Beliavsky_Mikhalchishin_Winning_Endgame_Technique.pdf

Winning Endgame Technique Alexander Beliavsky and Adrian Mikhalchishin

Translated by Laurence Webb

B. T. Batsford Ltd, London

First published 1995 © Alexander Beliavsky and Adrian Mikhalchishin 1995 Reprinted 1996 ISBN 07134 7512 9 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.

Typeset by John Nunn and printed in Great Britain by Redwood Books, Trowbridge, Wilts for the publishers, B. T. Batsford Loo, 4 Fitzhardinge Street, London WIH OAH

A BATSFORD CHESS BOOK Editorial Panel: Mark Dvoretsky, John Nunn, Jon Speelman General Adviser: Raymond Keene OBE

Commissioning Editor: Graham Burgess

Contents Symbols Introduction

5 7

1 Pawn Endgames Technical problems in pawn endgames The opposition Chess 'materialism' Overlooking typical counterattacks How to play with and against passed pawns Premature resignation Exercises Exchanging into pawn endgames Simple examples Complex examples Exercises

9 9 9 13 15 17 22 24 28 28 29 45

2 Rook endgames with two extra pawns

48

3 Typical rook endgames with a passed pawn on the wing The rook's pawn The knight's pawn The bishop's pawn The central pawn

55 55 60 65 72

4 Various rook endgames The riddle of Duchamp Botvinnik's puzzle Capablanca's puzzle Is the endgame J. Polgar-Spassky lost? A cunning king manoeuvre Prophylactic cutting off (cramping) moves

80 80 83 84 85 94 95

What is the best position for the rook? The 'shoulder budge' Exercises

96 98 99

5 Bishop Endgames Same-coloured bishops with an isolated pawn Same-coloured bishops with a passed pawn Same-coloured bishops with pawns on one wing

102 102 112 115

6 Which is stronger, the bishop or the knight?

122

7 An unusual endgame

134

8 %l, lb and 4~ vs:, i.. and 3~ on the same wing liJ and 4~ vs ~ and 3~ on the same wing With rooks

140 140 146

9 Rook against bishop: how to win an exchange up

158

10 Tactics in the endgame

182

Solutions to the Exercises Index of Games and Composers

185 191

Symbols +

++ # ~ (+)

± (+) +- (-+) = ! ? !! ?? !? ?! Ch Weh Z IZ Ct OL

(D)

Check Double check Checkmate Slight advantage to White (Black) Clear advantage to White (Black) Winning advantage to White (Black) Level position Good move Bad move Outstanding move Blunder Interesting move Dubious move Championship World Championship Zonal Interzonal Candidates Olympiad Diagram follow

Introduction Endgame theory was first analysed a long time ago. The legendary Stamma, Ponziani, Polerio and others analysed many specific positions. Recently computers have had an impact upon the scene by generating databases for various 5- and 6man endgames, a classic example of which is John Nunn's Secrets of Rook Endgames. All chess books, including those on the endgame, are valuable to the extent to which chess players can extract useful information from them and then apply this information to tournament praxis. The Yugoslavian Encyclopedia of Chess Endgames (referred to as ECE) only partly serves this purpose, as it comprises a collection of different positions without explaining the various methods of play. As the well-known trainer Mark Dvoretsky never tires of saying, endgame study consists of analysing both theoretical positions and playing methods in complicated positions. The aim of playing complicated positions is to simplify them, by means of exchanges or in other ways, to a familiar theoretical position. This is endgame technique. Of course endgame reference books are important for the tournament player, but even more valuable are books such as

Speelman'sAnalysing the Endgame, or Dvoretsky's series of books. This book discusses problems encountered by tournament players, and is targeted at the average to strong club player. Firstly we discuss methods of play in positions with various combinations of pieces and pawns (of course only touching a small part of all different chess positions). In particular we focus on the mistakes made by strong chess players. In endgames the choice of possible moves can usually be narrowed down to two or three continuations, and by knowing what is incorrect, the correct move may be reached more easily. The same mistakes crop up time and again, repeated from game to game and from player to player. It is not by chance that the maxim 'you learn from your mistakes' was coined long ago. We have specialized in endgames for a long time, publishing articles in magazines such as New in Chess, Schach, and Torre e Cavallo. These articles were of a more theoretical nature than our current approach to certain positions and problems. By studying endgames it is important not only to solve concrete problems, but also to fonnulate general principles in all types of position. We hope

8 Introduction

that all our readers will gain something from this book, and enjoy the wonderful and almost infinite world of chess analysis. We are sure that many readers will improve their endgame understanding and thereby achieve corresponding tournament successes. In other words, study the mistakes of grandmasters to avoid

making such mistakes yourselves! The authors of this book are certainly not immune to error, and would welcome any improvements found by our readers. Alexander Beliavsky Adrian Mikhalchishin Lvov 1995

1 Pawn Endgames Pawn endgames have a reputation for being straightforward, both strategically and tactically. Yet anyone who has tried to solve positions set by the famous endgame composer Grigoriev, or who has practical experience in these endgames, will know that they belong to a world with its own special characteristics. For the tournament player these endgames can be theoretically divided into two types: 1) Technical problems of the endgames themselves; 2) Exchanging into pawn endgames. We shall discuss them in that order.

Technical problems in pawn endgames In our very first chess playing days we learn about such concepts as simple and diagonal opposition, rule of the square, corresponding squares and other principles. In some ways pawn endgames are richer than other endgames, involving all these playing methods, the lengthy calculation of variations, and pawn endgame technique. Pawn endgames are also a battle of the kings! Let's look at

how and why first class chess players make mistakes in these endgames. The material is presented as a series of topics.

The opposition Kupreichik - Mikhalchishin Lvov 1988

Black, in slight time trouble, decided to play actively. 1 ••• ~d5? An easy draw was at hand with 1... ~f6 2 ~f2 ~g6 3 xh6 ~f6 with a draw.

5 b5 4 a3 ~cS 5 a4 as 6 g3 g6 7 e5 cwf.1d5 8 ~d2 ~d4 9 'ittc2 c3 10 ~d 1 'it>d3 and Black wins easily. 2 h5 f6? Why make it easier for White to create a passed pawn? After 2... ~d4 3 e5 \t>c3 4 f5 g4 5 Cit>b4, etc.

(See diagram on/allowing page)

In this position White resigned, having calculated the following variation: 1 d6

21 cii;g7

1-0

G. Garcia - Vladimirov Havana 1986

4 h5! is correct, after which the pawn ending would be drawn.

4

h5!

5 ~d3 e6 13 ~g6 b3 14 f7 b2 15 f8fi' bI'ii'+ 16 ~g5; here White maintains some winning chances. (Typesetter's note: The database proves that this position is a draw). ~e7 4 ~d7 5 ~e3

6 d6! 7 dS

Pawn Endgames 37

resource. Correct was 1...cS ~b6 ~xb7

'ittg3 xh3 'it'xg4

a6 a7 a81*'

h3 h2 hl'ii'+

~b8 ~xa8

~xa8+

23 fS White resigned, since he is unable to prevent a second pawn race followed by another queen exchange and then a third black pawn promotion.

Black could play the simple 1... lbc8, after which he should fully equalize. However he instead forced the game into a pawn ending: 1 liJdS? 2 :xe8+ l:xe8 3 ltxe8+ 'fixe8 4 ltJxdS 'iVel+ S g2 \re4+ 6 'ii'f3 fixdS 6...cxd5 loses even more quickly: 7 'ii'xe4 dxe4 8 g4 '1t'h7 (or 8... g5 9 hxg5 hxg5 10 f3 +-) 9 h5 g6 10 d5 cl;g7 11 d6 d6 11 ~f5 c4 (11 ...fS ~d5 A draw was agreed. Beliavsky - Karpov Brussels 1988

5 ~e4? The right move is 5 h4!. 5

6 a4 7 ~d3 8 Ciitc3

d7 8 f4 ~c7 9 ~f2 ttJxb7 10 i.xb7 :xb7 11 1:xb7+ ~xb7 12 ~e3 c6 13 d6

3 .th3! 4 lDh6

lbd6

hxg4 gxh3+ ~d5

fxg4+

g6

5 !f:Jxf7! l2Jxf7 6 c7 1:.e8 Black now has a lost position.

Pawn Endgames 45

The simplest way to win was 7 :b8 lLJd6 8 .td7.

Exchanging into pawn endgames: exercises Furman - Zhukhovitsky Leningrad 1969

This was the position Karpov believed to be lost. However, Black can play the simple:

1 ... f6 It is not clear how White can make progress. Karpov overestimated the strength of his opponent's more active king. As the reader will already have realized from the large number of wrong decisions taken by grandmasters, simplifying into pawn endgames is no easy matter. We must learn from their mistakes. These endgames should not be a lottery of guessing or not guessing the correct move, rather involve working out the right move by drawing on one's theoretical knowledge and accurately calculating the possible variations. In conclusion to this chapter there are several exercises for readers to test themselves.

Is 1... ltJc5 or 1...ttJxd4 Black's best winning attempt? Kaunas - Mordvinov Tashkent 1969

White to move. Find the simplest win.

46 Pawn Endgames

P. Ricardi - G. Garcia Bayamo 1986

White to move. Which pawn endgame should he opt for: 1 :xg4 fxg4 2 b5 or 1 b5 immediately?

Piskov - Dvoirys Helsinki 1992

Is 1 b6 or 1 llJxc4 the best move?

Rozentalis - Smagin Odessa 1989

Berman, 1961

White to move. Is it better to play 1 i.f3 with the idea of 2 g6, or simply 1 .i.xt7?

How can White simplify into a won pawn endgame?

Pawn Endgames 47

Briedis - Timoshenko Riga 1991

A. Selesniev TIdschrift for Schack, 1923

Is 1 :'xh7 the best way to win?

How should White exchange into a pawn endgame?

Glek - Dautov Frunze 1988

Vaulin - Groszpeter Kecskemet 1993

Is 1 l:td3 the best move?

Should Black play the immediate 1....txe5 or wait before exchanging into a pawn ending?

2 Rook endings with two extra pawns In all other endings the advantage of two pawns nearly always guarantees victory - but it was no accident that Rudolf Spielmann commented jokingly that all rook endings are drawn. There are some well-known examples of drawn positions, such as those with the a- and f-pawns or two rooks' pawns, although drawing chances depend on the activity of the pieces. In this chapter we discuss less well-known positions in which the weaker side has excellent drawing chances two pawns down, analysing both positions which were successfully defended and games where the players did not use all the possible resources. Beliavsky - Kupreichik Yugoslavia 1992

Black has the obvious plan of sacrificing his h-pawn in order to bring his king to the d-pawn, thereby reaching a theoretically winning position. However, in this initial position White has the chance to make life difficult for Black by impeding this plan. t :d5! In the game White played 1 :d2? but after l ...fS 2 ~xhS he had missed the crushing 2...l:.e3 !!. After 3 :f2+ (or 3 :dS+ ~e6 4 I1a5 d5 5 g4 e5 6 l%aS ~e4 -+) 3... ~e4 4 ~g4 d5 5 %IfS d4 6 :eS+ ~d3 7 :a8 lIel he was forced to resign. With colours reversed this very position (after 1 l%d5) arose in Beliavsky's twenty-year-old analysis, published in Dvoretsky's Secrets of Chess Training! Now White can draw in the following variations: a) l .. Jlf6 2 :xh5 rj;f7 3 ~g4 g3 3 ~h4 4 :dl

5

l:e4+! l:te3+! :00 ~f5

:n+

5 ~xh5 .:Ie3! transposes into the game. 5 cst>eS ~dS 6 :el+ 7 :dl+ ~c6 8 :'cl+ Cit>d7 9 xh5 1Ie4! 10 g5 d5 11 eM5 ~d6 12 :dl ~cS!

13 :cl+ 13 :al is answered by 13....:Ie3! 14 11a5+ ~c4 15 %:ta4+ e5 lle3+ 16 ~f4 d4 and Black wins.

14 15 16 17 18 19

Chaunin- Friedman Moscow 1956

IIc3 :&5+

c4

:a4+ :al

~b5

lte3!

%:tdl

~c4

:'cl+

%lc3

Now the d-pawn is unstoppable. The only game plan in this type of position is to sacrifice one of the pawns in return for greater activity of either king or rook, although this is not always possible.

In this position White (to move) knew the winning theoretical position, which involves advancing the f-pawn to f6, cutting the black king off from its crucial defensive post of g7. White can then win using the tactical trick J:th8, meeting ....:Ixc7 by lth7+. He therefore played 1 hxg3?, with the intention of then pushing f4, etc. However White forgot that the position is drawn with the g-pawn, and one can imagine his disappointment after 1...g4+! 2 fxg4 and three pawns up White cannot win! White would win easily after

1 ~xg3! followed by the pawn sacrifice 2 h4! and only then smoothly advancing the f-pawn up to f6. It is always dangerous to think there is a choice between two equally good moves; usually only one move is correct and the other one is flawed. The next game is provided from one of the authors of this book.

50 Rook endings with two extra pawns

Mikhalchishin - Kluger Pecs 1978

This position looks easily won. There followed:

1 ~h8 2 :te2

10 l:g5! ~f7 11l:tfS+ h2 3 J:xa6 :f4+ 4 ~e3 :f5

3 1112? In fact the win was hard to find: 3 h7 %%.gl 4 ,:[f2 ~e7 (after 4...xeS S :fS there is no defence from 6 11g8) 5 :a2 :'g6 6 :a7+ f8 7 l:aS+ et;f7 8 l:1g8 :e6 9 %lg7+! ~f8

to queen. 3

Brooner - Hulak Berlin 1990

~xe5 ~e6

5 h7 I thought there was no defence to 6 :g8, but my adversary found a cold-blooded move:

5 :h3! 6 ~g7 l:Ig3+ 7 h6 l:h3+ 8 d7 ~e7

White checks the wrong way; 4 :e6+ ~f7 5 c6 was correct, with an easily won position.

:tg5

4

q;f7

as

5 c6 6 f3

l:el!

:te8 :el :e8 16 :e7 17 :a7 18 l:t7+

~C5!

19 ~xg3

d2

20 'lfi>f3 21 q;f2

:e3+ a3

a4 l:g4

lte4 '11tf4 ~e3

~g6

White's difficulty is that his king is unable to reach either of his two pawns. 7 ~f4 :e2

8 fl.dS 9 l:d6 10 17+

:c2 lIe2

11 ~5

~e7 ~e8

f4 16 ~f6 g3 17 l%h4+!, drawing.

9 10 11 12

:18+ e7 l:b8 :18+!

~g5

b2 f6

Of course not 12 :xb2l:te6+! and Black wins.

12

~g5

13 :'b8 ~f6 Draw

3 Typical Rook Endgames The most frequently occurring endgames involve an equal number of pawns on one flank (especially three versus three) and a passed pawn on the other flank. The Moscow chess player Kantorovich invented a very interesting statistical system to help evaluate this type of ending. The stronger side has the best winning chances with the knight's pawn, which he gives 0.7-0.8, Le. from ten games one would score 7-8 points. The bishop's pawn is evaluated at 0.65, a centre pawn at 0.7, and the rook's pawn at around 0.6. It should be noted that these evaluations are averaged out, and the result of any concrete position depends in the first instance on the activity of the pieces. Playing methods in these types of position are well-known; for the stronger side they involve combining the use of king and rook to push the passed pawn as far as possible, and for the weaker side they involve attacking the enemy pawns as much as possible, especially when the enemy king has to abandon his pawns on one flank to support the advance of the passed pawn on the other flank. We discuss these positions from both sides and look at reasons for the mistakes that often occur in practice.

The stronger side must try to maximize the co-ordination of his king, rook and passed pawn, and the player defending must achieve maximum king activity to attack the enemy pawns and create his own passed pawn. Of particular importance is the position of the stronger side's rook; behind the passed pawn is usually ideal, although sometimes it is very important to reach a position in which the rook can simultaneously defend the passed pawn from the side as well as pawns on the other wing, while cutting off the enemy king.

The rook's pawn Van der Wiel - Seirawan Haninge 1990

1

~hS!

60 Typical Rook Endgames

The knight's pawn Chekhov - Eingom USSR 1984

B

gxh5 9 :h7 l:c5 10 ~f3 b5 11 ~e3 and now 11 ...b4? fails to 12 ~d4 b3? 13 ~xc5 b2 14 ~c6! +-. S :c4! 6 gxh5 :'xh4+ 7 g4 Black won easily.

g5

Mikhalchishin - Losev Moscow 1974

1 •••

~e8

Black has the obvious plan of bringing the king to the b6-pawn. The other possibility 1...h5, also leads to nothing after 2 llb7 l:c6 3 g4 ~e8 4 l:th7 :tc3+ 5 ~g2 hxg4 6 :g7 ':c6 7 h6 8 l:b7+ 9 %lb8 l:b3 10 b6 rJitg7 11 cli>d5

16 17 18 19

~d6 ~d8

:dl+ l:cl+ l:bl

:d6!

1-0

Q;;c7

If the number of pawns on one wing is increased to four then the stronger side certainly has greater winning chances, but the position can still be defended. Piket - Tukmakov

Amsterdam 1990

1 h4? 1 g4! is better, immediately creating trouble for Black on the kingside.

Typical Rook Endgames 65

1 ...

~g7

1...h5? is weak due to 2 1:a8+ and 3 ltb8, activating Black's rook. 2 g3 h5 3 g2 :b2

4 1:a8

Itd2!

It is vital to transfer the rook to a more active position, from where it will defend its pawns and open the way up for the king. 5 llb8 6 e8 e5 12 :xfl 13 ~f4 14 g4 15 xh4 Draw

f2

n fi ~xfi ~f2 ~e3 ~f4

Balashov - Salov Lvov 1984

1 1%h3 2 :a3?!

f5

An inaccuracy, although not a decisive mistake. Simpler is 2 :lc3 or 2 l1d3 followed by playing the rook to the seventh rank.

2 3 :d3 4 c4?

a6 f6

White could have scored an easy draw with 4 :d7 1te7 5 :xe7 rJ;xe7

6 ~e3 'ite6 7 el 7 8 e4 J:g3

6 l1bS 7 l1d5 8 lIbS (D)

:gl :bl

After 8 l:td3 e6 9 J:[f3 ~f6 10 d4 ~f5 11 g4 Black draws easily, having activated his king. 8 ••• 'ifi>f6? He should play the immediate 8...l:.e1+ 9 ~d5 :til. 9 %lb6+ r3;f7 If 9... rtte7, then 10 b4 :e1+ 11 ~d5 %:f1 12 :xg6 :'xf4 13 b5 ~d7

84 Various Rook Endgames

14 :c6! Itf5+ 15 ~c4 is winning for White.

10 b4 11 d4 12 e5 13 e3 and a draw

Now Black is at a crossroads.

2 .•.

as

The other possibilities are: a) 2...Citi>f8 3 'it>b6 :d6+ 4 ~a5 rt;e7 5 a4 ~d7 6 :c2! and now

Various Rook Endgames 85 6 .. J~d3 7 b4 :Ixf3 8 'it>xa6 :Ixg3 9 b5 :b3 10 g4 11 b6 f5 12 ~b7 f4 13 a6 f3 14 a7 l%a3 15 as'if ':xa8 16 ~xa8 g3 17 b7 winning, or 6...f5 7 b4 f4 8 g4 :Id3 9 ~xa6 llxf3 10 b5 and White should win easily. b) 2...g4 3 fxg4 :'xg3 (D):

as

• • ••• i. i. •.i.

w•

••

•.=•• .. . .~. ~.

.~.

• a •

••••

5 6 :a2 7 ~b5

8 ~xa4

a4

:Ixf3 :Lxg3 fS

9 b5 The white pawns are more dangerous. Black has the most chances in variation 'b', although even this line is clearly in White's favour. Capablanca's commentary is strange: why did he refuse to play actively with his king when in all these positions Black is playing virtually a king down?

Is the J. Polgar Spassky endgame lost?

as

bl) 4 :f4 1:e3 5 :f6 6 :a6 f5 7 gxf5 gxf5 with sufficient counterplay for Black. b2) Better is 4 ':f6 ':xg4 (a better try for Black is 4 ...aS! 5 ~b5 ':xg4 6 xaS cJ;g7 7 :f3 f5 8 b4 f4 9 b5 g5 10 b6 l:g3 11 ':f2 :a3+ 12 ~b4 l:a6 13 cJi>b5 :ta8 14 a4 but again White has serious winning chances) 5 ':xa6 f5 6 b4 f4 7 b5 f3 8 :'f6 :g5+ 9 ~c4 and White wins. 3 Ciitb6

1:.dS

4 a3!? Also interesting is 4 a4 with a furin mind. ther b4 and

as

4 ... 5 b4

:d3

Even stronger is 5 :b2 a4 6 bxa4 :xa3 7 a5 ':xf3 8 a6 ':'xg3 9 a7 1:ta3 10 l%b5 :xa7 11 rJ;;xa7 f5 12 cat>b6 g4 13 ~c5 f4 14 ~d4 and White wins.

In the final game of the J.PolgarSpassky match there occurred a very interesting position: rook ending with two pawns against one on the same flank. It seems to be completely drawn but praxis shows that this configuration (f- and g-pawns versus h-pawn) is quite dangerous for the weaker side. In 1983 this kind of ending was extensively analysed by the Soviet master Kuzminykh in Shakhmatny Biulleten no. 7, and there also exist some analyses of IM N.Minev and GM Ftacnik. However, in ECE there are no final conclusions on this subject. Given the dearth of actual games involving this position and the lack of any definite theoretical assessment we have reviewed current praxis and

86 Various Rook Endgames tried to draw some conclusions. The first example shows the correct way of defending against the advanced pawns. Brodsky - Magerramov Helsinki 1992

1 :tb2+ 2 ~f3 hth2 3 :18+ Of course not 3 :'h8? falling for the standard trick 3...:xh3+. 3 ••• ~g6 4 l:g8+ 4 h4? 1%h3+! is very dangerous for White. 4 et>f6 5 :h8 1;g7 6 :hS ~g6 7 J:h8 Ciitf6 8 g4! Wrong is 8 1:f8+? (or 8 lIh5? ~f5 9 :h8 :xh3+! -+) 8... ~g7. 8 .•. :tg2+ 9 ~f3 9...:g3+ 10 ~f2 gets nowhere as it is impossible to penetrate White's position with the king.

:a2

10 11 12 13 14 15

1:th6+ :h8 et>g2 l:te8+ l:td8+ :e8+

~f5

1:.a3+ ~e4

d4 ~e3 ~d2(D)

16 J:eS! White's only chance is to combine an attack on the g5-pawn with checks. 16 ... :'e3 17 :a5! Activity is much more important than a pawn: after 17 ':xg5? ~e2 18 :f5 f3+! 19 g3/g1 :e8! Black wins immediately. 17 ... :tg3+ 18 )f;h2 18 ~f2 is preferable. 18 ..• 19 'it>g2 ~e2 20:82+ ~e3 21 :f2! :'cl 22 :0+ ~e4

:c3

23

:f2 (D)

A very important theoretical position: White holds a draw. 23 •••

:c3

Various Rook Endgames 87 after 32...:d2+ 33 1txd2+ ~xd2 34 h2 l:tc3 39 ~g2 is similar to a position already reached. 37

:a2

38 hxg4+

Once more the players have reached the main position. 27 :dl 28 :f3+ ~e4

:d3

g4

xg4

39 '3ifi!

l:h3

40 :g2+ 41 %181

'tt'rs 1:£3+

Draw

l:el 1:.dl

Vaiser - Djuric Szirak 1985

This is a useful check in a position with many good possibilities, such as 31 ~f2.

31 ...

~d3 (D)

1 ~h3!!

32 l:81! e3 The pawn ending which arises

Psakhis and Vaiser give this move two exclamation marks. A logical possibility was 1 h5, which could lead to the well-known ending with f- and h-pawns.

88 Various Rook Endgames This kind of position with the pawn of the weaker side on his fourth rank is considered to be lost, except the case when the rook of the weaker side is on his eighth rank, because it is very importantto check from behind. There are various other possibilities. In the game Gliksman-Novak, Stary Smokovec 1976 there followed 1 h5 g5 2 l:b6:f7 3 :a6 ri;g7 4 f5 :b7 5 h6+ (or 5 lta5 ~h6! 6·'iitxf6 J:tbl 7 'iiif5 ~xh5 -+) 5...c3 31 ~cl ~d4 32 d2lLJe4+! 33 .ixe4 CJt>xe4 34 ~e2 and the pawn ending is easily winning for Black. 30 ~xa3

as!

31 ~c3 32 cxbS 33 .tg6

34 b5 .ie4 5 lbe6+

a6 b8 11 ~b6 winning.

4 l2Jb5+

d7

After 4 ... ~c6 White plays 5 a7

d7 6 ~a5 and 7 xh2 ~e3.

lbg3

12 ... 0-1

There is no defence to the h-pawn queening.

An Unusual Endgame 139

A very interesting and similar endgame arose in the following game. Suetin - Stein Kislovodsk 1972

bI) After 2...lbf4 3 e3lbxh3 4 i.b7 there arises one of the aforementioned positions. b2) Black has another possibility which we had missed when publishing this position in New in Chess magazine, namely 2...lbb6! 3 ~e3 lL\c4+ 4 ~d3 ~f4! and Black wins. The game continued: 1 ••• f6

2 ~gl 2 g3 is met by 2... g5, handicap-

ping White with chronic weaknesses on f3 and g3. 2 ••• ~eS

This game continued 1 .id7 (it is stronger to play 1 g7 White would respond with 2 ~c6, threatening to bring his king to the e8square, hitting the f7 -pawn.

2 IS!

pIS

White was threatening both 3 f6 and 3 fxg6 fxg6 4 ~d6 followed by e6.

3 lbf6+

q;g7

Simpler was 8 e6 fxe6+ 9 '1txe6 f4 10 Cite5 i.g3 11 llJf5! ..th2 (or 11 ...f3+ 12lDxg3 fxg2 13liJe2 +-) 11l2Jh4+ ~h5 12llJf3 ..tg3 13 ~f5 followed by the knight tour l2Jd4e2xf4. ..td4 8 ... 9 ltJc4 i.c3 10 ~d5 gS

11 12 13 14 15 16 17 18

lbd6

~g6

ltJbS lDd4 g3

i.b2 dS

9 ~e3

1-0

The most interesting positions occur when the weaker side has a passed pawn on the other wing.

Eingom - Mikhalchishin Simferopol 1983

i.c4? 1 •.• Black should have bravely played 1...i.g4! 2 lIxb5 Cifte6 3 ':b6+ 'it;e7 4 .1%b7 + g3 34 :hl +-.

1 h3 2 .tc3

~e6

f4?!

This seems dubious; 2... ~d5 3 ~f3 Citc4 and 4... ~d3 was prefer-

able.

3 f7! 27 ~h8 ~f8! draws. After White's actual move, Black was forced to resign.

A more complicated position occurred in the following game. Mikhalchishin - Bareev USSR Ch (Lvov) 1987

~c2

i.el i..c3 ~b3

17 ~S

a2

18 19 20 21 22 23

e4 :a5.

26 f4 27 i.xrs 28 e4

:g7 ~e7

:g3!

Black now has enough time to bring his king across, having intercepted the white king. 29 .ic8 rJ;f7

30 .id7 31 i.e8 32 .id7 33 ~ 34 ~g3 35 .ih3 36 hh3

10 •••

~d2

9;;g7 ~h6

:g4! Axh4 :hI 1hh3+ e4 13... i.c 1 is better. 14 l:d7 15 ~dS 16 :h7 17 lth8 18 :g8

~g5

.ih6 ~g6

.i.g7?! ~f6

.th6 f6 +-. Back to the game:

16 :c6! .tb2 17 :a6 i.c3 18 :a4! J.e5 The bishop cannot go to g7 due to 19 :a8+ ~f7 20 :a7+ cJtg8 21 h5

1 •.• 2 i.d3!

:Ie3+

If the white king leaves the third rank then Black would win easily by occupying the d4-square with his king, thereby restricting the white king to the second rank. This plan

Rook against Bishop 179 was executed in the game De FotisPodgaets, San Juan 1971 (on the other wing). In this position Black must try to push the white king back and advance with his own king.

2 •••

~f4

3 ~c4

eS

4 i.c2

l1e2!

4...Cit>d6 fails to 5 b4.

5 ~dl 5 i.d3 is met by 5...:'d2 6 iLg6 ~d6, as in the game. S ••• l:d2 6 i.f3 ~d6 7 ~c3

I:.d4

8 i.g2

:g4

Another possible plan is to play the king to a5 with 8.. .'~c7 9 i..fl c4

i.aS

8 ~d5 9 e8 .ic5 12 ~d8 Although White is trying to penetrate with his king, he rejected 12 ~f8 because of 12...d5+ 13 et>g8 d4 14 h7 g5 15 dS 3 ~xh5 4 g6 S ~xg7

~e4 ~d3

xc4 f4 3 g4! 4 ~c5 hS 5 Cittb6 h4

Black is winning easily.

-+.

:a8

4 ~e6

:a8

1Jr1

:a7

~f8

%la8+ :xe8+

5 6 7 8

P14

~7

1 ::txh7! 2 ~d7 3 e7

eB1i

1ta7+

Ciitxe8 +-

E3 1 bS! 1 ':xg4? fxg4 2 b5 ~d5 3 c6 bxc6 4 bxc6 cJi>xc6 5 ~xd4 lPb5 6 ~c3 ~c5! 7 d3 ~d5 8 ~e3 c4 0-1

Ricardi-G.Garcia. 1 2 3 4 5

gxf4+ c6 bxc6 xd4

6 a4

:xf4 ~dS

bxc6 ~xc6 ~d6

as=

E4

1 i.xf1! 1 .tf3? b5! 2 g6 fxg6! 3 ~xe6 a5 4 e5 d3 5 ~e4 ~c3 6