29 ...l:!b7 30 i.c6 2 g4 :g,33 l:!c7 31 i.d5 .!ld7 32 i.e6 !Ib7 33 One of the most important el i.c5 l:!bl 34 i.d7 .!lcl 35 i.f8t; (c) ements is playing for the restriction 29 ...l:!b l! 30 i.c5 l:l.'.cl 31 i.f8 c;t>g8 of the king's movements. 32 i.a3 l:l.'.al 33 i.e6+ 'it>h7 34 i.c5 3 i.f2 h6 4 h4 .l:Ib3 5 h5 .!lcl, and Black succeeds in defend White plans the transfer of the ing himself. bishop to d4 and organises the break 29 ... .l:Ia5 30 i.c4�a3 31 i.e6 g4-g5. As before 31 g5 fxg5 32 fxg5 5 ....l:Ib4 6 �g3 :gb3 7 i.d4 �d3 8 does not work in view of 32 ... :gh3 i.b2 .:i.b3 9 i.cl.l:Ic3 10 i.d2 J::rc2 33 gxh6 J::rh4+ 34 'it>e3 l:!xd4! with a More principled is 10 ...l:!d3, hold draw. ing the third rank and preventing 31 ... I:l.a532 i.d7.l:Ia2 33 g5 White's coming manoeuvre. Reconciling himself to the inevi 11 i.e3 lic3 12 Wf2 �a3 13 i.e4 table draw. On 33 'it>d5 the balance .l:!a4 is maintained by 33 ...1::r d2. Now defence by the rook moves 33 ... fxg5 34 fxg5 I:!'.a6 on to the fourth rank. Two minor pieces against a rook 169 Also possible is 34 .. Jlh2 35 gxh6 king to f6 or h6. However it should J::[.h4+ 36 @e3 l:rxd4 37 c,t>xd4 gxh6 not be enough to win. with a draw. 11.. . .!le7 12 ctJb4 J:ie8 13 i.g6 35 g6 .!id6 J:ib8 14 ctJd3 Drawn. It is difficultto reach the h6 pawn, On 36 i.g4 there follows since Black does not allow the white 36... l:!'.xd4+ !. king admission to f5 . All his hopes lay in 'humanitarian aid from the West'. Ponomariov-Plaskett Hastings, 1999 14... c,t>d5 15 'it>e2 'it>d4 16 'it>d2 f3? He can't keep himself waiting! 17 ctJel .l:!b2+ 18 lt.Jc2 'it>e5 19 'it>e3 J::rb3+ 20 i.d3 .!lc3 21 lt.Jel :!cl22 ltJxf3+ And White won shortly. Changing the material balance (obtaining a position with rook and pawns against two minor pieces) is very often exploited by chessplayers of the highest rank. Petrosian-Tal The well-known Scotch ending Moscow, 1964 with two white minor pieces against a black rook. The game smoothly transferred to this drawn position, there only remains to make a solid move beforethe time control... 1... .!la2? In accordance with his active style. 2 i.xd5J::r d2 2 ...'it>xd5 3 lt.Jb4+. 3 i.e4 i;t>e6 Despite the win of the pawn, as before it is not apparent how White wins. 4 h4 @d6 5 i.f5 .!le2 6 'it>f3 I:!'.e8 l ... CZJxe3!? 2 i.xb7CZJxdl 3 l::txdl 7 i.g6 .!le7 8 lt.Jcl J:iel 9 ctJd3 J::re7 .!lad84 CZJc3 l:!'.xdl+ 5 CZJxdl J::i.d8 6 10 i.e4 l:!e8 11 h5 i.f3l:id2 Not an obvious decision. It was For the two pieces Black has not possible all the same to exchange on quite enough material, but activat g5, attach himself to the black ing the rook gives Black sufficient pawns, place the bishop on f5 , the play. knight on e4, and try to run with the 7 @fla5 8 i.e2 e5! 9 a4 i.d4! 170 Two minor pieces against a rook After creating his passed pawn, Formanek-Mikhalchishin Black can rivet White's forcesto the Hastings, 1985 struggle against it. 10 .txd4 exd4 11 'it>el lta2 12 g4 'it>t'8 13 f4 'it>e7 14 lllf2 .l:!al+ 15 .tdl lia2 16 h4 f6 17 h5 <;fte6 17 tlld3 g6 18 hxg6 hxg6 19 .te2 .l:!c2 20 'it>dl :f.c3 21 @d2 g5 And Black quietly held the draw. Beliavsky-Geller Erevan, 1975 1 ... tllxb3! 2 'ihe7! tllxcl! 3 'it'xt'8+ There· is nothing else for White. 3 ... @xt'84 :xcl .txa2 Black has a rook and two edge pawns for two pieces. To win it is sufficient for him to exchange the bishop. 5 lllg5 f6 ! 6 llle4 After 6 lllxh7+ 'it,;Jf:-1 7 tllxf6 Zhc3 8 11xc3 'ifi>xf6 the edge pawn must 1 ...J:.b4 ! decide the outcome of the game. Black stands to win the a2 pawn, and White finds himself faced with ,problemsbeyond his control. Now 2 tlJd2 does not work because of 2 ...l:tb2. Therefore White has to confine his bishop to passivity. 2 .td3J:.a4 3 .Jlbl g5! A pawn advantage on the flank is only real when its exploitation gets under way. 4 @e2 @g7 5 lll d2 f5 6 @d3 .l:!f4 7 'it,;le2 After 7 f1 .l:.h4 8 h3 h5 and 9 ...g4 a weakness arises for White on g2. 7 ....l:!h4 8 h3 J:.a4 6 ... f5 7 tllf6 :f.a7! 8 lll xh7+ i:r.xh7 Anda draw was agreed. 9 .tb4+ l:!c5! 10 .txc5+ bxc5 11 Two minor pieces against a rook 171 lixc5 �a7 12 i..c4 i..xc4 13 �xc4 14 lic2 a3 15 l:l'.a2 @e7 16 @fl a4 'it>d6 17 We2 'it>c5 18 'it>d3 'it>b4 19 Again there has been a transform 'it>c2 e5 20 h4 e4 21 @bl lid7 22 ation of the advantage. This time the @c2 lic7+ 23 �bl 'it>b3 24 l:id2 game passes on to a rook ending l:!h7 White resigned. with positional advantage forBlack. 11 Rook against knight and pawn The rook is a little stronger than a @g7 13 c,t>xf4 @f6 14 g3+- or knight, and it would seem that with 10... lt:lxg2 11 'it>g5 f3 12 It.a2 and pawns on one flank the knight, in then 13 J:if2, winning the f3 pawn, view of its capacity to jump, could and then also those on g6 and h5. compete with a stronger piece. 11 l:ras lt:Jg4+ 12 @n @h7 13 However, practice shows that in this lla7! 'it>h614 'it>f8! type of position the knight is even Avoiding 14 'it>g8 ctJf6+ 15 @h8 weaker than the bishop in the g5! struggle against a rook. 14 ... ctJf6 15 J::!f7 lt:Jg4 16 J::!g7! Black resigned. Farago-Csom Budapest, 1984 We don't need to talk about a big alignment of forces, when miniatures provide brilliant examples ... Romanishin-Rodriguez Mo scow, 1985 1 'it>h3! Black is in a 'little zugzwang' and he is forced to let out the white king. 1...ctJd6 2 l:1c7 lt:Je4 3 �a7 ctJf6 4 'it>g3 lt:Jg4 5 'it>f4 lt:lf6 6 �e5 lt:Jg4+ 7 We6 f4 Is it possible, perhaps, for a Bad is 7 ...g5 8 'it>xf5 gxh4 9 �a6+ grandmaster to lose this position? 'it>g7 10 'it>g5 lt:Je3 11 J:ia2, and As it turns out,yes-it is possible! White attacks the pawns. 1 ...h5 8 @n 'it>h7 9 l:!a5! lt:le3 10 'it>f6! Simpler is 1...ctJg7 and 2 ... ctJf5, 'it>h6 creating a typical drawn setup. There is no saving himself by 2 'it>e4 lt:Jg5+ 3 \t>f4 lt:le6+ 4 �e3 10... tll f5 11 l:!xf5! gxf5 12 'it>xf5 lt:Jg5 , Rook against knight and pawn 173 Now, however, after 4 ...'Llg7 5 When defending one should re We4 'Llf5 6 'it>f4 and 7 l:ta6+ duce the amount of remaining ma weakening the g5 square where the terial, therefore he should prefer 6 white king will infiltrate. h5!. 5 h4 'Lle6? 6 ...f5 7 h5 .lla3 8 h6 Black loses his nerve. After the White intends to attack the h7 correct 5 ...'Llt7 6 Wf4 'Llh6 7 .l:Ia6+ pawn. Wg7 8 'it>e5 'Llg4+ 9 We6 g5 ! it is 8 ... @c4 9 @e2 :gb3 10 'Llhl still a draw. Now, however, a rook Trying to provide a covering de against pawn ending is reached. tachment forthe black king. 6 We4 g5 7 1If5+ Wg6 8 'it>e5 10... l:rb2 11 'it>f3 i;t>d3 12 CLlf2+ gxh4 9 'it>xe6 h3 10 We5 h2 11 �fl Wd2 13 e4 l:!b3+ 14 i;t>g2 fx e4 15 i;t>g5 12 lihl Wg4 13 I:!'.xh2 h4 14 'Llxe4+We3 16 'Llc5 lig2+Wf3 15 l:ra2Black resigned. After 16 'Llf6 Wxf4 17 'Llxh7 Black plays 17... :gb8 ! followed by h8, and all the white pawns are Hulak-Beliavsky attacked. Bled, 1998 16 ... lib4 17 Wg3 l:!xf4 18 'Lld7 .llf3+19 'it>g4l:tfl! 20 'Llf6 l:!gl+21 'it>h3'it>f3 White resigned. Incidentally, Vidmar lost the same classical endgame against Alekhine. Salov-Kamsky Buenos Aires, 1994 Defence in such positions is not easy and requires, in the first in stance, an understanding of the best distribution of the functions be tween his pieces. Therefore White should place his knight on h3 and wait for the black king. Instead of this he decides to post his knight on g3, which proves to be incorrect. 1... 'Llh7! 1 'Lle2 lial 2 'Llg3 'it>e5 3 'it>g2 A correct transfer of the knight to l::ta4 4 f4 +? g6 in order to defend the h4 pawn Too active, but, you know, this is and attack the e5 point. not a rook ending. He should prefer 2@h3 4Wf3 . After 2 g5 'Llxg5 3 l:!xh4 the 4 ... Wd5 5 i;t>of6 ! 6 g5? knight all the same goes to h7! 174 Rook against knight and pawn 2 ..•®f8 3 Ite4 ®g6 4 gs Wg8 5 Beliavsky-Kupreichik Wg4 Wh7 6 @n It is well known that with pawns on one flank the knight is rather stronger than a bishop. But against a rook the knight proves very weak, since it oftencannot escape fromthe pursuit of a more mobile rook. Beliavsky-Short Linares, 1995 1 ... li>e6? Despite all the suffering, he should allow l...®xe3 2 Itb5 li>f6 3 li>g3 Wg6 4 Ite5 or 3 ...tlJfl + 4 li>f2 tlJh2 5 :lf5 ! followed by 6 Wf2-g3. 2 .l::Ib3 WdS 3 :i:t'.bS+ lit>c6 4 :S:xgS tlJxe3 5 Wg3 tlJd5 6 :ifs e3 7 Suba-Chiburdanidze Dortmund, 1983 1 ltc3! Inferior is 1 We3 g5 ! with the threat of ...f5-f 4. 1.•. e4 2 b6 d2 Forced. 3 @e2 ext3+4 l:ixt3 'it>c6 If 4 ...®e4, then 5 l:.b3!, 5 Itxf5 ®e4 6 .l:te5! ®xg3+ 7 ¢'xd2 g5 After 7 . ..@xb6 8 @e3 g5 9 hxgS hxgS 10 :Xg5tlJfl + White plays 11 lit>d3 followedby t:tg2 and @e2. 8 hxg5 hxg5 9 Wd3 g4 On 9 ... 'it>xb6 again follows 10 One of the few positions where, llxg5 tlJfl 11 !Ig2. with equal pawns, the knight can IO l:te6+ Wb7 11 On 2 .lla6 there is 2 ...lt::\ f5 , while A.Petrosian-Rashkovsky on 2 'it>h3 tt::le4 3 �a5+ g5 4 g4+ Minsk, 1976 fxg3 5 k!e5 tt::ld6! 6 hxg5 lt::\f7 with a draw. 2 •.. 'it>g43 'it>gl lt::\fS 4 i;t>f2 lt::\g3! Of course not 4 ...tt::l xh4? because of5 g3 !. S llb4 If 5 �a6, then 5 ... g5! 6 hxg5 tt::le4 and 7 ... lt::\g5 with a drawn setup. S ...tt::l fS 6 �e4 gS 7 hxgS WxgS 8 1:1'.eS 'it>g4 9lle8tt::lh4 Also possible is 9 ...lt::\ d4. 10 J:Ig8+ @rs 11 'it>e2 tt::lg6! 12 i;t>d3tt::l eS+ 13 'it>d4 f3!and a draw! Is there a basis for Black to play for a win? To this question the Stein-Bobotsov Moscow, 1967 grandmaster himself answered in the affirmative and to start with went after one of the pawns. There followed: 1...l:l.'.d3 2 e4 l:!a3 3 @f2 .:i.h3 ! Black suggests that his opponent leaves the g-pawn. His argument ap pears to be 4 i;t>g2 l:th4 5 tt::le3 l:txg4+! 6 tt::lxg4 i;t>xg4 with a win. The Erevan grandmaster agrees with the evidence and tries to organise a defence with an equal number of pawns. 4 'it>e2 lig3 S @f2 l:rxg4 6 i;t>f3 l:!h4 7 'it>e3 l:th3+ 8 'it>d4 lla3 9 1...'it>fS 2 lid4bS 3 axbS axbS 4 'it>c4 .!Ia6 10 'it>d3 'it>g4 11 i;t>e2 lids 'it>e4! On l l ...'it>g3 there is the reply 12 Activity! We3, and it is not apparent how he S �xbS tt::ld3+ 6 'it>e2 f3+ 7 'it>d2 can improve his position. Rash f2 8 i;t>e2 i;t>d4 9 .:i.b8 'it>c4 10 @fl kovsky in fact allows the second 'it>d4 ll lib7 argument, the essence of which con In the event of 11 b4 'it>c4 12 b5 sists of transferring the turn to move 'it>c5 13 b6 tt::lb4 Black holds on. to his opponent. 11...i;t>c4 12 'it>e2 i;t>d4 13 J::i.d7+ 11. .. lla2+! 12 'it>e3 l:ta3+ 13 'it>d4 'it>e4! .!Ia6 14 'it>e3 i;t>g3 lS lt::\f4 l:!a3+ 16 The only, but sufficient,move. tt::ld3 l:l'.b3 17 'it>e2 J:tb8! 18 'it>e3 14 l:!b7'it>d4 Drawn. lle8 19 tt::lf2 fS 20 tt::lhl+ 'it>h2! 21 176 Rook against knight and pawn lLif2 'it>g2 22 'it>f4 'it>xf2 23 exf5 6 h4! g5 7 hxg5 h4 8 .l:Ixh4 .l:If8! And Black resigned because of And here is the last argument, 8 ...tt.Jdl + 9 �d2 liJf2 10 g6 @f6 11 bearing in mind that the ending has .l:If4. assumed a study-like character. 24 'it>g5 �e3 The zugzwang also plays an im On 24 @e5 Black would have portant role in the next example. proceeded on the other side-24 The extra pawn on the other flankis ...� g3. usually lost aftera few moves. 25 f6 @e4 White resigned. Adorj an-Morozevich Chemin-Geo rgiev Alushta, 1994 Dortmund, 1991 1... .lldl! 2 lLig6+ @h7 3 lLie5 The defence, even with play on :!fl! one flank and an extra pawn, isn't A 'little zugzwang'-White has to easy. Here the most unpleasant release the black king or start push thing is zugzwang. ing his pawns somewhere. 1 lie8! lLib2!? 4 c4 g6 5 h4 @g7 6 Ci:Jg4 h5 7 Or 1...�d6 2 l:!.g8 lt.Je5 3 @e4 g5 lLie3 :!:!'.bl! 4@f5 winning easily. The white king is riveted to the h4 2 .:i.d8+ �e63 lid4 pawn and his black counterpart is Now White's idea is to hunt for ready to start his 'promenade'. the runaway knight. 8 tt:'ld5 @f7 9 lLie3 @e6 10 'it>f4 3 ...�f6 4 �e4 h5 White trieshis last chance. After 4 ...@e6 White wins by 5 10 •••.l:Ixh4+ 11 g4 :!:i.b4 liJdl 6 1::rb6+ 'ii'f7 7llb3! Ci:Jf2+ Or 11 'it>g5 I:!'.e4 12lt.Jd5l:!xc4 - +. 8 'it>e5 @g7 9 .:i.f3 lLihl 10 'it>f4 g5+ 11. . . .l:Ihl 12 gxh5 lixh5 13 'it>g4 11 �g4 �g6 12 lifl h5+ 13 'it>f3 �e5 14 c5 �e6 White resigned. and the knight is caught. 5 �e3! @e5 Now two very similar positions Or 5 ...g5 6 �e2 followed by 7 with just one small difference-the llb4. white pawn. Rook against knigh t and pawn 177 Sokolov-Khalifman llJg6+ 2 'it>e4 'it>g5 3 .l:Ig7 'it>h6 4 Minsk, 1986 .!la7 'it>g5 5.!la3 CZJe5 6 .l:Ig3+'it>h4 7 Wf4 ClJc6 8 !Ic3 ClJe7 9 .!id3 'it>h5 10 .!id7. 1... @h4 2 �h8+ 'it>g53 'it>d5? This king transfer isn't correct. The king has to go towards his own pawn when White has serious win ning chances. 3 ... QJf34 'it>e6 It was still possible to return to his pawn. 4 ... CZJe5 5 'it>d5 ClJf3 6 Wd6 CZJe5 7 'it>e6 .!i:Jf3 8 'it>e7 White has definitely decided not Black played to go back to his pawn. 1...'it>g5 8 . . . CZJe5 9 .l:Ia8 CZJg6+ 10 @f7 ClJf4 ...tr ying to activate his king. The 11 .l:Ia3f5 12 l::tc3 ClJxh3! game continued Black has calculated this known 2 'it>f3 'it>f5 3 l:tal 'it>g5 4 .li(a5+ endgame very carefully. CZJe5+5 'it>e4 Wh4 13 lhh3 \t>g4 14 .llh6 f4 15 .:i.g6+ Black tries a counterattack on 'it>h3 16 I:!'.f6 Wg3 17 'it>e6 f3 18 White's pawn, but he hasn't enough 'it>e5 f219 We4 'it>g2 Drawn. time. 6 'it>f5 'it>h3 7 .!la2 ClJd7 A knight is generally more useful 7 ... ClJf3 doesn'twork because of 8 than a bishop when the pawns are .!la3! 'it>g2 9 .l:Ixf3 Wxf3 10 h4+-. all on one flank, but against a rook 8 .l:Ia7Black resigned. the knight has a more difficulttime. Dokhoian-Shirov Horvath-Sherzer Klaipeda, 1988 Brno, 1993 The game continued 1 'it>g3 'iite6 2 .!la3 g5 1 'iite4?! Black has to build some sort of This is not the most exact way to wall against White's king, but it win-more direct would be 1 \t>f4 cannot be done. 178 Rook against knight and pawn 3 l:!a5 Wf7 4 'it>f3 'it>g6 5 We3 7 @e3 h4 8 l:!.f3 'Lixg2 9 @f2 h3 'Lih5 6 J:Ia6+ Wg7 7 'it>e4 'Lif4 8 10 'it>g3 'Liel 11 l:!c3 'Lig2 12 'it>xh3 @f3 'Lif4+ 13 Wh4 'it>f6 14 J:Ia3 'Lig6+ Another good possibility was 8 15 'it>h5'Lif 4 16 @h4 l:!.a2 'it>f6 9 h4 'Lig6 l O h5 'Lif4 11 There is nothing in 16 @h6 'Lid5, g4 and after a check the king ob followed by the advance of the tains the very important f5 SCI!lare. f- pawn. 8 ... 'Lih5 9 'it>g4lt.Jf4 10 g3 tlle2 11 16 ... 'Lig6 17 'it>g3 e5 18 h4 'Lif4 l:!e6 19 .!la6+ Wg7 20 'it>f3 'it>f7 21 J:Ia5 Another method would be 11 '\t>f6 22 .!la6+Wf7 Drawn. .!id6, 12 l:!d2 and 13 'it>h5. 11•.. 'Lid4 12 .l:!e4 'Lib5 13 'it>h5 Zaichik-Sorin 'Lid6 14 .!le7+ Black resigned. Erevan, 1996 Two instructive and I would even say amazing endings played by an international master from Argentina against experienced grandmasters. Sorin-Alterman Erevan, 1996 For the exchange White has three pawns and an easily winning position-for example: 1 'Lig6 l:!.a62 f5 .!lb63 f6+! .!lxf6 4 h6+ rt;f? 5 h7 .l:!xg6+ 6 'it>h5 .l:!g8 7 hxg8='it'+ c;t>xg8 8'it>g6. But Zaichik did not see this plan and chose another one. 1 c;t>f5 .l:!f6+ 2 'it>e4.!la6 3 g5 :!.al White threatened to play 4'\t>f5 J:ig5-g3-c3 and, after capturing the Not bad was 4 f5 :a4+ 5 'it>f3 b2 pawn, winning the well-known l:!a3+ 6 Wf4 .!la4+ 7 'it>g3 l:!a3+ 8 ending of rook against knight with 'it>h4 l:ta4+ 9 'Lig4, and there is no three pawns on the flank. But Alter defenceagainst f5 -f6+ and g5-g6. man finds an elegant way to save 4 ... �a2 5 g6 'it>h6! the game. The threat was 6 'it>g5 and then 7 1...'Lie3+! 2 'it>xb2 'Lid5 3 'it>c2 h5-h6. If 3 g3?, then 3 ...'Lif6, and the 6 'Lid7 rook is in a trap after 4... h6. 6 'Lig4 led to roughly the same 3 ...'Lixf 4 4 Wd2 Wf6 5 l:!g8'it>t7 6 thing. llg3 1!5! 6 ....!la5+ 7 '\t>f6 l:ra6+ 8 Wf7 .!la7 Activity comes first! 9 'it>e7 .!la5 10 'Lie5 Rook against knight and pawn 179 Better really was 10 ll'if6, prepar Martinovic-Mikhalchishin ing the returnof the king. Jgalo, 1994 10 ... l:!.a7+ 11 'it>f6 l:ta6+ 12 @f7 lia7+ 13 @f6 .l:Ia6+ 14 'it>f5 @xh5 15 g7 .l:ta8 16 ll'ig4 There is nothing in 16 'it>f6 'it>h6. 16... .l:ta5+! Now Black's main task is to keep the king as far away as possible from thef4 pawn. 17 @e4 l;Ia4+ 18 'it>d3 �a3+ 19 'it>e2lia2+ 20 @f3 If the white king heads towards the rook, then at a necessary mo ment follows lia8, ll'if6 'it>g6, g8='iV l:txg8, ll'ixg8 'it>f5,and a draw. 20 ... lia3+ 21 ll'ie3 .l:Ia8 22 ll'if5 1 b5? @g6 23 g8='iV+ lixg8 24 ll'ie7+ @f7 Nothing is gained by 1 @fl! 25 ll'ixg8 Wxg8 26 'it>e4 l:l.'.xb4 2 @g2. 1...'itid5 2 ll'if5 @e4 3 ll'id6+ There is no saving himself by 3 ll'ie3 'it>f3. 3 •..'it>f3 4 'it>d 1 In the event of 4 ll'ie8 l:l.'.b1 + 5 'it>d2 l:txb5 6 ll'ixf6 after 6 ...h5 with the idea of 7 ...�f5 Black has a de cisive advantage. 4 ...l:txt1 4 ...'it>xf2? 5 ll'ie4+ 'it>f3 is mis- taken because of 6 ll'ixf6 lixb5 7 'Llxh7. 5@cl 26... 'it>fS! The result is not changed by 5 A clear-cut move--and a draw! ll'ie8 @xg3. 5 .•..l:ta2! 6 b6 In the following position White Or 6 ll'ie8 £5 7 ll'if6 h6. has good compensation for the ex 6 •.•lia8 7 @c2 lib8 8 b7 'it>xg3 9 change, but he should immediately 'it>d3'it>xh4 10 'it>e4 'it>g4 11 'it>d5 h5 start building a fortressfor his king. White resigned. 12 Queen Endings In queen endgames the main prin 2 gxf7�e2+ 3 �cl �xf7 4 h5! ciples are slightly different from White must exploit his chance to other endings. And, to be precise, the utmost-the passed h-pawn is though material advantage (for also ready to be exchanged for the example, a pawn) naturally has b5 pawn. great significance, much more im 4 ...�c4+ 5 'it>d2! portant is a passed pawn. Thus the Inferior is 5 'it>b1 b4, and after the main principles can be placed in the advance of the b-pawn the white following order: king is threatened with mate. White does not fear the transfer to a pawn (a) Passed pawns ending. (b) Activity (approach) of the king 5 ... 'iVc6 6 'iVh8! (c) Activity of the queen White wants to advance the h (d) Combinational exploitation of pawn further. various factors 6 ..• 'ifd5+ 7 @cl 'ifc4+ 8 �b2 'ifd4+ 9 '>t>b3! Passed Pawns Meeting the pawn half-way! 9 .•• 'iVa4+ 10 'itb2 'i¥b4+ 11 'it>a2 Mikhalchishin-Gufeld 'iVd2+ 12 �b3 'iVd3+ 13 'itb4 Nikolaev, 1981 'ifd4+14 'it>a5'iVa4+ 15 Wb6 The white king is up in the 'clouds' and Black reconciles him self to a draw. Ed.Lasker-Marshall New York, 1923 There followed: I g6+! 'it>g7 If l...'it>xg6, then 2 'iVg5+ with perpetual check or l ...fxg6 2 'iVe7+ with the same result. Queen Endings 181 A difficult struggle with a passed Again not 7 'i¥xg6? in view of pawn hoping to queen but with the 7 ...�h 2+!. It is important to activate king unable to get over to help it. his king via the g5 square. 1. ..'ii'e2 2 'ii'cl g5!? 7 ...'ii' a2 8 Wg5 'iVf2 Striving to create counterplay. No help is 8 ... 'i¥f7 9 'iVf6 !. However 2 ...'ii' e5 deserved the pref 9 �f4 \Wb6 10 iif7+ Wc8 11 erence, intending to transfer the 'iVxg6 king to the queen's flank with an Black can resign since l l ...'it'e3+ exchange of the h7 and g6 pawns 12 Wxh5 \Wxf3+ is no use because forthe pawn on a5. of 13 \Wg4+. 3 a5 f4 4 'i'fl 'ii'a2 5 'it'b5!�g6 6 'iib6+ 'it>h5 7 a6 g4 8 a7 g3 9 fx g3 fx g3 10 'iVc5+ 'it>g6 11 iid6+ @f7 Activity (approach) of the king 12 iif4+ �e6 13 'it'e3+ �d7 14 @fl! 'ii'a6+ 15 �el '\tics 16 'iVxg3 One of the most important factors 'iVxa7 17 'ii'g8+ 'it>b7 18 �xh7+ and in queen endings. 19 �xa7 +-. Ribli-Hertneck Taimanov-Gligoric Dortmund, 1986 Santa Fe, 1960 1 h3 h5 1 'ii'd3! Otherwise White will play g3-g4 Defending against ...'i¥c2+ and obtaining real winning chances. preparing a4, therefore Black de 2 'i¥d7+'\t>f6 3 'iVe8 �a6? cides on a pawn sacrifice with the The only defence was 3 ...'it>g7 4 aim of activating his forces. 'ii'e7+ @h8 5 iif8 'it>h7 6 Wiffl+ �h6 1...'it>f6 2 'ii'xd5 'i¥c2+ 3 �h3 7 'ii'g8 'iWh2 and White must return 'ii'cl to the plan with g3-g4, creating a Chances of saving the game re passed e-pawn. mained with 3 ...'ii' f2! 3 f4 fifl+ 4 4 \Wf8+'it>e6 5 'it>f3! 'iVg2 iid 1. Suddenly there has appeared the 4 'iVd4+! 'it>e6 idea of penetrating to the king's Not possible is 4 ...@f7 because of flank with the white king-very 5 'it'f4+, nor 4 ...'it>g5 because of 5 typical in fact. 'Yi'e5+. 5 ... �a4 6 'ii'g8 'it>f6 7 'iVd8+ 'it>e6 5 'ii'e4+ Wd7 6 �h4 \Wd27 �e5! 8 'ii'd4! 182 Queen Endings And Black resigned because 1 'it>c3!'ii' cS+ 2 'ii'c4 °iVe3+ 3 Wb4 there is no defence against Wf4-g5. �d2+ 4 'it>c5 'ii'a5+ 5 'it>d4 'ii'd2+ 6 For example 8 ... 'ifa2 9 'ifg7 or 'it>e5 �d8 7 'ifd4 'ifh8+ 8 'it>e4 8 ...� c2 9 e4!. 'iVhl+ 9 'it>d3 11Vt'3+ 10 'it>d2 'ii'g2+ 11 'it>cl 'ii'fl+ 12 'it>b2'ii' b5 13 'ifd8 Mikhalchishin-N esto rovich Black resigned. Budva, 1994 Hoi-Mikhalchishin Budapest, 1989 1.•. 'it>e7 ? In queen endings the king must either be ultra-safe or ultra-active! Better here would be l...'ii'a3! cut The king usually approaches its ting offthe white king. own pawn, but very often it also 2 e5 'ii'b6 heads towards the opponent's king If 2 . . .'if a3 then 3 'ii'd6+ 'it>e8 4 to create mating threats. e6+-. 1 'iVrs+ 'it>g6 2 °iVgs+ 'it>rs 3 3 'it>t'3 'ifc7 4 'it>f4 'i1Nb6 5 'it>g5 'ii'h7+ 'it>e5 4 'i'e7+ Wd4 5 1!Vd8+ 'it>e86 'it>h6! Black resigned. 'it>e3 6 'ii'e7+ Wf2 7 'i'a7+ @fl 8 'ifgl+ Rittner-Bernstein As a result of the approach of the corr, 1968 black king White has had to mis place his queen. 8 ... We2 9 'ii'd4 c2 10 'il'g4 'it>d3 11 'ii'g_3+ 'it>d2 12 'il'f4+ 'it>c3 13 'iVcl �d6+ 14 'it>h3 'iVd3+ 15 g3 'ii'd7+ 16 g4 'ii'd4! The black queen is centralised to the utmost, there is no threat of per petual check, and the march of the a6 pawn is decisive. 17 'it>g3 'ife5+ 18 'it>h3 a5! 19 a4 'it>b3 20 'it>g2 'ii'b2 21 'ife3+ 'it>a2 White resigned. Queen Endings 183 The active position of the queen Marshall-Alekhine Bradley Beach, 1929 This allows the king to be utilised in the struggle against enemy pawns. Gufeld-Minev Sofia, 1967 1...°iVa6+! 2 @g2 Witd6 In the first instance Black block ades the opponent's pawn. 3 iid4 @gs 4 11Vc4@f7 It turnsout that the white d5 pawn The white queen controls the can be attacked by the king. position and attacks all Black's wea s Wilbs °iVd8 6 f4 11Vc7 7 @f3 knesses. It is only necessary to in 1!Vc3+ 8 'it>g4 clude the king. Clearly not 8 @e4 'ir'b4+!-the 1 @fl fS 2 h4 'i'c7 3 @e2 'iib6 4 pawns in this formation usually win 'ikf7+ @h6 S b3 'i'd6 6 'i'c4 'it>h7 7 forBlack. @el 'iib6 8bs 'ii Wi/c7 9 @d2 'i'eS 8 ...'it>e7 9 'iib6 hS+! 10 'it>xhS 10 @c2 iial 11 'i'xcS 11Va2+ 12 'i'h3+ 11 @gs 'i'g2+ 12 @rs @c3 iixf2 13 11Vc7+ 'it>g8 14 'it>d4 'i'xdS+ 13 'it>g6 'i'd6+! 14 'i'xd6 'i'b2+ @xd6 lS fS a4 16 @xg7 a3 and If 14 ...11Vd2+, then 15 �e5 'i'xe3+ Black won. 16 @f6 with a decisive invasion. lS 1!Vc3 'i'f2 16 fixaS 11Vxg3 17 Keres-Aronin iid8+ @h7 18 aS Wi/el 19 'iib6 f4 Moscow, 1951 20 'iVc( 'it>g8 21 a6! 'i'd2+ After 21...'Wixe3+ 22 'it>e5 one advanced passed pawn is quicker than two. 22 'it>eS'ii' h2+ 23 @e6 'ii'xb3+ 24 �f6 'ii'h2+ 2S 'iVeS Black resigned. In the following position, who stands better? He who has the further advanced pawn or he who advances it further?Let us see. 184 Queen Endings 1 ...@e6? Combinational exploitation of Better is 1 .. . h5 or 1...e4, trying various factorsin queen endings somehow to exploit his pawn major ity on the king's flank. Ballon-Mikhalchishin 2 g3 'iVdl+ 3 @g2 'li'd7 4 h4 Berne, 1995 'itb7+ 5 @h2 'il'd7 6 'iVf8! h5? A decisive mistake. Correct is 5 ... f5, endeavouring to advance his pawn. 7 IVgs+ @rs 8 c5 'ii' d4 9 'iVc8+! @e4 10 Wg2 @d3 11 \'Wa6+@d2 12 'iVd6 Also not bad is 12 'il'xf6. 12 ...@e2 13 1.Wxd4 exd4 14 c6 d3 15 c7 d2 16 c8='iV dl=\li' 17 'ii'e6+ Wd3 18 'li'd7+ We2 19 'ii'e7+ @d3 20 'ii'xa7 and White won easily. Black has a protected passed Romanovsky-St�hlberg pawn plus a much better pawn Moscow, 1935 structure. 1 ...'ii' f5! 2 @gt Vi'd3! By speculating on a transfer to a winning pawn ending, Black threatens to advance his passed pawn. 3 'li'f4 c3 4 'ii'g4! White wants to put the black king in a box, but it boldly ploughs on. 4 ... f5 ! 5 'ii'h5+ @e6 6 ik'e8+ @d5 7 'ii'e5+ 'i!tc4 8 d5 c2 9 ik'f4+ 'it>b3 10 Wh2 'iVc4White resigned. Botvinnik-Donner 1 \'Wd5! Noordwijk, 1965 After this centralisation White threatens to transfer his king to e4, therefore Black must hurry with his counterattack. 1 ...\'Wb2+ 2 @e3 Yl'cl+ 3 @e4 'iiel+ 4 @f5 'iixg3 5 iixc5+ @g8 6 'i'c4+ @h8 7 'ii'g4! With the threat to exchange White obtains the g6 square for his king. 7 ...li'el 8 'itig6! 1.Wct 9 Vi'd7 V!\Ygl+ 10 @f7 @h7 11 'iVf5+ g6 12 'it'd7 Black resigned. Queen Endings 185 Black has the better structure, and 2 c5 e5 therefore White's only chance lies On 2 ...�xa3 would have followed in the creation of a passed a-pawn, 3 1!fc7+ @f6 4 c6 'li°cl+ 5 'itih2 e5 6 which will be quicker than the '@'d6, leading to a position occurring passed h-pawn. in the game. 1 'iVd7+ @gs 2 WVcs+ 'itih7 3 3 'iVc7+'itie6 4 �d6+ @f75 c6 'iVxa6 'ili'g5+ 4 @fl 'tlkxh5 5 'ili'd3+! The pawn is already on c6 and 'itigs there is no apparent perpetual check. A trap-after 5 ...'iVg6? 6 'i¥e4! the 5 ... 'i!Vcl+ 6 'itih2 'i!Vf4+ 7 'itigl pawn ending is winning for White. 'i!Ve3+ S @fl �f4+ 9 'itie2 �e4+ 10 6 a4 'iVhl+ 7 'itie2 h5 8 'iVd5+ 'itid2 'itih89 'iWdl! The king will boldly go behind Typical-by speculating on the enemy lines, the pawn already has transfer to a pawn ending, White no significance. forces back the opponent's queen. 10 ...'i!Vf 4+ 11 'itic2 VWe4+ 12 'itib2! 9 ...'iVh2? 'i!Vxg2+ 13 'itic3 'i¥a2 14 'i"d7+ 'itif6 Better is 9 ...�h4, though even 15 c7 then White's position is preferable. White has achieved his objective 10 b4 h4 11 �fl! -a pawn on c7. Now he needs to The queen is trapped! shelter his king from the checks. 11...g5 12 a5 bxa5 13 bxa5 g4 14 15 ... 'i!Vxa3+ 16 'itic4 'iVcl+ 17 a6 Black resigned. 'itib4'i!Vel+ 18 'itic5 'i!Va5+ 19 'itic4 Of course, not possible is 19 Korolev-Palm @c6? because of 19. ..'iVh5+ with a corr, 1980 perpetual check. 19 ...'ili'a2+ 20 'itib4 'i!Vb2+ 21 'itia5 �a3+ 22 'itib6'YWe3+ 23 'it>b7 'iVb3+ 24 @cs i¥b4 25 'i\Vd5 Black resigned, since the c7 pawn will in evitably go on to queen. Karpov-Timman Buenos Aires, 1980 Thanks to his far advancedpassed c-pawn and strong queen position on d6, White has a slight advantage. 1... 'iVb 2?! On l ...g5 there would have fol lowed 2 1!fd3 'itif6 3 'i!Vd4+ @f7 4 'it>h2 'iVc7 5 'itigl with a slight ad vantage. Stronger is 1...'ili'e4!, cen tralising the queen with chances of a White has problems since the draw. Black king is in play while White's 186 Queen Endings is placed far from the scene of Correct is 3 ..."i/Vb2 4 l'Wf5+ 'it>e7 5 action. l'Wc5+ @e6 6 �c6+ 'it>e7 7 'iVe4+ 1. ..b6 2 'ifh7 @d6 8 f4 , and all the same White If White had sacrificed a pawn to has a dangerous passed pawn. create a passed pawn by 2 h4, then 4 �c8 We7 5 �c7 @e8 6 \t>f6! Timman gave the variation 2 ...gxh4 Fantastic domination by the white 3 g5 h3 4 g6 h2 5 1!fe3+ �£3 6 pieces, leading to a winning pawn 1!fg5+ 'it>t2 7 1!fd2+ l'We2 8 'iVf4+ ending_. @g2 9 'ilkg5+ Wh3! 10 l'Wh6+ \t>g4 6 ...1\fb2 7 �e5 Black resigned. 11 g7 'iVd l+ 12 \t>a2 1lfd5+ 13 b3 1!fg2+ followedby 14 ...hl =�. 2 ...�d4 Marosi-Yudasin The exchange of the a7 pawn for Budapest, 1982 that on h3 favours Black. 3 'iVh6 'iVdl+4 @a2 l'Wd5+5 'it>al a5 6 �h7 1!fd4 7 l'Wh6 l'Wdl+ 8 'it>a2 �d5+ 9@al b5 Threatening a mating attack after ...b5-b4-b3. 10 h4 gxh4 11 g5 b4 12 axb4 axb4 13 b3 After 13 g6 b3! 14 ife3+Wg4 15 l'Wgl+ \t>h5! Black wins. 13 ...�e5+ 14 \t>a2 �e2+ 15 @bl ife3White resigned. Yusupov-Gerusel 1...1!fd5! Moscow, 1981 This centralisation is better than simply playing to eliminate White's pawns. 2 ife8+ @h7 3 ifb8�d7 He should not give the opponent counter-chances after 3 ...'Wxf5 4 1!fxa7 't\Vbl+ 5 @g2 �xb2 6 l'Wxf7. 4 l'Wf4 'iVc65 g4 White's only chance to change the course of the struggle is to open up the position of the opponent's king. 5 ... a5 6 h4 b5 7 g5 ifc4 8 �g3 �e4! 9 gxh6 gxh6 10 �c7 �xf5 11 l'Wxa5'ilfhl+ 12 'itig21!f xb2 Despite the material equilibrium, As a result of the exchange of Black has problems in connection blows Black's positional advantage with the excellent positions of has been transformed into a material White's king and queen. one. 1 l'We6+@f8 2 'it>g6 �g2+ 3 'ifg4 13 'ifc7 b4! 14 @fl 'Wi>l+ 15 �a2? , \t>e2�e4+ 16 Wd2 1!ff5 17 iVh6 Queen Endings 187 · . On 17 'ii'g3 Black\IUr achieves his Sometimes a com phcated method i · = ob/ect1ve after 17... wd5+ 18 @c2 has to be adopted to create a passed ,wc4+w 19 ,-.h.�2@ 'i/Nc3+! 20 wxc3 pawn. bxc3+ 21 xc3 �g 6 followed by 1 h6! 'i/Nc4+ '"°"' • .....,.,h5 and ...�xh4 . .1 ...'i/Nxh6 2 'ii'c8+ and 3 'i/Nxc6 17.•. 'i/Nf4 18 @d3 'ii'xh4 winning. Further comment 'ilNis u nnecess ary 2 @ 'ii' 6 3 e5! 'i¥a3+ 4 @ 'Wr f3 It>19 we3 'ii'f6 20 a7 b3 21 "iV�S "iVxb4 5 ex;: xh6 6 'i'c8+ �h g6 22 °iVf8 "iVxf2 23 'i¥g8+ @h5 'i1Nd7+@g6 8 fg7+ @h5 9 f7 'il'e4+l� White resigned. -���-=/. �- � ·, �.@ Mikhalchishin-C vetkoVIC .. ,.� 0�%�rut � "'i rnava, 1988 ., •. - -• . � . . ��·� ·� �• • ., •� -·�. ·�·. • 8 *"�, if:·w lffiljD •�. � p,1 '.t>h2 'ii'f4+ 11 @gl 'ii'c 1 + 12 @!�'. As a result of the 'triangulatio dance' Black h more check� 1;: ' and so he resign�� e game. 188 Va rious Endings Exercises: Various Endings 1 3 Indicate Wh ite 's winning plan. Find a plan to realise White 's extra pawn. 2 4 Find the right continuation Can Wh ite win? fo r Black. Various Endings 18. 5 8 What is right: l ... ilblor l ...ii.e6? How does Black realise his extra pawn? 6 9 Ho w does Black defend af ter Can Black make a draw? I 0.xa5 ? 7 10 How does Black continue? How does Black defend? 190 Va rious Endings 11 14 Find the right line fo r Black. Demonstrate a plan to realise the extra pawn. 12 15 What is Wh ite 's winning plan? Point out the correct plan of defence. 13 16 Whit� to move. What should be Can Wh ite win by I g3 hxg3 2 h4, th e result of the game? creating a passed pawn? Solutions to Exercises PAWN ENDINGS 1 dl=°iV 7 �c8+ and a queen ending was reached with an f-pawn for 1 Wf2 �rs 2 @f3 '\ties 3 g4 White, Janvarev-Schcherbakov, hxg4+ 4 'it>xg4 'it>e4 S hS fS+ 6 Moscow, 1994. Wh3! f4 7 h6 f3 s h7 f2 9 �g2 1-0, Botvinnik. s 2 (a) l ...g6? 2 'it>e4 'it>d6 3 h5 g5 4 'it>f5 b5 5 cxb5 c,t>d5 6 b6 'it>c6 7 White has a winning position, but 'it>e4 cJi>xb6 8 'it>d3, 1-0, Dreev in the game he lost. Anastasian, Tbilisi, 1989; 1 'be3?? (1 a4 'iii>d6 2 aS 'it>xdS 3 (b) 1... �f6 ! 2 'it>e4(2 h5 g6 3 a4 a6! +-) l...�d6 2 We4 c4 3 a4 c3 4 g5=) 2 .•. 'it>e6 3 hS 'it>f6 4 'it>f4 g6 S 'it>d3 h6, 1-0, Padevsky-Latinov, Sofia, h5 g4 6 h6 g3 7 'it>e3 d4+ 9 @f3 1973; d3=) 4 ...�d6 5 g5 fx g5 6 fxg5 @e6! (b) 1...�g5! 2 �e4 �g4 (2... f5+? 7 g6 @f6 8 @d5 @g7=. 2 ... @d6 3 3 �f3 !+-) 3 'it>e5 f5 4 h3+ �g5 5 h4 h5 'it>e6! 4 'it>c5 f5 5 g5 hxg5 5 h6 'it>g4 6 h5 f4=. @fi! 7 fx g5 f4=. 9 14 (a) l...g4? 2 We3 ! f4 + (2 ...gxh3 3 Yes, but not by 1 axb4 a3 2 @c3 @f2) 3 'ii>e2! gxh3 4 �f2 �f5 5 cxb4+ 3 �b3 �d5, and White re @gl @e5 6 @h2 @d4 7 �xh3 cJi>c4 signed, Sofia Polgar-Smyslov, Lon 8 �xh4 @xb4 9 'it>g4, 1-0, don, 1996, but 1 c4!!, creating an Rufenacht-Orseth,corr, 1996; impregnable fortress. (b) 1 ...@f6! 2 @e3 'ii>e5 3 f4 + gxf4 4 @f3 @d5 5 @xf4 @c4 6 15 'it>xf5 @xb4 7 @g4 @c5, and the king hurries to f8. 1 h6! (1 cJi>c6 f5 2 @d5 'it>f4 3 a4 'it>g3=, Filipescu-Citron, Romania, 10 1955) 1...'it>xh6 2 �c6 @g5 3 'it>d5 f5 4 a4 f4 5 'it>e4 +-; 1. ..gxh6 2 a4 1 c4!! cxd4 (l...dxc4 2 dxc5 bxc5 f5 3 a5 f4 4 a6 f3 5 a7 f2 6 a8='YW 3 'ii>e2 +-) 2 cxd5 bS 3 'ii>e2 b4 4 fl='i!V 7�g8+! +-. a4! �e5 5 h5 �e4 6 d6 b3 7 �d2, 1-0, Degraeve-Hansen, Germany, 16 1998. (a) 1 @e3? d4+! 2 cxd4 @d5 3 b5 11 'it>c4, 0-1, Havsky-Yuferov; (b) 1 @d2 d4 2 c4 d3 3 c5 @d5 4 (a) 1 'it>f4? g6=, Ilyin-Zhenevsky a5=. Botvinnik,Leningrad, 193 8; (b) 1 �f3! g6 (1...@e7 2 \t>f4 @e6 17 3 g3) 2 hxg6+ �xg6 3 �f4 h5 4 g3 +- 1 'it>e6! (1 @g6 �g3 2 'it>f5 @f3=) 12 1. ..@h3 2 'it>f5! @h4 (2 ...g6+ 3 �g5! +-) 3 c4! g5 4 c5 g4 5 'it>f4 g3 Yes, 1... cS!! 2 'it>g2 cxd4 3 exd4 6 'it>f3 @h3 7 c6 g2 8 c7 +-, Hoch, exd4 4 'it>f2 @e7 5 @e2 @e6=, 1972. Kozakov-Asparuhov, Pernik, 1972. 18 13 (a) l...e4 2 fxe4g4 3 �xf7! g3 4 (a) l...'it>e6?2 'it>c5f5 3 h3 fxg4 4 f6 g2 4 @e8 gl='if 5 fl+-; hxg4 d4 5 'it>xd4 �d6 6 f5 , 1-0, (b) 1...g4 2 fxg4 e4 3 g5 e3 4 Nimzowitsch-Chigorin, Carlsbad, �xfl e2 5 g6 +-; 1907; c) 1... 'it>g8 2 f6 g4!! 3 fxg4 e4 4 (b) 1 ...'it>c6 2 h4 2 h3 'it>d6 3 h4 g5 e3 5 g6 e2 6 gxfi 'it>h7 7 f8='if @c6 4 h5 ( 4 g5 fxg5 5 fxg5 hxg5 5 el=if+ 8 'it>ti+!! �e6 stalemate. Solutions 193 19 1-0, Anastasian-Khalifman, Erevan, 1996. 1... c,t>e2! (1...'i¥i>f2 2 'i¥i>e5+-) 2 h6 @d3=. 24 20 1 c4? led to defeat. White should give preference to 1 b4! cxb4 2 (a) 1 @g5? a4 2 h6 a3 3 g7 a2 4 cxb4 @f'l 3 b5 'i¥i>e7 4 g3 �f'l 5 h7 al='fl' 5g8='1'W 'iYcl+, 0-1, Wells @d6 @f6 6 'it>c7 'i¥i>e7 7 c,t>b7 'it>d78 Hector, Oxford, 1999; @xa7 'i¥i>c7 9 g4! e5 10 gxh5 gxh5 (b) 1 'i¥i>f4!!=. 11 'i¥i>a8 e4 12 a7, and a draw. 21 The game ended 1...'i¥i>f7 2 @d6 @f6 3 �c6 'i¥i>e7! 4 �b7 'i¥i>d7 5 (a) 1...'i¥i>h5?2 b4 'i¥i>xh4 3 b5 g5 4 �xa7 'i¥i>c7 6 'i¥i>a8 e5 7 g4 hxg4 8 h5 a5 g4 5 b6 axb6 6 axb6 g3 7 @f3! gxh5 9 b4 cxb4 10 c5 b3 11 a7 b2 +-· 12 cxb6 �xb6 13 �b8 bl='l'W 14 Cb) 1. ..c,t>g7!=, Uhlmann- a8='1'W 'l'Wh7!, 0-1, Mortensen-Plueg, Robatsch, Marienbad, 1965. Hamburg, 1997. 22 25 l...axb5? (after the correct 1...a5! Here White resigned, calculating a Black holds. On h3-h4 he replies variation with a necessary exchange ...h6-h5, while on g2-g4 forced is on f6 However, correct is not to ... g6-g5 !) 2 axb5 @d7 3 @f3�e7 4 take the pawn. h4 h5 5 'i¥i>e4 'it>d7 6 �d4 'i¥i>c7 7 1 �f5 'i¥i>xc4 2 h4! @b4 3 'i¥i>g6 'i¥i>c4 'i¥i>d7 8 �b4 'it>c7 (matters are 'i¥i>xa4 4 'it>xh5 'it>b3 5 'i¥i>g6, and the not changed by 8 ...�c8 9 c,t>a5 (9 pawns queen at the same time, while c6) 9 ...�c7 10 g3!) 9 c,t>a5 @c8 10 the f6 pawn 'helps' White, 'iitb6 e5 11 fxe5 fxe5 12 'i¥i>a5! 1-0, Klovan-Elistratov, Moscow, 1963. Speelman-Saltaev, Las Palmas, 1998. 26 23 (a) l...'i¥i>xb4 2 a6 'i¥i>c3 3 'iitel! l...�b5? (1...h5! 2 f3 [2 h3 'i¥i>b53 'i¥i>d3 4 a7 f2+ 5 @fl e2+ 6 'i¥i>xf2 g4 hxg4 4 hxg4 'it>c6 5 g5 fxg5 6 'i¥i>d2 7 a8=11Vel= �+ 8 'i¥i>g2 �xh4=, fxg5 lt>d6=] 2 ...@b5 3 g4 hxg4! 4 Papendieck-Geiger, Austria, 1998. fx g4 lt>c6 5 h4 'it>d6=) 2 g4 'i¥i>c6 3 (b) 1...'i¥i>c4! 2 a6 'i¥i>d3 3 @el e2 4 g5 e5+ 4 fxe5 fxg5 5 f3 h5 6 'i¥i>c4, a7 c,t>e3 -+ J 94 Solutions TRANSPOSITION TO A PAWN ENDING 1 5 No, 1 ... llle5?? After 1...b5? 2 J:ixb5 J::!xb5 3 I...llld6 2 g5 (2 h4 llle4) 2 ...fxg5 axbS+ 'it>xb5 4 e4 'it>c6 5 e5! fx e5 6 3 lllxg5 c4+ 4 bxc4 lllxc4 5 lllf3 g5 hxg5 7 f6 !, Black had to resign, llle3 6 g4 lllg2= Averbakh-Bebchuk, USSR, 1963. 2 lllxeS fxe5 2 ...@xe5 3 h4 \t>f4 4 h5 @g5 5 6 'it>c2 'it>h6 6 'it>b2 'it>g5 7 g3 @h6 8 a4 bxa3 9 'it>xa3 A 'it>a3-b2-c2-d3 No, it is not worth it. I i.g2? (1 b3-b4 +- i.xg6 'it>xg6 2 b5! @rs 3 'it>d4 +-) 3 gS e4+ 4 @e2! 'it>e5 l...�e5 2 i.f3b5=, Adams-Xie Jun, 4 ...d3 5 'it>dl!+ - France, 1994. s h4 'it>f5 6 g4+ 'it>f4 7 g6 d3+ 8 'it>dl e3 9 g7, 1-0, Ivanov 7 Sagalchik, Kramatorsk, 1989. No. 1 llld3?? (I 'it>f4=) 1.... llle5 2 2 lllxeS 'it>xe5 3 @gs b5 ! -+, Bronstein-Timman, Tallinn, 1973. No. 1 .l:.xfi+? @xii 2 �f3 @f6 3 @e4 gS! (3 ...@g5 4 �xd4 �g4 5 8 'it>e5 !=) 4 bS d3 S 'it>xd3 @es 6 'it>c4 g4 7 'i!ics 'it>e4 8 'it>c6 'it>d4!!, 0-1, (a) 1. ..i.c5 !, transposing to a Renet-Olafsson, Reykjavik, 1993. drawn rook ending; Instead I :c4 .l:Id7 2 @f2 d3 3 (b) I...'it>f8? 2 J::rxd6 J::rxd6 3 i.c5 @el d2+ 4 'it>dl @f6 5 .l:Ic5 .l:Id3 6 'it>e7 4 'it>e3 'it>e6 5 i.xd6 'it>xd6 6 .!la5 draws. �d4 'it>e6 7 'it>c5 h5 8 g5 @f5 9 'it>xd5 'it>g4 IO 'it>e5 'it>xh4 11 f5 , 3 1-0, Ricardi-Sorin, Buenos Aires, 1995. 1. .•l:l'.d5 2 .!la2 �g6 3 Zlf2 fS 4 J:id2 l:l'.xd2 S 'it>xd2 'it>g5 6 'it>e2'it>g4 9 7 @f2 'it>f4! 0-1, Balashov Tiviakov, St.Petersburg, 1993. No, he cannot. 1... l:l'.gxd7? 4 1...f4 ! 2 l:l'.xd7+ .!lxd7 3 .!lxd7+ �xd7 4 1 cS! .!lxc5 (I...bxc5 2 �d3 A f4 ! g4 :Ixa5+-) 2 J::rxc5 bxcS 3 'it>d3 eS 4 4 ...gxf 4 5 'it>f3 'it>e6 6 'it>xf4 \t>f6 @c4 fS 5 'it>xc5 h5 6 b4 axb4 7 7 g3 +- 'it>xb4 f4 8 a5 e4 9 @c3 e3 10 'it>d3, 5 g3 gxh3 6 gxh4 @e6 7 'it>g3 1-0, Kuzmin-Petrosian, USSR, 'it>f6 8 hS! 'it>g7 9 �xh3 �h7 10 1971. 'it>h4 @h6 11 b3 !, 1-0, Estrin Gusev, Moscow, 1963. Solutions 195 10 lS (a) 1...ll:\g5? 2 ..tf5 'it>f6 3 'it'e3 1 'it>f2? (1 it>g2! J:ic2+ 2 .!lf2 ll:\e6 4 i.xe6 'it'xe6 5 it>e4 +-, .!lxf2+ 3 'it'xf2 i.gS 4 'it'g3=) Petrosian-Larsen, San Antonio, l . . . .!lc2+2 .!le2 i.c5+3 i.e3 i.xe3+ 1972; 4 'it>xe3 J:Ixf3+ 5 'it'xf3 J:Ixe2 6 (b) 1...ctJeS 2 'it'e3 ll:\xg6=. it>xe2 it>g6 7 'it>d3 it>f5 8 'it>d4 a5, 0- 1, Almasi-Portisch, Budapest, 11 1994. 1...gS? (1.....te6!?) 2 ll:\xdS 'it>xdS 16 3 g4! (a typical breakthrough) 3 ... gxh4 4 gxhS 'it'e6 s it>g2 'it'rs 6 1... ..tf4! 2 ..te3 'it'gS 3l:!.xf4 J:ixf4 f4 !, Black resigned, Deak-Horvath, 4 i.xf4+ 'it>xf4 S b4 b6 -+, Hungary, 1994. Szelaig-Pinski, Poland, 1997. 12 17 There followed l.....tc4+?? 2 No, it is not possible. ll:\xc4+ 'ifxc4 3 'iVd3+, 1-0, Tal 1 ti:Jf6+? ll:\xf6 2 'iVxf8+ 'it>xf8 3 Suetin, Sochi, 1973. exf6 gS ! 4 f4 'it>g8 S fx gS 'it'h7 -+, Correct was 1 ...'iVf 4!=. Adorjan-Chemin,Budapest, 1995. 13 18 1 ...l:1dS? (1...'it>f6 2 .l:Ixb6+ it>e53 (a) l....l:i'.h5? 2 'it'b4 .!lxh2 3 c5±, l:!.c6 �el 4 b4 lixe3 5 b5 .!ld3=) 2 Cruz-Seirawan, Moscow, 1994; .l:IxdS 'it>xdS 3 'it>a3! (going round (b) 1... .l:IaS+! 2 'it>b3 .l:Ixa2 3 the flank) 3 ... 'it'cS 4'it>a4 Sg6 h4 hS 'it>xa2 it>d6 4 it>b3 'it'cSS'it'c3 eS 6 (5 ...g5 6 hxg5 hxg5 7 g4 +-) 6 g3 it>d3 rs 7 it>c3 e4 (7 ...h5 8 'it'b3 f4 9 'it>c6 7 b4! (7 it>b4? b5 8 'it>a5? 'it>c5 it>c3 f3 10 'it>d3 'it>b4 -+) 8 'it>b3 hS 9 it>a6 'it>b4 -+) 7 ... it>c7 8 'it>bS 9 it>c3 f4 ! 10 gxf4 e3! 11 fx e3 h4 'it>b7 9 'it>c4 'it>a6 10 it>c3! it>b7 11 12 rs it>d6!! (12... g3 ? 13 hxg3 hxg3 'it>d4 'it>c6 12 it>xe4 it>bS 13 it>d4 14 f6 it>d6 15 c5+!) 13 'it'b4 it>eS! it>xb4 14 e4, 1-0, Ryumin - 14 cS 'it>xfS lS c6 'it'e6 16 'it'cS g3 Ilyin-Zhenevsky, USSR (ch), 1931. 17 hxg3 h3! 18 it>b6 h2 19 c7 it>d7!, and the pawn queens with 14 check! 19 1....l:Ixg2+? (1... hxg2 -+) 2 l:1xg2? hxg2 3 'it'xg2 a5, 0-1, Valvo-Levitt, (a) l...ctJxe4? 2 fxe4 it>d6 3 e5+! Chicago, 1992; it>c5 4 'it'c7=, Mestel-Chekhov, Necessary was 2 'it'xh3! l:Ixg4 3 Tj entiste, 1975; 'it>xg4 a5 4 'it>f5! 'it>c6 5 e5 a4 6 e6 (b) 1 ...it>d6 2 it>xa7 it>cS 3 i.c6 a3 7 'it>g6!=. ctJfl -+. J 96 Solutions 20 g4 bS 7 rs gxfS 8 gxfS @d6 9 @xe4, Solozhenkin-Anchesi, (a) l...i.c4? 2 i.xc4 dxc4 3 e4 Reggio Emilia, 1998. @b5 4 f4 @c5 5 g4 @b5 6 e5! fxe5 7 f5 !! 'it>c5 8 g5 +-, Heinig-Liebert, 22 DDR, 1979; (b) 1...d4+! 2 exd4 �d6=. l...@e5? 2 i.xd6+! 'it>xd6 3 'itf4 cJi>d5 4 b5! axb5 5 a6 'it>c6 6 @xe4 21 b4 7 f4 , 1-0, Larsen-Uhlmann, Las Palmas, 1971. 1...lZJbS!=. 1 i.xb6! i.xb6 2 axb6 axb6 3 @f2 @e6 4 @e3 'it>dS S c4+ @cs 6 Solutions 197 ROOK ENDINGS 1 18 �b6 1:!.c5 19 1:!.h5 llc8 20 1:!.h6 +-, Lesiege-Ivanov, Bermuda, 1995. In the game followed l...l:ral? 2 J::rd5+ 'it>e6 (On 2 ...'it>c6 follows 3 s �g5+) 3 �d4! J:ifl 4 1::re4+ 'it>d7 5 �xa4, 1-0, Gurevich-Rechlis, Israel, (a) 1 Sb6?? laa4 2 Wh3 l::txg4 3 1989. 1:!.xf6 J:ih4+! ! , 0-1, Gulko-Gurevich, He should play 1. ..J::r f4 ! 2 �g7 Parsipanny, 1996; l:!.g4+ 3 'it>f6 l:1f4+ 4 'it>g6 'it>e6 s (b) 1 J::rc4 'it>h42 1:!.c6=. l::ta6+ 'it>e7 6 .l:!a7+ 'it>e6 7 'it>g7 J::rg4+ 8 �f8 �d6! 9 'it>e8 1:!.e4+ 10 6 �d8 J::rf4 11 l::ta6+ 'it>cS 12 rj;;e7 llxf7+! 13 �xf7 'it>b4 with a draw. 1 'it>c6! (1 c6? e5 2 fxe5 fxe5 3 Wb6 e4 4 .lld7+ �f6! 5 l:rd8 e3 6 2 bi.e8 l:rb1 + 7 rj;;a7 .l:Ial+ 8 �b7 J:ib 1 9 'it>c8 11el !=, Short-Topalov, Nov (a) 1 1:!.d6? �e3 2 J:ie6+ '\t>f2 3 gorod, 1996) I. . . eS 2 fx eS fx eS 3 J:id6 'it>e2 4 lie6+ 'it>dl 5 �g5 d2 6 �dS �f6 (3 ... e4 4 Itd4 +-) 4 c6 e4 \t>f4 I:!'.a2, 0-1, Topalov-Kasparov, S 11f2+ �e7 6 Ite2 .:.ct1+ 7 'it>xe4 Geneva, 1995; 'it>d6 8 J::rc2 +-. (b) 1 l:te8+! �f3 2 .l:If8+ We2 3 1:!.e8+ 'it>dl 4:i::!.a8 d2 S a7=. 7 Correct was 1. ..llc8! 2 c3 �f6 +. 3 In the game followed 1...:i::!.b8?. 2 .llb3 .:.xb3 After 2... 1:!.c8 3 c3 d4 4 (a) l...'it>f4? 2 h4 l::ta2+ 3 �g l ltb7+ 'it>f6 Black holds the rook l:1c2 4 h5 1:!.c5 5 h6 �h5 6h7 f5 7 ending without problems. 3 axb3 �a7, I/i-I/i, Leko-Akopian, Ubeda, �f6 4 �gl 'it>eS S 'it>f2 �e4 6 We2 1997; 'it>d4 7 'it>d2 aS_(7 ...�e4=) 8 �e2 eS (b) 1...a4 2 h4 l:1a2+ 3 'it>g3 a3 4 9 �d2 e4? (9 ...'it>c5=) 10 c3+ @cs hS J:ial S .:.a7 a2 6 'it>g2 'it>gS, and 11 'it>e3 'it>c6 12 'it>d4 'it>bS 13 h3 h6 the f- pawn advances. 14 h4 �c6 lS c4 1-0, Kupreichik Zheliandinov, Gome!, 1968. 4 8 (a) 1...�e7 2 l:te6+ 'it>f7! 3 :xd6 .l:td4 4 'it>g3 'it>e7 S l::te6+ 'it>f7 6 :!:!.es 1 �f6 'it>g8 (l...'it>e8 2 e6!+-) 2 Wf6 7 l:thS 'it>g6=; 1:!.g4+ 'it>f8 3 l::ta4 l:i'.d8? (3 ...'it>g8 4 (b) l...'it>e7 2 lie6+ 'it>d7? 3 �f6 l::ta8+ 'it>h7 5 �f8 lldl 6 l!txf7+ 'it>g8 �e7 4 1:!.f5 l!tc3 5 h4 llb3 6 h5 .l:Ia3 7 :a7 �fl+ 8 rj;;e6 l::tel !, with a 7 h6 l:ta8 8 h7 l:th8 9 �h5 'it>f6 10 theoretical draw) 4 l::ta7? ( 4 bi.h4! 'it>g3 'it>g7 11 �g4 �g6 12 1:!.hl 'it>f6 'it>g8 5 �e7 bi.d5 6 l:tg4+ �h7 7 13 'it>f4 'it>g6 14 'it>e4 'it>f6 15 �d4 'it>xf7 +-) 4 ... l:tb8? (4 ...�g8 5 .llxf7 'it>g7 (15... 'it>e7 16 'it>c4 �d7 17 1:!.dl 6 .l:!a7, again draws) S l::txf7+ �al! +-) 16 'it>c4.l:tc8+ 17 'it>b5 �h8 'it>g8 6 :i::!.g7+ �f8 7 e6 l::tb6 8 l:!:a7 J 98 Solutions l:tb8 9 lih7 'it>g8 10 .l:tg7+ 1-0, 1!z-1!z, Karpov-Yudasin, Madrid, Kirov-Kosic, lnformator 59. 1992; (b) 1 gxh5 1ih3 2 .l:!xa5 f5 ! 9 (2 ...I:!'.xh4 3 'it>d3 :Ixh5 4 cJi>c4 f5 5 'it>c5 fxe4 6 'it>c6 +-) 3 exf5 1ixh4 4 1 lib7!! (1 'it>g6 :g3+ D. 2 ...J:!.b3) 'it>e3 'it>e7! (±, Yudasin) 5 l:ta6! i..J:Ixh5 2 'it>g6 'it>rs 3 h6 l:tb1 4 :xh5 6 'it>e4+-. l:!.b8+ 1-0, Levenfish-Lisitsin, Moscow, 1936. 14 10 1...l:td2? (1...l:!b2!=) 2 b4! l:td4 (2 ...I:!'.b2 3 l:te4 +-) 3 l:tb3 'it>e7 4 b5 1 f5 !!-a typical breakthrough 'it>d8 5 b6 'it>c8 6 �c3+ 'it>b7 7 �c7 -1. .. exf5 2 e6! fxe6 3 'it>xg6 D. 'it>xb6 8 .l:!xf7 1id3+ 9 'it>f2 l:!d4 10 h5=. 'it>f3 l:td3+ 11 'it>e4 l:!g3 12 f6 The game continued 1 'it>f6? 'it>b5 lixg4+ 13 'it>f5 .Ugl, 1-0, Timman 2 l:!.al a4 3 f5 exf5 4 e6 fx e6 5 Van Wely, Amsterdam, 1995. 'it>xg6 f4 6 h5 f3 7 h6 e5! 8 l:tel (8 @f5 !Ih7 9 l:!.hl f2 -+) 8 ... a3 9 15 l:txe5+ 'it>c4 10 l:tel a2 11 h7 J:!.a8 (1 l...J:ixh7? 12 'it>xh7 'it>b3 13 J:!.fl = (a) 1 g4? l:tdl ! 2 l:!c8:!.b l 3 I:!'.f8 or 11...£2 12 :!:!'.fl Wd3 13 :!:!'.al!) 12 l:txb4 4 .l:txf7 l:tc4 5 .l:!xh7 b4 6 .llhl 'it>g7 (12 :!.al f2 13 J:Ixa2 1ixa2 14 b3 7 .l:.fl b2 8 l:tbl .l:!b4 9 f5 cj;c7, h8='fl' .l:ta6! -+) 12... f2 13 Ital 1!z-1h, Kasparov-Short, Novgorod, 'it>b3 14 :!:!'.fl al='ili' 15 l:!xal J:!.xal 1994; 16 h8='ili' J:igl+ 0-1, Lasker-Leven (b) 1 l:tc8! f6 2 gxf6 :xf6 3 g4 fish, Moscow, 1935. J:ic64 :hs+ 'it>a6 5 f5 h5 9 l:!g8!+- . 11 16 1 'it>b6? (1 l:te3! l:!.e8 2 'it>c6 +-) (a) 1 l:td5+? f5 !! 2 J:Ixf5+ 'it>h6 1...e4 2 b5 J:ie8 3 Wa7 e3 4 l:thl e2 -+· 5 .l:tel Wxh6 6 b6 Wg5 7 b7 'it>g4!, (h) 1 I:!'.dla2 2 l:tal:a3+ 3 'it>g2! I/z-1h, Bologan-Kramnik, Germany, l:ta4 (3 ...'it>xg4 4 h6=) 4 'it>f3=, 1994. Permiakov-Petkevich, Riga, 1988. 12 17 1 h7? (1 l:!.e8 l:!.f72 'it>c3 'it>f23 (a) 1 'it>d4?b4 2 'it>d3 'it>b5 3 'it>d2 'it>d3 l:.d7+ 4 'it>c2 e2 5 l:tf8+ 'it>el 6 b3 4 :!:!'.cl l:.c4! -+, Piampuu- l:tf6=; 1 l:tf8! ?) 1...1if7! ' 0-1, Lancava, Leningrad, 1992; Lautier-Y ermolinsky, Tilburg, (b) 1 :!:i.a2! b4 2 :c2=. 1993. 13 18 (a) 1 g5? fxg5 2 hxg5 h4 3 d6 (a) 1 d6? 'it>g7! 2 b5 @f6 3 d7+ l:ta2+! 4 @f3 h3 5 'it>g3 h2 6 �a8+ 'it>e7 4 I:!'.d6 'it>d8 5 I:!'.f6 'it>xd7 6 'it>d7 7 .!:l.h8 'it>xd6 8 l:txh2 a4 ... J:ixf7 'it>.e6 7 I:!'.xh7 :b3 8 �b7 l:tb2! Solutions 199 9 b6 e4 10 li:tb8 e3 11 @fl �f6 12 23 h4 Wg7 13 g4 'it>h7 14 h5 gxh5 15 gxh5 Wg7 16 .l:Ib7+'it>h8 17 h6 'iitg8 1 'it>c4 l:ta4+ 2 Wb3 .l:Ia3+ 3 'it>c2 18 .l:Ig7+'it>h8 19 b7 �b l+! 20 'it>e2 .l:Ic3 4Wb2! +-, Salwe . .l:Ib2! 21 Wxe3 .!lb3+ 22 Wd4 1Ib4+=, Yudasin-Kramnik, Candi 24 dates (m), 1994; (b) 1 .!lc8! 'it>g72 b6 .l:!b33 d6 +-. (a) l...'it>e3?2 Wg3 l:!al 3 .l:Ie8! (3 a7? l:tgl+ 4 Wh2 'it>f2=) 3 ... 'it>d4 4 19 'it>f4 l:!xa6 5 l:!.d8+! @c5 6 @g5, 1-0, Ulibin-Nevednichi, USSR, 1986; (a) 1....!lxh4? 2 .l:Ixg5 Wc3 3 (b) 1.. . �xf3!! 2 a7 .l:!a3!!=, .!id5!! Wd3 4 'it>cl .l:Ihl+ 5 'it>b2 l:!el Gelfand. (5 ...We3 6 @c2 .l:Ih2+ 7 Wd l ! d3 8 .lle5=) 6 .l:Id8, Draw, Dreev 2S Beliavsky, Odessa, 1989; (b) 1... .!lgl+ 2 �e2 (2 @d2 l:!g2+ (a) l...h3? 2 �g3 +-, Bykova 3 We 1 g4 4 �fl .l:!h2! 5 l:!xg4 'it>c3 Rubzova, m, 1958; 6 'it>g 1 l:!a2 D. d4-d3 -+) 2 ••.d3+ 3 (b) 1... 'it>h l! 2 Wg4 h3 3 Wxh3 Wd2 .!lg2+ 4 Wdl g4 s hS Wc3 6 .l:Ixg7!, stalemate. J::!c8+ 'it>d4 7 h6 .l:Ih2 We3 9 l:!e8+ 'it>f3 10 lih8 'it>g3 11 h7 .l:!h4!, and 26 there is no defence against the march of the g4 pawn. (a) 1 a8='ii'xa8 .ll 2 l:l'.xa8 Wf5 3 .l:Ih8 �g4 4 �c5 h3 5 'it>d4 �g3 6 20 We3 Wg2!=, Dammes-Sosonko, Leningrad, 1963; 1 �f6! z:!'.c6+ 2 @es z:!'.c8 (2 ....l:!c5 (b) 1 'it>b7! J::!ti+ 2 @a6! l:!f6+ 3 �d6 l:!c84 l:!el+-) 3 .l:!g6!! Wh7 4 (2 ... .S.f8 3 .l:.b5! h3 4 l:Ia3! +-) 3 .l:Ic6! l:Ia8 S Wf6 t:>. lie6, Cvitan. @bS li:tf8 4 as=iw .l:Ixa8 s .!lxa8 'it>fS 6�c4 h3 7 �d3 +-. 21 27 1 g6+ 'it>f6 2 l:!f8'it>eS 3 f6 !! (3 .!1£7? .llal=) 3 ....l:!xf 6 4 l:!ti! 'it>e6 (a) 1 'it>f4? 'it>c4 2 @e5 'it>d3! 3 ( 4 ...l:!f5 + 5 'it>g4 l:If6 6 'it>g5 +-) s Wf6 'it>e4 4 Wg7 l:!.a8 5 h7 Wf5 6 g6 l:txg7 :n 6 1Ia7 +-, Bayer-Polasek, Wg5=, Gutman-Tseitlin, Riga, 1976; Luxemburg, 1986. (b) 1 Wg4 Wc4 2 'it>hS 'it>ds3 g6 +-. 22 28 1...z:!'.aS! (1...Wg7? 2 a5 'iitf6 3 'it>f4 We6 4 @e4! h5 5 �f4! @f6 6 (a) l...Wg2? 2 li:tf6 f3 3 g6 .!la54 Wg3 @g5 7 .!la4+-, Holmov-Hasin, :i.£7, r;H/i, Morovic-Agdestein, Minsk, 1962) 2 �f4 @g7 3 We4 hS Havana, 1998; 4 'it>d4 h4 S 'it>c4 h3 6 �b4 l:!'.hS 7 (b) 1... .!la l+ 2 'it>d2 l:!.gl 3 g6 .:!.al h2 8 .l:Ihl Wf6 9 aS @rs 10 a6 l:!g3!! D. 4 ...�g2 -+ . .l:Ih7!=. 200 Solutions 29 (b) 1... �dl+! 2 �e6 :!al 3 l:id8 J:Ixa74 l:td7+l:ixd7 5 @xd7 @h7=. (a) l...f5? 2 h4 ! a6 3 e3 e4+ (3 ... a5 4 e4 f4 5 gxf4 exf4 6 e5) 4 34 fxe4 fxe4+ 5 c;t;>xe4 :Ixc4+ 6 l:txc4 c;t;>xc4 7 @e5=, Wirthensohn (a) 1 �xa4? l:!e3+ 2 c;t;>d8 f5 ! 3 Mikhalchishin, Lenk, 1998; gxf5 'it>f6 4 l:ta6 @xf5 5 @c7 g4 6 (b) 1... h4! 2 gxh4 f5 3 e3 a6 4 h5 d7 l:te7 7 @d6 �xd7+ 8 Wxd7 g3=, (4 �b3 J:Ixc4 5 �xb6 e4+!) 4 ... gxh5 Olafsson-Tal, Portoroz, 1958; 5 h4 a5 -+. (b) 1 d7 J:Ie3+ 2 @d8 a3 3 I:!'.a8! +-. 30 35 (a) 1 h6? @d3 2 @£3 l:th3+ 3 @g4 (a) 1 @gl? J:ig3 2 c;t;>f2 c;t;>e6 3 :ga5 l:rhl 4 @f5 e2 5 �f6 l:th3! 6 c;t;>f7 J:Ixg4 4 l:!.a4 @f5 5 .!Ib4 l:!e4 6 l:if3+ 7 @g6 �e3=, Piket-Sokolov, :gb5+ c;t;>f4 7 l:rb6 f5 8 l:id6 g4 9 Dortmund, 1995; J:id8 g3+ 10 @fl h3, 0- 1, Ioseliani (b) 1 �g4! @d3 2 g6 e2 3 @g5 Onischuk, Lucerne, 1997; el='Yi' 4llxel J:Ixel 5 g7 +-. (b) 1 lif5! l:tg32 d5 J:Ixg4 3 d6+ 'it>e6 4 lid5! @d7 5 .llf5 �f4 6 .llxf4 31 gxf4 7 �h3 c;t;>xd6 8 Wxh4 @e5 9 c;t;>h5!=. (a) 1...�b3? 2 J:ig7+! Wh8 3 l:!b7 l:rxb4 4 @g6, 1-0, Salov-Yudasin, 36 St.Petersburg, 1997; (b) 1...I:!'.f3+! 2 c;t>xe5 I:!'.b3 3 c;t;>d6 (a) l...We8? 2 g4 Wd8 3 @g6 @e8 1ixb4 4 e5 l:td4+ 5 @e7 l:ta4!! 6 4 c;t;>f6 Wd8 5 �g7 l:!b5 6 l:l'.xg5 J:ib7 @g7 7 e6 lia8 8 'iitd6 c;t;>f8 9 c;t;>xg5 a5 8 @h6 a4 9 g5 a3 10 g6 a2 @d7 @g7=. 11 g7 al=ii 12 g8=�+ '\t>d7 13 'iVg7+, 1-0, Lein-Suetin, Bad Wild 32 bad, 1997; (b) 1...l:!a3!! 2 �xg5 (2 g4 l:!xh3 Only 1 �a3!! makes a draw. 3 �xg5 �g3 4 @f6 @g8 5 g5 :£3+ (a) 1...gxh4 2 l:!c3+ c;t>d5 3 J:ic2 6 Wg6 �f8!=) 2 ... J:Ig3+ 3 @f6 �g8! �e4 4 @h5 c;t;>f3 5 llc3+ c;t;>e2 6 4 l:!xa6:gxg2=. l:!c2+ �d3 7 �h2=; (b) 1...g4 (L.c;t;>d5 2 lia5+ @e4 3 37 �a4+) 2 J:Ic3+ (2 c;t;>g5? �c5!) 2 ... c;t;>d5 3 :gc2 c;t;>e4 4 @g5 Wf35 1 llc2! �e6 2 :c3!! a3 3 f7 l:l'.xf7 J:ic3+ @g2 6 c;t;>xg4 l:Igl 7 l:.a3 4 l:te3+ 'iitd5 5 Wxf7 a2 6 I:(a3 1-0, al='iV 8 �xal l:.xal 9 h5=. Konstantinopolsky-Fridman, Lvov, 1940. 33 38 (a) 1....llal? 2 I!c8 �xa7 3 l:!c7+! (a) 1...1:1'.bl? 2 �h5 ligl 3 g5 fxg5 J:Ixc7 4 �xc7 c;t;>h7 5 Wd7, 1-0, 4 f5 ! c;t>f8 5 f6, 1-0, Zaitsev Benko-Gereben, Budapest, 1951; Hiibner, Busum, 1969; Solutions 201 (b) i. . . �b4 ! 2 rs .!lbl 3 'it>hs .!lgl (b)1 .!lh4+! Wxb3 2 aS +-. 4 'it>g6 .!lxg4+ S 'it>xr6 .!lgl=, Larsen. 42 39 (a) 1....!lb7? 2 .!lg4! We8 (2... b3 (a) l...'it>h4 2 .l:!c8 .l:!h7+ 3 lt>e6 l::ta4 .!lb8 4 .!lf4 We8 5 d7+ +-) .l:!xd7 4 'it>xd7 g4 5 'it>e6 g3 6 'it>f5 .l:!c4 l:tb8 4 .l:!c7 b3 5 .!lxg7 lt>f8 g2 7 'it>f4 +-; .!if?+ 'it>g8 7 d7 b2 8 :!.fl Wg7 (b) 1...'it>g4 2 'it>r6 :i.f8+ 3 c;t>e6 .!lbI! l:.b6+ 10 c;t>e7 .!lb7 11 'it>e! .!id8! 4 l:rdS (4 .l:.c8 .l:!xd7 5 lt>xd7 1-0, Gelfand-Lerner,Norilsk, 1986 \t>f5! 6 .!lc5+ lt>f4 6 lt>e6g4 7 .!lc4+ (b) 1...�f2! 2 .l:!xg7 .l:!e2+ 3 lt>d 'it>g5 8 lt>e5 g3=) 4 ...�r4 ! S .!lrs+ b3 4 I:i:b7 b2 S g4 'it>c8 6 J:ib 'it>g4 6 .!in 'it>h3 7 �rs g4 s �r4 g3 lt>d7=. 8 lt>f3 lt>h4, 11z_11z, Keres-Mikenas, Stockholm, 1937. 43 40 It seems that White has succes' fully completed a difficult defenc( 1...d3!! 2 cxd3 (2 'it>xe4 dxc2) and Black must agree a draw afte 2 ... �c4!! 3 bxc4 c2 4 'it>xf4 cl=i#+ l...a3 2 Wg6 .!lg l+ 3 Wh6 Ji(hl S lt>e4 °i'dl!, 0-1, Munios-Salazar, with a perpetual check or l...�fl Dubai, 1986. .!lxb7 l:i.f8 3 .l:!a7. But, as it turn out, there is a third possibilit) 41 There followed the disheartenin 1.. . .!lb6!!, and, playing on throug (a) 1 lt>f2? b5 2 axb5 (2 a5 inertia, 2 lt>g6 cS+ 3 'it>rs cxd4 'it>a3! !=) 2 ... axb5 3 'it>e2 lt>a3! ! 4 :!::txdS dxe3, White acknowledge Wd2 b4 5 Wc2 .!lc8+! 6 lt>d2 J:ih8, defeat, Lopyshnoi-Dreev, Maikor Draw, Levy-Peev, Cienfuegos, 1998. 1973. 202 Solutions VARIOUS 1 5 (a) 1 'i£tb3? ctJb8!! 2 ctJd6+ (2 (a) I....te6? 2 'it>e4! ;:,,, 3 ctJf5 +-, Cbxc5 ctJc6 3 ctJd3 ctJd4+ =) 2 ...'i£te6 Vuki6-Vujosevic, Vmjacka Banja, 3 ctJe4 Cbc6 4 gxf4 Cbd4+ 5 'i£tc3 1998; ctJxf3 6 ctJxc5 �f5 7 fxe5 Cbxe5 8 (b) 1... .tbl!=. ctJd3 Cbxc4! 9 'i¥txc4 �g4=, Sale-Mikhalchishin, Sibenik, 1990; 6 (b) 1 'i¥td2! ctJb6 2 @d3 ctJd7 3 'i¥tc3 @e6 4 'i¥tb3 ctJb8 5 gxf4! Cbc6 1 Cbxa5 'i¥tb2! 2 'i¥tc4! (2 b4 6 Cbxc5 @rs 7 Cbd3 Cbd4+ 8 'it>c3 @a3!=) 2 ....txa5 3 b4 .tb6!! 4 a5 Cbxf3 9 fxe5 ctJxh2 10 c5 +-. .tf2 5 a6 (5 b5 'it>a3=) 5 ... .ta7 6 'i£tb5 'i¥tc3=, Parma-Gligoric, Bled, 2 1961. 7 (a) 1... .txg2! 2 axb5 'i¥te6 3 Cbxf7 @xf74 'i¥txg2'i¥te6 5 @f3 'it>d5 6 b6 1 ... g5+! 2 hxg5 h4 3 ctJc6 Cbg6+4 @c6 7 b7 'i¥txb78 'it>e4 g4 - ; + Markowski-Gdanski, (b) I....td3? 2 axb5 �e7 3 Cbc8+ 'i¥te4 h3 -+ Ksiaz, 1998. 'i¥td7=, Van der Wiel - Larsen, Brussels, 1987. 8 3 He should choose 1...f6! 2 'i¥f8+ 'i¥th53 'ii'g7 h6 -+. 1 g4!! (1 ctJf2 'i£tf6 2 �f4 'i¥tg7 3 In the game White played l...f5? g4 hx_g4 4 Cbxg4 Cbxc4!) 1...Cbxg4 2 2 'l'We7 ! �d2+ 3 'i¥tg3 �el+ 4 'i¥th3 f4 tt:lf6 (2 ...@f6 3 e5+!! Cbxe5 4 'iWhI+ 5 �g3 'i!Vg l+ 6 'i¥th3 f4 7 Cbxe5 dxe5 5 'i£te4, and the pawn 'ii'f8+ 'it>h5 8 'ii'xf4, tfi -0., ending is won) 3 e5+ dxe5 4 tt:lxe5 Mikhalchishin-Kavcic, Slovenia, �d6 5 Cbxg6 Cbxd5+ 6 cxd5 c4 7 1997. �e4 cxb3 8 ctJe5, 1-0, Razuvaev Ostoji6, Berlin, 1988. 9 (a) l...�c2? 2 'ii'xe6 �xf2+ 3 4 'i¥th3'l'Wfl + 4 'i¥th4 'iVbI+ 5 'i1Yh3! +-, Polugaevsky-Bronstein, Tallinn, (a) I .te3? 'it>c7 2 'i¥tc5 g2 3 'it>d4 1964; 'i¥tc6 4 'i¥te4 'it>b5! 5 @f3 'i¥tc4 6 (b) 'i¥txg2 'it>d5 7 .tf4 'i¥te4 ! intending 1... 'i!Vf 5!=. ...'it>f5 , f6=, Kudrin-Cebalo, Berne, 10 1988; (b) 1 .ta5!! g2 2 .tb6 'i¥tb83 'i¥tc6 1... 'ii' xe5? (l...fxe5 2 'l'Wxh5 c;i;>c8 4 c;i;>d6+-. 'ii'xa5=) 2 'i!Vg3+! @f7 (2 ...'ii' xg3+ 3 Solutions 203 'it>xg3 +-) 3 'fl'xe5 fx e5 4 i;t>f3@e7 5 i.c5 �xh5 7 'it>b6 'it>g4 8 'it>xa6 @e4 'it>e6 6 f3@f6 7 f4 +-. Wf3 9 'it>xb5 'it>xe3 10'it'b6 ct:Jxb4!, 11z_11z, Chiburdanidze-Maric, Bel 11 grade, 1996. But better is 1 i.c2 ii.as 1...'it>c8? 2 'fl'g4 and 3 'fl'd l=, (I...i;t>xh62 i.dl j_c6 3 i.g5+ i;t>h7 Ehlvest-Topalov, Novgorod, 1994 4 h6 A 5 j_h5 +-) 2 e4!! dxe4 1 ... 'it>c7! 2 'fl'e7+ i;t>c6 3 'i!Ve8+ (2 ...fxe4 3 i.dl'it>xh6 4 i.g4j_c6 5 i;t>cs 4 iif8+ 'it>d4 -+. i.g5+ 'it>h7 6 i.xe6 +-) 3 i.b3 i.d5 4 i.xd5 exd5 5 i.d6! (5 j_g5? 12 ctJb8! 6 'it>e5 ctJc6+7 'it>xd5 ctJxb4+ 8 �e6 ct:Jc6 9 d5 ct:Jxa5=) 5 ...Wxh6 1 h7! �xh7+ 2 Wg6 :i.h4 3 rs+ (5 ...ctJf 6 6 'it>xf5 ctJxh5 7 'it>g5 +-) 6 �d6 4 ctJh6 �hl 5 ct:Jf7+! 'it>c7 6 i;t>xf5 c;t>xh5 7 'it>e6 e3 8 i.f4! e2 9 CZJg5 @d6 7 f7 l:!.h8 8 @g7 lla8 9 i.d2 lbf8+ 10 i;t>xd5 i;t>g4 11 i;t>c6 f6 ! (9 f8=11Y+? l:txf8 10 'it>xf8 Wf3 12 d5 'it>rz 13 i;t>b7! +-. 'it>e5=) 9 ... @e5 10 lLih7 I:!'.a7 11 c;t>g6, 1-0, Krumpachnik-Maksimen 15 ko, Ptuj, 1998. (a) l...e5? 2 l:txc6 exd4 3 exd4 13 �d3+ 4 @f4 �dl 5 �e4 �g6 6 d5 :!:!'.el+ 7 j_e3 'it>g7 8 d6 I!dl 9 i.f4, 1 l:!.xe5 �xe5 2 g3, Black re� and at last he resigned. Osterman signed, But he could have saved Mikhalchishin, Bled, 1995; himself by 2 ... i;t>g6! 3 j_xe5 @h5 (b) 1...e6! 2 �xc6 i.d8! with the and after4 ...f4 -draw. idea i.d8-a5-el with a sufficient defence. 14 16 The game continued 1 i.xf5+? exf5 2 'it>xf5 @xh6 3 'it>e6 lLib8 4 No, because of 2 ...i.a 4!! 3 �e2 (3 'it>d6 ctJc6 5 i;t>c7 i.a8!Black's only h5 i.xb3 4 h6 i.xc2 5 h7 �b3+ 6 chance, which White underesti @e2 g2 -+) 3 ... i.xb3 4 i.d3 mated. Losing is 5 ...ct:Jxe7 6 �xb7 �xe2+ 5 Wxe2 i.dl+!!, 0-1, i;t>xh5 7 'it>xa6 lbc6 8 'it>xb5 +-. 6 Timman-Salov, Amsterdam, 1991. 204 Solutions CHOOSE THE BEST CONTINUATION IN ROOK ENDINGS 1 Alatortsev-Chekhover, USSR, (b) !...'it>h5? 2 .llxf3 l:1h2 3 @gl 1937: l:.txh3 4 Wg2 +-. (a) I Wb5? :Ixa72 .!lxa7 'it>xf2=; (c) I...@g5!! 2 �xf3 .!lh2 3 h4+ (b) 1 Wd5? Wfl 2 'it>c6 .!lxa7! 3 'itg4=. l:.txa7 Wxf2=; (c) 1 @c5! J:Ic8+ (1..Jha7 2 3 Cuartas-Zuidema, Skopje, l:.txa7 'it>xf2 3 @d4 +-) 2 @b6 .U.e8 3 1972: @c6! (3 'it>b7 J::re7+ =) 3 ...@fl (a) l...f3 ? 2 c7 f2 3 J:id2 D. 4 (3 ....!lh8 4 @b7 J:ih7+ 5 @b6 .!lh8 6 l::!.b2+-. .!lc2 +-) 4 @b7 J:ie7+ 5 @b6 .!le8 6 (b) 1....l:Ie1? 2 c7 f3 3 .l:td8+'3;; e7 J:ic2! @g2 7 @b7 .!le7+ 8 @b8! 4 'it>b7+-. J:ie8+9 .!lc8. (c) 1....l:Icl! 2 c7 l:.tc4!! 3 J:id5 'it>e7 4 Wb7 @e6 5 .l:Ib5 @d7 6 2 Karner-Renter, USSR, 1955: �d5+ We6=, Karpov. (a) 1...l:th2? 2 h4 'it>h5 (2... f2 3 l:.tf3+-) 3 .l:Ia5+! @g4 4 .!lg5+ @h3 5 h5 �a2 6 h6 .!la6 7 h7, 1-0 Index of Players and Composers Abramovic-Nikolic 85 Beliavsky-Short I 74 Adams-Almasi 148 Beliavsky-Spraggett 59 Adams-Lautier I 7 Beliavsky-Sveshnikov I 7 Adams-Lutz I I Beliavsky-Tratar 54 Adams-Xie Jun 3 I Benko-Gereben 106 Adorjan-Chemin 33 Bogoljubow-Rubinstein 93 Adorj an-Morozevich 176 Bojkovic-Kakhiani 80 Akopian-Almasi 90 Bologan-Kramnik 103 Alatortsev-Chekhover 97 Botvinnik 25 Alekhine-Bogoljubow 1 IO Botvinnik-Balashov I 24 Alexandria-Marie 53 Botvinnik-Donner I 84 Almasi-Portisch 32 Botvinnik-Kan I 13 Alterman-Chemin 142 Botvinnik-Rabinovich I I 2 Anastasian-Khalifman 29 Botvinnik-Simagin 4 I Anastasian-Romanishin I 9 Branicki-Sefc 96 Andersson-Hilbner 68 Bronstein-Rantanen I I 8 Arbakov-Gurevich 76 Bronstein-Timman 31 Averbakh-Bebchuk 31 Brunner-Korchnoi 97 Azmaiparashvili-Kupreichik 8 I Bykova-Rubzova l 05 Azmaiparashvili-Novopashin 44 Capablanca-Tartakower 76 Bagirov-Berzinsh 60 Chaunin-Friedman 72 Bagirov-Kraidman 57 Chekhov-Karsa 129 Bagirov-Veingold 64 Chemin-Georgiev 176 Balashov-Korchnoi I 19 Chiburdanidze-Galliamova 92 Balashov-Tiviakov 30 Chiburdanidze-Maric 190 Balashov-Ulibin 94 Cruz-Seirawan 19 Balashov-Vaganian 145 Cruz-Seirawan 33 Ballon-Mikhalchishin I 84 Cuartas-Zuidema 97 Bareev-Farag6 139 Cvitan 104 Barle-Mikhalchishin 89 Dammes-Sosonko I 05 Barlov-Abramovic 45 Dan-Pytel 168 Barlov-Schiissler 78 Dao Thien Hai-lvanchuk 120 Barlov-Seirawan 128 Dautov-Alterman 65 Bayer-Polasek I 04 Deak-Horvath 32 Beliavsky-Azmaiparashvili 95 Degraeve-Hansen 26 Beliavsky-Dolmatov 163 Dokhoian-Shirov I 77 Beliavsky-Gelfand I 3 I Drasko-Vratonjic I 3 Beliavsky-Geller I 70 Dreev-Anastasian 26 Beliavsky-Hodgson 6 I Dreev-Beliavsky I 04 Beliavsky-Kupreichik I 74 Ehlvest-Polugaevsky 65 Beliavsky-Mikhalchishin I 32 Ehlvest-Rausis I 36 Beliavsky-Miles I 54 Ehlvest-Shirov I 0 Beliavsky-Neverov 143 Ehlvest-Topalov 190 206 Index Eliskases-Levenfish 81 Hiibner-Polgar 89 Emma-Riemersma 56 Hiibner-Spassky 116 Estrin-Gusev 31 Hulak-Beliavsky 173 Euwe-Alekhine 59 Huzman-Mikhalchishin 61 Evreinov 26 Ilincic-Abramovic 52 Farag6-Csom 172 Illivitsky-Taimanov 84 Fedotov-Arkhipov 34 Ilyin-Zhenevsky-Botvinnik 27 Fercec-Cvitan 13 I Ioseliani-Onischuk 107 Fercec-Mikhalchishin 83 I vanchuk-Eingom SI Filipescu-Citron 27 I vanchuk-Kasimdzhanov 20 Filipov-Kopatsny 90 lvanchuk-Kramnik 164 Finkel-Mikhalchishin 20 lvanchuk-Lautier 88 Fischer-Bisguier 23 Ivanov-Sagalchik 30 Fischer-Geller 94 Janvarev-Schcherbakov 25 Fischer-Letelier 21 Kamsky-Karpov 59 Fischer-Larsen 23 Kamsky-Cvitan 144 Flohr-Vidmar 18 Kan-Keres 115 Formanek-Mikhalchishin 170 Kamer-Renter 97 Gausel-Agdestein 142 Karpov-Hort 145 Gelfand 105 Karpov-Kasparov 12 Gelfand-Lautier 131 Karpov-Kasparov 157 Gelfand-Lerner 108 Karpov-Timman 185 Gelfand-Shirov 135 Karpov-Yudasin 103 Genba-lrzhanov 143 Karpov-Yusupov 117 Georgiev-Khalifman 114 Kasparov-Andersson 127 Gligoric-Fischer 22 Kasparov-Short 103 Godena-Lalic 44 Keller-Mikenas 66 Greenfeld-Golod ll Keres-Aronin 183 Gretarson-Magerramov !33 Keres-Mikenas 107 Grunberg-Brunner 85 Keres-Szab6 166 Gufeld-Grigorian 39 Kirov-Kasie 102 Gufeld-Minev 183 Klovan-Elistratov 29 Guliev-Tukmakov 25 Kochiev-Lemer 43 Gulko-Gurevich 102 Konopka-Shcherbakov 67 Gulko-Sveshnikov 144 Konstantinopolsky-Fridman 107 Gurevich-Andersson 134 Korchnoi-Kengis 99 Gurevich-Rechlis 101 Korchnoi-Ljubojevic 15! Guseinov-Beliavsky 70 Korchnoi-Pinter 125 Gutrnan-Tseitlin 105 Korolev-Palm 185 Havsky-Yuferov 27 Kovacevic-Tosic 114 Hector-Speelman 9 Kozakov-Asparuhov 27 Heinig-Liebert 33 Kozlov-Mikhalchishin 77 Hellers-Eingom 12 Kozul-Mikhalchishin 87 Herrera-Vasquez 57 Kozul-Nikolic 68 Hertneck-Narciso 151 Kramnik-Beliavsky 79 Hoch 28 Kramnik-Lautier 17 Hoi-Mikhalchishin 182 Krasenkov-Iskusnik 87 Holmov-Hasin 104 Kremenetsky-Razuvaev 37 Holmov-Timoschenko 74 Krumpachnik-Maksimenko 190 Horvath-Sherzer 177 Krumpachnik-Polak 8 Hiibner-Ftienik 127 Kudrin-Cebalo 188 Index 207 Kupreichik-Mikhalchishin 14 Miles-Kindermann 167 Kupreichik-Zheliandinov I 02 Miles-Van der Sterren 160 Kuzmin-Petrosian 19 Milov-Pelletier 142 Kuzmin-Petrosian 30 Mitrofanov 110 Larsen 107 Mokry-Pribyl 73 Larsen-Browne 76 Morovic-Agdestein 105 Larsen-Uhlmann 33 Morovic-Yusupov 154 Lastin-Cvitan 144 Morozevich-Balashov 129 Lasker-Levenfish 102 Mortensen-Plueg 29 Lasker-Marshall 180 Munios-Salazar 107 Lau tier-Y errnol insky 103 Natapov-Schuravlov 8 Lein-Suetin 107 Neverov-Dreev 124 Lek6-Akopian 101 Nezhmetdinov-Luik 123 Lerner-Dorfman 67 Nezhmetdinov-Romanovsky 121 Lesiege-Ivanov 101 Nikolaevsky-Gufeld 43 Levenfish-Lisitsin 102 Nikolic-Ftacnik 63 Levy-Peev 108 Niko!ic-Movsesian 83 Lilienthal-Smyslov 77 Nikolic-Portisch 138 Lj ubojevic-Ivanchuk 116 Nimzowitsch-Chigorin 27 Lombardy-Fischer 23 Novikov-Beliavsky 64 Lopyshnoi-Dreev 108 Novikov-Kaidanov 162 Lputian-Tukmakov 122 Novikov-Lalic 58 Macieja-Grabarczak 8 Olafsson-Tal 106 Madsen-Hansen 60 Oil-Benjamin 18 Magerramov-Kohlmeyer 139 Ostenstad-Kuzmin 11 Manukovsky-Pustovalov 109 Osterman-Mikhalchishin 190 Marie-McNab 86 Owen-Morphy 78 Maric-Zaitseva 52 Padevsky-Latinov 26 Markovic-Ivanovic 9 Panno-Donner 137 Markowski-Gdanski 189 Papendieck-Geiger 29 Marosi-Yudasin 186 Parma-Gligoric 189 Marshall-Alekhine 183 Pelletier-Arencibia 147 Martinovic-Mikhalchishin 179 Pelletier-Rozentalis 82 Martinovic-Yudasin 114 Pelling-Miles 137 Matlak-Tseshkovsky 13 Perrniakov-Petkevich 103 Matu!ovic-Uitumen 49 Petrosian-Larsen 31 Mednis-Gurevich 71 Petrosian-Rashkovsky 175 Mestel-Chekhov 33 Petrosian-Tal 169 Mikhailov-Volchok 141 Piarnpuu-Lancava 104 Mikhalchishin-Azmaiparashvili 110 Piket-Sokolov 106 Mikhalohishin-Beliavsky 119 Plaskett-Rowson I 0 Mikhalchishin-Cvetkovic 187 Polgar-Smyslov 27 Mikhalchishin-Eslon 66 Polnareva-Akhsharumova 45 Mikhalchishin-Gufeld 180 Polugaevsky-Bronstein 189 Mikhalchishin-Kavcic 189 Polugaevsky-Parma 73 Mikhalchishin-Khme!nitsky 71 Ponomariov-Plaskett 169 Mikhalchishin-Nestorovich 182 Portisch-Kramnik 118 Mikha!chishin-Stangel 88 Portisch-Pietzsch 92 Mikhalchishin-Savchenko 151 Portisch-Ribli 112 Mikhalchiishin-Sveshnikov 150 Psakhis-Bonsch 115 Miles-Adorjan 165 Rausis-Farag6 137 208 Index Razuvaev-Kirov 155 Tairnanov-Chekhov 78 Razuvaev-Ostojic 188 Tairnanov-Gligoric 181 Renet- Olafsson 30 Tal-Balashov 146 Reti- Bogoljubow 159 Tal-Suetin 32 Ribli-Hertneck 181 Tavadian-Tseshkovsky 164 Ribli-Mikhalchishin 88 Tiets-Forsberg 91 Ricardi-Sorin 31 Tikhornirova-Morozova 125 Rittner -Bernstein 182 Tirnrnan-Ree 46 Rornanishin-Markowski 152 Tirnrnan-Salov 190 Rornanishin-Nunn 121 Tirnrnan-Van Wely 103 Romani shin-Rodriguez 172 Toothill-Heernsoth 116 Rornanovsky- Stahlberg 184 Topalov-Beliavsky 126 Rossolirno-Fischer 24 Topalov-Kasparov 101 Rozentalis-Christiansen 133 Torre-Portisch 51 Rublevsky-Sh ar iazdinov 128 Tosic-Gyirnesi 57 Rufen acht-Orseth 26 Trabattoni-Barlov 93 Ryurnin-Ilyin-Zhenevsky 32 Uhlrnann-Robatsch 28 Sajtar-Benko 48 Ulibin-Nevednichi 105 Sakaev-Novikov 141 Vaganian-Portisch 15 Sale-Mikhalchishin 188 Vagani an-Schlosser 84 Salov-Karnsky 173 Vaganian-Srnirin 124 Salov-Khalifrnan 123 Valvo-Levitt 32 Salov-Malaniuk 70 Van der Doel-Klovan 82 S alov-Yudasin 106 Van der Sterren-Douven 44 Salwe 105 Van der Wiel-Larsen 188 Schandorff- Speelrnan 10 Van Laaturn-Mikhalchishin 148 Schlechter-Lasker 79 Van Wely-Ad arn s 50 Schrnittdiel-Mikhalchishin 87 Vasiukov-Tirnoschenko 42 Serrnek-Hulak 54 Vaulin-Voikhovsky 86 Shirov-Krarnnik 73 Vladirnirov-Novopashin 40 Shirov-Lautier 132 Vladirnirov-Rashkovsky 58 Shirov-Morozevich 86 Vujala-Srnith 69 Shirov-Tirnrn an 7 Vukic-Pietzsch 113 Shirov-Van Wely 129 Vukic-Vujosevic 189 Short-Kasparov 158 Vukovic-Eingorn 47 Short-Top alov 102 Vyzhrnanavin-Chiburdanidze 45 Srnagin-Mikhalchishin 150 Vyzhrnanavin-Lerner 58 Srnagin-Naurnkin 80 Wa rd-Baburin 95 Srnyslov-Epishin 82 We lls-Hector 28 Sokolov-Khalifrnan 177 Wirthensohn-Mikhalchishin 106 Solozhenkin-Anchesi 33 Yates-Alekhine 117 Sorin-Alterman 178 Y errnolinsky-Ivanov 25 Speelrnan-Saltaev 28 Yerrnolinsky- Seirawan 69 Stein-Bobotsov 175 Yudasin-Krarnnik 104 Suba-Chiburdanidze 174 Yudasin-Osnos 9 Sulipa-Gricak 14 Yurtaev-Ternirbaev 136 Sveshnikov-Sokolov 51 Yusupov-Gerusel 186 Svidler - Lobron 8 Zaitsev-Hiibner 107 Szabo-Keres 62 Zaichik-Sorin 178 Szabo-Penrose 138 Zinar 109 Szelaig-Pinski 32 Zotkin-Kudrin 55