Kingwalks: Paths of Glory 9781949859386, 9781949859393, 194985938X

Citation preview

Kingwalks Paths of Glory by Yasser Seirawan and Bruce Harper ISBN: 978-1-949859-38-6 (print) ISBN: 978-1-949859-39-3 (eBook) © Copyright 2021 Yasser Seirawan and Bruce Harper All Rights Reserved No part of this book may be used, reproduced, stored in a retrieval system or transmitted in any manner or form whatsoever or by any means, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without the express written permission from the publisher except in the case of brief quotations embodied in critical articles or reviews. Published by: Russell Enterprises, Inc. P.O. Box 3131 Milford, CT 06460 USA http://www.russell-enterprises.com [email protected] Cover by Fierce Ponies Special thanks to John Donaldson Printed in the United States of America

Table of Contents Introduction Foreword by Hans Ree Chapter 1 Kingwalks to Prepare an Attack Chapter 2 Kingwalks in Anticipation of an Endgame Chapter 3 Kingwalks to Defend Key Points Chapter 4 Kingwalks to Attack Key Points or Pieces Chapter 5 Mating Attacks Chapter 6 Escaping to Safety Across the Board Chapter 7 Escaping to Safety Up the Board Chapter 8 Kingwalks in the Opening Chapter 9 Kingwalks in the Endgame

Chapter 10 Double Kingwalks Chapter 11 Unsuccessful Kingwalks Chapter 12 Wilhelm Steinitz Chapter 13 Aron Nimzowitsch Chapter 14 Tigran Petrosian Chapter 15 Duncan Suttles Chapter 16 Yasser Seirawan Chapter 17 Recent Examples Bibliography

Introduction Getting off to a good start In a hole in the ground there lived a hobbit. Not a nasty, dirty, wet hole, filled with the ends of worms and an oozy smell, nor yet a dry, bare, sandy hole with nothing in it to sit down on or to eat: it was a hobbithole, and that means comfort… We wanted to start this book with a compelling opening sentence, but it was pointed out to us that this one was already taken. So we tried again: When Mr. Bilbo Baggins of Bag End announced that he would shortly be celebrating his eleventy-first birthday with a party of special magnificence, there was much talk and excitement in Hobbiton. Sage legal advice forced us to also give up on this equally striking start, so we can only hope that the third time’s the charm: We celebrate not the cautious, sedentary monarch who remains at home, counting his treasure while fearful of its loss, who sends his minions into battle against the enemy king while cowering in his fortress; but rather the brave, inspiring leader who asks nothing of his followers that he would not do himself, and who will risk all to prevail. This starting sentence, while overly long and somewhat inaccurate – kingwalks often involve movement away from, not into, danger – at least avoids a lawsuit. But surely The Hobbit and The Lord of the Rings are the epic recounting of long, dangerous and ultimately successful kingwalks? Did Tolkien consciously or otherwise base his underlying theme on a memorable chess game he played in his distant past? There is no evidence of this, either way, as far as we know, so we will never know. But this book, which explores the subject of kingwalks in chess, might

well be placed on your bookshelf next to Tolkien’s masterpieces. Or not. The fascination of kingwalks There is something fascinating about a kingwalk. What is essentially a strategic or tactical maneuver charms us, perhaps because of its inherent contradiction and even implausibility. One of the weakest, and certainly the most important and vulnerable, chess piece does something, apart from trying to remain safe. Checkmate ends the game, so king activity carries with it the seeds of total disaster. Make a mistake with the other pieces or pawns, and the game goes on. If a kingwalk goes awry, that’s it. But it’s not just the risk of kingwalks that attracts us. The chess king may move only one square at a time, which means that it moves at a snail’s pace compared to a bishop, rook or queen. Even the short-range knight can move twice as fast as the king. The deliberate and regally serene pace of the king ensures that kingwalks are indeed walks – in a game where a single tempo may often decide the outcome, kingwalks are fascinatingly incongruous. A king may be said to race to catch a pawn that threatens to promote in the endgame, but the term kingwalks somehow seems apt when the board is still full of pieces. More could be said on the aesthetic and psychological attraction of kingwalks, but it would be hard to justify an entire book on the subject of kingwalks based only on their beauty, undeniable as that beauty is. As we will see, in some instances kingwalks can be hideous as well. In chess, form and function merge in a manner that would delight the most devoted member of the Bauhaus school. If something works, it’s beautiful; if it doesn’t, it’s ugly. “Ugly moves,” in this sense, cannot be strong moves. Equally, an unsound sacrifice must at least have a psychological component that can cause the defender to make a mistake before it can be called “beautiful.” Tal made many beautiful combinations, and some of them were even sound. For the others, what hidden resources his opponents might have had often remained hidden until after the game. Kingwalks must have a point in order to be interesting and attractive – simply moving your king around for no reason is just weak play. This is not to say that the goal of every kingwalk has to be strategically justified or objectively

sound. We may justifiably be entertained by dubious or unsuccessful kingwalks, just as we are amused when someone slips and falls into a muddy puddle. When a king walks off a cliff, it’s funny – as long as it’s not our king. Still, a good kingwalk has to have some justification, if only in the mind of the player concerned. Our emotions will then likely be engaged, one way or the other. Practical applications While our hope is that readers of all strengths will enjoy this book, there is also much to be learned from the study of kingwalks. They are a legitimate part of chess, and can transform the nature of the position to a great extent. Given the difficulty people, including chess players, have in coping with change, the psychological effect of kingwalks cannot be overestimated. The extent to which a kingwalk will throw the opponent off his or her game depends in part on the position and the nature of the kingwalk, and in part on the opponent. There are two aspects to this. One is that king location and safety are fundamental elements of every chess position. Moving the king to a different area of the board alters the position in an objective sense. The second aspect is that a kingwalk may distract the opponent, either by offering a real or imagined target for attack (depending on the soundness of the kingwalk), or by making a direct attack against the king impossible, more difficult, or just different. This subjective impact may cause an opponent to lose his or her bearings, in the same way an unexpected offer of material may induce “sacrificial shock.” When a player starts to wonder if an opponent’s kingwalk is an act of desperation or a colossal blunder, his or her focus may shift to thoughts of mate. If this shift is unjustified, bad things may happen. Of course, if the shift to a mating attack is justified, bad things may also happen… In the pages that follow, we not only give examples of different types of kingwalks, but we try to explain the positional, tactical or psychological basis for each example. A legitimate kingwalk doesn’t come out of the blue, any more than a combination arises by chance. By exploring the preconditions for the different types of kingwalks, we hope the attentive reader will recognize positions in his or her own games where a kingwalk might be the path to

victory. Equally, this type of analysis will help players in coping with opposing kingwalks. What is a kingwalk? The entry for “kingwalk” in the Caissa-Webster Dictionary reads: kingwalk /'kiŋwȏk/ Etymology: From Middle English king, kyng, from Old English cyng, cyning (“king”); and from Middle English walken (“to move, roll, turn, revolve, toss”), from Old English wealcan (“to move round, revolve, roll, turn, toss”), ġewealcan (“to go, traverse”). In chess, a voluntary, purposeful maneuver with the king, part of a deliberate, planned strategy occurring usually in the opening or middlegame, whereby the king seemingly is put in danger as it moves near or among enemy pieces, seeking a strategic or tactical goal. Our definition of the term “kingwalk,” which in turn defines the scope of this book, is slightly different from the dictionary definition. Clearly every kingwalk involves the movement of the king, but not every movement of the king can be termed a kingwalk (a book on games where one player’s king never moved at all might be interesting, but that’s for another time). The term “kingwalk” as used in this book refers to a king maneuver with the following characteristics: •

•

•

It must be purposeful, in that the king maneuver must have a strategic or tactical goal (although, as discussed above, this goal need not be justified). It should be voluntary. A kingwalk may be made under duress, in that it is the only viable defense, but it should still be a deliberate, planned strategy on the part of the player doing it, having anticipated the opponent’s attack, as opposed to an unplanned blasting of the king out into the open. It generally occurs in the opening or middle game, as opposed to the endgame (although we have been flexible on this – compelling and clear king maneuvers in endgames have been included).

Kingwalks are therefore clearly (and in some cases, not so clearly) different from “king hunts,” where an attack results in the exposure of the defender’s king. There is also no minimum length for a kingwalk, although naturally the longer the kingwalk, the better. But it is wrong to emphasize this factor too much, just as it’s wrong to say “the longer the combination, the more impressive it is.” A kingwalk of even one or two squares can be compelling. The length of a kingwalk may be a measure of its audacity, but this will depend on the circumstances. The organization of this book The chapters in this book deal with different subjects, and may be grouped as follows: Chapters 1 and 2 discuss kingwalks made in preparation for later developments: chapter 1 for an attack (usually by positioning the king on the opposite flank to the location of the attack) and chapter 2 for the endgame (positioning the king in an advanced square so that endgame activity is facilitated). Chapters 3, 4 and 5 involve a more active role for the king: chapter 3 discusses the king’s role in shouldering the burden of defense; chapter 4 looks at the king’s role in pressuring enemy weak points and attacking or capturing enemy pieces; chapter 5 explores the most dramatic and extreme example of an attacking king – participation in a mating attack. Chapters 6 and 7 are quite different – they are concerned with kingwalks to safety (sometimes under pressure and sometimes not). The difference from chapters 1 and 2 is that the king relocates because of enemy pressure, not to prepare an attack or prepare the activation of the king for the endgame. Chapter 6 looks at kingwalks across the board, while chapter 7 looks at kingwalks up the board, which tend to be even more spectacular. Chapter 8 turns to the opening, giving a variety of examples of kingwalks early in the game, which is when you would least expect to see them. Chapter 9 follows naturally, as its topic is kingwalks in the endgame. King activity is a fundamental aspect of all chess endgames, some more than

others. As the number of other pieces drops, the king tends to come into its own more and more, with the ultimate example being king and pawn endgames. At times the king’s role in an endgame is so striking that it can properly be termed a kingwalk, and the examples given will help remind us of how important it is to have an active king in the endgame. Chapter 10 addresses a particularly curious animal – games that feature kingwalks by both players. Such games are uncommon and are highly entertaining. Chapter 11 might be even more entertaining, while being instructive as well. This chapter deals with failed kingwalks – the kingwalks off a cliff as alluded to earlier. Exactly how one defines a failed kingwalk is a judgment call, as even during successful kingwalks there can be a moment where the right response from the opponent could have radically changed the outcome. In some cases, though, the failure is indisputable. Chapters 12 to 16 look at examples of kingwalks from five different players. To some extent the selection of players (Steinitz, Nimzowitsch, Petrosian, Suttles, Seirawan) was somewhat arbitrary, being based in part on the authors’ familiarity with the subject matter, but all five players are arguably “positional players” (and perhaps “extreme positional players” at that). At the very least, these chapters provide additional interesting and instructive examples. Where games involving these players appear in earlier chapters, they are not repeated.. Finally, chapter 17 gives a selection of recent kingwalks, which show that kingwalks are alive and well and are an integral part of modern chess. Sources This book contains a range of games, with the sources ranging from world championships, candidates’ tournaments and matches, and classical time control events, to rapid, blitz and bullet games. The sole criteria for inclusion were whether the kingwalk in question was instructive and entertaining (not necessarily in that order). Sometimes a kingwalk didn’t even occur in the actual game, and is found only “in the notes,” as they like to say.

The examples have been gathered from various sources. A list of those is provided in the bibliography. However, two works deserve special mention. The first is Mednis’s interesting 1999 book The King in the Middlegame, which contains a number of well-known and not-so-well known examples of kingwalks. The second is Chess Curiosities by the Dutch writer Tim Krabbé, as the third chapter (Steel Kings) is devoted to kingwalks. His website is a treasure trove and well worth visiting: https://timkr.home.xs4all.nl/chess/chess.html. Where analysis has been borrowed from other sources, it is acknowledged. Where the authors have used their previously published analyses, we have been less generous. All analytical errors are most definitely someone else’s fault. Yasser Seirawan, Amsterdam Bruce Harper, Vancouver May 2021

Foreword Bravery and Profundity Paths of Glory, the subtitle of this book, is the bitterly ironic title of a movie by Stanley Kubrick from 1957 in which French soldiers in World War I are being sent to their death by ruthless commanders, in missions that serve no military aim, except to provide glory to the generals. But here the expression is used sincerely. True, there is a chapter in which the kings walk to their doom, but mostly the kingwalks that are shown here are really glorious: horizontal, vertical or, for intrepid walkers, even diagonal, to run to safety or help in a mating attack or in some spectacular instances even both. A few times the authors stress the educational power of their examples. Readers may develop a feeling for the required conditions for a successful kingwalk. It’s always good to learn something, but surely their main aim must have been to inspire awe and amazement. And provide entertainment sometimes, when things go wrong. There is no bravery without the possibility of failure, so obviously there is a chapter about unsuccessful kingwalks. The authors make no secret of the fact that these failures can be funny. Maybe they subscribe to the old maxim that all good jokes are an elaboration of a man slipping on a banana peel. They write: “There is nothing quite as entertaining as someone failing – provided it isn’t you (or someone you’re cheering for)!” Elsewhere they mention “the near universal human trait” of Schadenfreude, the pleasure with somebody’s misfortune. And they go on: “Chess can be a very cruel game. That’s why we play it.” Far be it from me to deny the fact that this innocent sadistic streak is no stranger to me. But the joyous admiration for spectacular and successful kingwalks prevails. Once you have seen one of the classic examples, you’ll never forget it.

Whenever a book has two authors, you may wonder sometimes who wrote what. I think I found a sentence about Nimzowitsch that may betray a small difference of opinion between the authors. They write: “It is therefore unfair to consider him a thief, and equally unfair to accuse him of fraud, although he was not as generous in acknowledging the discoveries of his predecessors as he might have been” I imagine an argument: “But Yasser, surely you can’t call the great Nimzowitsch a fraudulent thief?’’ “OK, Bruce, let’s say then that it’s unfair to call him so, but…” The book has a large and entertaining collection of kingwalks. But, among all these great kingwalks I missed the magnificent game Gashimov-Grischuk (Bursa 2010) where Black’s king marched all the way from e8 to b1, to victory. Of course, with a subject as vast as this, every reader will always miss one of his favorites. It cannot be otherwise. I have mentioned already that I am not quite immune to what the authors call Schadenfreude. Not to my credit, I experienced quite some Schadenfreude after I had executed the one and only successful kingwalk in my career. It certainly can’t be called a path of glory, but I found it hilarious. By coincidence it happened in a blitz game against the Great Sage of kingwalks Tim Krabbé, played at my home. My king walked from g1 to a5. He could have mated me on the kingside, he could have mated me in the middle of the board and he could have mated me on the queenside. Finally I found refuge near b2 and then he immediately blundered a piece. I won the game and Tim cried: “With my own theme, the winning king’s march!” Immediately after he left my place, I wrote down the moves of the game. Tim himself had once published a blitz game against me in which right after the opening he had mated me with his black knights on f3 and c3. I would have my revenge on Tim. Beaten with his own theme, hoist with his own petard. After reading Yasser’s and Bruce’s exhilarating book, I tried to find the notation of that blitz game, looking forward to the pleasure of playing it over move by move. Many years and a move to another apartment had passed

since I had played that game. I couldn’t find the notation. But no matter, the sweet vague memory was enough. Executing a successful kingwalk has the power to make a chessplayer happy and the same can be said about playing over the many beautiful examples in this book. Enjoy! Hans Ree Amsterdam May 2021

Chapter 1 Kingwalks to Prepare an Attack The “classic” kingwalk of this type involves marching the king from the kingside (its usual position, given the high frequency of kingside castling) to the queenside, in anticipation of opening lines on the kingside. The king is removed from a potential, rather than an actual, area of danger, the idea being that the opening of lines on the kingside will expose the king to enemy counterplay. This sort of kingwalk is an example of extreme “prophylaxis” (Nimzowitsch) or “anticipatory defense,” and while in many instances it is unnecessary or impossible to achieve, it is a very effective strategy when the opponent is subject to a bind which gives the attacker time for such a step. As will become apparent from the following examples, several conditions must be met before anticipatory defense by means of a kingwalk comes into play: • •

•

One side of the board must be completely safe; on the other side of the board attacking lines must exist or be capable of being created. The side conducting the kingwalk must have a sufficient positional superiority that it can dictate the pace of play, leaving the opposing side without counterplay. The opposing player must not only be unable to impede or exploit the dominant player’s kingwalk, but must also be unable to emulate it, so that his or her king must remain on the side of the board where it will be exposed to attack.

It is no accident that the quintessential example of anticipatory defense by means of a kingwalk comes from world champion Tigran Petrosian (Petrosian-Unzicker, Hamburg 1960), the recognized master of anticipatory

defense. But in fact our first example is from a game involving another world champion: Alexander Alekhine. Given Alekhine’s justifiably renowned attacking style, it’s slightly surprising that he adopted such a positional approach. The fact that this game is from a simultaneous exhibition might be relevant – is it cynical to think that Alekhine might have been treading water while finishing off some other opponents? We prefer to think that, even in a simul, Alekhine was quite capable of demonstrating his deep understanding of chess and his flair for beauty. All three of the prerequisites for a kingwalk exist, including that White’s king is trapped on the kingside.

Kevitz & Pinkus – Alekhine New York 1929 (consulting clock simultaneous exhibition)

In this position, Black has a clear advantage. White’s pieces are tied to the defense of his weak e4-pawn and the light-square weaknesses around White’s king are a constant source of concern. The engine of course recommends direct action, but Alekhine, who was no stranger to that type of play, first takes the time to reposition his king. 28...Kf7 29.Bf2 a5 30.Bg1 Re5 31.Bf2 Qg4 32.Bg1 R8e7 33.Bf2

33...Ke8! White’s helplessness has been demonstrated by the last five moves. There is no way he can improve his position or create counterplay.

Black now continues the kingwalk he started on his 28th move. Interestingly, one source rearranges the moves in this game so that Black’s kingwalk is one continuous maneuver. Presumably this was done for dramatic effect, or perhaps (more charitably) for pedagogical value. But if changing moves in completed games is allowed, your authors also both want licenses to do this. 34.Bg1 Kd8 35.Bf2 Kc8 36.Bg1 Qh5 37.Bf2 Bg4 38.Kg2 Qh3+ 39.Kh1 a4 40.Bg1 Kb7 41.Bf2 g6 42.Bg1 Ka6

Black has completed his kingwalk. His king is absolutely safe on a6, and he can proceed with his breakthrough in the center. 43.Bf2 f5 44.exf5 gxf5 45.Rxe5 dxe5 46.Qe3 46.Rxe5 Rxe5 47.Qxe5 Ne4 leads to mate. 46...e4 47.d6 cxd6 48.Bg1 f4 0-1

Black’s extremely safe king means that White doesn’t have a hint of counterplay. The next example is from Nimzowitsch. This game features anticipatory defense on a grand scale. Nimzowitsch’s kingwalk prevents any future king problems on the kingside, but he is equally careful to ensure that nothing untoward happens at the king’s new place of residence, on the queenside. Meanwhile, all Black can do is wait for the ax to fall, as his king’s route to safety is blocked by his own pieces.

Nimzowitsch – Asztalos Bled 1931

Nimzowitsch has achieved a dream position – White has an advantage in space and only he can think of opening lines. The potential fly in the ointment is that this can only be done on the kingside, which means that both sides know where the action will eventually occur. With this in mind, Nimzowitsch transfers his king to the queenside. It is not so easy for Black to do the same. 30.Reg1!? Ra7 31.Kf1 Qb8 32.Ke1 Bd8 33.Kd1 Be7 It’s usually a bad sign if all you can do is shuffle your pieces back and forth, but Black is so cramped it’s impossible for him to imitate White by walking his king to the queenside. This means that when White breaks through on the kingside, he will have an additional target. 34.Kc2

White has finished the first step in his preparations, and now goes about his business on the kingside, although with a typical Nimzowitsch eye to prophylaxis. 34...Qc7 35.Bc3 Rea8 36.Qe2 Qd8 37.f4 f6 38.Be6+ Nxe6 39.dxe6 Bb7 40.Qd3 Qe8 41.f5 Bc6 42.g4 Kh8 43.Rg3 Rb8 44.Rhg1 Rab7

45.Bd2 An interesting moment. The engine is quite happy with 45.g5, which is certainly strong, but Nimzowitsch prefers to prevent any shred of counterplay before starting the final attack. 45...Rd8 46.Nd1!? Rbb8 47.Nc3 Rdc8 48.Qe2 Rb7

Now everything is ready. 49.g5 Rd8 50.g6 Bf8 51.gxh7 Kxh7 52.Rg6 1-0

Smyslov – Panov Moscow 1943

It is hardly necessary to know that Black’s last move was 30...Nh8 to figure out that Black’s position is completely passive and that White’s spatial advantage on the kingside will inevitably lead to a breakthrough. 31.Kg2! White begins a kingwalk, not only to ensure the safety of his king, but also to get his king out of the way of his rooks. Black, for his part, has the same problem as in the previous example – his position is so cramped that he can’t emulate his opponent and transfer his king to the queenside. This means that White’s positional expansion on the kingside will lead to a direct attack against Black’s king. 31...g6 32.Kf1! White ignores the attack on his f5-knight, as 32...gxf5 33.exf5 just brings more white pieces into the attack. 32...Bc8 33.h5 Rab7 34.Rh2 Ra7 35.Ke1 Not for the first time we see one player conduct a kingwalk while his opponent literally can only move pieces back and forth.

35...Rab7 36.Kd1 Ra7 37.Kc2 Rab7 38.Kb3

White’s king has arrived at its destination, while Black can’t improve the position of his pieces. White now begins the final assault. What happens next isn’t pretty. 38...Ra7 39.Qg3 Rab7 40.f4 exf4 41.Bxf4 Ra7 42.Nd1 Reinforcements. 42...gxf5 43.exf5 Ng7 44.Ne3 h6 45.f6 1-0 Black couldn’t stand it any more. We now come to the Petrosian-Unzicker game mentioned earlier. There are many aspects of this game that make it a masterpiece, quite apart from White’s kingwalk, which is marked by a particular purity (White’s king moves straight along its first rank, from g1 to b1). One important difference from the previous examples is that there is play on both sides of the board, and Petrosian was masterful at combining kingside and queenside play.

Petrosian – Unzicker Hamburg 1960

White has a big edge, mainly because of his iron grip on the c-file. There is no immediate way for White to exploit his advantages on the queenside, however, and attempting to do so would likely result in Black being able to activate his pieces. With all this in mind, Petrosian concludes that a second front needs to be opened on the kingside, which leads to ... 29.Kf1! Kg8 30.h4 h5 This facilitates White’s later attack, but ultimately Black couldn’t prevent White from opening lines on the kingside. 31.R1c2 Kh7 32.Ke1 Kg8 33.Kd1 Kh7 34.Kc1 Kg8 35.Kb1 Kh7

Once again we see the passive defender reduced to moving a piece back and forth to mark time during his opponent’s kingwalk. The transfer of White’s king from g1 to b1 has improved his position, both because White’s king is safer on b1 if lines are opened on the kingside, and also because if lines open on the queenside it is better placed to exploit them and to reduce Black’s counterplay. 36.Qe2 Qb7 37.Rc1 Kg7 38.Qb5!? An interesting move, the discussion of which has nothing at all do with the subject of this book. Because White’s queen was on b5 just two moves earlier, going back with 38.Qb5!? might be said to be a “typical Petrosian move.” Actually it is the type of move any strong player might make. The first point is that the players are approaching the time control, so it is natural for White to repeat moves in order to defer his kingside breakthrough. This is a consideration not only if White was short of time, but also because in 1960 games could be adjourned and then analyzed in depth. The second point is that, rather than continue to conduct a passive and enervating defense, Black might have cut the Gordian Knot and traded queens. He would then have lost, because White’s passed b-pawn would cost him material: 38...Qxb5 39.axb5 a4 40.b6! Rad7 (40...Raa8 41.b7 Rab8

42.Na5 is no better) 41.Na5 Ra8 42.Rxd6! Rxd6 43.b7 Rb8 44.Rc8 Rd8 45.Rxd8 Rxd8 46.Nc6, and White ends up a piece ahead. There is a psychological aspect to 38.Qb5!? as well, quite apart from giving Black a losing option. By demonstrating his control of the position, White may accelerate his opponent’s demoralization. 38...Qa8 39.f4 Kh7 It is worth pointing out that Black can’t prevent the opening of lines on the kingside with ...f7-f5 (either now or later), because the resulting weaknesses along the sixth rank would prove fatal. For example, 39...f5 loses to 40.Rb6! Kf7 41.Rcc6, and Black’s position is hopeless whether or not he trades queens on b5. 40.Qe2 Qb7

41.g4 The strength of 41.g4 is increased by the unfortunate position of Black’s king on h7, but this doesn’t alter the fundamental aspects of the position. 41...hxg4 42.Qxg4 Qe7 43.h5 Qf6 44.Ka2!

The “!” is not only because White completes his kingwalk by reaching the edge of the board, but also because White wants to avoid a queen trade after 44.hxg6 Qxg6+. 44...Kg7 45.hxg6 Qxg6 46.Qh4 After pointing out that White has opened two lines against Black’s king, maintained his control of the c-file and his king is completely safe (all of which is true), Mednis adds “Such positions are always won” (which is definitely not true – better positions than this one have been thrown away). But there are no miracles today. 46...Be7 47.Qf2 Kf8 48.Nd2 Rb7 49.Nb3 To be a Petrosian fan, you have to accept such repetitions. 49...Ra7 50.Qh2! Bf6 Black succumbs to a mating attack after 50...Bd6 51.Rxd6! Rxd6 52.f5 exf5 53.Rc8+ Ke7 54.Qh8. 51.Rc8! Rad7 It is mate after 51...Rxc8 52.Rxc8+ Ke7 53.f5 Qxf5 54.Qb8 Rd7 55.Re8. 52.Nc5! b3+

There is nothing to be done: 52...Rxc8 53.Nxd7+ Ke7 54.Rxc8; 52...Rd6 53.f5 Qxf5 54.Qxd6+. 53.Kxb3 Rd6 54.f5! Rb6+ 55.Ka2 1-0 It’s fitting that the final move of the game is made by White’s king. The next example has some striking similarities to the previous one – White has an advantage, although not a decisive one, because of his control of the cfile and pressure on the light squares. While a kingside pawn advance is part of White’s strategic plan, it alone isn’t decisive, and the play ranges across the board. These similarities aren’t necessarily coincidental – Petrosian was again playing the white pieces...

Petrosian – Peters Lone Pine 1976

White’s advantage is tangible, but not overwhelming. Each of White’s pieces is better than its black counterpart, and Black has an annoying weakness on b5, but it’s far from clear how White can increase his advantage. 30.Kf1! Petrosian’s solution is to transfer his king to b1, both to allow him to advance his kingside pawns and to prepare for a king advance on the light squares. While this is going on, White continues to probe Black’s defenses, which gradually weaken almost of their own accord. 30...Re6 31.Qb5 Na7 32.Qb3 Nc6 33.h5 Ne7 34.Ke1 Nd5 35.Qb5 Nf6 36.Kd1 Nd5 37.Be5 Ne7 38.g4 Nc6 39.Bg3 Na7 40.Qb3 Nc6 41.Kc1 Re4 42.f3 Re3 43.Kb1 Ne7

White’s kingwalk is complete, and in the meantime, Black’s defensive prospects have deteriorated. White now seizes the opportunity to eliminate Black’s best defender of the light squares. 44.Bh4! Qd6 45.Bxe7 Rxe7 46.Rc8+ Kh7 47.Rf8! Qc7 48.f4 Bc5 49.Qd5

The fatal blow. The threatened 50.Qf5+ is more than just an abstract exploitation of Black’s light-square weaknesses. 49...Re5 50.Rxf7 1-0 Of the various winning moves, this is the simplest. The following example also has an easy flow, illustrating the kingwalk as part of a more general, positional plan involving a pawn advance on the kingside. This comes as no surprise, given that it involves an encounter between two positional heavyweights – Anatoly Karpov and Ulf Andersson.

This game could easily have been included Chapter 10 (double kingwalks), but White’s first (!) kingwalk fits in so well with the theme of this chapter, we will look at it now.

Karpov – Andersson London 1984

White’s advantage is very real, because Black’s weakness on b7 ties down his pieces. In addition, Black’s queenside pawn majority is crippled, while White’s corresponding kingside majority is mobile. This is all long-term stuff, mind you, and winning from this position is not a simple task, especially against a tough defender like Ulf Andersson. However, Karpov was near the height of his powers when this game was played, and he shows how it is done. White starts with a kingwalk, for the reasons that are by now familiar to us. White’s kingside pawns will ultimately have to advance, so White’s king leaves the danger zone. There is also a possible invasion route into Black’s position in a distant ending (a5-b6-c7). 24.Kf1 Ng7 25.Ke2 Rce7 26.Kd1 Qc8 27.Rab1 h5 28.h3 Ne6 29.h4 Kh7 30.Qh3 Qe8 31.Kc2 Rd7

White has nearly completed his kingwalk, and has also induced the advance of Black’s h7-pawn, giving him the possibility of a g2-g4 pawn break. But first things first! 32.Kb2 Ng7 33.Nf4 Rfe7 34.Ka1 Qf7 35.Rg1! At the moment White can’t make any progress on the queenside, but Black’s weakness on b7 won’t go away. 35...Ne6 36.Nd3 It’s important for White to avoid a trade of knights until he has weakened Black’s kingside defenses. 36...Ng7 37.g4! hxg4 38.Rxg4 Nh5 39.Rb1 Qe6 40.Qf3 Rg7 41.Rbg1 This is a good illustration of the “two weakness theory,” which holds that it is necessary to burden the defender with two weaknesses, rather than just one, before progress can be made. 41...Rde7 42.Kb2! Kh6 43.Kc3

White’s king is perfectly safe on c3, and it has more options on that square – including ... returning to the kingside! 43...Qf7 44.Nf4! Now the exchange of knights eliminates a key defender of Black’s king. 44...Nxf4 45.Rxf4 Re6 46.Kd2! A remarkable concept! White’s king returns to the kingside, in order to better position itself for the ending. It is clear that king safety will not be an issue for White. 46...Qe7 47.Ke2 Kh7 48.Kf1 Kh6 49.Rg3 Kh7 50.Rfg4 Qf7 51.Rf4 Kh6 52.Kg1 Kh7 53.Kh2

53...Kh6 54.Qg2 Kh7 55.Kg1

White’s king perambulations have almost taken on a life of their own. As is often the case, Karpov isn’t really making threats, but his opponent is also unable to defend against them. 55...Re8 56.Qf3 Rf8 57.Kf1 Qe7 58.Qd1 Qe8 59.Qb1 Kh6 60.Ke2

Back to the queenside? 60...Qd8 61.Rfg4 Rfg8 62.Kf1 Nope! Just taking a look around. 62...Qe8 63.Qd1 Qe6 64.Qf3 Rf7 65.Kg1 Rfg7 66.a3 Re7 67.Kh2

Back again. 67...Rf7 68.Rf4 Kh7 69.Qd1 Kh6 70.Qd3 Qe8

71.e4! After a suitably long period of tormenting his opponent, Karpov picks an opportune moment to open more lines. 71...dxe4 72.Rxe4 Qd7 72...Re7? is impossible because of 73.Qe3+. 73.Qe3+ Kh7 74.Re6 Rgg7 75.Rf3 f5 Black “cracks,” but it is unfair to criticize this move, given that Black has been under pressure for a very long time. 76.h5 gxh5 77.Qh6+ Kg8 78.Rfe3 Qc7+ 79.Kh3

79...Re7? 80.Rxe7 Rxe7 81.Qg6+ Kf8 82.Qf6+ Ke8 83.Qh8+ Kd7 84.Rxe7+ Kxe7 85.Qg7+ 1-0 White now picks up Black’s f5-pawn with check after 85...Kd8 86.Qf8+ Kd7 87.Qxf5+. By now the reader may well have gotten a misleading impression of how kingwalks in anticipation of action on a flank work in practice. It can fairly be said that in all six of the previous examples, things have gone smoothly for the winning side, with the only issue being how long the defender can hold out. Examples of this sort are instructive, but in most games the opponent has something to say about the outcome as well. We will finish this chapter with a somewhat different example, where the result of the kingwalk is far from clear.

Vadasz – Bronstein Tallinn 1977

This complicated position is roughly equal, despite (or perhaps because of) the material and positional imbalances. Bronstein decides that he can best improve his position by bringing his king closer to queenside, although a kingwalk of the type we are discussing seems like a distant dream. 41...Kg8 42.Qg5 Kf8 43.Kh2 Ke8 44.Qe5+ Hold it! White asks Black a serious question with this move – does he really want to go through with his plan? For the moment, at least, Black considers discretion the better part of valor. This is the way chess is actually played most of the time, by the way. A player may have an idea, but against a decent opponent there are often doubts, and back tracking and interruptions frequently occur. 44...Kf8 45.Qg5 Ke8

46.h5?! Kd7 47.Nd1?! This retreat is intended to free White’s e1-bishop from its defense of the c3pawn, but the knight was well-placed on e3 and should have stayed there. The engine recommends the logical 47.Qh6, to which Black has several reasonable replies, but not 47...Qxf6?, as White continues 48.Qxh7+ and 49.hxg6. 47...Kc8 48.Bg3?! Another imprecise move. White should first play 48.hxg6!, avoiding a weakness on h5. 48...gxh5 49.gxh5 Be8

50.Qf5+?

For neither the first nor last time, an unnecessary check spoils everything. 50.Qf4! was much better, as then Black’s king can’t cross the b-file without allowing Qb8+, while 50...Qxh5+ 51.Kg1 doesn’t lead to anything for Black. Once Black’s king gets to a6, the picture changes. 50...Kb7 51.Qf4 Ka6 52.Kg1

52...Rb6?! The engine’s recommendation of 52...Qe6! was stronger. The play at this point has all the signs of a time scramble. 53.Rf2 e3 54.Qxe3 Qxh5

55.Qe5?

It is sometimes true that the winner is the player who doesn’t make the last mistake. White had a surprising resource – after 55.f7! Bxf7 56.Nxb2! Bxb2 57.Qe7!, the “safe” position of Black’s king on a6 is suddenly shown to be anything but safe, as White has Qa3+ in many variations. This would have eliminated Black’s dangerous b2-pawn, solving White’s problems. In hyperdynamic positions like this, there often isn’t any safe square for the king. 55...Qg6 Attacking White’s b1-rook. 56.Rbxb2 Nxb2 57.Rg2 0-1

Chapter 2 Kingwalks in Anticipation of an Endgame Another category of “positional” kingwalks involves improving the position of the king in anticipation of an endgame. One principle that every chess player knows (or should know!) is that when the endgame is reached, kings play a much more important role. The clearest example is a king and pawn endgame, where all the other pieces have been exchanged. Unless a pawn can simply promote on its own, such endgames are decided by the kings. But the principle applies in minor piece and rook endgames as well, and often in queen endgames too. Of Reuben Fine’s fifteen “rules for the endgame” set out in his classic Basic Chess Endings, the pithiest and easiest to remember is, “The king is a strong piece. Use it!” Centralizing the king is often a priority once an endgame is reached, as from the center the king can go to either flank, or block the opponent’s king from doing the same thing, depending on the position. The kingwalks we are about to examine are directly connected to this principle. Kings move slowly compared to the other pieces, and if there is time to improve your king’s position prior to the endgame, this may make a big difference to the outcome of the game. It may also make it impossible for the opponent to go into at least some types of endgames, and limiting the opponent’s options is an important part of chess. The positions where this type of kingwalk is justified are harder to define than is the case for many of the other categories of kingwalks discussed in this book. Moving the king even one square closer to the action (usually the center) qualifies, but the more dramatic kingwalks we will be looking at could, in theory, put the king at risk.

We would therefore suggest that the following conditions must be met before a kingwalk in anticipation of an endgame can usefully be made: •

•

•

•

The transposition from the middle-game to the endgame must be a realistic possibility, which implies that some exchanges have already taken place. The repositioning of the king must be relevant, generally because a potential invasion route into the opponent’s position in the endgame either exists or is likely to exist. The time spent on the kingwalk cannot be better used in other ways, which implies a positional superiority by the side making the kingwalk. The kingwalk must not unduly expose the king to danger, giving the opposing player counterplay.

In this chapter, curiously enough, all the examples are from Petrosian, who was justly renowned for his kingwalks. This also illustrates that this type of kingwalk is actually quite common – you’ve probably done some yourself, without even realizing it.

Petrosian – Medina Palma de Mallorca 1968

White has a winning position, as, in addition to his positional advantages, he has an (extra) outside passed pawn on the a-file. Petrosian doesn’t rush, though. It’s clear that the only area of activity for White’s king will be the queenside, and so before doing anything else… 37.Kf2 Ng8 38.Ke2 Ne7 39.Nxe7 Bxe7 40.Kd3 Kg8 41.Kc2 Kf8 42.Kb3

With his king optimally placed, White now reorganizes his pieces so he can advance his a4-pawn until something stops it. 42…Qb7 43.Qa1 f6 44.Kc2

To unpin his b5-knight. 44…Qa6 45.Bc3 Bd8 46.a5 Ke7 47.Qa4 Bc7 48.Kb2 Kd8 49.Na3 Bb8 50.Qb5 Qc8 51.a6 1-0 Black had seen enough. White’s pieces, including his king, will just keep advancing. The next example is similar to the first, and the reader should have no difficulty in seeing Petrosian’s shorter kingwalk prior to the exchanges leading to the endgame. The biggest difference between the two is that White’s advantage in the second example is more slight – for one thing, material is equal.

Petrosian – Camara Nice 1974

For exactly the same reasons as in the previous example, White brings his king closer to the invasion route into Black’s queenside before doing anything else. 38.Kc2 Kd8 39.Kb3 Bxa4+ 40.Qxa4 Qxa4+ 41.Kxa4 Na7 42.Nb5

White now has a classic spatial advantage, and his hopes of breaking through rest on the pawn break c4-c5. 42…Nc8 43.Ne2 Kd7 44.Nc1 Be7 45.Nd3 Bf8 46.Nc3 Na7 47.Kb3 Nc8 48.Na4 Be7 49.c5 bxc5 50.bxc5 dxc5 51.Naxc5+ Nxc5+ 52.Nxc5+ Kd6 53.Na6 c6 54.Kc4 cxd5+ 55.exd5 1-0 (83)

As a result of his kingwalk, White’s king arrives in time to retain his advantage. There followed a long, difficult and, if one accepts the assessment of the engine, not particularly well-played endgame. White eventually won. For the remainder of the game, with Kotov’s at times suspect, pre-computer, notes, see Informant 17/657. The next example illustrates that this type of kingwalk is not as clear cut as some others discussed in this book. This is typical of maneuvers in “quiet positions” – improving the position of a piece (in this case the king) improves the position as a whole (provided the time spent isn’t better used by the opponent), but in itself that improvement may not be decisive.

Petrosian – Nikolaevsky Kiev 1957

After a sedate opening, White achieved a positional advantage by weakening his opponent’s light squares. Whether the position after the ensuing exchanges should be termed an “endgame” or a “queenless middlegame” is almost a matter of semantics. For our purposes, it is enough to see that once the queens and a pair of rooks are exchanged, Petrosian immediately brings his king into play, both to control the e4-square and so that his king is prepared to exploit any opportunities should the position simplify further. 25.Qxh7+ Kxh7 26.Rf3 Rf8 27.Rxf8 Bxf8 28.Ke2! Kg7 29.Bd2 Nf6 30.Ng3 Kf7 31.Ba5 Ke7 32.Kd3 Kd7

White’s king is ideally placed – as is Black’s, for that matter, although it doesn’t look nearly as impressive. White has retained his advantage, but should it be enough to win? 33.Rh1 Rb8 34.Bc3 Ng8 35.Ne4 Be7 36.f3 Ra8 37.Ra1 Bf8

38.cxb5 The engine dislikes this move, preferring 38.g4, after which White certainly has locked down some key light squares on the kingside. Just how White can exploit this is a different question. Petrosian, who obviously considered 38.g4, makes the practical decision, shortly before the time control is reached, to partially open the queenside. 38…axb5 39.a4 b4 40.Bd2 Ne7

Attacking White’s d5-pawn. The engine judges this position as equal, but it is much more pleasant to play the white pieces. 41.Nf6+ Kd8 42.Ke4 White’s king can’t get into Black’s kingside, provided Black’s e7-knight continues to guard the f5-square 42…Bg7 43.Ng4 Kc7 44.Ne3 Rf8 Preventing 45.Nf5. 45.g3 Rf7? Black falters. 45…Ng8!, with the idea of 46…Nf6+, was right. 46.Rc1 Kd7 47.a5! Not so much to queen, as to distract Black from the kingside. 47…Rf8 48.a6 Ra8 49.Ra1 Ra7 50.Ra5 Kc7 51.Nc4 Nc8 52.Be3 Nb6?

With every piece on its best square, White now breaks through. 53.Nxd6! Kxd6 54.Rxc5 Rc7 55.Rxc7 Kxc7 56.d6+ Kc6 57.Bxb6 Kxb6 58.Kf5 The culmination of White’s strategy.

58…Bf8 59.Kxe5 Kxa6 60.Ke6 1-0 Petrosian always made it look easy. But Black had his chance on move 45… The next example illustrates two additional points. One is that chess games are usually a fight, and objective analysis shows that things don’t always unfold smoothly in favor of the eventual winner. The other point is that Petrosian’s games can be hard to understand!

Petrosian – Reshevsky Lugano 1968

This example starts with a very odd move for a kingwalk. 25.0-0-0 This move wouldn’t be worth a comment in any other book. White castles, so he can bring his rook to h1. Interestingly, the engine recommends a more direct approach. So why did Petrosian castle? There is no obvious answer to this question. White might have changed his mind, deciding after playing 25.0-0-0 that his king wasn’t well placed on the queenside. But we’re talking about a former world champion here, not a club player, so that’s a pretty unlikely explanation. It’s more likely that White didn’t want to commit his king to the kingside until the configuration of Black’s pieces became more settled. But there is a third possibility – White had a kingwalk in mind when he castled, knowing it would confuse his opponent. Since it confused your authors, we are going with this reasoning. 25…Nf6 26.Qe2 Kd7 27.Rh1 Ng4 28.Nf3 Re8

29.Kd2!? White begins his kingwalk, which objectively makes perfect sense, other than that only a few ago moves the king was even closer to the kingside. The engine suggests this very move, which gives you a good idea of just how subtle our silicon friends have become. 29…Rb8 30.Ke1 b5 31.Kf1 a5 32.Kg2

Now White’s king is where it wants to be. It is not only completely safe, and defends White’s f2-pawn – it is in the best possible position to exploit Black’s weak h6-pawn. 32…a4 Locking up the queenside seems odd at first, but after 32…b4? 33.axb4 axb4 34.Ra1, Black has only succeeded in giving White a winning attack down the

a-file. 33.Ne5+!? With Reshevsky in his inevitable time trouble, Petrosian decides to go into an endgame. This turns out to be the correct decision, but objectively the position is equal. 33…Nxe5 34.Qxe5 Qxe5 35.fxe5 Rh8 36.f3 Ke7 37.g4 fxg4 38.fxg4

38…Rg8?! 39.Kh3 39.Rxh6 Rxg4+ 40.Kf3 Rg7 41.Rh8 gave better chances, but even after White wins Black’s b5-pawn, things are far from clear. 39…Rf8? This is Black’s fatal mistake. After 39…Kf8! 40.Rf1+ Kg7 41.Rf6 Rf8!, the king and pawn ending that arises after 42.Rxf8 Kxf8 is drawn, while 42.Rxe6 Rf2 gives Black strong counterplay. 40.Kg3 Rh8 41.Kh4 Rg8 42.Rf1 1-0 The final, exceptionally interesting and instructive, example in this chapter could have been put in several other places in this book, as White’s kingwalk(s), both in the game and in the notes, fall into several categories, some of them overlapping. Not only that – it wasn’t Petrosian, but rather the

Hungarian super-grandmaster Lajos Portisch, who was making the kingwalks.

Portisch – Petrosian Manila 1974

Let’s start with a short kingwalk by Black, which is almost (but not quite) an example of “castling by hand” (see chapter 8). 16…Kf8 17.Nec3 Kg8 18.b3 f5

Now it’s White’s turn. He undertakes a pawn advance on the queenside, and brings his king to b3. 19.Kb2 Nf7 20.a3 Bf6 21.Bf2 Rh7 22.b4 Rg7 23.Kb3 Be7 24.Qe2 Qc8 25.Be3 Qb7 26.Bd2 Bf6 27.Rb1 Kh7 28.Qd3 Kh8 29.Ne2 axb4 30.axb4 The engine assesses this position as being somewhat better for Black, but it soon changes its mind.

30…Rgg8 31.Nec3 f4 32.Ra1 Rgc8 33.Qe2 Rxa1 34.Rxa1 Bxh4 35.Ra7 Qb8

Portisch has sacrificed his weak h4-pawn in order to make progress on the queenside. Eventually Black’s extra kingside pawn will give him counterplay, so the question is whether White can accomplish anything in the meantime. 36.Qf1 Be7 37.Qa1 Nd8 38.Be1 Kg7 39.Bf2 Kf8 40.Qa6 Be8 41.Na3 g5 42.Nab5 h4 43.Qa1 Nf7?! Black’s d8-knight, while passive, was covering some important squares (b7 and c6). Petrosian opts for active defense, but this gives White his opportunity. 44.Ra6 Bd7 45.Na7 Re8 46.Nc6 Qc8 47.Ra7 Bxc6 48.dxc6 White now has a decisive advantage, as his queenside play is more important (and faster!) than Black’s kingside pawn attack. But to demonstrate this, ultimately another kingwalk is required! 48…g4

49.fxg4? White lets himself be distracted. The immediate 49.Nd5! was right, and if neither side blinks with 49…gxf3 50.Nxc7 fxg2 51.Nxe8, White wins, as 51…Qxc6 52.Nc7 Qxe4 53.Ne6+ Kg8 54.Qd1! covers all the key squares, and Black’s three connected, passed pawns on the kingside aren’t as important as White’s extra rook and the related threats to Black’s king. If Black responds to 49.Nd5! with the natural 49…Bd8, he loses more prosaically after 50.Ra8 Qe6 51.Bxh4!, collapsing the defense of the c7pawn. 49…Qxg4 50.Qh1 Bd8 51.Ka4!? White continues to press. Portisch’s idea is to prepare 53.Nd5 by bringing his king to b5, as the immediate 51.Nd5 gives Black counterplay after 51…Qe2. The engine now gives the variation 52…Kg8 52.Kb5 d5!? as equal. 51…Re6? All this must have occurred during a time scramble, as the second time control approaches. 52.Nd5 Qe2

The key moment of the game, as it turns out. White can now win with an astounding move, 53.Kb5!, defending White’s c4-pawn. After 53…Qxf2 54.Qh3!, White has an overwhelming position, because his king is perfectly safe and White’s active pieces and advanced cpawn are more important than Black’s sizable material advantage. For example, White wins convincingly after 54…Rg6 55.Nxc7 Bxc7 56.Rxc7 Kg7 57.Rxf7+!, and White gets a second queen. This is an excellent illustration of how one type of kingwalk (preparation for an ending) can morph into another (escaping up the board to safety – the subject of chapter 7). Of course there was no way that even a player of Portisch’s power could have anticipated this when he brought his king to b3, and in fact he missed this opportunity in the game. 53.Nxc7? This is actually a blunder that changes the assessment of the game completely. 53…Bxc7? Black returns the favor! 53…Re7! 54.Bxb6 Qxc4 was crushing, as White’s entire position falls apart. 54.Rxc7 Qxc4

Now the game ends in a not-so-peaceful draw. A fair result, if there is such a thing in chess. 55.Qxh4 Qa6+ 56.Kb3 Qd3+ 57.Kb2 ½-½

Chapter 3 Kingwalks to Defend Key Points We feel we should apologize at the outset of this chapter, because it covers one of the more mundane and least dramatic types of kingwalks – moving the king to defend a key point. But sometimes His Majesty just has to pitch in and help, performing defensive tasks just like any other, lower-born, piece. If one uses a broad definition, it would be difficult to find a chess player who has not conducted this type of kingwalk – probably without even realizing it. The king, after all, defends not only all the friendly pieces and pawns next to it, but also the squares. Given that kingwalks to defend key points can be thought of almost as a standard technique, we have tried to present a relatively small number of examples. The reader should keep in mind, however, that the fact that kingwalks to defend key points are common is strong evidence that such kingwalks are also an important weapon that every player should use. The question, as is almost always the case when discussing kingwalks, is just how far one can take this concept. The chapter on unsuccessful kingwalks will be one of the longest in this book… Our first example is from a classic, and justly famous, heavyweight encounter between the young Sammy Reshevsky and the mighty Capablanca. The kingwalk in question appears only in the notes, which shows that maneuvers of this sort are by no means self-evident.

Reshevsky – Capablanca Margate 1935

After an eventful and tense middlegame, in which both sides missed chances, White has a definite positional advantage, because of Black’s weak d5-pawn. Having turned down a draw offer on the previous move, White now felt obliged to do something active, despite his customary time trouble. 39.Kh1?! Reshevsky condemned this move, recommending the kingwalk Kg1-f2-e2d2-c2, to defend White’s c3-knight. This would allow White to continue with Rb4-b5, winning Black’s d5-pawn, without being concerned about the reply …Qa3!, creating counterplay on the queenside. This is easy to visualize from the diagrammed position, because Black can’t do anything until White tries to break into his position. Leave Black’s position as it is and put White’s king on c2, and the strength of the kingwalk is apparent. 39…Bd8 40.g4? The “?” is from Kasparov, who analyzes this game in Volume IV of his Great Predecessor series and terms 40.g4? “reckless.” Let’s follow the game to its conclusion and see why.

40…fxg4 41.hxg4 Qd6 42.Kg1 Bc7 43.Kf2 Rf7 Threatening 44…g5. The unfortunate position of White’s king is obvious. 44.g5 Bd8 45.Ke2?!

White is carrying out something resembling the previously mentioned kingwalk, but now each tempo counts and Black has serious counterplay. Kasparov makes no comment on 45.Ke2?!, but the engine points out that 45.Qg4! would have both defended against the threatened 45…Bxg5 and prevented Black from transferring his queen to the h3-c8 diagonal. 45…Bxg5?! Reshevsky terms this move “desperate.” Actually, it is worse than both 45… Qe6!, followed by 46…Qf5 looking for counterplay on the d3-square, or Kasparov’s 45…Rf5 46.Qh3 Qd7 47.Rh1 Kg8, covering all the bases. 46.Rxb6 Qa3 47.Kd2 White’s king arrives in the nick of time. 47…Be7 48.Rb7

48…Rxa4? A fatal mistake, understandable even for a Capablanca after a difficult defense. Kasparov gives the following astounding variation: 48…Rf5 49.Rd7 Rh5 50.Qg2 Kh6! (Black’s king takes a short walk of his own, to unpin his e7-bishop) 51.Rbb7 Rb5!! 52.Rxb5 Bb4 53.Ke2 Qxc3, and Black’s counterplay is sufficient for a draw. An absolutely amazing find.

Position after 53…Qxc3 (analysis) 49.Qxd5! A powerful refutation of Black’s last move. Black’s d5-pawn is much more important than his a4-rook, the capture of which loses immediately in any event (49.Nxa4? Qd3+ 50.Kc1 Ba3+ 51.R7b2 c3). As many fine authors have noted, knights can be invaluable defenders of their monarch.

49…Ra5 50.Qxc4 Rh5 51.Kd3

White’s king has completed its journey from h1 to d3, where it is completely safe. 51…Qa8 52.Qe6 Qa3 53.Rd7 Rhf5 54.Rb3 Qa1 55.Rxe7 Qf1+ 56.Kd2 1-0 Things turned out well for White in the end, but he could have saved himself a lot of trouble had his king gone to c2 before the excitement started! In the next example, Petrosian comes up with a very deep repositioning of his king, but after an eventful time scramble ultimately fails to bring home the point. The tactical possibilities, which to some extent were generated by White’s kingwalk, reveal both the richness of chess and the level of precomputer engine commentary.

Petrosian – Boleslavsky Moscow 1949

White has a significant positional advantage, because of his spatial advantage and better minor pieces. Black’s d5-bishop is a tower of strength, but the same can’t be said of his forlorn g7-bishop, which is seriously restricted by White’s powerful d4-e5-f4 pawn wedge. Petrosian now begins a short, but very interesting, kingwalk, in order to not only defend some important squares on the kingside, but to prepare an attack against Black’s king. Before we get too carried away, however, it is worth pointing out that the cold-blooded engine recommends the materialistic 24.Bd2!?, in order to exploit the other main flaw in Black’s position – his disrupted queenside pawn structure. It should come as no surprise that strategic maneuvers such as kingwalks must often contend with completely different, but equally or more compelling, ideas. 24.Kh2 Kh8 Probably not the most precise reply, but 24…gxh3 hardly solves Black’s problems. 25.hxg4! fxg4 26.Kg3

This was White’s idea. White’s king defends the f4-pawn, without creating the weaknesses that would arise after g2-g3; blockades (and attacks) Black’s g4-pawn; and clears White’s first rank and the h-file so that White’s rooks can operate without impediment. We think that explains why Petrosian found the maneuver so attractive. 26…Rf8 27.Rh1 Rf5 28.Rh4! One commentator recommended 28.Qc3 instead of 28.Rh4, but Petrosian’s judgment was correct. 28…h5 29.Rch1 Bf7

30.Qc3 Curiously, the engine recommends 30.Kh2, followed by 31.Kg1 – walking White’s king back to where it started, now that White’s rooks are on the h-

file! One thing you have to say about computers – they play “without prejudice,” as they say. 30…Bh6

Unpinning Black’s h5-pawn, to prevent 31.Rxg4. 31.Rf1? The young Petrosian loses the thread of the game (to be fair, his ascension to the chess throne was nearly 15 years in the future). 31.Qxa5? was bad because of the reply 31…Rxd4!, but 31.e6! was strong, as the recommended 31…Qc7!? was not to be feared. Rather than 32.exf7? Rxf4!, White can answer 31…Qc7!? with 32.Rxg4! hxg4 33.exf7, because Black’s h6-bishop is both pinned and hanging with check. If Black replies to 31.e6! with 31…Bg6, then White takes on g4 anyway, as after 32.Rxg4! hxg4 33.Rxh6+ Kg7 34.Rh4, Black must spend a move to prevent White from playing d4-d5+, which allows White to take another important pawn (34…Rdd5 35.Rxg4, and White stands considerably better). The calm 31.Qd2 was also possible. After 31.Rf1?, White should lose most of his advantage. 31…Qc7? 31…Rf8! was much better.

32.Qc2?! 32.Ne4! would have exploited Black’s mistake more convincingly. 32…e6 33.Qc4 Again, 33.Ne4! was stronger, although White is still on top. This must have been a time scramble. 33…Bf8?!

34.Rfh1? Amazingly, the tactics work after 34.Nxe6! – a move that only a beginner or a computer would play: 34…Qd7 35.Nxf8; 34…Qe7 35.Nxf8 Bxc4 36.Ng6+ and 37.Nxe7; 34…Qc8 35.d5!). 34…Kg8? 34…Bxc5! would have eliminated most of the danger. 35.Qe2? White misses his last chance. 35.Rxh5! Rxh5 (35…Bxh5 36.Nxe6) 36.Rxh5 Bxh5 37.Nxe6, and Black will be buried under an avalanche of white pawns. 35…Bxc5 36.dxc5 Rd5 37.Qxg4+ ½-½

To avoid falling into a disadvantageous position, White forces a draw by perpetual check: 37…hxg4 38.Rh8+. A game of missed opportunities, thanks to the potent mix of stress and the chess clock. The next example is much simpler and smoother than the PetrosianBoleslavsky game – and less nerve wracking as well. The fact it comes from a Karpov game is perhaps not entirely coincidental.

Karpov – Kuzmin Leningrad 1973

After overly passive opening play by Black, Karpov has obtained a “typical advantage,” with more space and greater piece harmony as the most prominent features. Karpov now expands on the kingside, further restricting Black’s d7-knight and h6-rook. The engine offers other, more direct, suggestions, but Karpov has his own ideas. 27.g4! Qc7 28.g5 Rh8 Black’s rook had to go back or sideways, and neither alternative is particularly attractive. After 28…Re6 29.Rxe6 fxe6 30.Kg3, Black’s kingside is an accident waiting to happen. 29.Kg3!

White’s king defends his f4-pawn and clears the second rank for his e2-rook, leaving Black to figure out how to organize his defenses. A spectator walking by the position might wonder why Black didn’t castle, but in fact he did, on move ten. Of course, with the h-file open, it wouldn’t help Black much to have his h8-rook in the center. Ironically, Black was the one who opened the h-file in the first place, to weaken White’s king position. The rest of the game is an effortless Karpovian exploitation of White’s advantage. 29…Nc5

30.Bf5 Karpov was fond of geometric moves, and seeks to dominate Black’s c5knight with his f5-bishop.

Interestingly, the engine suggests 30.Rh2!?, with the idea of 30…Rxh2 31.Qxh2 Ne6 32.Kg4!, reinforcing the defense of White’s f4-pawn by unblocking the diagonal for White’s queen. White’s kingwalk would then have gone one square further up the board, although Black could also suffer by declining the exchange of rooks (30…Rg8 31.g6!, with a strong attack). 30…g6 31.b4! Ne4+ Retreating with 31…Nd7 is no better, as White breaks through with 32.Bxg6 fxg6 33.Qxg6. 32.Bxe4 dxe4 33.Qxe4 Kg7 34.b5 Na5 35.Qe7! With control of the e-file and a superior knight, Karpov simply exchanges Black’s only good piece. 35…Qxe7 36.Rxe7

36…Rd3 37.Rc7 Nb3 38.Kg4 Rf8 39.Ree7 1-0 Black’s f7-pawn will soon fall, leaving him with nothing to play for. The next example illustrates the role the king can play in consolidating an advantage in the middle game, but with an eye to the endgame.

Petrosian – Butnorius Riga 1975

White has a healthy extra pawn, but his position is airy and his rooks aren’t connected. Petrosian solves both problems with a short kingwalk that the reader is sure to have spotted. 23.Kf2! g5 24.Ke3 Is this going too far? White retains an advantage with 24.Ke3, but the question of whether White’s king is defending key points in the center, as opposed to simply being exposed, is a bit in doubt. 24…Qf7?! This helps resolve the question. 25.Rhf1 g4 26.Rfd1 With the kingside closed, White’s king looks more comfortable on e3. Of course, White still has to be careful not to advance his central pawns too soon. But who better than Petrosian when patience is required? 26…Rfe8 27.c4 Re6 28.Qb5 Rh6 29.Rd2 Rh3 30.d5 h5 31.Qb4 h4 32.Rg1 Kg7

Black is doing his best to open the kingside and create counterplay against White’s king. But soon the queens will come off and White’s centralized king will be ideally placed for the endgame. 33.e6 Qf6 34.Qb2 hxg3 35.hxg3 Qxb2 36.Rxb2 Kf6 37.Kd4

37…Rdh8 38.Re2 R3h7 39.c5 Re8 40.Kc4 Rhh8 41.Rge1 bxc5 42.Kxc5 Rh3 43.Rg2 Reh8 44.Ree2 R3h7 45.Kc6 1-0 The next example is radically different from the previous games in this chapter. It still features a kingwalk to defend key points, but this time the king’s life hangs by a thread. And no wonder – the game is a titanic clash between arguably the greatest attacking and greatest defending players of all time.

Kasparov was deeply impressed by Petrosian’s play in this game and annotated the game in detail in Volume III of My Great Predecessors.

Kasparov – Petrosian Tilburg 1981

White has just played 30.a4, and was looking forward to locking down the queenside with 31.a5, then tripling on the b-file with 32.Rcb2 and 33.Qb1, collapsing Black’s b7-pawn. Petrosian reacts with a startling move. 30…b5!? Kasparov writes: “This move staggered me.” We are all taught that moving the pawns in front of the king weakens its defenses (as indeed it does), so what does Petrosian have in mind? 31.axb5 cxb5 32.Ra2! Kb7! And here Kasparov writes “But this move staggered me even more: can such play really go unpunished?” That remained to be seen at this point, but Petrosian considered a kingwalk to be clearly preferable to the obliging 32…Bd6 33.Rxb5 Rxb5 34.Nxd6 Qxd6 35.Qxb5. 33.Bb4?

33.Na3! wins, as the detailed analysis in My Great Predecessors shows. But 33.Bb4? shouldn’t lose for White… 33…Qe8! 34.Bd6 Kasparov endorses Petrosian’s suggested 34.Nd6+, but the engine is not impressed. 34…Ra8 35.Qb1? Kasparov suggests 35.Rb1, while the engine recommends 35.Ba3. Whatever the exact truth might be, the move played is a mistake.

35…Kc6!!

Here is it worth quoting Kasparov in full:

“A fantastic defense! This move, which Petrosian made instantly, threw me into complete confusion: how is it possible to move the king forward with a board full of pieces? After Steinitz, who had done such a thing? The psychological effect of the 30…b5!? thrust and the king march …Ka7-b7-c6 was so strong, that I was unable to gather my thoughts and I quickly lost.” As will be clear to the reader by the time he or she has read this entire book, others besides Steinitz had done such a thing, but the psychological effect of a well-timed kingwalk can’t be over emphasized. We are talking about Garry Kasparov here, after all. 36.Rba3? White can escape into a drawn ending with 36.Bxc7. 36…bxc4 37.Rxa6+ Rxa6 38.Rxa6+ Bb6 39.Bc5 Qd8 40.Qa1 Nxc5 41.dxc5 Kxc5 42.Ra4 0-1

The final example in this chapter is another kingwalk from Petrosian, against Karpov, a no less formidable opponent than Kasparov (or, at least, only slightly less formidable). In the starting position, Petrosian is in trouble, as he is unable to castle and his center is under pressure.

Petrosian – Karpov Moscow 1981

White has just erred by exchanging rooks on b5, and now finds his king trapped in the center of the board. Petrosian’s reasoning is simple – if his king is going to be stuck in the middle of the board, at least it should try to do something useful. 21.Kd2!? You can be sure that White wasn’t thrilled with this move, but the engine confirms that it’s White’s best chance. 21…c5! 22.Ke3!?

In for a penny, in for a pound. White at least doesn’t fear 22…cxd4+ 23.Nxd4, as he can answer 23…Qg5+ with 24.f4. Karpov of course plays with more finesse. 22…Rd5 23.Rd1 g5!? 24.h3 It isn’t easy finding a move for White. He isn’t getting checkmated, and in fact his king isn’t really being attacked, but Black has a lot of pressure on White’s center. 24…Qc7 The engine recommends 24…Ba6. 25.Qb2 a5 26.a3

26…cxd4+?! This lets White off the hook. Belov’s recommendation of 26…f5! leaves Black with a big advantage. 27.Rxd4 Rc5

28.Kd2! Covering the invasion squares on the c-file. 28…Bf1! A nice reminder that White’s king can’t be everywhere! 29.b4 axb4 30.axb4 Rb5

31.Rg4 Belov gives 31.Rg4 an exclamation mark and awards White a slight advantage, but the ruthless engine is having none of it, and gives Black a near-decisive edge after 31…Qb7!, threatening 32…Bxg2! 33.Rxg2 Qxf3. 31…Rd5+?! 32.Ke1

White’s king is back where it started. 32…Bd3 33.h4 Bf5 34.Rd4 Rxd4 35.Qxd4 Qc2 36.hxg5 Bd3 ½-½ Draw agreed, as one side or the other will give perpetual check (37.Qe3 Qb1+ 38.Kd2 Qc2+; 37.Qd8+ Kh7 38.g6+!).

Chapter 4 Kingwalks to Attack Key Points or Pieces It’s time to switch from “defensive” kingwalks to “aggressive” kingwalks, in which the king joins the battle in support of his army. Before switching to mating attacks in the next chapter, we will deal with kingwalks which see the king attacking and capturing a trapped piece, or at least threatening to do so. As should be clear already, it is not all that easy to categorize kingwalks, because often they have several purposes. A king that moves forward to attack might also be moving to a safer area of the board, while at the same time it is getting ready for the endgame. A good example of the multipurpose aspect of kingwalks is Hao-Bologan, the final game in this chapter. If attacking kingwalks are overlooked, it is only because players are used to kings remaining passive, especially in the middlegame. And rightly so, as we will see in the later chapter on failed kingwalks. But once players think of the king as a potentially powerful piece, much of the difficulty disappears. The first example in this chapter sees the debut of Wilhelm Steinitz, the first official world champion and a name that is often associated with kingwalks. In fact, it isn’t that easy to find Steinitz games featuring kingwalks, other than his notorious variation of the King’s Gambit in which White’s king rockets to the third rank before the players have managed to get comfortable in their seats. But, curiously, here are two Steinitz games that end in almost the same way.

Steinitz – De Vere London 1866

White has an edge in this endgame, thanks to his better minor piece and more harmonious rooks. The first thing to note is that White’s a2-pawn is immune from capture: both 32…Rxa2 and 32…Nxa2 are answered by 33.c3, and Black loses a piece. After the natural 32…d5 33.c3, while White would retain his advantage, but there would be no clear path to victory. However, Black conceives of an idea – to cut off the escape route of White’s b5-bishop and then capture it with his a7-rook. There are only two problems with this plan. One problem is that the plan might fail. The other problem is that the plan might succeed. 32…c4? 33.Rgd1 Logical, but White had better: The tactically alert 33.Re2! won a pawn, as 33…Rxe2 is met by 34.Bxc4+ and 35.Bxe2, while 33…d5 fails to 34.Rxe4+ dxe4 35.a3!, followed by 36.Bxc4. 33…d5 34.a3

This was Steinitz’s idea. 34…Ra5? Black must resist with 34…c3! 35.Rd4 (35.bxc3?! Rxa3! 36.cxb4? Rxb4+, and Black checkmates!) 35…cxb2 36.Kxb2 Nxc2 37.Kxc2 Rxa3, leaving White with only his fragile f2-pawn. 35.axb4 This leads to a pretty (and thematic) finish, thanks to Black being oblivious to White’s idea. The engine instantly reveals an easier win, however. After 35.Bd7+! Kxd7 36.axb4 Ra7 37.Rxd5+ Ke8 38.R5d4, it is difficult for Black to resist. 35…Rxb5 36.c3

36…Ke5?! Of course Black should play 36…d4, giving up a pawn to free his imprisoned b5-rook. 37.Ka2! Kf4? From bad to worse. This was Black’s last chance for 37…d4. 38.Rd4! Kf3 39.Ka3 1-0

At this point Black realized what White was up to – White’s Kb1-a2-a3a4xb5 kingwalk will cost Black an entire rook. The next example is rather simple, but simply had to be included because of the striking resemblance to the previous game.

Steinitz – Tinsley Hastings 1895

Black, having boldly essayed a Benoni Defense well ahead of its time, actually outplayed Steinitz for the first 25 moves, before dropping his crucial a6-pawn without any compensation. This was no mean feat, as Steinitz had only lost his world championship title the previous year. Move 25 was a long time ago, of course, as Black is completely lost in this position. The way he loses, though, gives us an amusingly short example. 40…Rb5? 41.b4! f5 42.Kd3! 1-0

To borrow a lovely expression from Nimzowitsch, White’s king “travels by express,” as it attacks Black’s d4-bishop on its way to c4, after which it will capture Black’s trapped b5-rook.

The reader will remember the previous example, where Steinitz also captured a forlorn black rook on b5. We can be sure that Steinitz hadn’t forgotten his earlier game! The next example has a similar scheme (a kingwalk in a simplified position leading to the win of a rook), but on a much grander scale.

Alekhine – Yates London 1922

White has a clear advantage, as he dominates the dark squares and his e5knight is much stronger than Black’s a6-bishop. Black should exchange a pair of rooks, which would at least avoid what follows. 21…Reb8 22.f3! b3 23.a3 h6 24.Kf2! White activates his king, since this is an endgame, or close to it. But 24.Kf2! also is the start of what turns out to be one of the most striking kingwalks in chess history. 24…Kh7 25.h4 Rf8 26.Kg3 Rfb8 A common, but not universal, theme when it comes to kingwalks. Black has no play and can only mark time. White now improves the position of his rooks before continuing to advance his king. 27.Rc7 Bb5 28.R1c5 Ba6 29.R5c6 Re8 30.Kf4

White could bring his knight to the queenside, winning Black’s a4-pawn, but he has something else in mind. Patience is also a virtue in the endgame – Black’s a4-pawn isn’t going anywhere. 30…Kg8 31.h5! Bf1 32.g3 Ba6 33.Rf7! White doubles rooks on the seventh rank, setting up the decisive combination. 33…Kh7 34.Rcc7 Rg8

35.Nd7! With the threat of 36.Nf6+, winning the exchange. 35…Kh8 36.Nf6!

White’s knight goes to f6 anyway! 36…gxf6 is obviously impossible, because of 37.Rh7 mate, but doesn’t Black have a way to resist? 36…Rgf8!? 37.Rxg7! Rxf6 38.Ke5! 1-0

Another kingwalk, another rook bites the dust. Black can neither retreat his rook to f8, nor defend it with his other rook, as White then has 39.Rh7+ Kg8 40.Rcg7 mate. Since Black’s f6-rook has no other safe squares, it is lost, after which White quickly checkmates anyway. Of all the ways to win, Alekhine found the most elegant and memorable. The next example – another Petrosian game – is also justly famous, and is a very good example of a king wreaking havoc, this time in a middlegame.

Dückstein – Petrosian Varna 1962

18…a5! Faced with a pawn advance against his king, Petrosian strikes at the enemy pawn formation by attacking the base of White’s b4-c5 pawn chain. We have seen this sort of “active defense” before, in the famous Kasparov-Petrosian game. This is one of several striking features of this game, not the least of which is the creative level of Petrosian’s play in an early round of the chess Olympiad, when dry, technical play would likely have been enough to defeat an overmatched opponent. From our point of view, 18…a5! also prepares a dramatic kingwalk. 19.Rad1 Rxd1 20.Rxd1 Rh4!? 21.bxa5 Bxc5 22.a6?! 22.Qc2! would have confronted Black with some difficult problems. 22…b6 23.Re1?

23…Ka7 The start of the kingwalk. The engine recommends the more accurate 23… Nd4, followed by 24…Ka7. 24.Be5 Qd7 25.Ne4 Bd4!? 26.g3

26…Bxe5!? Petrosian imaginatively, and unnecessarily, sacrifices the exchange. But to a player like Petrosian, who placed a special emphasis on pawn structure, the temptation of inflicting two sets of doubled, isolated rook pawns on his opponent was too much to resist. 27.gxh4 Nd4 28.Qd1 Qd5 29.Re3?! 29.Qd3 was better, but by repeating Petrosian allows this anyway.

29…Nf5 30.Re1 Nd4?! 31.Qd3! f5 32.Ng5? White’s knight should say in the center, with 32.Nd2! 32…c5 33.Re3 c4 34.Qd1

Black had made considerable progress, thanks to White’s g5-knight, which is on the wrong side of the board. Both Black’s king and his dangerous passed c4-pawn are on the queenside, and White’s knight will soon be missed. Black now captures White’s a6-pawn, activating his king. 34…Kxa6 35.Ra3 Bf6 36.h3 f4 37.Qg4

White defends his g5-knight and creates the possibility of Qc8+, but now most of White’s army is on the kingside. Petrosian reacts immediately. 37…Ka5!

Black’s kingwalk gathers pace. 38.Nf3 Kb4!

White’s a3-rook is overloaded, as it can’t be defended, can’t move along the third rank, and can’t retreat without allowing 39…Nxf3+. 39.Nxd4 Kxa3 40.Nc2+ Kxa4 0-1

Black’s king ends its walk by destroying the last remnant of White’s queenside. Black’s Kb8-a7xa6-a5-b4xa3xa4 kingwalk leaves a vivid impression. The next example ends tragically. White’s kingwalk works and he wins a pawn, but he forgets that a king on the fourth rank has its drawbacks.

Sax – Castro Budapest 1977

White has a comfortable advantage, and would be clearly winning after the natural 20.Bc1, followed by an eventual b2-b3, driving away Black’s c4knight. Instead, White spies a chance to set a trap in order to win material, with his king playing the leading role. Black falls for the trap, but tragedy strikes later when White fails to follow up correctly. 20.g4

20…Qe7? After 20…Nxd2 21.Qxd2 Ne7, the position is equal.

21.Nxd5! Nh4+ 22.Kg3 exd5 23.Kxh4

So here we are. White has an extra, passed e5-pawn, and “only has to consolidate.” However, this proves to be difficult, as White’s king is on the fourth rank. 23…Qe6 24.b3 Nxd2 25.Qxd2 Rf7 26.Rxf7 Qxf7 27.Re1 Rf8 28.e6?! White has convinced himself that his king is safe on h4, or he has just forgotten that it might not be. At any point he could have withdrawn the king to g3, where it would be safer. 28…Qe7

29.c4?! dxc4 30.bxc4 h6!

Obviously White saw this possibility, and was counting on his next move to preserve his advantage. 31.Re5 Rf3!

A strong move, preventing White’s king from retreating. Objectively this position is about equal, but White can hardly accept this assessment. 32.Rf5 Ra3! 33.d5

33…a5? 33…g6! wins immediately. We have to assume mutual time trouble at this point. 34.Qb2 Re3 35.Qc1 Rd3 36.Qf1 Rc3 37.Re5?

White had to play the sad 37.Rf3, giving back his extra pawn after 37… Qxg5+. 37…hxg5+ 38.Rxg5 g6!

With the terrible threat of 39…Qh7+. 39.Qe1 Rd3 40.Qe4? Qh7+ 41.Rh5 g5+! 0-1 Not the first or last kingwalk to go wrong. We will have an entire chapter on them. Our final example has a much happier outcome, in which White’s king fearlessly advances up the board to support a pawn advance, while at the same time attacking Black’s weak pawns. Hao’s kingwalk could have gone in any one of several chapters, but there wasn’t any doubt that it had to appear somewhere in this book!

Hao – Bologan Bilbao 2014

White has, at most, a very slight advantage. White now finds an inspired idea. He walks his king up the board, along the h-file, both to tie Black’s queen to the defense of Black’s king, and to support a pawn breakthrough on the kingside. White doesn’t threaten checkmate, but his active king sets Black’s annoying problems, which he fails to solve. 29.Kg3! Rb6 30.Kh4 Rd6 31.Kh5

Now White has a threat, or threatens to have a threat. After 31.Kh6, White’s queen invades, so Black is obliged to retreat with his own queen, so he can answer 31.Kh6 with 31…Qf8+.

31…Qb8 32.Qe2 Kh8 33.h4 The second part of White’s plan. 33…Qg8? 34.Rb2! Seizing the open b-file, and then the seventh rank. 34…Rd8 35.Rb7 c4 36.g4!

White’s king is strangely safe on h5. 36…Qg7 37.g5

White’s king is perfectly placed to support this pawn advance, which allows White to use his queen for other purposes.

37…fxg5 37…Rg8 38.Qxc4! fxg5 39.Rxf7! is a comical and instructive variation – Black’s queen is trapped, because White’s king covers h6! 38.hxg5 h6 39.g6 Qf6 40.Qe3 fxg6+ 41.fxg6 Kg8

42.Rf7 The engine would calmly take an important pawn with 42.Rxa7. 42…Qd6 43.Qf3 Rd7 44.Qf1! Attacking Black’s vulnerable c4-pawn. 44…Rc7 45.Rxc7 Qxc7 46.Kxh6

The triumphant culmination of White’s idea. 46…Qg7+ 47.Kg5 Qe7+ 48.Qf6 1-0

Chapter 5 Mating Attacks The ultimate form – and goal – of aggression in chess is to checkmate the opponent’s king. From this point of view, the ultimate kingwalk is one that delivers mate. Of course a king itself can’t even check, much less checkmate, the opposing king, because to attack the other player’s king, your own king would (illegally) move into check. A king can only help another piece to give checkmate. This is so common in the endgame that most players don’t even notice it. You have undoubtedly explained to a beginner “this is how you mate with a queen,” even though in fact a lone queen can’t checkmate on its own – it needs help from its king. Since there is always a king on the board, this is taken for granted. It may be hard checkmating with a bishop and knight, but try it without a king to help! Since many games, if played out to the bitter end, would finish with a king helping to checkmate, there is no need to deal with this trivial type of kingwalk. This chapter instead deals with kingwalks in the middlegame that have, as their goal, a mating attack against the opposing king. Normally other pieces combine to deliver mate, but on occasion the king must help out. The requirements for such kingwalks can be set out: • • •

The king must have a path to the opposing king, usually along a complex of weak squares. Other pieces must be in the vicinity to deliver mate. The defender must either be completely helpless or the kingwalk must deliver mate before the perambulating king itself comes under attack.

The third point is particularly interesting. A kingwalk that finishes off a helpless enemy king is one thing, but a kingwalk that imperils both kings is something else. Kingwalks to deliver mate may also fall into the category of “kingwalks up the board to safety,” which are dealt with in Chapter 7. In this chapter, we break with tradition and don’t start with the famous ShortTimman game, classic though it is. Instead we will first give some earlier examples, to demonstrate the continuity of startling ideas. For our first example, we go all the way back to Paris in 1858, when Adolph Anderssen demonstrated that not only Morphy could play interesting chess. Since it’s hard to know where to start, we give the entire game – the swashbuckling opening of which doesn’t bear close examination… Anderssen – Arnous de Rivières Paris 1858 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Bc4 Nf6 5.Ng5 Ne5 6.Bb3 h6 7.f4 hxg5 8.fxe5 Nxe4 9.0-0 d5 10.exd6 f5 11.Nd2 Qxd6

Needless to say, the engine has a lot to say about the way the first ten moves of this game were played. It’s enough to point out that White’s position is every bit as bad as it looks. 12.Nxe4!? Qxh2+ 13.Kf2 fxe4 13…Qh4+ was more precise.

14.Qxd4 Be7?! 15.Qxe4 Bf5!? This is the way they used to play. Why play a normal move (15…Rf8+) when a tricky move can be played instead? 16.Bf7+! Kxf7 17.Qxf5+ Kg8 18.Qd5+ Kh7 19.Qe4+

19…Kh6? The question mark is because 19…Kh6 is objectively bad. Black should accept a draw after 19…Kg8, although one suspects that White would also have refused to disgrace himself with such a result and would have found another continuation. 20.Be3? The engine doesn’t have to take any time to find 20.Rh1! Qxh1 21.Bxg5+!, winning on the spot (21…Kxg5 22.Qxe7+, then 23.Rxh1). But White is still winning, because of the threat of 21.Rh1. Black tries to avoid the worst with a counterattack. 20…Rhf8+ 21.Ke2 Qh5+ 22.g4 Qh2+ 23.Rf2 Rxf2+ 24.Bxf2 Rf8 25.Rh1!

Forced, but strong. White has seen that his king escapes, but it turns out to do more than that. 25…Rxf2+ 26.Kd3 Rd2+ 27.Kc4 Rxc2+ 28.Kd5

28…Bf6? After 28…Bd6! 29.Rxh2+ Rxh2 30.Ke6, White has a better endgame, but no clear win. 29.Ke6! 29.Rxh2+ Rxh2 30.Ke6 also wins. 29…Bxb2 30.Kf7 Naturally Anderssen avoids taking Black’s queen.

30…Rf2+ 31.Kg8 1-0

Now the importance of 28…Bd6! is clear – with Black’s bishop on d6, 31… Rf8 would be mate, so White’s king couldn’t have gone all the way up the board to force mate. Our second example is much quieter, and demonstrates that the oft-maligned Siegbert Tarrasch had a creative, artistic streak that is often not recognized.

Tarrasch – Réti Vienna 1922

White has a dominating position – the question is how best to win it. Tarrasch comes up with a grand conception, reminiscent of the Alekhine-Yates game given in an earlier chapter. The main difference is that Alekhine’s kingwalk resulted in the win of material, while Tarrasch aims for mate. With this in mind, Tarrasch’s first goal is to trade queens. The engine, which plays aesthetically only by chance, calmly announces that 29.g4! was the faster road to victory. Once Black’s f6-knight is dislodged, his position collapses. 29.Rc7 Qe6 After 29…Qb6 30.Qe7, Black is forced to play 30…Qe6, leading to the same position as in the game. 30.Qxe6 fxe6 31.Rg7+ Kh8 32.Re7 Kg8 32…Ne4 offered more counterplay, but this wouldn’t change the fundamental aspects of the position. 33.f3 Ne8

34.Kh2! There’s no secret as to where White’s king is headed. 34…Nd6 35.Rg7+! It’s important to prevent 35…Nf7. 35…Kh8 36.Rd7 Nb5 37.Kg3

Now White’s king simply follows the dark-square road. 37…Nxc3 38.Kf4 Nb5 39.Ke5 Re8 40.Kf6 1-0

The threat of 41.Kf7 and 42.Bg7 mate can be delayed, but not prevented: 40…Kg8 41.Rg7+ Kh8 (41…Kf8 42.Rxg6 mate) 42.Rb7! Nd6 43.Rd7, and after Black’s d6-knight moves, White plays 44.Kf7. Our next example is included mostly for entertainment value, because no kingwalk ever actually happened (Black resigned first). The kingwalk in question was only one of many ways to win (although it would arguably have been the prettiest), and it’s hard to believe this game is even real (although apparently it is). It has been featured on several websites and lauded as a thing of beauty. You can judge for yourself – to us it seems a bit too one-sided.

Kupferstich – Andreassen Denmark 1953

Black has a wretched position and White has many ways to win. The prettiest involves a kingwalk that never happened. 29.Rg7+ Kf8 30.Rxb7+ Kg8 Running to the center doesn’t help: 30…Ke8 31.Nd6+ Kd8 32.Bg5 mate. 31.Rg7+ Kf8 32.Rxa7+ Kg8 33.Rxa8 Bxa8 34.Nd6 1-0

Black resigned at this point. He can only move his bishop or give away his pawns. The simplest, although perhaps not fastest, way for White to win is with a kingwalk, which Black may only watch:

34…Bc6 35.Kd2 Ba8 36.Kc3 Bc6 37.Kb4 Ba8 38.Kc5 Bf3 39.Kb6 Ba8 40.Kc7 Bf3 41.Kd8 Ba8 42.Ke7 Bc6

Position after 42…Bc6 (analysis) Now that White has completed his kingwalk, covering f7 with his king, his d6-knight can checkmate if it can get to f6. There are two routes to f6: Nd6e4-f6 and Nd6-e8-f6. Both e4 and e8 are covered by Black’s c6-bishop, so White only needs to dislodge Black’s c6-bishop by advancing his b-pawn. When White plays b4-b5 and Black moves his c6-bishop, including by capturing White’s b5-pawn with Bxb5, White will respond by moving his d6knight to either e4 or e8, followed by Nf6 mate. Finally, if Black lets White capture his c6-bishop with his b5-pawn, White can either promote his c6-pawn to a queen (or rook) or mate with his knight on f6. Black has no stalemate possibilities, because even if he could give away his e- and g-pawns in time, his dark-square bishop would never be captured. In fact, if Black’s e- and g-pawns weren’t on the board, Black would be in zugzwang and would lose right away. All of these variations (none of which are necessary for White to win) are possible because of White’s kingwalk from e1 to e7. This brings us to the famous Short-Timman game, where the kingwalk leading to mate was conducted with the queens still on the board.

Short – Timman Tilburg 1991

After some adroit play, White has obtained a big advantage. He has control of the d-file and his queen is menacingly posted on f6, where it is ideally placed to exploit the weaknesses in Black’s king position. This is the type of position that the Old Ones had in mind when they condemned kingside fianchettoes because they weakened the king’s defenses! Black’s only solace is that his queen-b7-bishop battery ties down White’s f3knight and g2-pawn, making a “standard breakthrough” with g2-g4 more difficult. This cuts both ways, though – if Black reduces the pressure on the h1-a8 diagonal, White’s options on the kingside will increase accordingly. With all this in mind, White begins his fabled kingwalk, although he is not committed to completing it. As for the engine, after considering this position for a few hours, it simply announces mate in 29 moves. We hardly needed to be told that. It’s pretty obvious… 31.Kh2! Rc8

This move leaves a bad impression, because it does nothing to disrupt White’s plan. The critical variations arise after 31…Bc8, Ian Rogers gives 32.g4 hxg4 33.Ng5! Bxd7 (?) 34.h5! “with a winning attack,” and in fact White mates in this variation. But Black can offer more resistance with 33… g3+! 34.fxg3 Bb7!, restoring his queen-bishop battery on the h1-a8 diagonal. Since 35.Re4? fails to 35…Qxd7, White must consolidate with 35.Ne4 or 35.Qf1, after which he is still winning, but there is no mate in sight. In response to 31…Bc8, the more precise response for White is 32.Ng5! Bxd7 33.g4! It’s amazing that White can attack so calmly when he is down an entire rook, but Black’s pieces are so jumbled up that there is no defense against the methodical opening of lines against Black’s king. After 31…Bc8, White could therefore have won impressively, but the game wouldn’t have achieved immortality. Timman, a super-grandmaster, might have realized this and opted to allow White to demonstrate his idea. 32.Kg3! Rce8

33.Kf4! Bc8 34.Kg5! 1-0

Black resigned, as 34…Kh7 fails to 35.Rxf7+, while the desperate 34…Qxf3 would be demeaning. There is therefore no way to prevent White from completing his kingwalk with 35.Kh6, after which Black is checkmated on g7. Our last three examples move into the realm of tragedy, with the losing side not just allowing, but positively forcing, the opposing kingwalk that sets up checkmate. This furnishes further proof that almost anything can happen on the chess board. There is of course an underlying logic to this, as driving the opposing king up the board often involves a commitment of forces deep into the enemy position. The pieces used in the attack, often including the attacker’s queen, can be out of position to assist in defending their own king. The attacker,

instead of mating the opposing king, then gets the worst of both worlds – he weakens his or her own defenses, while bringing the enemy king into the counterattack. The next example is one of the most tragic in this entire book.

Filip – Darga Oberhausen 1961

After a good opening, Black misplayed the transition into the middlegame and has ended up in a hopeless position. But it is really hopeless? 29.h4? After 29.Rc1!, threatening 30.Rxc8, 30.Rc7+ and even 30.Qc7+, transposing into a easily winning endgame, Black might well have resigned. The engine’s recommendation of 29.Qh8! is pretty good as well. 29.h4? actually doesn’t spoil anything, but it sets White up for more mistakes. 29…Qf3!

Black seeks counterplay. His first hope is that White blunders with 30.Qxc8? Qxd1+ 31.Kh2 Qd2 32.Qb7+ Kg8 33.Qf3 Qxa2 34.Qxf6, with a drawn queen and pawn ending. 30.Rc1!? 30.Qc7+ Kg6 31.Rd8 wins. For example, after 31…Bb7, 32.Rg8+ Kf5 33.g4+ forces Black to give up his queen. 30…Bb7

Now Black has mate threats, so White is under pressure. 31.Rc7+ 31.Qc7+ Kg6 32.h5+! Kh6 33.Qf4+ would force a trade of queens, as would 32…Kxh5 33.Qxh7+ Kg5 34.Qg7+ Kh5 35.g4+ Kh4 36.Qh6+ Kxg4 37.Rc4+ Be4 38.Qg6+ Kh3 39.Qg3+. But it’s fair to ask: “how on earth did White end up having to calculate variations like this?” 31.Rc7+ also wins, if followed up properly… 31…Kg6 32.Qg8+ Kf5

Now White has two ways to win: (a) 33.g4+ (a theme in this sort of position – it’s critical to prevent Black’s king from going to g4) 33…Ke5 34.Rc5+ Kd6 (34…Bd5 35.Qb8+ Ke4 36.Qg3 wins for White, because Black’s king on e4 blocks the h1-a8 diagonal, preventing 36…Qh1 mate) 35.Qf8+ Kd7 36.Qf7+ Kd8 37.Qc7+ Ke8 38.Qb8+ Kf7 39.Rc7+ Kg6 and now not 40.Qxb7 (as suggested by Mednis, although it also wins) but 40.Qg8+ Kh6 41.Qxh7 or 41.Rxh7 mate. (b) 33.Rc5+ e5 34.g4+, and Black must allow the queen trade, because 34… Ke4 runs into 35.Qc4 mate. The move played, in contrast, is a colossal blunder, because it not only allows, but forces, Black to continue his kingwalk to h3. 33.Qxh7+? Kg4 0-1

Now White realized, to his horror, that 34.Rxb7 doesn’t solve his problems, because after 34…Kh3!, mate can’t be prevented. The next illustration is similar to the previous one, in that Black’s kingwalk starts as a king hunt, then becomes a race to safety, and ends with the king participating in a decisive attack – again from the h3-square. The main difference is that the play is more balanced and complex than in the FilipDarga game.

Jelen – Larsen Portorož 1977

The position is dynamically balanced. Black’s extra pawn isn’t that important, and White has play against Black’s king. After White’s next move, Black has no option but to go forward with his king. 28.Qg8+ Kf6 29.Qh8+ Kg5 30.Qf8 Rxe2 31.Ng8

A most unusual position. The engine, which is often criticized for being a soulless demon with no appreciation for beauty, now recommends 31…Kg4! 32.Qd8 Kh3!, judging h3 to be the safest square available for Black’s king. Had Black played 31…Kg4!, this game would still have been in this book, but in a different chapter.

31…h5? 32.Nh6! Bd2

33.Nxf7+? Mednis, who includes this game in his book The King in the Middlegame, gives 33.Rxd2 Rxd2 34.Re1 as winning, as White threatens 35.h4+ Kf6 36.Qh8 mate. He gives the variation 34…Re2 35.Rxe2 Qxe2 36.h4+ Kf6 37.Qxf7+ Ke5 38.Qe7+ and 39.Qxe2, overlooking 36.f4!+ Kf6 37.Qxf7 mate. White has a stronger move, though: 33.Qh8!, threatening 34.h4 mate, with 33…h4 running into 34.f4+. Aren’t chess engines nice? Despite these missed opportunities, after 33.Nxf7+? White is still on top. 33…Kg4 34.Rxd2? Mednis gives 34.Rxd2 an exclamation mark, but in fact this move is a serious mistake. After 34.Qd8! Qxf7 35.h3+!, White should win: 35…Kxh3? 36.Qh4 is mate; 35…Kf3 36.Qd3+; and 35…Kf5 36.Qd3+ Kf6 37.Qxe2. 34…Rxd2 35.f3+ Kh3

Black’s kingwalk has reached its apogee, but in contrast to our previous examples, is it attacking or trying to avoid being attacked? Or both? 36.Qc8+! The only move to prevent mate. 36…Qg4! Black finds an equally good “only move” in reply. 36…Rxc8? 37.Ng5 mate was not an option, and there was no other reasonable way to get out of check. 37.Ng5+! Also forced, because 37.fxg4 Rg2+! 38.Kh1 Rxh2+ 39.Kg1 Rxc8 leaves Black with an easily winning endgame (40.Ng5+ Kxg3 41.Ne4+ Kh3 42.Ng5+ Kxg4 43.Kxh2 Rc2+ and 44…Kxg5, which is why Black had to take White’s h2-pawn before capturing White’s queen). 37…Rxg5 38.fxg4

After the previous forced sequence of moves, it’s time to assess. Black certainly has a perpetual, but can he try for more? 38…Rg2+ 39.Kh1 Rc5!

An excellent try, especially in time trouble. Black risks nothing with this move, and White may respond incorrectly… 40.Qd8? White had to keep his queen on the h3-c8 diagonal, with 40.Qd7 or 40.Qe6. After 40.Qd7, Black draws by 40…Rxh2+ 41.Kg1 Rg2+ (perpetual check) or 40…hxg4 (to which Mednis mistakenly gives two questions marks) 41.Qh7+ Rh5 42.Qb7 Rc2 (a positional draw). 40…Rxh2+ 41.Kg1 g5!

Now White is lost! 42.Qe8 and 42.Qh8 both fail to 42…Rcc2 43.Qxh5+ Kxg3, and White is forced to go into a lost rook and pawn ending with 44.Rb1 Rxh5 45.gxh5 Rc6. Similarly, 42.gxh5 Rcc2 43.Qa8 (or 43.Qd5) 43…Rcg2+ 44.Qxg2+ Rxg2+ 45.Kh1 Rh2+ 46.Kg1 b3 also leads to a lost rook and pawn ending, because Black easily stops White’s passed h-pawn. 42.Rb1 0-1 Finally, a recent game where the same principles are seen. The example starts with White’s queen on h6, but ends with his king on the same square, difficult as that it to believe, given the starting position.

Hillarp Persson – Laurusus Batumi 2018

White has some compensation for his pawn deficit, but that’s about all. It is difficult to see how, in only ten moves, this game will become the best of the 2018 Olympiad. Since there really isn’t anything happening in this position, White’s motif brings to mind the theme song for the Mary Tyler Moore Show: “Who can take a nothing day, and suddenly make it all seem worthwhile?” Of course we’re referring to Joan Jett’s cover version, not the original… White gets some help at this point. The engine recommends the supremely logical 24…Rfd8!, allowing 25.Rxh5, so that 25…Qg7 forces the exchange of queens, leaving Black with a favorable endgame. Black instead surrenders the wrong pawn. 24…Qg7?! 25.Qg5 Rcd8?! 26.Qxe7

Black is still not worse, and with his next move he begins an ambitious attack against White’s seemingly vulnerable king. After all, with White’s rooks in front of his kingside pawns, the back rank is open. 26…Rd1+ 27.Kg2 Qa1 28.Bxc6?! White understandably takes the opportunity to obtain a material advantage, but 28.Rf4! was better. 28…Rg1+ 29.Kf3 Qf1? Up to this point, White’s kingwalk has been purely defensive in nature, and Black could keep the balance with 29…Bc8!, bringing his bishop back to the kingside, while keeping his queen on the a1-h8 diagonal. Instead Black completely commits his queen to the attack, which leaves his own king wide open to a counterattack. 30.Kf4!

Stopping 30…Qg2+ and 31…Qxc6, but also bringing White’s king into the attack, which had gone dormant for the last few moves. 30…Qxf2+? The fatal blunder – Black can resist everything except temptation. But what could be better than taking a pawn with check? 30…Qa1! is the answer, admitting his mistake on the previous move. The position then remains balanced. 30…Qxf2+? not only drives White’s king towards h6, but also gives White a tempo for Rh4-f4. If you look only at the upper right portion of the board, it is immediately clear that Black’s king is about to be overwhelmed by superior forces. 31.Kg5

31…Kg7 32.Rf4 Qxh2

Black now threatens to capture White’s g3-pawn, but the cavalry arrives too late. 33.Qf6+ Kh7

34.Qxg6+! A very pretty, although not difficult, queen sacrifice. Black is mated after 34…fxg6 35.Re7+ Kg8 36.Bd5+ Rf7 37.Re8+! and 38.Rxf7 mate. 34…Kh8 35.Kh6 1-0

White’s king has the last word. Black is checkmated on the next move, no matter what he does. A beautiful creative achievement by White, with some help from his opponent at the critical moment. The most remarkable aspect of this game is that there wasn’t really anything to make one think that something like this was even possible ten moves ago.

Chapter 6 Escaping to Safety Across the Board We now turn from aggressive kingwalks, whether to prepare, exploit or participate in attacks, to defensive kingwalks. In military terms, a retreat in the face of enemy pressure is one of the most challenging operations, and chess is true to its origins in this area. Kingwalks under pressure are difficult to execute and require both nerves of steel and the ability to calculate variations. They also require a good sense of timing, as waiting too long may be fatal. In this chapter, we will consider lateral kingwalks, which see the king escaping across the board. In the next chapter, we will look at more dramatic kingwalks up the board, some of which we have already seen in an earlier chapter. The two types of kingwalks are very similar, as the only real difference is the king’s destination. Our first example is fairly simple – White commits his major pieces to an attack that wins an unimportant pawn and drives Black’s king to safety.

Illescas – Andersson Ubeda 1997

White has a slight advantage, because of his more active pieces and Black’s isolated d-pawn. The one thing White doesn’t have, however, is a kingside attack. But, perhaps misled by Black’s earlier …Kg8-h8, he chases a mirage and launches a direct attack, which is refuted by a timely walk by Black’s king. 20.Rh3?! Kg8 21.Bxf6 Bxf6 22.Qxh7+ Kf8

White has won an unimportant pawn, and could maintain the balance by bringing his rook back to d3 or his queen back to h5. Instead he decides to chase Black’s king into the center, which is exactly where it wants to go.

23.Qh8+? Ke7 24.Qh5 Rd7! 25.Re1+ Kd8 26.Rxe8+ Kxe8 27.Qh8+? It’s not often you get to make the same mistake twice! 27…Ke7

White’s pieces have lost their coordination, while Black is well positioned to exploit White’s weak back rank. 28.Nd1 d4! There’s another weak point in White’s position – his g2-pawn. 29.Qg8 Qe5 29…Qxf5 was even stronger. 30.Qxf7+ Kd8 31.Rh8+ Kc7 32.Qe8 Qxf5 33.Qe1 Re7 34.Qg3+ Kb6

Black’s king has moved all the way across the board, and is perfectly safe. The same can’t be said for White’s king. 35.Bc4 Qxc2 36.Bb3 0-1 White either resigned (36…Re1+ 37.Qxe1 Qxg2 is mate) or lost on time. The next example, on the other hand, is anything but simple, at least from Black’s point of view. His kingwalk doesn’t solve his problems, but it confused his opponent enough to let him escape from a difficult position.

Efimenko – L’Ami Plovdiv 2008

White has the advantage – he has the bishop pair and can open lines on the kingside, while Black’s earlier attack on the queenside hasn’t led to anything definite (the fact that White controls the b-file, which Black opened, is not a good sign). Black doesn’t panic – he starts to walk his king to the safer side of the board. 24…Kd8!? 25.h5 gxh5 26.Bxh5 Rf8 27.Bg4 27.Ng5! was more incisive. 27…Kc7 28.Rh7 Rg8 29.Bh5 g6 30.Bg4 Qe8 31.Rbh1

There’s no doubt that Black is on the defensive. 31…Kb6!? 32.Ng5 Bxg5 33.Bxg5 Ka5

Black’s king has gone as far from the kingside as it can, but there’s a limit to how far it can go! 34.Rb1?! White is right to switch his attention to the queenside, but wrong to withdraw his rook from the h-file. Two retrograde bishop moves would have increased the pressure even more: 34.Bc1! Qf8 35.Bd1! Nb6, and Black is in trouble. 34…Rh8 35.Rxh8 Qxh8 36.Rb7 Qf8

The position has changed a lot – now there aren’t any rooks on the kingside, and White has a rook on the seventh rank. But Black hangs on. 37.Qc1 Ka6 38.Rc7? A miscalculation, presumably in time pressure. White’s rook is suddenly transformed from an attacking piece into a target for Black’s marauding king. 38…Kb6 39.Rd7 Qe8 40.Rd6 Nb2 41.Bd1? Kc7

42.Bc2 Na5! 43.Bf4 Nb7 44.g4 Nxd6 45.exd6+ Kc6

The position is still confusing, but an exchange is an exchange, and Black is winning. 46.Bg3 Qd8 47.Qh6 Na4

Now Black’s other knight comes back to take something on d6. The d6square has become the Verdun of White’s position. 48.Qe3 Nb6 49.Bd1 Nc8 50.Bf3 Nxd6 51.Qc1 g5 52.Qb1 Qa5 53.Bxd6 Kxd6 54.Qb7

At first this looks scary, but by now Black knows how to respond – with one last kingwalk! 54…Qa7 55.Qb4+ Kd7 56.Bd1 Qa5 57.Qb7+ Ke8 58.Qc6+ Ke7 59.Qb7+ Kf8 60.Bc2 Kg7 61.Bb1 Qd8 0-1

Spectators who arrived late can be forgiven for thinking that Black castled kingside and bravely withstood an assault there. In fact, Black’s king moved 14 times in this game (almost one-quarter of Black’s moves), while Black’s a8-rook still hasn’t moved. But it’s about to,

and that was enough to convince White to resign. The next two examples caused some soul-searching. Should our examples be limited to tournament games under classic time controls, or was it permissible to use games under faster time controls? Ultimately, we decided in favor of inclusion, for several reasons. First, it is impossible to know the conditions under which every game was played, especially games from the nineteenth century. Some of Morphy’s finest games were played in coffee houses, often without a chess clock being used. It would be difficult to exclude games based on the conditions under which they were played when, in some cases, the conditions aren’t known. Second, as we have already seen, many of the examples in this book occurred under time pressure, sometimes in severe time scrambles. If it is acceptable to include a game where players had only seconds per move in a tournament game because of time pressure, why not include games played under faster time controls right from the start? Third, the very first game in this book (Kevitz & Pinkus-Alekhine) was played in a simultaneous exhibition, and we shall see some games from rapid tournaments. As long as the reader is aware of the conditions under which the game, why not include those games? Fourth, the following two games, despite being 3-minute games (or maybe because of it), are at least entertaining, and maybe even instructive. So enjoy the next two examples, but don’t spend more time looking at them than the players spent playing them. Anon – Nakamura ICC 2012 (three minute) 1.c4 b6 2.Nf3 Bb7 3.d3 e6 4.Nc3 d5 5.Bf4 d4 6.Nb1 f6 7.Nbd2 e5 8.Bg3 Nh6 9.a3 Nf5 10.Rc1 a5 11.Nh4 Nxh4 12.Bxh4 h5 13.h3 g5 14.Bg3 Bh6

We could have littered the first 14 moves with punctuation marks, but what would be the point? This is a 3-minute game. We’ll just point out that Black had a big advantage, and now he doesn’t. The engine assesses the position as almost exactly equal, which surely beats the odds. 15.e4?! dxe3? 16.Qxh5+ Ke7 17.fxe3 g4?! 18.Qxg4 Bxe3?

Black either missed White’s next move, underestimated its strength, or just wanted to see what he could get away with. 19.Qg7+! Ke6 20.Be2! Rg8 21.Bg4+ Kd6

22.c5+ This isn’t as strong as it looks. 22.Qf7! was the calm killer. 22…Kc6! 23.Bf3+? 23.Qf7! was still the move. 23…Kb5!

Believe it or not, Black is now winning! 24.Qf7 Bxf3! 25.Qc4+ Kc6

Have you ever seen a safer king? 26.cxb6+ Kxb6 27.Nxf3 Rxg3 It took seven moves for the skewer Black started with 20…Rg8 to do its job, but in chess you have to be patient. 28.Rc3 Na6 29.Qe4 Rxf3 After what Black’s king has been through, he hardly cares about a check on c6. 30.Rc6+ Ka7 31.gxf3 Bb6

Now Black’s king really is safe, and soon it will be Black’s turn to be the aggressor.

32.Rxf6 Nc5 33.Qc4 Nxd3+ 34.Ke2 Nf4+ 35.Rxf4 exf4 36.Qxf4 Qe8+ 37.Kf1 Qb5+ 0-1 In the previous example, Black at least invited the danger. That is not the case in the next example.

Anon – Harper ICC 2012 (three minute)

White has just played 28.Rd6, hoping to generate some counterplay in a lost position. Of course Black should take on d6 and win a second pawn: after 28…Rxd6 29.cxd6 Rd8 30.g3 Rxd6 31.gxf4+ Kxf4, Black has increased his advantage. Instead, Black decides to embark on a “kingwalk to safety” up the board. He gets the “kingwalk” part right, but misjudges the “safety” part. 28…Bf8? 29.Rg6+ Kh4

30.g3+?

After 30.Be1!, Black’s king is in a mating net. But after 30.g3+?, which White played instantly, despite having nearly two minutes left, Black’s kingwalk works. What a difference a single move can make! 30…Kh3 31.gxf4 exf4 32.Rg1 Bxc5

33.Ke2 33.Rg5 fails to 33…Rd5! 34.Bxd5 g3+! 35.hxg3 Ng4+ 36.Kf1 Nh2 mate. It goes without saying that Black would have seen all this. After 33.Ke2, things are simpler. 33…Rxe4 34.fxe4 f3+ 35.Ke1 Ng2+ 36.Kf1 Rd1+ 0-1 All right – back to work. To paraphrase Napoleon, it is but a step from the ridiculous to the sublime, and White’s first move in the next example may be justly termed “sublime.”

Khismatullin – Eljanov Jerusalem 2015

44.Kg1! Forced moves can be beautiful! With his rook hanging with check and a dangerous black pawn on d3, White calmly takes care of his biggest problem – his king’s safety. It is more than fair to categorize this as a “kingwalk to safety,” even though it’s a short one, but White’s king also ends up playing a small, but ultimately important, role in White’s attack on Black’s exposed king. 44…Qxd1+ This loses, although it would be heartless to give it a question mark. The best try for Black was 44…Rd5! 45.Kh2 Kf6 46.e4 Rc5 47.Qd6+ Kg7 48.Rxd3 Rxc6 49.Qe5+ Rf6 50.Rf3 Qc6, after which Black is all tied up, but there is no clear win for White. 45.Kh2

45…Rxc6 46.Qe7+ White can afford to repeat the position by giving this check before taking on f7, as 46…Kh5 47.g4+ Kh6 48.Qf8+ Kg5 49.Qxf7 wins. 46…Kh6 47.Qf8+ Kg5 48.Qxf7! Threatening 49.Qf4+ Kh5 50.g4+, mating. Surprisingly, Black’s disorganized forces can’t stop the attack. 48…Rf6 49.f4+ Kh6 49…Kf5 is met by 50.Qd5 mate. 50.Qxf6 Qe2 51.Qf8+ Kh5 52.Qg7 Threatening Qxh7 mate. 52…h6 53.Qe5+ Kh4 54.Qf6+ Kh5 55.f5!

Breaking down Black’s defenses. White’s king is still completely safe. 55…gxf5 56.Qxf5+ Kh4 57.Qg6 1-0 It’s over after 57…d2 58.Qxh6+ Qh5 59.g3 mate. In the end, White’s king provided critical support! We end this chapter with another Nakamura game, this time played under a classical tournament time control. This example is an excellent segue into the next chapter, as White’s king goes across the board, then goes up the board.

Nakamura – Robson St. Louis 2012

White had an advantage, but it has slipped away. The game remains complicated, however, despite the fact that queens have been exchanged. 28.c5 White activates his queenside pawn majority, and fixes Black’s b7-pawn as a target. What keeps the position balanced is Black’s counterplay, but that balance can easily be disrupted in favor of either player. 28…Nb5 The alternative was 28…Nf5. 29.Rd7 Kf7 30.Rxb7 White ignores the attack on his e6-knight, and attacks Black’s b5-knight. 30…Rg1+ 31.Kc2 White’s king starts to move. 31.Kb2? would lose to 31…e4+, so there’s no choice for White at this point. 31…Na3+ 32.Kb2 Nb1

This isn’t something you see every day, but b1 is the only safe square for Black’s knight. Black is now finally threatening to capture White’s e6-knight. 33.Nd8+ Ke8 34.Nc6 Now there is a series of forced moves for both sides. 34…e4+ 35.Kc2 Na3+ 36.Kd2 Nb1+ 37.Ke3

White’s king isn’t just trying to escape the checks – it threatens to capture Black’s main source of counterplay, his e4-pawn. 37…Re1+? An understandable error in time pressure. Black should defend his e4-pawn with 37…Nc3!, and after 38.Rxa7 play 38…Rd1!, which the engine assesses as absolutely equal.

38.Kf2! Rc1 39.Rh7!

White is more interested in Black’s e7-pawn than in his e4-pawn, because it’s the one that’s protecting Black’s king. 39…Rc2+ 40.Kg3? Having run across the board, White’s king now changes direction and starts to go up the board, but to the wrong square. 40.Ke3! was correct. 40…Rc3+? Black also errs on his 40th move. After 40…Nc3!, he’s back in the game. 41.Kg4! e3

Wait a minute! How is White going to stop Black from queening his e3pawn, other than by retreating his h7-rook? The answer is: he isn’t going to stop it. 42.Nxe7! e2 43.Nd5!

Now 43…e1Q 44.Nxf6+ mates, so Black has to try to queen with check. 43…Rg3+ 44.Kf4! Bg5+ 45.Ke5! e1Q+ 46.Kd6

Black has his new queen, and his other pieces as well, but what can he do about the mating threats? 46…Be7+ 47.Rhxe7+ Qxe7+ 48.Rxe7 leads to a hopeless endgame, so Black starts giving things away. 46…Bf4+ 47.Nxf4 Rd3+ 48.Nxd3 Qg3+ 49.Ne5 1-0

White’s king has much to be proud of, but the last three moves by White’s knight single-handedly destroyed Black’s army. Since Black has no useful checks, he now resigned.

Chapter 7 Escaping to Safety Up the Board This chapter is longer than the previous ones, because there are many impressive examples of kingwalks up the board to safety. There are, not surprisingly, many more instances in the literature of kings going up the board to their doom. It might be that this is why successful kingwalks of this type strike a chord with chess players. Another reason might be the irony of such kingwalks. Normally the king is safe on the back rank, behind his pawns, preferably in a corner. But at times the king goes out into the world, and seeks safety elsewhere. There is a very logical reason for this, which we encountered earlier in Chapter 5 (kingwalks to deliver mate). If a king is being attacked from the side or rear, significant enemy forces may be committed to an area of the board from which they can’t easily redeploy. In addition, the fact that one player is using pieces to attack may mean that the opponent can counterattack. We will see many examples of this. This chapter, like all of Gaul, is divided into three parts. The first part deals with the main topic – kingwalks up the board to safety. The second part illustrates that sometimes such kingwalks are an illusion. This is a useful thing to remember when confronted with a seemingly “impossible” kingwalk. It might actually be impossible. The third and final part gives several examples of kings walking up the board, then returning to the comfort of home. Escape up the board Our first example is from a very famous game, in which Botvinnik made his name by defeating the mighty Capablanca (we have already seen a game

where Reshevsky did the same thing – Capablanca must have been pretty good if beating him meant you had arrived).

Botvinnik – Capablanca AVRO 1938

This position, and the ensuing combination, has appeared in a thousand books, and is justly famous. Detailed notes are found in Predecessors. We will look at what follows from a different perspective – the kingwalk that Botvinnik had clearly calculated when he played the winning combination. 30.Ba3! Qxa3 31.Nh5+! gxh5 32.Qg5+ Kf8 33.Qxf6+ Kg8 34.e7

As Kasparov points out, 34.Qf7+ Kh8 35.g3!, giving White’s king a safe square on h3, also won. Since White’s threats are unstoppable, the only remaining question is whether Black can save the game by giving perpetual check. 34…Qc1+ 35.Kf2 Qc2+ 36.Kg3 Qd3+ 37.Kh4 Qe4+ 38.Kxh5!

Now exchanging queens on g6 doesn’t help Black – after 38…Qg6+ 39.Qxg6 hxg6+ 40.Kxg6, White’s e7-pawn queens. 38…Qe2+ 39.Kh4 Qe4+ 40.g4 Qe1+ 41.Kh5 1-0

White’s kingwalk up the board, which has to be considered part of White’s combination, has succeeded and Black is out of useful checks. Escaping up the edge of the board is more common than one might think, because the defender’s king isn’t exposed to attack from all directions. The Great Patriarch would have nodded approvingly at the following, more recent, example.

Rogers – Berzinsh Reading 2007

White has sacrificed two exchanges, but the safety of the respective kings is what matters most. 35.g4! White's king prepares to escape. Capturing Black’s c2-rook would lead to little after 35.Nxc2? Qd3+ 36.Ke1 Qxc2. 35…Qa1+? A typical mistake that no one who has read this book would make. Attacking along the eighth rank will just drive White's king up the h-file to safety. Black’s only chance was to give his king some air with 35…f6!, then bring his queen back to the defense with 36…Qd3+. This wouldn't solve all Black’s problems, of course, as White’s c-pawn is dangerous. 36.Kg2 Rc1 37.Nd7! White's b6-knight joins in the attack. Now all Black can do is give some checks, but White’s king knows exactly where it wants to go.

37…Rg1+ 38.Kh3 Qf1+ 39.Kh4 Qxf2+ 40.Kh5 The resemblance to the previous game is striking. Since 40…Rf1 is met by 41.Nf5!, there is nothing to be done. 40…f6 41.Nxf6+ Kf7 42.Nxe8 1-0

The next two games form a remarkable pair. Both games were played in the 1959 Candidates tournament, which was won by Tal. In both games, Bobby Fischer was White and Tigran Petrosian was Black. The same opening was also played in both games. But, most surprisingly, in both games Black executed a dramatic kingwalk.

Fischer – Petrosian Bled 1959

Fischer has been outplayed and Black’s only problem is how to avoid having his king harassed by White’s queen. Petrosian’s solution is simple – since White’s queen is attacking from the rear, Black will walk his king in front of his advanced queenside pawns. 57…b3! 58.Qb8+ Ka5 59.Qa8+ Kb5 60.Qb8+ Kc4! 61.Qg8 Kc3 62.Bh5 Nd8 Black could give up his f7-pawn, but why should he? 63.Bf3 a3 64.Qf8 Kb2

Black’s king is well-protected, and there’s no longer anything White can do to stave off defeat. 65.Qh8 Ne6 66.Qa8 a2 67.Qa5 Qa4 68.Rxd2+ Ka3 0-1 Here is the companion game, in which Black’s kingwalk is even more striking – and more dangerous.

Fischer – Petrosian Bled 1959

In what becomes an increasingly intense struggle, Black has an extra pawn, although Fischer has his bishop. Black should now oppose rooks on the f-file, but Petrosian has a different idea in mind, similar to that in the previous game, which was played earlier in the same event. 23…Kb7!? 24.h5 Qxb4 Black could still play 24…Rhf8! 25.Rf7+ Kb6 Black didn’t play 23…Kb7 in order to retreat with his king. 26.Qf2! a5? 26…Qb2! was better, to answer 27.c4 with 27…Qxf2+. 27.c4 Nc3 28.Rf1 a4 29.Qf6 Qc5 29…Qd6! was better, but then we would have been deprived of an extraordinary creative achievement. 30.Rxh7! Rdf8! 31.Qxg6?

31.Rh8 Rxf6 32.Rxf6 was stronger. 31…Rxf1+ And here 31…Rxh7! was correct. But we will pass over these details, as the position is extremely complicated and the game has been analyzed elsewhere. 32.Bxf1 Rxh7 33.Qxh7 a3 34.h6 a2 35.Qg8 a1Q 36.h7 Qd6 37.h8Q Qa7 38.g4

Here is where the action really begins, although Petrosian started his kingwalk some 15 moves earlier. After some inexact moves, Black has not only lost his advantage – he’s losing. In four-queen positions (a good topic for a book?), king safety is crucial, for obvious reasons. In this position, White’s king is safer than Black’s, because Black’s queens (!) are passive. White also has a passed g-pawn that may become a fifth queen… Black’s king now begins a desperate flight to safety, trying to simply get as far away from White’s queens as it can. 38…Kc5!? 39.Qf8?! The engine rejects the suggestion of 39.Qh2, in favor of the more logical 39.g5. 39…Qae7 40.Qa8 Kb4! 41.Qh2 Kb3!

42.Qa1? After 42.c5!, White would still be on top, although you need an engine to prove it. 42…Qa3 The only move, but a good one. 43.Qxa3+ Kxa3 44.Qh6 Qf7! 45.Kg2 Kb3 Threatening 46…Nd1, which was impossible without first moving the king (45…Nd1? 46.Qc1+). 46.Qd2 Qh7! 47.Kg3? A famous blunder, brought on by the strain of such a titanic struggle. And yet… 47…Qxe4 48.Qf2 Qh1 ½-½ Here Petrosian agreed to a draw, having trouble shifting gears now that Black stands better. The next example is a simpler, but no less famous, game. It is analyzed in detail in Predecessors, Volume II. We shall focus only on White’s kingwalk.

Bronstein – Ljubojević Petropolis 1973

19.Ke2? An error in a very complicated position, as Kasparov explains – 19.0-0-0! was better. Black has an extra rook, but White’s attack is very strong, despite the pin on his c3-knight. But without this mistake, this game would not be in this book. 19…Bc5? 19…Qc5! gave Black a clear advantage. 20.Ne4 N8d7 21.Rc1 Qc6 22.Rxc5! Nxc5 23.Nf6+ Kh8 24.Qh4

Black’s extra rooks don’t matter – White is now mating. Black’s only chance is a counterattack against White’s exposed king, but it escapes by running at its attackers. 24…Qb5+ 25.Ke3! h5 26.Nxh5 Qxb3+ After 26…Qd3+ 27.Kf2, White’s king escapes less dramatically. 27.axb3 Nd5+ 28.Kd4! Ne6+ 29.Kxd5

29…Nxg5 30.Nf6+ Kg7 31.Qxg5 1-0 (41) Black played on until White reached the time control, which wasn’t a given, since White had only seconds left on his clock. The next example is included for two reasons. One is that it is always good to have a Kasparov game. The second is that Short’s kingwalk foreshadows his subsequent achievement against Timman, which we saw in an earlier chapter.

Kasparov – Short London 1987 (rapid)

Black has good compensation for his slight material disadvantage, and could consolidate with 40…Be8, defending his b5-pawn. Instead Black chooses to imbalance the position further – a natural course of action in a rapid game! 40…Bxf3!? 41.Rxb5 Rc7 42.Rb8+ Kh7 43.Qf8 Qa7+ 44.Kf1 Re7 45.R1b2 Kg6!?

Short advances his king, perhaps with a premonition of his game with Timman four years later? If White does nothing, Black’s king will work its way into White’s position along the light squares h5-g4-h3. 46.Bc1?

Kasparov misses the key, and (in hindsight!) obvious idea of exchanging Black’s remaining rook with 46.Re8! This opportunity arises several more times in the next few moves, but instead White chases Black’s queen into his position, with disastrous results. 46…Kh5! 47.Ra8 Qc5 48.Rc8 Qxa3 49.g4+ Bxg4 Safety first. Black doesn’t need his king for the decisive attack – he just needs it to be safe (49…Kxg4 50.Rxc4+ Kh5 was still winning, but there is no reason for Black to play that way). 50.Rxc4 Qa1 0-1

Sometimes kingwalks of this type go more smoothly than other times, as is shown in the next example.

Bellón – Gagarin Gothenburg 2004

White has outplayed his opponent and has a winning position. He is not only two pawns up, but Black’s king is more open to attack. Black tries for counterplay. 29…Re8!? 30.Qc6 30.Qa6 would have defended White’s e2-knight, but instinctively White prefers to keep his queen closer to the center. He also presumably had calculated the attractive kingwalk that ensues. 30…Rxe2 31.Rxf6! Rh2+ 32.Kg4

White’s iron grip on the light squares ensures his king’s safety, although this idea is rarely demonstrated as clearly as it is here. 32…Qd1+ 33.Rf3 Qxd4+ 34.Kh5 Qe5+ 35.Kg6

Black’s queen is now outmatched, as Black’s h2-rook contributes neither to offense or defense. 35…Qb8 36.Qd7 Qb6+ 37.Be6 Qb1+ 38.Rf5 1-0

Two nice interpositions on the light squares e6 and f5 have made White’s king invulnerable. The same cannot be said for Black’s king. This recent example shows that kingwalks to safety are alive and well, and will be as long as chess is played.

Ider – Hou Gibraltar 2017

Black has played a brilliant game, and now she has to find the path to victory. 43…Nxh2! The only move to win. 44.Rxe5 Bf3+ 45.Kg1

45…Nxf1! 45…Kh6 comes out the same, but it’s much better for Black to sacrifice a rook with (double) check. The most important factor in the position is that Black’s king has a path to safety, whereas White’s king is trapped.

46.Rxe7+ Kh6 47.Qg7+ Kh5

48.Qh7+ Taking the second rook is immediately fatal, as 48.Qxf8 h2+ 49.Kxf1 h1R is mate. White therefore has to keep trying to give checks, but Black’s king has a clear path to safety. 48…Kg4 49.Re8 Rxe8 50.Qd7+ Kh4 51.Kxf1 Rd8! 52.Qh7+ Kg4 0-1

There is nothing to be done. A modern classic. Occasionally a kingwalk only occurs in the notes – but “what might have been” kingwalks can be just as instructive.

Kasparov – Anand Linares 2002

After some powerful play, White has a winning attack, but instead forced a draw. 39.Bg6+ Kg8 40.Bh7+ Kf7 41.Bg6+ The win was there with 41.Re5! Qc1+ 42.Kg2, and now there are a number of tries for Black, the most interesting of which, from our standpoint, is 42… Rb1: (a) 42…Ke7 43.Rxe6+ dxe6 44.Qxg7+ Kd8 45.Bc2; (b) 42…Nf6 43.Rf4 Rd2 44.Bd3; (c) 42…Rd2 43.Bg6+! Kxg6 44.Rg5+! Kxg5 45.Qxg7+ Kf5 46.Qg4+ Kf6 47.Rh6+ Ke7 48.Qg5+ Kf7 49.Rh7+ Kf8 50.Qg7+ Ke8 51.Qg8 mate; or (d) 42…Rb1 Black goes all out for a counterattack against White’s king. 43.Rxd5! Qf1+ (43…Qh1+ 44.Kh3 Qxd5 45.Rf4+ Ke8 46.Qxg7) 44.Kf3 R1b3+ 45.Kg4 cxd5 46.Qf4+ Ke7 47.Kg5!

Position after 47.Kg5 (analysis) White’s king advances up the board, not only to ensure its own safety, but to clear the fourth rank so his h4-rook can come to the f-file. 47…Qa6 48.Bg6 Qc6 (48…Rf8 49.Qxf8+ Kxf8 50.Rh8+ Ke7 51.Re8 mate) 49.Rh7 Qxc5 50.Rxg7+ Kd8 51.Rg8+ Ke7 52.Rxb8 Rxb8 53.Qxb8 Qxf2 54.h4 Qf6+ 55.Kh6

Position after 55.Kh6 (analysis) The final step in the kingwalk. Sadly, this variation remained only in the notes. 41…Kg8 42.Bh7+ ½-½

The last example in this section, also analyzed in detail in Predecessors, Volume II, leads nicely into the next section. A kingwalk can even flummox the mighty Tal – at least in a time scramble!

Geller – Tal Moscow 1975

In a difficult position, Tal has just played 29…Bd4!?, hoping for complications in the impending time scramble. 30.Qe2? As Kasparov points out, White should first drive Black’s queen off the first rank by offering an exchange of queens with 30.Qc1!, switching to the attack only after 30…Qxa2 31.Qe1! Then Black would have no play against White’s king. 30…Ne7! 31.Nb5 If White takes Black’s e7-knight, either before or after checking on e6 with his queen, Black wins: 31.Qxe7 Qg1+ 32.Kg3 Qf2+ 33.Kg4 Qxg2+ 34.Bg3 h5+ 35.Kh4 Qe4+! 36.Qxe4 Bf6 mate. 31…Bg1+ 31…Qg1+ is also sufficient to draw. Now White has no choice other than to begin a kingwalk, and the road to victory for White and for a draw for Black lies along the White king’s path, which is wide enough only for one… 32.Kg3 Nf5+ 33.Kf3

33…Nh4+? Black had to retreat his bishop along the g1-a7 diagonal with 33…Bb6 (or even 33…Bc5), after which White would have had to give perpetual check. The move played allows White to continue his kingwalk to safety. 34.Kg4 Nf5 35.Qe8+ Kg7 36.Qd7+ Kh8 37.Nxd6

The road ahead is clear – g5-f6-f7, and safety. 37…Qd1+ 38.Kg5 Qh5+ 39.Kf6 Bd4+ 40.Ke6 Ng7+ 41.Kf7 g5+ 1-0 Here Black resigned, without waiting for 42.Kf8.

Position after 42.Kf8 (analysis) Smoke and mirrors In the previous game, Geller’s king had to shoot the rapids – at one point Tal could have secured a draw. But in the examples that follow, the kingwalks, while ultimately successful, have holes in them that you could drive an 18wheeled truck through. As always, it is easier to see these things afterwards…

Heidenfeld – Roele Amsterdam 1954

Black must move his queen, but he has a choice of squares. The safest move is 15…Qg7!, after which White must retreat his d3-bishop, giving Black a comfortable advantage. Instead he invites a sacrifice. 15…Qa3 16.Bxh7+! Kxh7 17.Ng5+ Kg6! Black’s king laces up his shoes and starts his journey. 18.Ne6

There is no alternative.

18…Bxe6? Much safer was 18…f3 19.Nxf8+ Qxf8 20.Qd2 f2+ 21.Qxf2 Qxf2+ 22.Kxf2 Bf5 23.Re2 Be4, and while the engine assesses the position as equal, Black will find it easier to play. 19.Rxe6+ Kf5 20.Rh6!

Black has a piece for a pawn, but his king is on the fourth rank and there is no going back. White is winning, but he still has to prove it. 20…Qxc3 There’s no other move. 21.Qh5+ Ke4 22.Rd1! Rae8 22…Nxd4 runs into the elegant 23.Re6+! Nxe6 24.Qxd5+ Ke3 25.Qf3 mate. 23.Qg6+ Ke3

Black’s king has gone on quite a journey! How does White finish what has, up to this point, been a king hunt, not a kingwalk? The answer may surprise you. 24.Rh3+? 24.Kf1! was the winning move. Whether or not you want to call this a kingwalk, after 24.Kf1!, White’s king covers e2, and Black can’t escape. The variations are actually quite beautiful (and almost incomprehensible). For example, 24.Kf1! Nxd4 25.Rh3+ f3 26.Qg7!, and White wins, as he threatens to continue calmly with 27.gxf3 and 28.f4+. 24…f3? This is a mistake, because it allows White to transpose into the variation mentioned above, with 25.Kf1! No one can fault the players for missing such a move. 25.Rxf3+ Ke2! 25…Rxf3? 26.Qxe8+ is clearly bad for Black. 26.Rxc3 Kxd1

The worst is over for Black. His king has managed to get all the way up the board, and White missed his chance to win. Materially, White is ahead, but Black has an active king (!), and threatens mate on the move. 27.h4 27.h3 would have prevented Black’s next move, because after 27…Nxd4? 28.Rd3+! leads to a favorable endgame for White. But Black maintains the balance with 27…Re1+! 28.Kh2 Ne7. 27…Nxd4 28.Kh2 Now 28.Rd3+ cxd3 29.Qxd3+ Kc1 30.Qxd4 Re1+ 31.Kh2 Re4 32.Qxa7 Rxh4+ 33.Kg3 Rff4 34.Qxb7 d4 is equal. 28…Ne2 29.Rf3? Rooks belong behind passed pawns! 29.Rh3! was right. 29…d4! 30.h5 Rxf3 31.gxf3

31…Re3? 31…Rh8! would have held up White’s h-pawn long enough for Black to have broken through with his own passed pawn. 32.Qf5? 32.h6! was equal. 32…d3! 33.cxd3 c3! 34.Qc5 Rxf3 35.h6 Rf6! 36.Qg5 Rf2+ 37.Kh3 Nf4+ 38.Kg4 Rg2+ 39.Kf5 Rxg5+ 40.Kxg5 c2 41.h7 c1Q 42.h8Q Nxd3+ 0-1 It was interesting to see how well-placed Black’s king was on d1. Once it got there, it never had to move again! In the next example, the survival of Black’s king seems truly miraculous.

Kerkhoff – Marcus Leeuwarden 1970

After an up and down struggle in which the advantage shifted from one side to another, and back again, White now initiates what should be a decisive attack. 29.Rd8+ Rxd8 30.Rxd8+ Kh7 31.Qd3+ f5 32.Qg3

32…g5!? Since 32…e5 33.Qd3 Qe4 (33…e4 34.Qc4, threatening mate on g8) 34.Qxe4 fxe4 35.Nxc5 loses a piece, Black has nothing to lose by trying the text. 33.fxg5 Bd6 34.g6+

Pretty much any move wins, but 34.Qh4, with the idea of 35.Qd4, was fastest. 34…Kg7 35.Qd3 Qc7 36.Qd4+ e5 37.Qd5

White’s attack has broken through, so Black plays his last card – a kingwalk to the Land of Hope. 37…Kf6!? 38.Qg8? 38.g7! was better – White’s queen now occupies the queening square of his g6-pawn. Black keeps running with his king. 38…Kg5! 39.g7 Qxc3 Black begins his counterattack. Of course it shouldn’t work, but passive defense is out of the question. 40.Qf7? An important loss of tempo. After 40.Nc1! Rb6+ 41.Qb3, White is still winning. 40…Qe1+ 41.Nc1 Rb6+ 42.Qb3 Ba3?! Having salvaged his position, Black falters. 42…Rxb3+ 43.axb3 Ba3 44.g8Q+ Kh4 was equal. White’s king escapes to a2, so he isn’t mated, but nor does White have a material advantage.

In order to avoid giving White’s king an escape square on a2, Black cleverly avoids taking White’s queen with his rook. But there’s a big difference between a pinned queen and no queen at all! 43.g8Q+ Kh4

44.Qgc4+? 44.Rd1! won, because 44…Qxd1 fails to 45.Qg3+!, Kxh5 46.Qh3+, and one of White’s queens will end up capturing Black’s a3-bishop: 46…Kg5 (46… Kg6 47.Qxb6+ and 48.Qxa3) 47.Qe3+ f4 48.Qxb6 axb6 49.Qxa3, or 48… fxe3 49.Qxe3+ and 50.Qxa3. 44…Rxb3+ is answered by the simple 45.Qxb3 Qxd1 46.Qxa3. A fascinating demonstration of the power of a pinned piece! 44…e4 45.g3+ Kh3 46.g4+ Kh4

Now White has no checks and no way to prevent …Qxc1 mate. 47.Rd1 Qxd1 0-1 In the previous example, the pendulum swung back and forth, and there were some tricky tactics involved. In the next example, though, White is just lost, for move after move (sometimes more lost than at other times), until Black blunders and the game turns around completely. Once again, the power of the kingwalk to confuse can’t be underestimated.

Grooten – Ligterink Leeuwarden 1981

Black has a dominating position, with an extra pawn and a developing attack against White’s king. White can’t retreat his e5-bishop, because then Black has 22…Nxf4. Giving in with the dismal 22.Bf3 is hardly any better, so White starts a speculative counterattack. 22.Qa5?! c6? 22…Nxf4! was not only still possible – it destroys White’s position. 23.Bf1 Qc1+ 24.Kf2 fxe5 25.Be2 Qb2 26.Bxh5 Rf8

Despite his missed opportunity a few moves earlier, Black is still winning – White’s king is simply under attack, while Black’s king is safe. Rather than

resigning, White’s king goes for a walk. 27.Kg3!? exf4+ 28.exf4 Qd4 29.Rf1 Be6 30.Rf2 Bd5 30…Bxh3 clips another pawn, but 30…Qd6!, pinning White’s f4-pawn, was even stronger. 31.Bg4?! g5! 32.Qc7 gxf4+ 32…h5! 33.Bxh5 Rxf4! was better (34.Rxf4 Qg1 mate). 33.Rxf4 Qg1+ 34.Kh4 Qe1+ 35.Kg5

35…Qxd2? Now was the moment for 35…Rxf4, and White has no good way to recapture: 36.Kxf4 loses to 36…Qf2+! 37.Kg5 h6+! 38.Kxh6 Qf6+ 39.Kh5 Bf7+, while 36.Qxf4 Qe7+ 37.Kf5 Be6+ forces White’s king to the e-file, after which 38…Bxg4+ wins a piece. Now, in contrast, White’s king is suddenly safe and Black’s king is in danger. 36.Bf5! Seizing his opportunity. 36…Bf7 37.Qe5 Re8

In what must have been a time scramble (it’s interesting how often kingwalks happen at this stage of the game), Black misses a difficult defense: 37…Qd5! held the position. After 38.Bxh7+! Kxh7 39.Rh4 Bh5! (there is no other move, since 39…Kg8? 40.Rh8 is mate) 40.Rxh5+ Kg8 41.Rh8+ Kg7 42.Rh7+ Kg8 leads to either a draw by perpetual check or a drawn endgame if White exchanges queens. 38.Bxh7+! Kf8 39.Qh8+ Ke7 40.Qf6+ Kd7 41.Qxf7+ Re7 1-0 The previous examples have been a bit complicated, so let’s turn to some lighter fare, where the ultimately successful kingwalks were every bit as bad as they looked. West – Rogers Melbourne 1976 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.d4 exd4 5.Nd5 Nxd5 6.exd5 Nb4 7.Bc4 Qe7+

8.Kd2?

White invited this position, and his judgment was correct. After 8.Kf1!, he would stand better. 8.Kd2?, on the other hand, looks very dubious, and indeed it is. 8…Qc5! 9.Re1+ Be7 10.Qe2 b5! 11.Bb3 Bb7

Black’s vigorous play has exposed the folly of White's eighth move. If White’s king were safely on g1, Black would be in trouble because of the pin on the e-file, but White’s king is not safely on g1… 12.Qe5 Bxd5? Black had many alternatives, including the calm 12…Kf8. 13.Qxd4?

Missing 13.a3!, which wins two pieces for a rook after 13…f6 14.Qxe7+ Qxe7 15.Rxe7+ Kxe7 16.axb4. 13…0-0 Now Black is just a pawn up with a good position, so White decides to go down fighting. 14.Rxe7?! Qxe7 15.Bxd5

15…Rae8 This isn’t bad, but 15…c5! 16.Qe4 Qd6 picks up a piece. 16.Bb3 White saves the piece, but now his king goes for a very dangerous walk. 16…Qe2+ 17.Kc3

17…Nc6 17…c5 would end it. 18.Qd3 b4+ 19.Kc4

19…Re4+? 19…d5+! was crushing, as after 20.Kxd5, both the human 20…Rd8+ followed by 21…Rxd3, and the even stronger 20…Qxf2!?, as recommended by the engine, leave White’s position in ruins. 20.Kb5 was no better, because of the geometric 20…Rb8! 21.Ka4 Qe8! But after 19…Re4+?, White's king actually finds a haven on a4. 20.Kb5 Rb8+? 21.Ka4

21…Qxd3? The last chance was 21…Re5! 22.Nxe5 Qxe5 23.Bd5, with dynamic equality. Now White’s two bishops will have their say. 22.cxd3 Rg4 23.g3 Rb6 24.d4 Kf8 25.Bf4 d6 26.d5 Nb8 27.Rc1 Nd7 28.Rxc7 Nc5+ 29.Rxc5 dxc5 30.Ne5 Rxf4 31.gxf4 Ke7 32.Ka5

It’s alive! White’s king emerges from its crypt. 32…Rf6 33.Kb5 Rxf4 34.Kxc5 Rxf2 35.d6+ Kd8 36.Nxf7+ Ke8 37.Kc6 Rd2 38.Be6 1-0 The next two examples are from 3-minute games, which at least partly explains the missed opportunities. But kingwalks themselves are often inherently difficult, for both sides, so there is some instructional, as well as entertainment, value to these examples.

Harper – Anon ICC 2011 (three minute)

The position is equal, with both players having about a minute and a half left. 20.Rf3 maintained the balance. However, after a considerable amount of thought, White sees a tactic, but so does Black. 20.Bxe7 Rf2+ 21.Kh3 h5 22.Kh4 The problem with White’s kingwalk is that his king isn’t heading into safety, but rather into checkmate. The first condition for a kingwalk to safety therefore isn’t met. But White’s only alternative, 22.Rh1, led to a dismal position. 22…Rxh2+ 23.Kg5

23…Bh6+? 23…d5!, opening the sixth rank for Black’s queen, won on the spot. Black sees this idea, but a move too late. 24.Kg6 d5+? Now was the time for the mundane 24…Rxe7, taking a piece and bringing Black’s rook to g7. 25.Bf6? 25.Rf6! was winning. To be fair, both players had less than 15 seconds remaining. 25…Re7? 25…c6!, covering the important d5-square, kept rough equality, if such a thing is possible in a game like this. 26.Nxd5? There always time for another mistake. 26.Qxd5+ won, but White had his heart set on the knight fork, and wanted to bolster his defense of f6. 26…Rg7+ 27.Kf5

27…Qd6? Both 27…Qc6 and 27…Rf7! won for Black. 28.Ne7+ Rxe7 29.Qxd6 Good enough, especially since the game was now going to be decided on time (in White’s favor), but 29.Qb3+ and 30.Rxd6 was better. 29…cxd6 30.Bxe7 Ne3+ 31.Ke6 1-0

Black lost on time. The next illustration is a little less frightening for White.

Harper – Anon ICC 2011 (three minute)

Black has a slight advantage and the most sensible continuation is 28…Qf7, getting out of the pins on the e-file as well as the h4-d8 diagonal. 28…Rag8?! 29.f4! Rxg3 30.fxe5 Qf7 31.exf6 Now White has gained the advantage, and Black’s next move makes things worse. 31…Rg1+ 32.Kd2 Qb3

White now has two winning options: 33.Re7 and 33.f7. He chooses a third alternative.

33.Bf5? Qxb2+ 34.Ke3 The start of a kingwalk, to be sure, but Black is winning. 34…Qc1+ 35.Kd4

35…Rd8+? Black overlooks the best way to defend against the threatened mate on h7 – 35…Rh1!, with a winning advantage. White also missed this when he played 33.Bf5? 36.Ke5 Qxc3+? This was Black’s last chance for 36…Rh1! 37.Ke6 Qb3+ 38.Ke7

Consistent and photogenic, although 38.Rc4! mated as well, because Black can’t defend h7. 38…Qg8 39.Qxh7+ Qxh7+ 40.Bxh7 Rc8 41.Bf5 White now is comfortably winning, especially since Black had less than ten seconds remaining. 41…Rf1 42.Bxc8 42.Rh4+ Kg8 43.Be6 was mate. 42…Kg8 43.Be6+ Kh7 44.Bf7 b4 45.Rh4 mate 1-0 There and back again The third sub-category in this chapter features kingwalks that end up back home. The kingwalks themselves can be either completely sound or not so sound – here we’re looking at things from the artistic perspective.

Cardoso – Petrosian Las Palmas 1975

Black is two pawns up, but his king is still in the center and White has annoying pressure. Black’s best is to consolidate with 27…Be7!, but Petrosian takes a characteristically different approach. In view of the inexact play that follows from both players, it’s reasonable to assume that both players were short of time. 27…Kd7!? 28.Bb4 Qc6 29.Qa7+ Ke6? 29…Qc7 was forced. After 30.Qxa6 Rb8, Black still has an edge, although White has regained one of his pawns. 30.Rg3? 30.Rh5! won immediately, as White threatens 31.Qf7 mate. After 30…Qe8 (30…Rf8 31.Bxf8) 31.Qxa6+, Black’s position collapses. 30…e4 30…d4!, covering the c3-square, was much better. 31.Qf7+ 31.Rc3! would have brought White’s rook into the attack.

31…Ke5 And now 31…Kf5! was more precise. 32.Rc3 Qe6 33.Rc7 Kf5? 34.Rd7! Re8 35.Rd6 Qc8 36.Rxd5+? 36.Rxf6+! gxf6 37.g7 would give White adequate counterplay. 36…Kg4!? 37.Rd1? Qe6 38.Rg1+?! Kh3

38…Kf3! was better, but Black’s kingwalk has still been impressive. 39.Qa7 e3 40.Qb7 Qe4 41.Qd7+ Re6 42.Qd1 e2 43.Qe1 Qe5

After many ups and downs, Black’s kingwalk has succeeded. His king is safe, his e2-pawn cripples White’s pieces, and he has threats against White’s king as well. There now follows a remarkable return journey.

44.Rh1+ Kg4 45.Qg1+ Kf5 46.Qf2+ Kxg6 47.Rg1+ Kf7 48.Re1 Kg8

Unnecessary, but Black could now afford a touch of artistry, after his lengthy near-death experience. 49.Kd2 Bg5+ 50.Kd3 Qd5+ 0-1 In the next game, both sides had chances, but ironically Black’s fatal error only occurred after White’s king had completed his journey.

Miles – Ljubojević Bugojno 1978

Having equalized in a quiet opening, Black now seeks to liven things up, especially since the alternative is a potentially dispiriting defense of his isolated d5-pawn. 13…Bg4 14.Bxg4 Bxh2+?! This was Black’s idea, but this standard sacrifice comes in two basic versions: sound and unsound. This one belongs in the unsound category. 15.Kxh2 Nxg4+ 16.Kh3

This is hardly a noteworthy kingwalk, in and of itself. The alternative, 16.Kg1?, was better for Black after 16…Qh4 17.Rfe1 d4!

16…Qg5 17.Qd4 Qh5+ 18.Kg3 Nh6 19.Rh1! Nf5+ 20.Kf4

Since the endgame after 20…Qxh1 21.Rxh1 Nxd4 22.cxd4 is pleasant for White, especially in view of his active king, Black keeps the queens on the board, in order try to exploit White’s advanced king position. In any case, we can be sure that Ljubojević didn’t sacrifice his bishop in order to go into a bad endgame. 20…Qg6 21.Qxd5 Re6

This was the move Black was relying upon. 22.Qxf5? It’s always fun to play with a bunch of pieces against a queen, but here White had a much stronger move – 22.g4!. After Black’s f5-knight retreats, White’s king slips away to g3, with a piece in hand.

22…Rf6 23.g4 Rxf5+ 24.gxf5 Qg2!

Annoying. Now White’s “kingwalk up the board to safety” has turned into “White’s king is on the fourth rank and Black has an active queen.” 25.Raf1 g5+! 26.fxg6 Forced, because continuing forward with 26.Ke5? leads to disaster after 26… Re8+ 27.Kf6 h6!, weaving a mating net. Black’s threat of 28…Qf3!, followed by 29…Re6+, is decisive. 26…fxg6 27.e4 Now the chances are equal. 27…Rf8+ 28.Ke3 Qf3+ 29.Kd2 Rd8+ 30.Kc2 Qd3+ 31.Kc1 Rc8 32.Rd1

White’s king has made it all the way back to the first rank, and no one could tell that White’s didn’t just castle on the queenside. The position would be dead even after 32…Rxc3+ 33.Bxc3 Qxc3+, but Black now blunders, presumably in time trouble. 32…Qxe4? 33.Rd8+! Kg7 34.Rd7+ Kf6 35.Rhxh7 Suddenly Black’s king is in the middle of the board, but the resemblance ends there. 35…Qe1+ 36.Kc2 Qxf2+ 37.Kb1 Qf1+ 38.Rd1 Qg2 39.Nd5+ Kg5 40.Bf6+ 1-0 Black loses his queen to a knight fork.

Chapter 8 Kingwalks in the Opening One of the first things beginners learn is to “castle quickly.” This isn’t bad advice, although with more experience players realize that there are advantages to delaying castling (which itself might be considered one of the most fundamental types of kingwalks). As long as the king remains in the center of the board, with the possibility of castling on either side, the opponent can’t launch a flank attack against your castled king. After all, it could wind up on the other side of the board! Delaying castling is one thing, but giving up the right to castle with an early king move is something else altogether. The other thing beginners are told is to “develop your pieces,” and the king usually isn’t considered to be a “piece” in this context. Nevertheless, there are times when an early kingwalk is justified, sometimes to avoid exchanges, sometimes to keep an intact pawn structure, sometimes to defend a key point. These factors, and others, may provide an objective basis for a kingwalk in the opening. In short, the goals of a kingwalk in the opening aren’t all that different from the kingwalks we’ve already looked at. But, as we shall see, often the psychological effect of an early kingwalk is even more pronounced than for later kingwalks. It is impossible to discuss kingwalks in the opening without starting with the (in)famous “Steinitz Gambit,” which allowed Black to drive White’s king into the center of the board at a very early stage. Steinitz had his own, quite fixed, ideas about this type of thing. Posterity, and modern chess engines, agree with Morphy’s understated comment regarding Steinitz: “His gambit is not good.” Steinitz – Neumann Dundee 1867

1.e4 e5 2.Nc3 Nc6 3.f4 exf4 4.d4!? Qh4+ 5.Ke2

Here it is – Steinitz’s idea of how to play the King’s (or, more accurately, the Vienna) Gambit. The first point to be made is that, in contrast to most King’s Gambits, at least back in the Romantic Era, White is not playing for checkmate. Perhaps more accurately, he is not playing to checkmate Black – whether he’s playing to checkmate himself is another question. Instead White has a positional focus – by diverting Black’s e5-pawn to f4, he has obtained an “ideal” central pawn duo, and he hopes that his central superiority will be enough to give him a permanent, structural advantage. As compared to the venerable Queen’s Gambit, there are two flies in the ointment. One is that White has to recover his sacrificed pawn by capturing Black’s f4-pawn. The other is that the cost of preventing a rapid Black counterattack in the center with …d7-d5 is that White’s king is forced into the center by …Qh4+. It is this kingwalk that marks the “Steinitz Gambit.” 5…d6?! Since this is not an opening textbook, we won’t concern ourselves with precise variations. Let’s just say that the engine would be quite happy to play

Black, since it has a slightly different set of priorities than did the first world champion. 6.Nf3 Bg4 7.Bxf4! Bxf3+?! Black is playing with charming naiveté. 8.Kxf3!

Interestingly, 8.gxf3?! is nearly playable, answering 8…Qxf4(!) with 9.Nd5. Steinitz’s move is stronger, although Black still does not stand worse. 8…Nge7 The engine recommends 8…g5. 9.Be2 0-0-0 10.Be3? Steinitz’s calculating abilities were not always up to the level of his ideas. But White consistently emphasizes protection of his center over protection of his king. 10…Qf6+ 11.Kg3

White’s king wanderings continue. 11…d5! 12.Bg4+ Kb8 13.e5 Qg6 14.Kf2 h5! 15.Bh3

15…f6?! 15…Nf5! was almost decisive. 16.exf6 Qxf6+ 17.Qf3 Qxf3+?! 17…Qh4+. 18.gxf3 1-0 (34)

Now the position is equal – a triumph for White’s opening, all things considered, This was enough for Steinitz, who won in another 16 moves. The next game is annotated by Kasparov in Predecessors, where he refers to 5.Ke2 as “nonsense in modern-day chess.” This is certainly true of the Steinitz Gambit itself, but, like all generalizations, it is not always true. It’s worth mentioning that Steinitz employed this variation against all comers, as Paulsen was one of the strongest players in the world at the time this game was played. Steinitz – Paulsen Baden-Baden 1870 1.e4 e5 2.Nc3 Nc6 3.f4 exf4 4.d4 Qh4+ 5.Ke2 d6 Kasparov analyzes 5…d5!? 6.Nf3 Bg4 7.Bxf4 0-0-0

8.Ke3 What could be more natural? 8…Qh5 9.Be2 Qa5? The right move was 9…g5!, as played in Barle-Portisch, Portorož/Ljubljana, 1975 – and as instantly recommended by the engine some 50 years later. 10.a3! Bxf3 11.Kxf3!

11…Qh5+ 12.Ke3 Continuing his natural play! 12…Qh4 13.b4 g5 14.Bg3 Qh6 15.b5 Nce7 16.Rf1!

It’s never too late to castle! 16…Nf6 17.Kf2 Ng6 18.Kg1 1-0 (36)

Not even the most eagle-eyed scout could find a trace of White’s early kingwalk. It is amazing to consider that one-third of White’s moves have been with his king, yet his remaining moves have been efficient and White is almost fully developed, with a strong attack against Black’s king. Steinitz won in another 18 moves. Steinitz also defeated Zukertort with his gambit. Steinitz – Zukertort London, 1872 1.e4 e5 2.Nc3 Nc6 3.f4 exf4 4.d4 Qh4+ 5.Ke2 d5 6.exd5 Bg4+ Steinitz won against a broken Zukertort in the final game of their world championship match in 1886, after 6…Qe7+?! 7.Kf2 Qh4+ 8.g3 fxg3+ 9.Kg2 Nxd4? 10.hxg3. In this opening, the direction of White’s kingwalk varies… 7.Nf3 0-0-0 The engine offers the calm 7…Nce7 as leading to equality. 8.dxc6 Bc5 9.cxb7+ Kb8 10.Nb5 Nf6

11.Kd3! In the spirit of the opening, and also best, according to the engine. 11…Qh5 12.Kc3? But this is too much of a good thing – 12.c3!, solidifying White’s center, held onto White’s advantage. 12…Bxd4+? 12…a6 or 12…Ne4+ was better. 13.Nbxd4 Qc5+ 14.Kb3 Qb6+ 15.Bb5 Bxf3 16.Qxf3 Rxd4 17.Qc6 Qa5 18.c3 Rd6 19.Qc4 a6 20.Ba4 Nd5 21.Ka3 1-0 (51)

White’s king has made it all the way across the board along the third rank. After 21…g5 22.b4 Qb6 23.Qd4, the queens came off, leaving White with a winning endgame. Steinitz also bequeathed us the following exhibition game, from a different opening, which featured a remarkably early – and rapid – kingwalk along the third rank. Steinitz – Steinkuehler London 1874 (exhibition) 1.e4 e5 2.Nc3 Bc5!? 3.Na4!?

What follows is so astounding that we have to conclude that not only was this variation known at the time the game was played, but that there was also a tacit consent to enter into it. 3.Nf3 gives White an effortless advantage. 3…Bxf2+!? This begins a sequence of forced moves. 4.Kxf2 Qh4+ 5.Ke3 Qf4+ 6.Kd3 d5

White is up a piece, and “all” he has to do is consolidate… 7.Kc3 Qxe4?! The engine gives the more positional (!) 7…d4+ 8.Kb3 Qxe4, after which Black has two pawns for the piece, and White’s (extra) a4-knight is awkwardly placed. 8.Kb3? Thematic, but bad. 8.d4! was right. 8…Nc6 It was still not too late for 8…d4, but now White starts to untangle. 9.c3 b5? Tactically justified, but it is a mistake to give White’s a4-knight the c5square. 10.Nc5 Na5+ 11.Ka3

Black’s queen and knight can’t get the job done on their own, and none of Black’s other pieces are developed. 11…Nc4+ 12.Bxc4 Qxc4 13.d4! a5 14.Qe2! Qxe2 14…b4+ is met by 15.Ka4!, and everything is covered. 15.Nxe2 e4 16.b3 1-0 (40)

White’s king soon slipped away to b2, leaving White with a decisive positional and material advantage. Leaping ahead a hundred years, and skipping many other examples of early kingwalks, we now look at some more modern games, starting with a curious contribution by super-grandmaster Victor Korchnoi. It’s not easy to categorize Korchnoi’s kingwalk, which removes the king to a safer file but also must have been intended to provoke his opponent.

Korchnoi – Plachetka Luhacovice 1969 1.d4 Nf6 2.c4 c5 3.d5 e5 4.Nc3 d6 5.e4 Be7 6.g3 0-0 7.Bg2 Nbd7 8.Qe2 a6 9.a4 b6

White has adopted a somewhat unusual setup against the Czech Benoni, combining a kingside fianchetto with the positioning of his queen on e2. All of this is aimed at Black’s usual plans (…Be7-g5, exchanging dark-square bishops, and an eventual …f7-f5 break). With “normal moves,” White would retain an advantage, but Korchnoi decides to change the nature of the position with a short kingwalk. 10.Kd1?!

The main point of 10.Kd1?! is to get White’s king off the e-file, since White’s idea is to play 11.f4, which will lead to an opening of the center. In some variations, Black also might gain a tempo with …Be7-h4+, which is no longer possible once White’s king has left the e1-square. This is deeper stuff than voluntarily giving up the right to castle, but that doesn’t mean it’s good. 10…Ne8 The engine suggests the more active 10…b5!? 11.axb5 Nb6, with counterplay. 11.f4!? f5?! Black overreacts to White’s aggression. After 11…exf4 12.gxf4 Nc7, Black’s position has a considerable amount of potential energy. 12.exf5 exf4 13.Bxf4 It’s a bit surprising that Korchnoi didn’t play the greedier 13.g4, but White ends up holding onto his f5-pawn anyway. 13…Ne5?! 14.g4 b5?! Black is trying too hard to punish White for his cheeky tenth move. 15.Bxe5 dxe5 16.axb5 Nd6 17.Nf3 Bb7 18.Kc2

White’s king arrives at its intended destination, and is delighted to find that Black’s position is in ruins. Korchnoi now goes on to win pawn after pawn, keeping an escape route for his king in reserve. 18…axb5 19.cxb5 e4 20.Nd2 Rxa1 21.Rxa1 e3 22.Qxe3 Bg5 23.Qxc5 Bxd2 24.Kxd2 Qg5+ 25.Kc2 Qf4 26.Re1 Nc4 27.Re2 Qxg4 28.d6! Bxg2 29.d7 Threatening 30.Qxf8+! 29…Nd6 30.Qxd6 Qxf5+ 31.Kb3 Qf7+ 32.Kb4 1-0

With the kingwalk completed, Black threw in the towel. Not all early kingwalks are as dramatic as the one in the previous example. In the next game, Karpov employs what is now a standard maneuver in what used to be a fashionable line of the Nimzo-Indian Defense. The purpose of his kingwalk is mainly to guard important squares, negating possible counterplay by his opponent. Karpov was always quite good at that. Williams – Karpov Nice 1974 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Bg5 h6 5.Bh4 c5 6.d5 d6 7.e3 Bxc3+ 8.bxc3 e5 Again, we will refrain from detailed opening analysis, as entire books have been written on this variation (and on every other opening under the sun…)

The strategic lines are clearly drawn – White has two bishops and a spatial advantage in the center, but in return Black has a compact position without weaknesses, and hopes to prove the truth of Canadian GM Duncan Suttles’s jocular maxim that “if you have two bishops, one of them must be bad.” As things turn out, both of White’s bishops end up being bad. 9.Bd3?! This inexact move plays into Black’s hands. 9…e4! 10.Bc2 g5 11.Bg3 Qe7 12.h4 Rg8 13.hxg5 hxg5 14.Ne2 Nbd7 15.Qb1

15.Qb1 might be considered an “all purpose move,” as it puts pressure on Black’s e4-pawn and also inhibits the development of Black’s queenside by attacking his b7-pawn. Black finds an elegant counter to White’s idea. 15…Kd8! 16.a4 a5 17.Ra2 Kc7!

Black’s kingwalk, which has brought his king to the queenside, has two ideas behind it. The first is that Black’s king defends White’s invasion squares along the bfile, nullifying White’s pressure in that part of the board in the most efficient manner possible. The second is that, by vacating the e-file, Black makes it possible to bring a rook to e8 to help protect his e4-pawn, should it be necessary. The black king’s walk isn’t over yet, though! 18.Rh6 Ra6 19.Qb5 Kb8! 20.Rb2 Ka7

Black’s king is now completely secure, without requiring more than his a6rook and c8-bishop for support. This leaves Black with a preponderance of force in the center. White recognizes this and quickly withdraws his queen.

21.Qb3 Ng4! Since Black’s a6-rook defends his d6-pawn, Black can play this move, evicting his opponent’s h6-rook and preparing to utilize his f7-pawn as a battering ram. 22.Rh1 f5! 23.Kd1 Rb6 24.Qa2 Rxb2 25.Qxb2 b6 The positional disaster begins to unfold – White’s c4-pawn becomes a target, and while White can defend it, his piece coordination is hampered by having to do so. 26.Bb3 Ba6 27.Nc1 Nde5 28.Qe2 Ng6 29.Kd2 Nf6 30.Qd1 f4 0-1

White resigned, without having yet lost any material. But his position is hopeless, the relative safety of the opposing kings being one of many factors that give Black a decisive advantage. The following example is more in the spirit of Steinitz than Karpov – Black’s kingwalk is bold, dubious and highly successful, all at the same time. It meets the psychological requirements of the position, although perhaps not the objective requirements. Let’s enjoy seeing Sveshnikov play the variation of the Sicilian Defense that bears his name. Ivanović – Sveshnikov Krk 1976

1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e5 6.Ndb5 d6 7.Bg5 a6 8.Bxf6 gxf6 9.Na3 b5 10.Nd5 f5 11.Bd3 Be6 12.c4 Qa5+ 13.Kf1 Bxd5 14.exd5 Nd4 15.cxb5 axb5 16.Nc2 Nxc2 17.Qxc2

White has played for this position and is happy. He has lost the right to castle, but so will Black, who can’t allow 18.Bxf5, and White is also about to win a pawn on the queenside. 17…e4 18.Qc6+ Ke7 19.Bxb5 Ra7

20.Qe8+?! The engine, which never succumbs to temptation, recommends the cool 20.a4! It is less impressed by Mednis’s suggestion of 20.a3. 20…Kf6 21.g4?!

White can’t resist trying to punish Black for having a bad king position. The problem is that Black’s king position isn’t actually all that bad, and White’s unjustified aggression is creating weaknesses in his own camp. 21…Re7 21…Bg7 was also good. 22.Qb8

22…Ke5?

This audacious move, which defends Black’s d6-pawn, is so thematic and aesthetically pleasing that it is hard not to think it’s also strong. This illusion led Mednis to award it an exclamation mark, but in fact it gives away Black’s advantage. While Black now threatens 23…Bg7, trapping White’s queen, this threat is easily parried by the withdrawal of White’s b5-bishop with 23.Be2!.

The drawbacks to Black’s audacious kingwalk then become apparent. Rather than Black’s king being “completely safe in the middle of the board,” in fact it is “under attack from all directions,” as a few variations make clear. In answer to 23.Be2, the natural 23…f4? fails immediately to 24.Qc8!, with the deadly threat of 25.Qf5+, while 23…fxg4 runs into 24.Qb3 (threatening 25.Qg3+) 24…Bh6 25.h3. Black’s best is 23…Bg7 24.Qb3 f4 25.Re1, and in contrast to the game, White has play against Black’s king, although any result is possible. 22… Qb4! was the right way to defend Black’s d6-pawn. 23.f4+?! Kxf4

24.Ke2? This is the fatal mistake. After 24.Be2!, White still had a playable position. 24…Ke5! 25.Rhf1 fxg4 White was hoping for 25…f4? 26.Rxf4! Kxf4 27.Qxd6+, but now it is clear that White has given up two pawns for the illusion of an attack. 26.b4 Bg7 27.bxa5 Rxb8 28.Rab1 f5 0-1 (38)

The endgame is completely winning for Black, as White’s queenside pawns are harmless, while Black’s central pawns, supported by his King, are unstoppable. Early kingwalks can also be conducted to keep your pawn structure intact, although the usual balance must be struck between positional considerations of that sort and the dangers involved. Here is a quick example. Miles – Rivas Tilburg 1978 1.c4 Nf6 2.d4 e6 3.Nf3 b6 4.Bf4 Bb7 5.e3 Be7 6.h3 0-0 7.Nc3 d5 8.cxd5 Nxd5 9.Nxd5 Bxd5

10.Bd3!? Bb4+ 11.Ke2!? Bd6 12.Bxd6 Bxf3+?! 13.Kxf3!? Qxd6

14.Qc2 f5 15.Rac1 f4?! 16.Qxc7 16.Bxh7+ was also good. 16…Qd5+ 17.Ke2 Nd7 18.Qc6?! 18.Bc4! would have given White a decisive advantage. 18…Qh5+ 19.f3?! And now 19.Qf3! was better. 19…Rad8 20.Qxe6+ Kh8 21.e4 Nf6 21…Rfe8! would have given Black good play for his two-pawn deficit, although White would still have an edge. 22.Qe5! Qg6 23.Rhg1 Qg3? Black’s last chance was 23…Rfe8. 24.Rc7 Rfe8 25.Re7 Qh2 26.Kf1 1-0 (37)

Having neatly consolidated everywhere, White won in a few more moves. The next two examples, from the same event, form a striking and instructive duo. They are from my own games (co-author Seirawan) and both feature the same mirrored path for their kingwalks (Ke1/e8-d2/d7-c2/c7). But the resemblance ends there, as the kingwalks have completely different objectives. Kovacević – Seirawan Wijk aan Zee 1980 1.e4 d6 2.d4 Nf6 3.Nc3 g6 4.Be2 Bg7 5.g4!?

It hardly needs to be said that this is one of the most aggressive lines in this variation of the Pirc Defense. Black follows the first principle that all Pirc/Modern players have internalized – Stay Calm.

5…c6 6.g5 Nfd7 7.h4 b5 8.h5 Rg8

9.hxg6?! An instructive error: White releases the tension on the kingside too soon. White could continue with 9.h6!?, content with having prevented Black from castling kingside, or just leave his pawn on h5. The further course of the game must have come as a big surprise to White. 9…hxg6 10.Nf3 b4 11.Nb1 a5 12.a4? White’s play is logical, but his premise, rather than his logic, is faulty. He reasons that he has the initiative on the kingside, and therefore closing the queenside will deny Black play on “his” side of the board, leaving White on top. The flaw in the argument soon becomes apparent – in fact White doesn’t have the initiative on the kingside. 12…c5 13.d5 Nb6 14.c4?!

White continues with his plan, but now gets a rude shock. 14…Kd7! The main purpose of this move is to clear the back rank, so Black can contest the h-file that White went to so much trouble to open. The secondary purpose of 14…Kd7! is to bring Black’s king to c7, so that when more lines open in the center, it is safely tucked away. 15.Nbd2 It is suddenly not easy to find moves for White. Given the topic of this book, it is only natural to see whether White can match Black’s kingwalk with a kingwalk of his own. Unfortunately, the equally imaginative 15.Kd2 Rh8 16.Kc2 Kc7 17.Kb3? ends in disaster after 17…Bd7, lining up to fire upon White’s a4-pawn, as …Qe8 will follow shortly. 17.Ra2, with the idea of 18.b3, comes too late in this variation, as Black can sacrifice his b4-pawn with 17…b3+!, and the “closed queenside” turns out not be so closed after all. White therefore concedes control of the h-file. 15…Rh8! 16.Rg1 Kc7

17.Rb1 Rh3! A powerful move, as the h8-rook makes way for the queen. The critical parts of the battlefield are the open h-file and the a1-h8 diagonal. Black’s queen will be an ideal guardian for both. 18.b3 Qh8! 19.Nf1 N8d7 20.Bf4 Ne5 21.Nxe5 Bxe5 22.Bxe5 Qxe5 23.f3 Bd7 24.Qc2 Qd4 25.Rg2 Rh1 26.Rf2

Black has a dominating position. There follows some nice chess geometry. 26…Qh8! At the time, I was quite proud of this queen retreat. As imposing as it appears on the d4-square, the queen doesn’t actually attack anything vulnerable. The redeployment in the game sparks a crisis in White’s hoped-for fortress.

The engine’s recommendations, 26…Bh3 and 26…Rah8, are a good example of the difference between human and machine “thinking” in chess. After both 26 Bh3 and 26 Rah8, as well as the more human 26 Qh8!, Black has a decisive advantage. 27.f4 Qh4! The point has become sharper. The queen now exerts a most unpleasant pin, making ideas like …Bh3 and …Rxf1+, as well as …Rh1-g1-g2, possible. Furthermore, with White’s f2-rook in need of a defender, White’s king can’t go anywhere and certainly can’t go for a walk. 28.Rd1 f6! Another open file is all it will take. 29.gxf6 exf6 30.e5 Desperation. 30…fxe5 31.fxe5 Rf8 32.exd6+ Kb7 33.Bd3 Re8+ 0-1

Black’s king ended up on b7, still completely safe, while White’s king, which never moved, will soon be mated. A gruesome punishment for prematurely opening up the h-file might have been 34.Be2 Rxf1+ 35.Kxf1 Qh1 checkmate.

The kingwalk in our final example has a completely different purpose – White’s king moves to cover a key invasion point, thereby refuting Black’s opening idea. In that sense, the game resembles Williams-Karpov, Nice 1974, discussed earlier in this chapter, but the kingwalk itself is much more jarring. Seirawan – Timman Wijk aan Zee 1980 1.c4 Nf6 2.Nc3 e6 3.e4 d5 4.e5 Ne4 5.Nxe4 dxe4 6.Qg4 Bd7?! A theoretical variation at the time, but it withers under the eye of the engine. 7.Qxe4 Bc6 8.Qe3 Na6 9.d4 Nb4

Black has sacrificed an important central pawn for which he has a lead in development and pressure against White’s center. The obvious threat is 9… Nc2+, and a move like 10.Qc3? is met by 10…Qxd4! 11.Qxd4 Nc2+ 12.Kd1 Nxd4, regaining the pawn, while 10.Bd3? fails because White’s g2-pawn is hanging. How to defend against the threatened fork? 10.Kd2!

Here is the answer! White’s king takes over, covering the c2-square. Frequently only moves can be very good ones as well. White has just enough time to consolidate, because Black has no quick pawn attacks against White’s center (10…c5 is illegal). 10…a5?! During the game I was mainly focused on piece sacrifices from two very different lines of attack by Black’s queen. The lines that most concerned me were 10…Ba4!? 11.b3 c5 12.bxa4 Qa5, as well as 10…Qh4 11.a3 O-O-O 12.Ne2 Bc5. In both cases, my king isn’t exactly safe in the center. Jan was quite skeptical of these sacrifices and so prepared a cautious retreat for his knight. 11.a3 Na6 12.Kc2

Now the first stage of White’s kingwalk is complete, and both sides develop. Black still hopes to exploit the position of White’s king, which would rather be on the kingside. 12…Qd7 13.Nf3 Be7 14.Bd2 0-0 15.Bc3 The engine recommends 15.h4!? 15…b5! 16.cxb5 Bxb5 17.Bxb5 Qxb5 18.Rhd1 Rfd8

Black has created counterplay and White’s advantage is minimal. So now White’s king retraces his steps, trusting to his extra pawn. 19.Kd2!? I definitely felt uncomfortable with my king hanging out on the queenside. It felt a bit too breezy over there. I reasoned that if I could play as in the game, my king would be a lot more secure on the kingside. Besides, the exercise was not only a good stretch of the legs – it gave me a concrete, active plan. 19…Nb4?! 20.Qe2 Qb7 21.axb4! axb4 22.Ke1! bxc3 23.bxc3 Qb3 24.Qd3 Rxa1 25.Rxa1 h6 26.Rb1 Qa3 27.Qc4 c5 28.Kf1! Qa8 29.h3 Rc8 30.Qb5 cxd4 31.cxd4 Qa2 32.Kg1

With each step my king took towards home, I became happier and more confident. Having completed one of the most attenuated examples of “castling by hand” you will ever see, I can now focus on the position, which is as simple as it looks – White is a pawn ahead. Jan was short of time and his position deteriorates with surprising speed. 32…Bg5? A bad time for active measures. Black’s best chance of resistance was to batten down the hatches with a sequence like …Be7-f8, …Rc8-d8, …Qa2-a8, controlling the d5-square in order to prevent White’s d4-pawn from advancing. There is no guarantee that progress could have been achieved against this defense. 33.Re1 Bf4? 34.Qd7 Rb8 35.g3 Qa8? Allowing a killer tactic. 36.d5! A dreamlike breakthrough. 36…Rd8 37.Qc6 Qxc6 38.dxc6 Bg5 39.Nxg5 hxg5 40.Kg2 Rc8 41.Rc1 f5 1-0 With the time control having been reached, the game was adjourned. I was looking at the adjourned position with Canadian GM Peter Biyiasas in a room adjoining the playing hall. I had found a way to spoil the winning position

when, much to my surprise, Jan came in to resign the game. After Jan had left, Peter and I shared a good laugh. Timing has always been important in chess.

Chapter 9 Kingwalks in the Endgame Kingwalks are hardly unusual in the endgame, as one of the most fundamental principles of good endgame play is to have an active king. In terms of fighting power, the king is often at least equal to a minor piece in strength, and may sometimes be stronger. We give the king a power of four in the material count. Since many endgames consist of extended kingwalks by both sides, often punctuated by moves by other pieces and pawns, we have kept this chapter short. It features a relatively small number of examples of particularly striking kingwalks that have dominated a game’s endgame to an unusual and striking degree. If instructional value is sought in these examples, it is that they may serve to remind the reader that the king can be a powerful piece in the endgame, and that it is capable of winning a game almost single-handedly. The most impressive kingwalks in the endgame have checkmate as their goal. We give two examples of this: the first example is a famous game from an elite tournament; the second is an unknown gem extracted from the mass of blitz games played online. The parallels between the games are disconcerting.

Topalov – Kasparov Linares 1999

Having missed an easier win earlier in the game, Black found another winning idea. Since Black’s rooks are doing all they can, Kasparov now embarks on a kingwalk along the light squares on the queenside, into the heart of White’s position. 44…Kc6 45.Kf3 Kb5 46.Bc5 Ka4 47.Bd4 Rd7 48.Ke3 Kb3 49.Ke2

Black’s king has been blocked and can’t make further progress. Kasparov now judges the position as ripe for a combination. 49…Rxg5 50.fxg5 Rxd4!?

The engine, which has no appreciation for the dramatic, recommends the more effective 50…f4, holding …Rxd4 in reserve. 51.cxd4 c3 52.g6 c2 53.g7 c1Q 54.g8Q

Black has a more active king and, importantly, it’s his move, but is it enough to win? 54…Qc4+ 55.Ke3 Kc3! The move Kasparov was relying upon. Black’s e6-pawn is defended and White’s d4-pawn is attacked. 56.Qd8? Natural and wrong, as White’s queen ends up out of play. The only defense was to play as actively as possible with 56.Qg5! 56…Qd3+! 57.Kf4 57.Kf2 is met by 57…f4. 57…Qd2+! 58.Kf3 Qd1+ 58…f4 was even better, although the finish wouldn’t have been as pretty. 59.Ke3 Qg1+ 60.Ke2 Qg2+ 61.Ke3 f4+ 0-1

White resigned at this point. The elegant point of Black’s play would be shown after 62.Kxf4 Kd3!

Position after 62…Kd3 (analysis) The threat of 63…Qg4 mate cannot be prevented by 63.Qg5, because then Black has 63…Qf2 mate. The role of Black’s king in this attack is striking, as White’s own pawns, in study-like fashion, prevent any counter checks. Unexpected mating patterns in the endgame need not involve a friendly queen.

Anon – Nakamura ICC 2015 (five minute)

The position is equal, but this is a 5-minute game and there was no question of agreeing to a draw until every possibility had been exhausted. Black starts a kingwalk, but surely it will reach a dead end? 37…Kd7!? 38.h4 Kc6 39.Rh2 Qa5 40.h5 Qd8 41.h6 Kb5 42.Rf2 Ka4 43.Qc2 b5

Here is Black’s first idea – his king supports his only pawn break (…b5-b4). But Black’s king doesn’t really accomplish much on a4 – it has to go farther, which seems very unlikely. 44.Rf6 Qe7 45.Bd2 a5 46.Kg2 Qb7 47.Kg3 Qa7 48.Rf1 b4!?

It’s either this or face reality, and in a 5-minute game the latter is a nonstarter. 49.axb4 Kb5 A sad necessity, in view of the threat of 50.Ra1+. 50.Ra1 a4 51.Kf2 Rh8 52.Ke2 a3 53.Bc1 Qa4 54.bxa3 This wasn’t really necessary, as the position was in a strange type of equilibrium, but 54.bxa3 shouldn’t lose. 54…Rxa3 55.Qxa4+ Rxa4

56.Rxa4? This exchange costs White the game, as he can’t stop Black’s king from continuing its invasion. After 56.Rb1, it is difficult to see how Black can make progress. 56…Kxa4 57.Bd2 Kb3 58.Kd1 Ra8 59.Ke2 This makes is easier for Black, as his king now gets to c2. 59…Kc2 60.Be1 Kb1 61.Bd2 Kb2 62.Be1 Kc2 63.Bd2 Ra1 64.Be1 Ra4 65.Bd2 Ra2 66.Be1 Kc1+ 67.Ke3 Kd1 68.Bg3 Ra3 69.b5 Rxc3+ 70.Kf4 Ke2!

An apt finish to the game, echoing Kasparov’s kingwalk against Topalov. White is checkmated, this time by Black’s rook! 71.h7 Rf3 mate 0-1 The purpose of a kingwalk is, of course, not always checkmate, other than that is the ultimate goal in every chess game. In the following game, the purpose of White’s kingwalk is to ensure the promotion of a passed pawn. To achieve this goal, White abandons his queenside entirely.

Seirawan – Quinteros Mar del Plata 1982

Being an exchange and a pawn ahead, White has a decisive advantage. Black’s knights are beautiful, and in a different position they would matter, but here they aren’t particularly relevant. The engine recommends the calm 48.Bd1, protecting White’s a4-pawn and maintaining White’s material advantage. White chooses a completely different approach, based on the assessment that, while Black’s knights can destroy White’s queenside, it will take them too long to do so. 48.Bxh5!? gxh5 49.Rxh5 Nxa4 Black seeks counterplay by creating a passed a-pawn. 50.Rh7+ Kg8 51.Rc7 Nxb2

52.Kh3 A very human solution, as opposed to 52.g4!? White’s king will go straight up the h-file, speeding the passage of his g-pawn to its queening square. 52…a4 53.Kh4 a3 54.Kh5 a2 55.Ra7 Nxc4 56.Kh6 1-0

Since 56…Na5 fails to 57.g6, threatening mate (57…Kf8 58.g7+), Black resigned. It may be, when the opponent is tied up, that a king invasion deep into the opposing position will cause a collapse. This is what occurs in our next example, and the reader should keep the whole board in mind as the endgame develops.

Nakamura – Kaidanov St. Louis 2012

White has the bishop pair, a better pawn structure and more space. This translates to a nearly decisive advantage, but White has to demonstrate good technique to bring home the point. The solution is a slow-motion invasion of Black’s kingside by White’s king. The kingwalk starts normally, but some obstacles have to be overcome en route. 50.Kg2 Bd7 51.Kf3 Bc8 52.g4! hxg4+ 53.Kxg4 Bd7 54.Kg5 Kb7 It is clear that White’s threat to penetrate Black’s kingside is getting serious. 55.Be4 Rh8 56.Bf3 Ra8 57.Rc4 Kc7 58.Bc5 Rh8 59.Rc1 Bc8 60.Bf2 Bd7

61.f5! Perhaps not the only way to win, but the most incisive. 61…Rg8?! For better or for worse, Black had to play 61…gxf5. 62.f6! Rh8 63.Rc5! Ra8

64.Kh6 Now White resumes his kingwalk, adding the extra pressure that collapses Black’s defenses. It is extremely useful that White’s f4-pawn has been magically transferred to the f6-square. 64…Nf4 65.Be3 Nh5 66.Bxh5 gxh5 67.Kg7 Be8 68.Kf8! The king’s march hasn’t concluded yet! 68…Kb7 69.Ke7 Kc7 70.Bd2 1-0

The journey of the white king from g1 to e7 is now complete, and Black gave his hopeless position a rest. We end this chapter with a fairly easy illustration of the power of the king in the endgame. But, as will be apparent from the final chapter in this book, it also shows that the kingwalk is alive and well in the twenty-first century as an important chess technique.

Roeder – Rozentalis Charleroi 2018

The position is roughly equal. White threatens 27.Rd8+, so the first move of Black’s kingwalk is innocuous enough. 26…Kf8 27.Rd8+? The decisive mistake, as White’s attack against Black’s b4-a3 pawn chain is too slow. After 27.Kf1, White’s king would keep Black’s rook out of his position, while White’s rook would do the same with respect to Black’s king. 27…Ke7 28.Rb8

28…Kd6

Objectively, 28…Re1+ 29.Kg2 Re2 was even stronger, as White’s a2-pawn falls to Black’s rook. But it is understandable that Black instead launches his king into the heart of White’s position, as it has a ticket for Nimzowitsch’s express, attacking White’s c6-bishop along the way. 29.Bf3 Kc5 30.Kf1 This fails, for while Black’s rook is constrained, Black’s king penetrates White’s position along the a1-h8 diagonal. Eating crow with 30.Rd8 wouldn’t have been any better, though. 30…Kd4! Of course. Black’s king shows that it is quite the bully. 31.Rxb4 Kc3 32.Rb7 Kb2

And that’s it. White tries to fight on by taking Black’s pawns along the seventh rank, but Black’s new queen is what matters. 33.Rxc7 Kxa2 34.Rxf7 Kb2 35.Rxg7 Ra5 36.b4 Ra4 37.Rf7 a2 38.Rxf6 a1Q+ 0-1 (49) White resigned some ten moves later.

Chapter 10 Double Kingwalks What could be better than a really good kingwalk? The answer to this trick question is easy – two kingwalks! The games in this chapter feature kingwalks by both players. In some cases, the kingwalks occur simultaneously, while in other cases one side completes its kingwalk before the opponent’s kingwalk starts. As we have discussed earlier, it may be that a kingwalk by one side can be met by a kingwalk by the other side – the last game in this chapter features an unusual example of this. But usually the kingwalks just happen, for different reasons, and with different degrees of success. So this chapter is mainly for fun.

Alekhine – Chajes Karlsbad 1923

This game has rightly been used as an example of how strong Alekhine had become by 1923, four years before he became world champion. He plays over the whole board to exploit the positional advantages that he has amassed. But we will look at the game from a slightly different point of view – our eyes will be trained on the king maneuvers carried out by both sides. White starts by bringing his king to the center. 35.Ke2 Kf8 Black, anticipating White’s ultimate breakthrough on the kingside, begins a kingwalk of his own. Both players sense what is coming so they abandon the kingside. 36.Kd2 Rb7 37.Bf3 Ke7 38.Rhe1 Nf8 39.Nb4 Kd8 40.Kd3 The engine spots the chance opportunity of 40.Bxd5 cxd5 41.Nxd5 Qg6 42.Nb4, followed by the advance of White’s c- and d-pawns. It is understandable that not even Alekhine would notice (or risk) such a “computer variation” – such things were rarely even considered in the precomputer era.

40…Rge7 41.Qd2 Ra7 42.Rh1 Rec7 43.Rh2 Bg6 44.Qe3 Kc8 45.Rch1 Kb7

Black has completed his kingwalk to the apparent safety of the queenside. The key word is “apparent.” 46.Kd2 Re7 47.Nd3 Nd7 48.Bh5 Ra8 49.Bxg6 hxg6 50.Rh7 Rae8 51.Ne5 Nf8 52.Rh8 Rg7 53.Nf3 Rb8 54.Ng5 Re7?! 55.Qe5! The exchange of queens eliminates an important defender. 55…Qxe5 56.fxe5 Ra8 57.Rg8 b4 Desperation. 58.Rhh8 Ree8 59.axb4 Ka7 60.Kc3 Ka6 61.Nf7 Rec8 62.Nd6 Rd8 63.Rh1! This is why White refrained from the more obvious 61.Nh7. Alekhine has seen that Black’s king, rather than being safe, is in a mating net. 63…Rd7 64.Ra1 1-0

We give the next game in its entirety, because if you skip even a few moves, you might miss one of the kingwalks! Tal – Simagin Leningrad 1956 1.e4 c6 2.d4 d6 3.Nc3 Nf6 4.f4 Qb6 5.Nf3 Bg4 6.Be2 Nbd7 7.e5 Nd5 8.0-0 Nxc3 9.bxc3 e6 10.Ng5 Bxe2 11.Qxe2

The young Tal was going to attack you, and it would probably work. This wasn’t well known in 1956. Simagin has seemingly provoked his opponent into attacking. To that extent, his strategy admirably succeeds. Simagin may have shared the viewpoint of Pacific Northwest master Victors Pupols, who staunchly maintained that “Youth is basically unsound!” Proving the truth of such a statement can be tricky.

11…h6 12.Nxf7! You don’t have to be Tal to make this sacrifice. 12…Kxf7 13.f5 dxe5?! 14.fxe6+ Kxe6

This is not a kingwalk – Black’s king has just been blasted into the center of the board, where it is extremely vulnerable. Tal now can’t resist complicating further. The engine assesses the position after the calm 15.Be3 as clearly winning for White. 15.Rb1!? Qxb1 16.Qc4+ Kd6 17.Ba3+ Kc7 18.Rxb1 Bxa3 19.Qb3 Be7 Black hangs onto his material for dear life. Otherwise he is just losing. 20.Qxb7+ Kd6 21.dxe5+ The remorseless engine again suggests a quiet move: 21.Rd1! 21…Nxe5 22.Rd1+ Ke6 23.Qb3+ Kf5 24.Rf1+ Ke4

Regardless of what you call it, Black’s king has been all over the place – in stark contrast to his rooks, which haven’t yet moved. 25.Re1+ Kf5 26.g4+ Kf6 27.Rf1+ Kg6 28.Qe6+ Kh7 29.Qxe5 Rhe8 30.Rf7 Bf8 31.Qf5+ Kg8

Let’s now call Black’s king wanderings a kingwalk, as His Majesty has gone from e8 to c7 to e4 and all the way back to g8. It is the next series of moves that earns this game a place in this chapter. 32.Kf2?! Bc5+ 33.Kg3 Re3+

34.Kh4

34…Rae8

Mednis cites Tal for the following drawing line: 34…Be7+ 35.Kh5 Rd8 36.Rxg7+ Kxg7 37.Qg6+ Kf8 38.Qxh6+ Kf7 39.Qxe3 Rh8+ 40.Qh6 Rxh6+

41.Kxh6, but there is a flaw in this analysis – White can provide his king with an escape square by first throwing in 35.g5!, thereby justifying his kingwalk. 35.Rxg7+! Kxg7 36.Qxc5 Black was unable to defend this endgame, especially with the time control approaching. 36…R8e6 37.Qxa7+ Kg6 38.Qa8 Kf6 39.a4 Ke5 40.a5 Kd5 41.Qd8+ Ke4 42.a6 Kf3 43.a7 Re2 44.Qd3+

44…R2e3

45.Qxe3+ 1-0 While many moves would have won, the accurate 45.Qf1+ Ke4 46.Qf5 mate does the job best.

In the next example, White’s king successfully escapes a speculative (and unsound) black attack, while Black’s ensuing kingwalk is doomed to failure. Gipslis – Plachetka Tbilisi 1977 1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nc3 d6 6.Be3 Nf6 7.f4 Be7 8.Qf3 e5 9.Nxc6 bxc6 10.f5 Rb8 11.Bc4 Qa5 12.Bb3 Nd7 13.Qh5 d5 14.Bd2 Rxb3 15.cxb3 Bb4

16.a3 Ba6?! 16…0-0 was safer and, far more importantly, would have kept this game out of this book. 17.Rd1 Nc5? But this is Black’s real mistake. 17…Bxc3 was still playable for Black. The attack initiated with 17…Nc5? is unsound. 18.axb4 Nd3+ 19.Ke2 Qxb4

20.Kf3?! A very human reaction to the threat of discovered check. The engine cannot see ghosts, and points out the stronger 20.Qg5!, when none of Black’s fearsome discovered checks actually accomplishes anything. 20…Qxb3 21.Qg5 dxe4+ 22.Nxe4 And here White could continue his kingwalk with 22.Kg3, which would simplify matters. 22…Ne1+ 23.Kf2 Nd3+ 24.Kf3 Mednis gives this as the only winning move, overlooking that after 24.Kg3 Nf4+ 25.Nc3! cuts off Black’s queen from the attack, after which it’s White’s turn. 24…Ne1+ 25.Kg4

This is also good enough, because Black’s army is too scattered and too weak to take advantage of White’s advanced king position. 25…Qxd1+ 26.Kh3 Qb3+ 27.Bc3 Now White’s king is completely safe, and Black’s king position collapses. Black’s e1-knight is hanging, so he doesn’t have time for a defensive move, even if he could find one. 27…Nd3 28.Qxg7 Rf8 29.Ra1! Qc2 30.Bxe5 Qe2

Black counterattacks, hoping beyond hope that his king can come up with a miracle kingwalk that will get it to safety. But where on the board could Black’s king possibly be safe? 31.Nf6+ Ke7 32.Bd6+

The engine is not attracted by this sort of move. It would have White play 32.Bg3!, solidly defending White’s king and asking Black “what are you going to play?” 32…Kxd6 33.Qxf8+ Kc7 34.Qxf7+ Kb6 35.Nd7+ Kb5 36.Qb3+ Nb4 37.Qa4+ Kc4

Black’s king isn’t any better off on the fifth rank than it was at home. 38.Rc1+ Kd5 39.Qd1+ The engine assesses the position after 39.Nf6+ Ke5 40.Ng4+ as something like +53 in favor of White, so presumably 39.Nf6+ wins even more convincingly. 39…Qxd1 40.Rxd1+ 1-0 The endgame is hopeless, especially since White’s f5-pawn is ready to run. Our next game is one of the most well-known in world championship history, and arguably cost Viktor Korchnoi the title. The game was full of drama, and also featured two quite different kingwalks.

Korchnoi – Karpov Baguio City 1978

White has played powerfully and now takes an important pawn, answering Black’s counterattack with a kingwalk. Even though the first time control had already been reached, both players were soon in intense time trouble again. 44.Qxf5 Qd2+ 45.Kg3 Nhf6 46.Rg1 Re8?! 47.Be4! Ne7 48.Qh3 Rc8 49.Kh4

49.Qh4 was also good, with Kh3 to follow. 49…Rc1 50.Qg3 As Kasparov and the engine point out, the endgame after 50.Qg2! was winning for White.

50…Rxg1 51.Qxg1 Kg8?! Karpov starts a kingwalk of his own, hoping to complicate matters and confuse his opponent. 52.Qg3?! Kf7 53.Bg6+ Ke6?

Not all kingwalks are good! After the more modest 53…Kf8, both sides would have chances in a complicated and extremely tense position. 54.Qh3+ Kd5 55.Be4+? A tragedy for Korchnoi. 55. Bf7+ Kc6 56.Qe6+ led to mate. 55…Nxe4 56.fxe4+ Kxe4 57.Qg4+ Kd3 ½-½ (124)

The second time control was reached and Black’s kingwalk had – miraculously – succeeded. White exchanged queens with 58.Qf3+ Qe3 and the game continued for another 67 moves, ending in a draw. Our final example in this chapter features two kingwalks occurring at more or less the same time, with one kingwalk triggering the other. The play is even more striking because it was a rapid game, with both players short of time.

Anand – Nakamura Paris 2018 (rapid)

After complications, a position has arisen in which there is a dynamic material balance (White has a rook, bishop and pawn for a queen) and any result is possible. Both sides are playing to win, and White begins with a kingwalk to the queenside. 33.Kf1!? Qb6 34.Nc5 Kh6!? Black moves his king to a safer square, but also has in mind a kingwalk of his own. Black reasons that if White is serious about transferring his king to the queenside, why not use his own king to invade White’s abandoned kingside? Thus we get an example, in a pure form, of one side’s kingwalk positively inviting an opposing kingwalk. 35.Ke1 Ng3? A miscalculation which costs Black two tempi. He should hustle along with 35…Kg5! 36.Kd1 Kh5 37.Kc2 Kh4 38.Kb1 Nh5 39.Ka2 Kg3

The effect of the two kingwalks is striking – in part because both kingwalks have been long ones. 40.Nd3! Qd4! 41.Rc2 Nf6 42.Nc5 Qd1 43.Kb3 g5 44.Nd3 Nd7 The complications that follow are almost unfathomable, and the fact that the players were able to navigate them as well as they did when so short of time is impressive. 45.Kc3 Nb6? 46.Nxe5? 46.d6! should win. 46…Qe1+ 47.Kb3 Qb4+ 48.Ka2 Nxa4 49.Bxa4 Qxa4+ 50.Kb1 Qd4 51.Nc4 a4 52.d6 Qd1+ 53.Rc1 Qd3+ 54.Rc2 b5 55.Nd2 b4 56.Kc1 b3 57.Rc3 Qxd6 0-1

Chapter 11 Unsuccessful Kingwalks There is nothing quite as entertaining as someone failing – provided it isn’t you (or someone you’re cheering for)! In this chapter we give a fine assortment of examples of kingwalks that turned out badly – usually because they never should have been started in the first place. One of the most entertaining aspects of failed kingwalks is that they are voluntary, at least up to a point. We have already seen a number of examples of kingwalks that were “successful” – in that they ultimately worked out – but which could well have ended badly. In this chapter, the kingwalks do end, most definitely, badly… It is worth mentioning that some of the examples in this chapter come perilously close to king hunts, and straying into those would considerably lengthen our book. The first example is charming in that Black shows astounding naiveté in playing as he did against one of the titans of chess.

Capablanca – Black New York 1916

White has an advantage, but a long fight looms. Black is now possessed with the idea of a kingwalk, which in this position is a truly terrible idea. Perhaps he was actually possessed by demons? 38…Kf8?! 39.Nc4 Ke7? 40.Na5 Kf6?

In three short moves, Black has destroyed his position with his ill-advised kingwalk. While it’s hard to see just what Black’s king would do on f6 at the best of times, the fact that he has walked into a mating net makes that question entirely moot. White now completes the transfer of his knight to c6, ensuring that Black’s king is beyond the point of no return, so Black loses a lot more than some

tempi. 41.Nc6 Rc8 42.h4 Bc7 43.f4 Bxe4 Black’s problem is that if he captures on f4, he gets mated by White’s queen on the a1-h8 diagonal. Such is the radiant power of a self-mate. 44.g5+ hxg5 45.hxg5+ Nxg5 Now both 45…Kg6 and 45…Kf5 fail to 46.Ne7+ and 47.Nxc8. Everything is perfectly placed to ensure that Black loses. 46.fxg5+ Kxg5 47.Bxe4 Nxe4 48.Qxf7 1-0 (63)

Black, perhaps not fully aware that he was playing the Capablanca, played on for another 15 moves before resigning. A good warmup game for our chapter. Here is another Capablanca game, more famous than the previous example. It is hard not to comment on the lower level of chess technique exhibited by non-elite players one hundred years ago – Black’s kingwalk was not necessarily bad, but the way it was carried out was fatally flawed.

Capablanca – Schroeder New York 1916

Capablanca has just played 19.Ng5, and his opponent has three ways of meeting the threat of 20.Qxh7+: 19…g6, 19…f5, and 19…h6 (in descending order of strength). Black chooses the worst of the three options, which commits him to an unnecessary kingwalk. 19…h6? 20.Qh7+ Kf8 21.Qh8+ White must have felt, with some justification, that he could play whatever he wanted and still win. Things might have ended differently against the engine, which recommends the simpler 21.Ne4!, with a decisive advantage. 21…Ke7 22.Qxg7 hxg5 23.Qxg5+ Kd6 24.Ke2

Interestingly, Mednis includes this game in his book, but because of 24.Ke2, not because of Black’s kingwalk. 24…Rac8?! Having gone this far, Black should continue with 24…Kc6. By delaying this move, Black ends up in a complete bind. The finish of this game is well known. 25.Rc4 Kc6 26.Rhc1 Kb6 27.h4 f5 28.Qg7 Re7 29.Qe5 Rc6 30.Rxc5 1-0 While the reader, by this point, may have an uncontrollable urge to try some kingwalks, we can’t repeat often enough that kingwalks are not a panacea. Sometimes a position is just bad, and there is nothing to be done about it. That’s the situation in which Kotov finds himself in the following game. He can’t be blamed for rolling the dice by trying a kingwalk, but the kingwalk itself is not to blame for his loss.

Smyslov – Kotov Moscow 1943

24.Nf5! A completely sound sacrifice that, at the very least, increases White’s positional advantage. The engine recommends the sad reply 24…g5. 24…gxf5 25.gxf5 Nc7 26.Rg1 Ne8 27.Rg6 27.Bxh6 Bxh6 28.Rg6 would have broken through, but White prefers to increase the pressure. 27…Rf7 28.Rbg1

Black now decides to run for safety with his king, since passive defense will fail. 28…Kg8!? 29.Rxh6 29.Bxh6! 29…Kf8 30.Rh7 Ke7 31.Qh5 Kd6

Almost there, but it was better for Black to move his d8-rook, to free d8 for his king. That wouldn’t have saved him, though. 32.Bf4+ Ne5 33.Bxe5+ fxe5 34.f6! A lovely move, because there is no good way for Black to capture. 34…Nxf6 35.Qxe5+ Kc6 36.Rhxg7! Kb5

Black continues his zombie-like kingwalk out of momentum. 37.Nxd4+! Kb6 38.b4 If you stop and think about it, Black’s king’s journey really hasn’t brought him any relief. 38…Rc8 39.Rxf7 Qxf7 40.Qd6+ Rc6 41.Nxc6 Nxe4 42.bxc5+ 1-0 A powerful reminder that a rain storm can turn to hail. Some kingwalks can go either way. There is an objective truth to their soundness, of course, and the merciless engine reveals this truth (afterwards). But a single error by either side can decide the outcome. And sometimes there is more than one error, even in games by very strong players.

Karpov – Korchnoi Leningrad 1971

Black now has the choice between entering a roughly equal endgame, in which Black’s bishop pair balances his weaker pawn structure, or going for complications by counterattacking. While this game was played in a training match, Korchnoi would likely have gone for the complications in any situation. 17…Bf5!? 18.Qe5+ f6 19.Qe7+ Kg6 Black’s king starts its journey. 20.Rd2 Be3 21.Rf1 Bxd2+ 22.Nxd2

22…Qd4? Mednis gives 22…Qd4 an exclamation mark, commenting that it “is an absolute must for defending.” In reality, 22…Qd4 is a blunder, and 22…Qb6! was correct, but the reason why is far from obvious. After 22…Qb6! 23.Rxf5 Kxf5 24.Bd3+ Kf4, the chances are equal. 23.Rxf5 Kxf5 24.Bd3+ Kf4 25.Qd6+? White wins after 25.Qe2 Rhg8 26.Qf1+ Kg3 27.Qf3+ Kh2 28.Qh3+ Kg1 29.Nf3+. This is why Black’s queen should be on b6, not d4, so that the knight fork is avoided. To be fair, this type of variation can only be found with an engine. 25…Qe5 26.Qb4+ d4 Mednis questions this move, although there’s nothing wrong with it. Korchnoi’s recommendation of 26…Kg3 is fine as well. After 27.Nf1+ Kxg2 28.Qd2+ Kh3 (28…Kf3 is also equal) 29.Qf2 Rhg8 30.Bf5+ Rg4 31.Nd2 and the game is still balanced. But after 29.Bf5+? Kxh4 is not “unclear,” as stated by Korchnoi, but instead is completely winning for Black. We take no credit for these corrections, rather we tip our hats and exclaim what a difference chess engines make! 27.Ne4 Kf5 28.Qxb7 Kg4 29.Be2+

29…Kxh4? Remarkably, it is only this move that costs Black the game. Both 29…Kf5 and 29…Kf4 were still all right for Black. 30.g3+ Kh3 31.Nf2+ Kh2

32.Qh1+! Did this lengthy retreating move escape Viktor’s attention? 32…Kxg3 33.Ne4+ Kf4 34.Qf3 mate 1-0 Generally, players don’t intentionally play to lose games, just as people generally don’t intentionally drive off a cliff in their car. It’s possible there’s a cliff up ahead, but the realization that there is may come too late. White’s kingwalk in the next example bears the epitaph, “it seemed like a good idea at the time.”

Torre – R. Byrne Leningrad 1973

In this tense position, White, focused on his own attacking play, underestimates the strength of Black’s exchange sacrifice on e5. He should play 23.Qb2!, retaining an advantage. 23.Rag1? Rxe5! 24.fxe5 Qxe5 25.gxf5 Rxf5+ 26.Ke1 Rf7 27.Qc2 g6 28.e4?! White continues to play optimistically, but his position is getting looser. 28…Ba6! 29.R1g2 Qa1+ 30.Ke2 Rf1 31.Qd2 Rb1 32.Qg5 Re1+

33.Kf3?

A suicidal kingwalk, which presumably was based on a miscalculation. After 33.Kf2! Rf1+ 34.Ke3 Qd4+ 35.Ke2, or 34…Qc1+ 35.Rd2, the chances would be equal, although the position is harrowingly complex. 33…Qd1+ 34.Kf4 Rf1+ 35.Ke5 Qa1+ 36.Kxe6 Bc8+ 37.Kd6 Rf6+ 38.Kc7 Be6

White’s king is deep in enemy territory, which is precisely why White is now lost. White’s tripling on the g-file, at first, second and third blush, seems strikingly impotent. 39.d4 Qxd4 40.Rd2 Rf7+ 41.Kb8 Qxe4 0-1

Since White cannot prevent mate for very long, he didn’t try. Not all fatal kingwalks are as long and dramatic as the one in the previous example. In the next game, White either relies on a kingwalk to solve his

problems, or suddenly finds it to be his last, forlorn hope. The final mating position is a good illustration of how things can go very, very wrong.

Evans – Dorfman São Paulo 1978

Black’s position is superior, because his king is safer and his major pieces are better coordinated. White can try to coordinate his own pieces with 38.Re5!?, preparing to play his rook to e2 or e3, depending on what Black plays. Instead he relies on a kingwalk to save himself, but White’s king is actually worse off after it moves up the board. 38.Rc2? Rd3 39.Qc6 Rd1+ 39…Qd4 was even stronger. Centralization! 40.Kg2 Qd3! 41.Kh3

White probably was not under any illusions as to how this was likely to turn out. 41…Rg1! 42.Qc5 42.Rg2 Ra1 wasn’t any better. 42…Qf1+ 43.Kh4 Rg2! 0-1 White resigned at this point, in view of the variation: 44.Rxg2 Qxg2 45.h3 Qf3 46.Qg5 f6 47.Qg4 Qe4 48.a4 Kh6, when an instructive mate follows after 49…g5+!

Position after 48…Kh6 (analysis) We now come to two absolutely classic games, which are in the grey area between kingwalks and king hunts. Since in both games the kingwalks were entered into voluntarily, our conscience is clear in including them in this book. Our conscience is also clear in keeping the notes to these games brief, as both have been extensively analyzed elsewhere, including by the winners.

Shirov – Short Las Vegas 1999

15.Rad1!? In Fire on Board II, Shirov terms this “an extremely brave decision based on a total miscalculation.” With 15.Ne2, White would retain an advantage. 15…exd4 16.Rxd4 f5! 17.Ng3 Bc5! This was the move Shirov missed. 18.Rd2 Qb6 19.Qe2 Shirov also questions this move, suggesting that 19.Nxf5 Kf6 20.Qf3 “still offered some hope” – but not much, according to the engine. 19…Nf6 20.Qe5 Kg8 21.h3!? Up to this point, Black has been equal to the task. White therefore offers more material… 21…Re8! …which Black correctly accepts. 22.Qxe8+

Shirov’s honesty is refreshing: “No talk about brilliancies, it’s merely a sad necessity to give up the queen and pretend that White can still resist.” 22…Nxe8 23.Rxe8+ Kg7 24.b4 Qc6? With 24…Be7!, Black would be on the verge of bringing home the point (25.Rxe7?! Be6!). The move played prepares a kingwalk. Black must have thought that his material advantage was so great that an opportunity to simplify the position was bound to arise. 25.Rdd8! Bb6 26.Rg8+ Kf6?! As Shirov points out, 26…Kh6 27.Rxc8 Qxc8! 28.Rxc8 Rxc8 29.Nxf5+ Kg5 gave Black the advantage. 27.Nh5+ Ke5

Now this is a kingwalk! On paper Black has a big material advantage, but his lazy a8-rook has trouble getting into the game. In addition, Black’s king is hardly safe on e5 – this has not been a “kingwalk up the board to safety.” 28.Rd3! Shirov gives this move two exclamation marks, commenting that “Of course … after 28.Rde8+ Be6 29.Rxa8 Qxc3 it’s all over.” But in fact White can fight on after 30.Rg3 Qxb4 31.Rf3. Still, from the practical point of view, Shirov’s move sets Black more problems.

28…Qh6? An error in a complicated position. 28…f4!, which hasn’t been suggested by any commentators, would have won, as Black’s king would have a flight square on f5, should one be needed. 28…f4! also allows Black to develop his c8-bishop to f5. 29.Be2 Qc1+ 30.Rd1 Qxc3 31.Bf3 Be6 Shirov writes this is “…probably the decisive mistake.” This qualification is appropriate. 31…Bd7!? 32.Rxa8 Ba4 also held the balance. 32.Rxa8

32…Qxb4? In fact it is this move that loses the game for Black. 32…Qb2!, targeting the f2-pawn, was right. But with both players playing on increments, anything can happen. 33.Re8! Bd4 34.Rd8 Bb6 35.a3 Qa5 36.R8d2

Here is the problem – with White’s f2-pawn still on the board, Black can’t use his dark-square bishop to interpose when White checks on the e-file. Well, actually he can, but White ends up with a material advantage. 36…f4 37.Bg4 Bc4 38.Re1+ Be3 39.Rdd1 Bb3 40.Rb1 Bc2 41.Rb4 f5 42.Bf3 Be4 43.fxe3 White has been very patient, and now starts taking material. 43…Bxf3 44.gxf3 Qxa3 45.Nxf4

A nice finishing touch – 45…Qxb4 is met by 46.Nd3+ and 47.Nxb4. Black’s kingwalk to e5 failed, but some spectacular tactics by White and missed opportunities by Black were required. This illustrates once again that kingwalks can be dangerous!

45…b6 46.Rd4 1-0 The next example is one of the most famous ever games played – and rightly so. It is annotated in depth in Predecessors.

Kasparov – Topalov Wijk aan Zee 1999

When contemplating what follows, it’s important to remember that Topalov embarked on the ensuing kingwalk both knowing the risks and aware that Black had viable options. Both players must share the credit for the resulting masterpiece. 24.Rxd4! cxd4? 24…Kb6! was better, giving Black a slight advantage. Taking the rook loses – eventually! 25.Re7+! Kb6 25…Qxe7? fails because of 26.Qxd4+ Kb8 Qb6+, mating. 26.Qxd4+ Kxa5 27.b4+ Ka4

28.Qc3? The world’s best players took several days to find 28.Ra7!, which wins immediately, as after 28…Nxd5, 29.Rxa6+! diverts Black’s queen: 29… Qxa6 30.Qb2 Nc3+ 31.Qxc3 Bd5 32.Kb2!, followed by 33.Qb3+! and mate. The engine finds 28.Ra7! instantly… Now the complications continue. 28…Qxd5 29.Ra7! Bb7 30.Rxb7 Qc4 31.Qxf6 Kxa3? The engine also takes a fraction of a second to find 31…Rd1+, which leads to an endgame in which Black has chances: 32.Kb2 Ra8 33.Qb6 Qd4+ 34.Qxd4 Rxd4 35.Rxf7 a5 36.Be6 axb4 37.Bb3+ Ka5 38.axb4+ Kb6 39.Rxh7 Rf8! 32.Qxa6+ Kxb4 33.c3+! Kxc3 34.Qa1+ Kd2 35.Qb2+ Kd1 36.Bf1! Because 36…Qxf1 loses to 37.Qc2+ Ke1 38.Re7+. But Black has one more move to try. 36…Rd2!? 37.Rd7!

And thus ended what is possibly the most brilliant failed kingwalk in chess history! 37…Rxd7 38.Bxc4 bxc4 39.Qxh8 Rd3 40.Qa8 c3 41.Qa4+ Ke1 42.f4 f5 43.Kc1 Rd2 44.Qa7 1-0 It was Napoleon who told us “There is only one step from the sublime to the ridiculous.” This seems like the right way to transition from KasparovTopalov to the following example.

Anon – Harper ICC 2015 (three minute)

White is down an important pawn and his f2-pawn is hanging. He could resist with 24.Ng5! Ra7 (to cover the seventh rank) 25.Nh3, but instead decides to force the issue by invading Black’s position with his king. This is a terrible idea that leads to immediate disaster. 24.Kb4? Nc6+ 25.Kc5? Ncxe5+ 26.Kd6? Rd8+ 27.Kc7 0-1

It’s a shame not to give White’s final move a question mark as well. White resigned in view of 27…Rd7 mate. We close this chapter with another example of a kingwalk that turns into a king hunt. As with other types of kingwalks, failures are alive and well.

Wei – Bruzón Danzhou 2015

21.Nd5! A powerful line-opening move, aiming to exploit Black’s weakened king position. Black could reply 21…Bxd5, accepting a disadvantage, but instead gambles with 21…Nxd5, hoping that the kingwalk that follows isn’t fatal. 21…Nxd5 22.Rxf7! Strong, but also forced, as 22.exd5 Bxd5 favors Black (his d5-bishop covers f7). 22…Kxf7 23.Qh7+ Ke6 23…Kf6 fails to the simple 24.exd5. 24.exd5+ Kxd5

This was Black’s idea, as 24…Bxd5 runs into 25.Bxg6, and Black’s king can’t escape to d5. 25.Be4+! Kxe4

Black is a rook and a bishop ahead, but his hope is really that White will run out of pieces. 26.Qf7 A powerful and elegant quiet move, the only drawback to which is that 26.c4! forced mate – 26…bxc4 27.Qxg6+ Kd5 28.Qf7+, and 28…Kc6 fails to 29.Qxc4 mate. 26…Bf6 27.Bd2+ Kd4 28.Be3+ Ke4 29.Qb3 Kf5 30.Rf1+ Kg4 31.Qd3

White just keeps making threats that Black can’t defend against. 31…Bxg2+ Desperation. 32.Kxg2 Qa8+ 33.Kg1 Bg5 34.Qe2+ Kh4 35.Bf2+ Kh3 36.Be1 1-0

Mate can’t be prevented. Black’s king did its best, but it didn’t have much help. The deceased kings in this chapter could relate to the immortal words of Levon Aronian, when forced to resign to coauthor Seirawan in a bughouse game: “I fight alone!” His offending, guilty partner will go nameless. To mollify Levon, I explained to him that the three of us, all working together, were ready to clobber him badly. He was cheered, responding: “That is the only way you can win!.”

Chapter 12 Wilhelm Steinitz In the next chapters, we look at the kingwalks of some historical chess figures. Naturally, we start with the first world champion, Wilhelm Steinitz. Chess players have a distinct image of Steinitz – old, wise, balding, hugelybearded, irascible – and an inveterate kingwalker. In fact, there is much to question in this image. Steinitz was not always old and bearded, and as a young man had a full head of hair. His wisdom, and irascibility, increased as he aged. Most of all, it turns out Steinitz’s fame as a kingwalker stems largely from his experiments with the Steinitz Gambit (see Chapter 8). But we dutifully present the other kingwalks in Steinitz’s games – and there are some – to set the record straight. And because every kingwalk is worth a look. In this and the following chapters, games which have already appeared in previous chapters are referenced. The first example from Steinitz is a wild game, played in London while the civil war raged in the United States. Games from that era drive the engine crazy. Steinitz – Wilson London 1862 1.e4 e5 2.f4 exf4 3.Nf3 g5 4.h4 g4 5.Ne5 Nf6 6.Bc4 d5 7.exd5 Bd6 8.d4 Nh5 9.Bb5+ Kf8 10.0-0 Qxh4

White’s position is actually as dangerous as it looks. 11.Bxf4 g3? The counterintuitive 11…Bxe5!, eliminating White’s e5-knight, gives Black the advantage after 12.Bxe5 (12.dxe5 g3) 12…g3 13.Rf3 Bg4. 12.Bh6+ Kg8 13.Rf3? 13.Nf3 was a simple win, but that would be a retreat, and that would never do! 13…Qh2+?! 13…f6! 14.Kf1 Qh1+? 15.Ke2 Qxg2+ 16.Kd3?

A kingwalk, but to the wrong square. 16.Ke3 was right. 16…f6 Black is in the game after the ridiculously complicated 16…Bxe5 17.Rxf7 Nf4+. 17.Bc4? 17.Be8!, weaving an unusual mating pattern, was winning. 17…fxe5? Suicide. Or should we say a chivalrous suicide? 17…Qh3 was better. 18.Rf8+! Bxf8 19.d6+ Be6 20.Bxe6 mate 1-0 The next example is from Steinitz’s bloodthirsty match with Adolf Anderssen, in which every game was decisive. There were no concerns about the death of chess back then. Steinitz’s kingwalk is a light-hearted affair that hardly required calculation.

Steinitz – Anderssen London 1866

35.Reh1?! Simpler was 35.Kf3 Rxg4 36.Rxh7+ Qxh7 37.Qxh7+ Kxh7 38.Rh1+ Kg6 39.Kxg4. But Steinitz has calculated that his king can escape to the queenside. 35…Rxg4+ 36.Kf3 Rg3+ 37.Ke2 R3g7 38.Rxf6 Rg2+ 39.Kd3 R8g3+ 40.Kc4

Steinitz was right. Black’s counterattack has reached a dead end. 40…Re3 41.Rf8+ Rg8 42.Rxg8+ Kxg8 43.Rg1+ 1-0

The following kingwalk against Andersson is quite different from the one examined earlier. First of all, it ends very badly. Secondly, it was completely unnecessary, as Steinitz had a winning position, which he failed to win, then lost. This was not uncommon for Steinitz, as the depth of his ideas tended to outrun his tactical abilities, great as those were on occasion.

Steinitz – Anderssen Baden-Baden 1870

Having frittered away an even better position (!), Steinitz now misses a winning combination: 27.Nh6+! gxh6 28.Qxf6 Rxb4 29.Qxh6. Instead he falls victim to a sudden counterattack. 27.Bxc5 Rb2+ 28.Ke3 Qa5! Whoops. The old double attack. Black threatens 28…Qd2 mate, as well as 28…Qxc5+. 29.Rd1 Qxc5+ 30.d4 exd4+ 31.Kf4

The aggressive positioning of White’s knights doesn’t really go well with White’s unexpected kingwalk.

31…h6! 32.Nh3 Black, who isn’t actually down much material, calmly develops his attack. 32…Re8 33.Qd3 g5+ 34.Kf3 g4+ 35.Kg3 Rxe4 36.Qf1 Qe5+ 37.Kh4

The king won’t be safe up here, either. 37…gxh3+ 37…Rxb1 mated more quickly, by a few moves… 38.Kxh3 Rb3+ 38…Re3+. 39.g3 Rf4 40.Nxh6+ Kf8 41.Qc4 A mate threat! But it is Black’s move, and things end sadly quite soon. 41…Rh4+ 42.Kg2 Rxh2+ 43.Kxh2 Qxg3+ 44.Kh1 Qh3+ 45.Kg1 Rg3+ 0-1

What makes it even sadder is that White’s king ended up back on the first rank, but could find no happiness: it will be mated on its original square after 46.Kf2 Qg2+ 47.Ke1 Re3+ 48.Qe2 Rxe2 mate. Steinitz was not one to be flummoxed by an unexpected kingwalk. Bird tried one against him and was quickly routed. We have seen similar kingwalks in Chapter 11.

Steinitz – Bird London 1883

Black’s opening has not gone particularly well – he may have been thrown off by the fact that Steinitz has played more or less normal moves. Bird decides that a kingwalk to the queenside is just what the doctor ordered. He should have gotten a second opinion. 12…Kd7? 13.Bc4 f5 14.Rd1 Bxc3 15.bxc3 fxe4 16.Ba3

Steinitz is not one to be thrown by a kingwalk – at least not this one. 16…d5 17.Qxe4! b5 18.Bxb5 cxb5 19.Bxe7 Kc6

We detect a fault in Black’s development – he has only moved his king. 20.Qe5 Qd7 21.Qxh8 Qxe7 22.Re1 Qd6 23.Re8 Qd7 24.Rae1 Qc7 25.Qf6+ Kc5 26.R1e6

It is pretty easy to figure out which category of kingwalks this one falls into. 26…Bxe6 27.Rxe6 1-0 There is nothing breezy about the next example, which is from the pivotal seventh game of the first Steinitz-Lasker match. The score was 3-3, and it was by no means pre-ordained that Lasker would take the world championship title and hold it for 27 years.

Lasker – Steinitz Havana 1894

Lasker, playing with his customary ingenuity, has made it difficult for his opponent to convert his material advantage. As Kasparov points out, neither 33…Qg3 nor 33…Rg3, as recommended by various commentators, solve Black’s problems. Steinitz resorts to a kingwalk. 33…Kf7?! 34.Rhg1 gxf5 35.Qh5+ Ke7 36.Rg8 Kd6?! 36…Kd7 was safer. 37.Rxf5 Qe6?! 38.Rxe8 Qxe8 39.Rxf6+ Kc5

Black’s king has escaped from the kingside, but it has not found safety, and White’s h7-pawn is very annoying. 40.Qh6 Re7 41.Qh2 Kasparov awards this move two exclamation marks, commenting that White fails to win after 41.Qd2 Qd8, although the engine awards White a decisive advantage after 42.Qb4+ Kd3 43.a4. 41…Qd7 42.Qg1+ d4 43.Qg5+ Qd5 44.Rf5 Now everything is clear. 44…Qxf5 45.Qxf5+ Kd6 46.Qf6+ 1-0 Steinitz’s predilection for ruining winning positions became more pronounced as he aged, which is entirely understandable. In the following game, it is not his kingwalk that is at fault, but rather his failure to take advantage of his opponent’s more dubious kingwalk. Steinitz – Fritz Cologne 1898 1.e4 e5 2.Nc3 Nf6 3.f4 d5 4.d3 d4 5.Nce2 Bd6 6.fxe5 Bxe5 7.Nf3 Nc6 8.Nxe5 Nxe5 9.Ng3 Nfg4 10.Bf4 Qh4 11.Qd2 f6 12.Be2

White has come out of a somewhat modern opening with an advantage, which Black greatly increases with his next move.

12…g5? 13.Bxe5 Nxe5 14.Bh5+ Ng6 15.Qf2 Why not 15.0-0? 15…Bg4? 16.Bxg6+ hxg6 17.Qxf6 Rd8

White is clearly winning, and can hardly lose – can he? White now helps his opponent castle by hand, although he wins another pawn in doing so. Interestingly, the engine recommends just the opposite: 18.Kd2!, preparing to castle by hand himself. 18.Qe5+ Kd7 19.Qxd4+ Kc8 20.Qe3 20.Qxa7? would allow 20…Qxg3+! 20…Rhf8 21.Kd2!? Now 21.Qxa7! was strong. 21…Rf4? 22.Raf1 Rdf8 23.Qxa7 R8f6 24.Qa8+ Kd7 25.Qa4+ c6 26.Qd4+ Ke6

Black’s king has found itself back in the center. 27.Rxf4 Rxf4 28.Qg7 Rf2+ 29.Kc3

White’s king is much better off than Black’s. 29…Bh5 30.Rf1 The greedy 30.Qxb7 was fine, but White can hardly lose – can he? 30…Rxf1 31.Nxf1 Qf2 32.Nd2 b5 33.Qd4 Qxg2 34.Nb3 Bd1 35.Qg7! Qxc2+ 36.Kb4 Qf2

37.d4? Time trouble? 37.Nc5+! forces Black to give up his queen to avoid mate (37…Kd6 38.e5+ Kd5 39.Qd7+! Kxe5 40.Qe6+ and 41.Qe4 mate). 37…Qe1+ 38.Kc5 Qxe4 39.Na5? Qd5+ 40.Kb6 Qd8+

41.Qc7? Qxd4+ 42.Ka6 Qd6 43.Qxc6 Be2 44.Kb6 g4 45.b4 Bf1 46.a3 Bd3 47.Qxd6+? Kxd6 48.Nc6 Kd5 49.Na7 Ke4 50.Nxb5 Kf3 51.a4? Now, the disaster is complete. 51.Nd4+ was the last chance … for a draw. 51…Kg2 52.a5 Kxh2 53.a6 g3 54.Nd4 g2 55.Nf3+ Kh1 56.a7 Be4 0-1

Chapter 13 Aron Nimzowitsch Aron Nimzowitsch, the “stormy petrel of the chess world” (the term “storm petrel” describes a family of birds), was renowned for his deep, profound positional play, including such things as “mysterious rook moves” and “overprotection.” These terms come from Nimzowitsch himself, who some see as one of chess’ great positional players, while others see him as a selfpromoter who failed when matched against the elite players of his time. The truth, as is so often the case, lies between these two extremes. All good chess players must have confidence, and Nimzowitsch certainly wasn’t lacking in that department. He didn’t invent all the concepts that he claimed as his own, but he didn’t exactly appropriate them either. The ideas were there already, and Nimzowitsch might be thought of more as an interpreter of those ideas. It is therefore unfair to consider him a thief, and equally unfair to accuse him of fraud, although he was not as generous in acknowledging the discoveries of his predecessors as he might have been. But chess players owe Nimzowitsch a debt, for without him would we defend key points more than might be necessary for an engine with silicon nerves, and would we centralize our pieces and blockade passed pawns as often as we do? Regardless of how much of Nimzowitsch’s reputation is myth, as opposed to reality, one would expect to see kingwalks in his games, if only because Nimzowitsch was definitely an iconoclast by nature, and the respectable thing to do is to keep your king safely at home. And one does find kingwalks in Nimzowitsch’s games, but not kingwalks as extreme as one might have expected. In our first example, played early in his career, Nimzowitsch undertakes what by now the reader will consider a “standard kingwalk” to move his king out of danger.

Alekhine – Nimzowitsch St. Petersburg 1914

Black is under pressure. White’s major pieces are more active and Black’s king is a target. There is nothing Nimzowitsch can do about the first problem, but he solves the second problem with a kingwalk. 29…Kf8!? 30.Rh4 Ke8 31.Qh6 Ke7 32.Rf4 Qh8 33.Re1 Rc8 34.Rh1 Qg8 35.Qg5+ Kd6 36.Qe5+ Kc6 37.a4 Kb7 0-1 (75)

Mission accomplished. White is still better, but at least Black’s king is relatively safe. After Black defended tenaciously for some 20 moves, White slipped and Black gained counterplay on the queenside. Eventually Black was able to activate his central pawn majority (!) and start a winning attack against White’s king.

Bear in mind, this was the Alekhine of 1914, who was still very much learning his trade. Ten or more years later, Alekhine would not have let this type of advantage slip away! The next example has some strong artistic elements, as White’s king steps out of the way of his rooks, which move with great agility.

Nimzowitsch – Bernstein Karlsbad 1923

White has played vigorously on the queenside, and after some inexact defense he has achieved a winning position. The first step is for White to move his king so his rooks are connected. 23.Kd2! cxb5 24.Ra1 Nc6 25.Bxb5? Right capture; wrong piece. After 25.Nxb5, White kept his advantage. 25…Na5! 26.Be2 Rfb8 27.Na4 Bxa4 28.Rxa4 Rxb6?! 28…Qe6! would allow Black to capture on b6 with his queen. 29.Bc3 Nb3+? 30.Qxb3! Rxb3 31.Rxa8+ Ne8 32.Bd1!

The key to White’s queen sacrifice. If Black retreats his b3-rook, White continues with 33.Ba4, so Black must give up the exchange. 32…Rxc3 33.Kxc3 Qc7+ 34.Kd2 Kf7 35.Bh5+! g6 36.Rha1 Qb6 37.Be2 Kg7

Now White continues his kingwalk, if that’s the right way to put it, by bringing his king to g2, where it will be safe from harassing checks by Black’s queen. 38.Ke1! Nc7 39.R8a5 Kh6 Black’s king takes a brief stroll as well, but there is nowhere for it to go. 40.Kf1 Qb3 41.h5! Ne8 42.Ra6 Qb2 43.hxg6 hxg6 44.R6a2 Qb7 45.Ra7 Qb2

46.Kg2! Very pretty. White’s king moves off the first rank, threatening 47.Rh1 mate. 46…Nf6 47.Rh1+ Nh5 48.Bxh5 gxh5 49.Rha1! 1-0

A beautiful switchback. Black succeeded in blocking the h-file, but only at the cost of opening the sixth rank. Rooks are particularly good at exploiting open lines of this sort. What kept Nimzowitsch from gaining the world championship title, apart from happening to play at the same time as Capablanca and Alekhine, was arguably his nerves. This led to inaccuracies, which could be fatal against the strongest opponents. The next game is a good illustration of this, although his opponent failed to exploit his opportunities and Nimzowitsch ultimately won the game.

Sämisch – Nimzowitsch Karlsbad 1923

After an eventful and by no means error-free start, the players have arrived at a complex and balanced position. White should now continue with 26.Rg1!, as 26…Rxf2? loses immediately to 27.Be4! 26.Bf3? Kf8? Black is right to begin a kingwalk away from the dangerous g-file, but 26… Kf7! was more accurate. Black would then have a decisive advantage, because he could answer 27.Rcd1 with 27…Nc2!, trapping White’s queen. 27.Bxd4? Missing his chance, as 27.Rcd1! Nc2? fails to 28.Bxg7+, taking Black’s queen with check. 27…Qxd4 28.Bxe2 Bxe2 29.Qh3 Qxd5+ 30.Kg1 Bxf1 31.Qh6+ Ke8 32.Kxf1

32…Kd7! This isn’t essential, but finishing the kingwalk removes even the hint of counterplay, leaving Black with a winning material advantage. 33.Qxh7+ Kc6 34.Qh3 Rg8 35.b4 axb3 36.axb3 Kb7 0-1 (71)

Black went on to win in another 35 moves, although at one point in the queen and pawn ending White could have drawn. As a proponent of the French Defense, including the “Winawer” variation, which is also called the “Nimzowitsch” variation, Nimzowitsch more than once resorted to kingwalks under pressure to transfer his king from the kingside, White’s natural area of play in the French, to the safer queenside. Sometimes this worked out better than other times. In our first example, he gets away with it – barely.

Becker – Nimzowitsch Breslau 1925 1.e4 e6 2.d4 d5 3.Nc3 Nc6 4.Nf3 Bb4 5.e5 Bxc3+ 6.bxc3 Na5 7.a4 Ne7 8.Bd3 b6 9.Nd2 c5 10.Qg4 c4 11.Be2 Nf5 12.Nf3 h6 13.Qh3

13…Kd7!? A creative attempt to solve Black’s opening problems, but White doesn’t panic. 14.g4 Ne7 15.Nd2 Qe8 16.Ba3 Kc7 17.f4 Bd7 18.Qf3 h5 19.Nxc4?! A dubious sacrifice, given an exclamation mark by Keene in Aron Nimzowitsch: A Reappraisal. 19…Nxc4 20.Bxc4 hxg4 21.Qg2 Nf5 21…dxc4? loses to 22.Bd6+. 22.Bd3 Bxa4 23.Bxf5 exf5 24.Qxd5 Bc6 25.Qd6+

25…Kc8? Black must have miscalculated. After the natural 25…Kb7! 26.d5 Bb5, he would be on top. 26.d5 Rh6 27.e6 Bxd5? Both 27…Ba4 and 27…Rxe6+ 28.dxe6 Bxh1 were better. 28.Qxd5 Qxe6+ 29.Qxe6+ Rxe6+ 30.Kd2 Kb7 0-1 (61)

Nimzowitsch’s assessment was “White now has a piece for two pawns, but his own pawns are weak.” This is optimistic, to say the least. Black won in the end, but only because White misplayed the ending badly, not because Black had the better position.

Against his arch rival, Siegbert Tarrasch, who was by no means as pedantic as Nimzowitsch made him out to be, Nimzowitsch’s kingwalk was not as sure a thing as it first appears. Nimzowitsch – Tarrasch Baden-Baden 1925 1.c4 e5 2.Nf3 Nc6 3.d4 d6 4.Nc3 Bg4 5.d5 Nb8 6.g3 Nf6 7.Bg2 Be7 8.0-0 00 9.Kh1 Ne8 10.Qc2 f5 11.b3 Nd7 12.Bb2 g5 13.Nd2 Ng7 14.e4 f4 15.f3 Bh5 16.g4 Bg6

After a surprisingly hypermodern opening by Black, a closed position has arisen in which White’s king is poorly placed on the h-file. The answer, of course, is a kingwalk. But the timing is important too. 17.Kg1?! h5?! The engine recommends either 17…Ne8, preparing to play 18…c6, or 17… c6 right away. The logic is clear – if White’s king is going to the queenside, Black should liven things up on the flank to ensure His Highness receives a warm welcome. 18.h3 Kf7?! Too slow – Black should be playing for an advantage with 18…c6!

19.Kf2 Rh8 20.Ke2 c6 21.b4

21…Nf6? Characteristic play in the early 20th century. Encouraged by unprejudiced computers, a modern player would have opened lines on the queenside with 21…a5! The ideas of interfering with the opponent’s plan, rather than just carrying out your own, and of keeping an open mind to take advantage of changed circumstances are perhaps the greatest changes in chess in the “modern” era. 22.Qd3 Rc8 23.Rac1 Qb6 24.a3 c5 This ends the game. 25.b5 Qc7 26.a4 Qd7 27.Kd1 Bd8 ½-½

There is no real chance of either side opening lines. In another French, Nimzowitsch’s king makes it out of danger, but only just – running more than walking.

Spielmann – Nimzowitsch New York 1927

Black has played the opening either “provocatively” or “badly,” depending on how you want to look at it. Either way, White has a decisive advantage, which can best be exploited with the paradoxical preparatory move 19.Ke2!, with 20.Nxf7 coming. Both Alekhine and the engine recommend 19.Ke2! Spielmann, who once commented that he could see Alekhine’s combinations well enough, but not how he got the positions from which the combinations came, proves his point by sacrificing right away. 19.Nxf7?! Kxf7 20.Nxh5? Alekhine suggests no fewer than four ways to follow up the sacrifice (20.Bg6+, 20.Ne4, 20.Nf5 and 20.Ne2), but only the first two are good. The text just plain loses. 20…Bxh4+! Opening an escape square for Black’s king. This is the very move that 19.Ke2! would have prevented. 21.Ke2 Nxh5 22.Bg6+ Ke7 23.Bxh5

23…Kd7! Apparently this was the move White overlooked. 24.Qg7+ Be7 25.Bf7 Rh2+ 26.Kd1 Kc7 27.Bf4 Rxb2 28.Qh7 Kb6 0-1 (41)

Black’s king has escaped White’s attack. Black has a decisive material advantage, which led to victory in another 13 moves. It would seem these near-death experiences weren’t enough to deter Nimzowitsch, but against Steiner his luck ran out.

L. Steiner – Nimzowitsch Berlin 1928

Once more, Black’s position is simply bad. This time, Nimzowitsch’s patented kingwalk doesn’t help him. 14…Kd7!? 15.Ng5 Nh6 16.f3 Kc7 17.g4

The problem wasn’t Black’s king – it is that his entire kingside is weak. 17…Re8 18.Bg7 Ng8 19.gxh5 gxh5 20.f4 Nce7 21.Bxh5 The harvest begins. 21…Nf5 22.Bxf7 Nxg7 23.Bxe8 Nxe8 24.Nf7 1-0 (40)

There is nothing to be done about White’s h-pawn, which will cost Black a piece. Black gave up the piece right away with 24…Ngf6, but it didn’t help. The next example is something quite different – White’s kingwalk consists of a single move, but a very nice one!

Nimzowitsch – List Frankfurt 1930

After an inexactly played middle game, White has ended up with an edge because of Black’s last move (the passive 33…Rc8-e8, rather than 33…Rc7). Nimzowitsch now begins a combination which leads to victory, although it shouldn’t have led anywhere. Likely the following moves were played in a time scramble. 34.Rxf5!? Qxh4+ 35.Rh3 exf5! 36.Qxe8 Qd8 36…Qxd4 was simpler. 37.Qf7+ Kh6 38.d5

38…Qd7? Either 38…Rd7 or 38…b5! held without much difficulty, as White’s h3-rook is not participating in the attack. White’s queen alone cannot prevail. 39.Qf8+! White alertly seizes the opportunity presented to him. 39…Kg5 39…Kh7 fails to 40.Re3!, and suddenly White’s rook is back in the game with decisive effect. 40.Kg3! 1-0

Now it’s White’s king that steps up to the plate to help deliver mate. An unexpected finish! The following exhibition game features a kingwalk by Nimzowitsch’s opponent, just for a change of pace. The kingwalk goes straight into mate, however.

Michel – Nimzowitsch Bern 1931 (exhibition)

White has fought hard, but his weaknesses on the kingside make things difficult for him. Making a virtue out of a necessity, White has brought his king into the center. Can Black exploit it? 35…Qh3 36.Ke3?! Qf1? Missing 36…Rc8!, with the idea of 37…Re8+, among other things. 37.Rd1! Qg2 38.Qd2 Qh3 39.Qd5 Rc2 40.Rd2 Rc3+

41.Rd3 This was a good time for 41.Kd4!

41…Rc1 42.Qxf7 Qxh2 43.Qf5 Rc4 44.Rd1?! Rc3+ 45.Kd4 Rxa3

46.Kd5? 46.Rd3! had to be played. 46…Rc3? 46…Qe2! 47.Ke6? Qe2+ 48.Kf7 White’s king is right on time for his appointment in Samarra, as he goes straight into checkmate. 48…Rc7+! 0-1

Like all strong players, Nimzowitsch knew how to use his king in the endgame, and we give one striking example of Nimzowitsch’s king at its best.

Colle – Nimzowitsch Bled 1931

An endgame has arisen in which Black has the advantage, because of the light-square weaknesses in his opponent’s position. We won’t analyze the position in depth, but instead will draw the reader’s attention to Black’s kingwalks. 32…Kf7 Black’s king starts things off by going to d5. 33.Kg1 Ke6 34.Kf2 Kd5 35.Ke3 c5 36.a4 b6 37.Be1 Ne7 38.Kd3 h6 39.h4 g6 40.g3 Nc6 41.Bd2 Nd8 42.Be1 Ne6 43.Bf2 Ng7 44.Be1 Nh5 45.Bf2 Ng7 46.Be1 Ne8 47.Bd2 Nc7 48.Be1 b5 Black’s knight maneuvers haven’t accomplished much, so Black has to advance his queenside pawns. A pawn exchange is unavoidable. 49.axb5 axb5 50.Bf2 Ne6 51.b3 g5 Each pawn exchange brings White a little closer to a draw, but Black has to open lines on the kingside for his king, as well as creating a target on f4. 52.hxg5 hxg5 53.Be3 gxf4 54.gxf4

54…Kc6! Black’s king heads for the kingside. This is the only possible way to make progress. 55.Kc3 Kd7 56.Kd2 Ke7 57.Kc3 Kf7 58.Kd3 Kg6 59.Kc3 Kh5 60.Kd2 Kg4

61.Kc3? White could hold with 61.Kd3! Kf3 62.Bc1. Black can’t take White’s f4pawn with his knight, because White’s e5-pawn queens, and further progress is impossible. 61…Kf3 62.Kd3 c4+! If White’s bishop were on c1 or d2, this move wouldn’t work.

63.bxc4 bxc4+ 64.Kd2 Ke4!

It took Black ten moves, and some help from his opponent, to (in effect) play …Kd5-e4. 65.Ke2 c3 66.Bc1 Nd4+ 67.Kd1 Kd3

68.Ba3 Nb5 69.Bc1 Nd4 70.Ba3 Nf3 71.Bc5 c2+ 72.Kc1 Nd2 0-1 Our final example, from Nimzowitsch’s last tournament, sees him seriously out of form. But he goes down in style.

Nimzowitsch – Henneberger Zürich 1934

White’s position is creaking, and after 23.exf5 comes 23…e4! Nimzowitsch therefore tries a kingwalk, but it would take more than that to save the game. 23.Kd1!? Bd7 23…b5! was strong. 24.Kc2 b5! 25.Qe1 bxc4 26.Rd2 Ba4+ 27.Kb2 Rb8+ 28.Ka3 Why not? At least White’s king can cross visiting a3 off its bucket list. 28…Reb7 29.Kxa4

29…Qf7 Black had a very pretty forced win with 29…Rb4+ 30.cxb4 Rxb4+ 31.Ka3 Qe3+ 32.Rd3 cxd3 33.Qc3 Rc4!, but the move played doesn’t give anything away. 30.Bxc4 Qd7+ 31.Ka3 Qc7 32.Bb3 Qa5+ 33.Kb2 c4 34.Qd1 Qa4 35.Rf2 Rxb3+ 36.Ka1

Back to the first rank. 36…Rb2! would now force mate. 36…Qa5 37.Rc2 Rxc3 38.Rhh2 Rcb3 39.Qc1 c3 40.Rhg2 Rb1+ 0-1 For the record, White did develop his g1-knight with 13.Ne2. He just moved it back to g1 three moves later.

Chapter 14 Tigran Petrosian “Iron Tigran,” the ninth world champion, was a master of defense and maneuver, although many players fail to realize just how good he was at calculating variations (probably because he often used this ability to deaden the position, rather than to enliven it). Petrosian is also considered to be one of chess history’s greatest masters of kingwalks, and this reputation is well deserved. We have already seen many examples of his kingwalks in the preceding chapters. Even so, this will be one of the longest chapters in the book, so enjoy! In our first example, a young Petrosian brings his king close to the action before trying to break through.

Petrosian – Arutiunov Tbilisi 1945

There are many ways to play this queenless middle game, in which White has a big advantage. Petrosian chooses a kingwalk from g2 to c4, leaving it for later to decide whether to replace his king with a knight. 40.Kf3 Rb7 41.Bc6 Rb8 42.Ke2 Nc8 43.Kd3 Bd8 44.g4 h4 45.Ra1 Kf6 46.Bd7 Ba5 47.Bf5 Ne7 48.Kc4

Mission accomplished, although the engine judges the position as now being equal. 48…Be8 49.Rb1 Ng6 49…a6!

50.Bxg6 Kxg6 51.Nd3 Rb7?! 52.Rb2 Rf7 53.Rf2 Rxf2 54.Bxf2 The trade of rooks has greatly assisted White, who is now winning. To be fair, Black had to watch out for sacrifices on c5, which limited his options. 54…Kf6 55.Be1 Bd8 56.Nb2 Ke7

57.Kd3 The changing of the guard. 57…Kd7 58.Nc4 Bf6 59.Bd2 Bd8 60.a5 Be7 61.a6 Kc8 62.Bxg5! Bf8 63.b6 Kb8 64.Bxh4 axb6 65.Nxb6 Ka7 66.Be7! As mentioned earlier, Petrosian had a keen eye for tactics, which he usually employed in the service of his positional goals. 66…Bxe7 67.Nc8+ Kxa6 68.Nxe7 Kb5 69.Kc3 Kb6 70.g5 Kc7 71.g6 1-0 In the next game, Petrosian is unrecognizable, but it’s also hard to believe this game was played in the middle of the twentieth century, as opposed to a hundred years earlier. Oltetsian – Petrosian Yerevan 1946 1.e4 e5 2.Nf3 Nc6 3.Bc4 d6 4.c3 Bg4 5.Qb3 Qd7 6.Ng5 Nh6

Best was 6…Nd8!, defending everything. 7.Nxf7?! White sacrifices to divert Black’s queen from the defense of his c6-knight, but the immediate 7.Qxb7 was better. 7…Nxf7 8.Bxf7+ Qxf7 9.Qxb7

9…Kd7!? A romantic move, which unfortunately is only good for a draw. After 9…Rd8 10.Qxc6+ Bd7, Black will get more than enough counterplay with 11…Qg6! 10.Qxa8 Qc4 11.f3 The only move, but it’s good enough. 11…Nd4

This attack must have taken Petrosian back to the days of his youth (he was in his mid-twenties when this game was played), but Black doesn’t have quite enough forces to break through. 12.cxd4?! White enters into the spirit of the occasion, with a kingwalk of his own. 12.d3! was the more cold-blooded defense, leading to a draw by perpetual check, but with White giving the checks: 12.d3! Nxf3+ 13.gxf3 Qxd3 14.Qd5 Qxf3 15.Qb5+ c6 16.Qb7+ Ke6 17.Qc8+ Ke7 18.Qc7+. 12…Qxc1+ 13.Ke2 Qxh1 14.fxg4 Qxg2+ 15.Kd3

A chaotic, but balanced, position has arisen, in which material is somehow almost equal. 15…Qf1+ 16.Kc2 exd4 17.Qd5 Be7 18.Qf5+?!

White is in too great a hurry to simplify. 18…Qxf5 19.gxf5 g6! 20.Na3 gxf5 21.exf5 Rf8 22.Rf1 d5 23.Nb5 Bf6 24.Re1 a6 25.Na7 Bg7 26.Re6 Rxf5 27.Rxa6 Rf3 ½-½ 27…c5! would have retained Black’s advantage. The game was drawn after further play, but the moves aren’t available. The following game demonstrates that surprising possibilities exist in even the most mundane positions – if a player is willing to take chances. Petrosian plays a move that few players would even consider. Konstantinopolsky – Petrosian Moscow 1950 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 d6 5.c4 Bg4 6.d3 g6 7.Nc3 Bg7 8.Nd5 h6 9.b4

In this more or less normal position, Petrosian comes up with an astounding idea. 9…Kf8!? Even though 9…Kf8 is objectively suspect, we can’t help but give it some applause with an “!?” punctuation mark. Black reasons that his position would improve if he were free of the obnoxious pin on his c6-knight. Black can’t block White’s a4-bishop or take it, and while he could interpose his

own bishop with 9…Bd7, there is another option – move the piece on which the pin is based. And so he does. 10.Bb2 Nge7 11.h3 Bxf3 12.Qxf3 Nd4 13.Bxd4 exd4 14.Nf4 White can’t resist threatening 15.Ne6+, but now any hope of an opening advantage is gone. 14…Kg8 15.0-0 Kh7 16.c5 Rf8 ½-½ (38)

Who would ever know, looking at this position, that Black actually castled by hand? The position is equal, and after an eventful middlegame, in which White often held the upper hand, the game was drawn. The next example is far less surprising. White’s brief kingwalk should be almost routine for the reader by now.

Petrosian – Smyslov Moscow 1950

14.axb7?! A bold decision. The position would be roughly equal after 14.Bf4. 14…Rxa1 15.Qe4 Nxf2?! Now it’s Black’s turn to gamble. After 15…Qb6, Black would have an edge. 16.Kxf2! d3+ 17.Kg3!

White’s king fearlessly takes up residence on the third rank. 17…Rxb1 18.Qxd3?

White should play 18.Bg5!, followed by 19.Rxb1. 18…Rxc1 19.Rxc1 Qc7

20.Qb5?! The engine’s recommendation – 20.Kh3! – would avoid tactics against White’s e5-pawn and retain a slight advantage. 20…Bf2+? 20…Rb8! was equal. 21.Kxf2 Qxc1 22.Qxd7 Qxb2+ 23.Nd2 Qxe5

White has emerged from the complications with a winning position, but Black gained enough counterplay in the time scramble to draw.

24.Bd1 Qb8 25.Qb5 Rd8 26.Ke1 Qd6 27.Qb2 Qxh2 28.Bf3 h5 29.Nf1 Qf4 30.Bxh5 Qh4+ 31.g3 Qxh5 32.b8Q Rxb8 33.Qxb8+ Kh7 34.Qf4 e5 35.Qe4+ g6 36.Ne3 Qg5 37.Kf2 Qf6+ 38.Kg2 Qd6 39.Nd5 f5 40.Qc4 ½-½ In the next game, it is Petrosian’s opponent, Smyslov, who undertakes a kingwalk in a winning attempt. This led to exceptionally complex play, given the reduced material, and the kingwalk came very close to success.

Smyslov – Petrosian Zürich 1953

For a number of moves, White has had a draw in hand, but now he plays to win. 41.d5!? Qa2+ 42.Kg3 Qd2? It was better for Black to bring his knight into the center with 42…Nd3. 43.d6 Qe1+ 44.Kg4

44…Nd3 45.Qd5+?! 45.Qf8+! Kh7 46.d7 won.

45…Kh7 46.d7 Qe5!?

46…Ne5+ and 47…Nxd7 was objectively better. 47.Qxd3+ The win was there with 47.Qd6!, as after 47…Nf2+ 48.Kh4 g5+ 49.Kh5

Position after 49.Kh5 (analysis) White would have successfully completed his kingwalk. 47…cxd3 48.d8Q ½-½ After 48.d8Q Qe2 49.Kh3 d2, White is forced to give perpetual check with 50.Qd7 d1Q 51.Qf5+.

Attacking Petrosian was never easy. More than most players, his king really was a moving target. Here’s one example.

Ragozin – Petrosian Gagra 1953

White has a superior position and should now force Black to sacrifice his king’s (!) rook with 27.Bd3 Rxd4 28.Qxd4. Instead White sacrifices his own king’s rook, but it’s not sound. 27.Rxe6? fxe6 28.Bg6+ Kf8 28…Kd8? loses to 29.Bb6. 29.g5 Bc5! 30.Qxh4? Bxd4 There is no reason not to take more material, as Black’s king escapes. 31.Rxd4 Rxd4 32.Qxh6+ Ke7 33.Qg7+ Kd6 34.Qxd4+ Bd5

Black has an extra rook. 35.Bd3 Qb6 36.Ne4+ Kc7 37.Qc3+ Kb7 38.b3 Rc8 39.Qg7 Rc7 40.g6 e5 01 Petrosian was partial to surprising king moves early in the game. The following example is similar to the Konstantinopolsky game given above, except here Petrosian’s king simply steps aside to get out of check, avoiding the simplification that would follow from the normal interposition. Did Petrosian win games like this because of, or in spite of, such moves? As long as chess is played by human beings, instead of by computers, it will sometimes be hard to differentiate between the two. Petrosian – Karaklajić Belgrade 1954 1.c4 Nf6 2.Nc3 c6 3.d4 d5 4.Nf3 e6 5.e3 Nbd7 6.Bd3 Bd6 7.e4 dxe4 8.Nxe4 Nxe4 9.Bxe4 Bb4+

10.Kf1!? Of course the natural 10.Bd2 was perfectly playable, and left White with a small, but lasting, advantage. 10.Kf1!?, on the other hand, keeps the bishops on the board and therefore complicates the play. Objectively Black is now equal, but more chess remains to be played. 10…Be7 A bit passive. 10…Nf6 and 11…Bd6 was better. Now Petrosian manages to play completely normally for another dozen moves before even starting to deal with the issue of his king. 11.Qc2 Nf6 12.Bd3 c5 13.dxc5 Bxc5 14.Bf4 Bd6 15.Bg5 h6 16.Bh4 Bd7 17.c5 Rc8 18.b4 Be7 19.a3 Nd5 20.Bg3 0-0 21.Ne5 a6 22.Qe2 Ba4

23.h3 23.h4! was better. 23…Nc3 24.Qe1 Nd5 25.Kg1 Bg5 26.Kh2 1-0 (84)

White finishes castling by hand, and went on to win this equal position in another 55 moves. The next one is easy – a short stroll away from the firing line, then a breakthrough.

Petrosian – Milić Belgrade 1954

White doesn’t have much to work with. Petrosian starts by bringing his king to a safer spot. 35.Kb2! Qc8 36.Ra1 Qa6 37.Kc2 Bd8 38.Kd3

38…Qc8? Black is too passive. Now was the time for 38…c6!, to at least make White’s king a bit uncomfortable. 39.Qb2 R5a6 40.Rb1 Rb6 41.Rb5 Qd7 42.a4 Qe7 43.Ra1 Rab8? 44.a5 Rxb5 45.Nxb5 Kh6 46.a6 Qh4

This counterattack is too late – White’s passed a-pawn is very strong. 47.a7 Ra8 48.Nc3 c6 49.Qb8 cxd5 50.Qxa8 d4 1-0

Black resigned without waiting for White’s reply. As Petrosian’s style developed and his playing strength increased, the theme of strangulation appeared more frequently in his games. Petrosian often implemented Nimzowitsch’s concepts of the blockade and inflicting weak square complexes on the opponent, and the resulting paralysis leaves a strong impression. Kingwalks played a part in this approach to chess, as the king can be a good blockader, and is admirably suited for exploiting weak squares of either color.

Petrosian – Pavey USA-USSR Moscow 1955

White has a tremendous positional advantage, and he could just “sit on” his opponent with a move like 32.Re1. Petrosian decides the time is right for a breakthrough. 32.Bxg6!? hxg6 33.h7 f5 34.Rh1 Ngf6 Black must give back the piece. 35.Nxf6 Nxf6 36.exf6 Rxf6 37.Re3 Re8 38.Rh4 Bd7

White’s combination has achieved clarity, above all else. Material is equal, but White has an iron grip on the dark squares.

The time is therefore ripe for a kingwalk, to relieve one of White’s rooks from its restraining duty. 39.Kf3! Kg7 40.Kf4 Bc8 41.Re1 Rff8 42.Ke5

White is now in no rush. 42…Rh8 43.Reh1 Re7 44.a4 Ree8 45.a5 Rd8 46.Rh6 Rde8 47.Nf3 Re7 Black can’t save his g6-pawn with 47…Ref8 because after 48.Kf4 Bd7 49.Ne5 Be8 50.Kg5 Rf6 51.f3, he will soon be in zugzwang:

Position after 51.f3 (analysis) 48.Nh4 1-0

Curiously, Petrosian had another chance for a kingwalk against the same opponent in the same match.

Pavey – Petrosian USA-USSR Moscow 1955

Having blundered a piece in an equal position, then unsoundly sacrificed a rook on g6 in an attempt to salvage the situation, White is completely lost. This gives Petrosian an opportunity for a kingwalk, almost just for fun. But since the only thing that can possibly go wrong for Black is if he stumbles into a mate, it makes sense to eliminate this possibility. 33…Kf8!? 34.Qh5 Ke7 35.Qh7 Kd8 36.h4 Kc7 37.Qxg7+ Kb6

There is nothing for White to do but resign. Or play a few more moves, then resign. 38.Qxf6 Qc7 39.h5 Qg7 0-1

Not all of Petrosian’s kingwalks were “walks in the park.” Sometimes they were carried out under duress, and the fact that they often had a happy ending shouldn’t obscure this fact. Consider the following game, in which Ståhlberg – an extremely strong player – sacrificed a piece for a great deal of positional compensation. He held the advantage for many moves, but then came the time scramble…

Ståhlberg – Petrosian Gothenburg 1955

White has a comfortable advantage, with more space and a better bishop. He now makes a creative and interesting, but unnecessary, positional sacrifice. 15.Bxf6!? Bxf6 16.Nxh5!? gxh5 17.Qxh5 Ng6 18.g3

18…Kf8?! White has positional compensation for his slight material disadvantage, and both sides have to be careful. Black’s kingwalk may not be the most accurate continuation, but psychologically it’s an astute choice. By depriving White of a target, Black takes the wind out of his sails. 19.Qf3 Ke7 20.Nd1 Bh8 21.h5 Nf8 22.Ne3 Nh7 23.Nf5+ Kd8 24.0-0

White has played well up to this point, and has considerable pressure. 24…Nf6 25.b4 Kc7

Black’s king is more comfortable on c7 than it was on the kingside, but it’s not completely comfortable. 26.Rfb1 Reb8 27.Rb3 Qe8 28.Rab1 Nd7 29.Nh6 Bg7 30.Nf5 Bh8 31.Kg2 Rb7 32.h6 Qg8 33.Qd1 Qe8 34.Qh5 Rbb8

35.Rh1? White must have been in time trouble, as there follows a series of mistakes. 35.Rh1? gives Black a chance to break the bind on the queenside. Naturally Petrosian seizes the opportunity.

35…cxb4! 36.axb4 a5! 37.Rf3? axb4 38.Ng7? Bxg7 39.hxg7 f6 40.Qg4 Qf7 41.Rh7 0-1 With the time control reached, White resigned. It’s one thing to execute a kingwalk “under duress,” but sometimes a kingwalk “to safety” is an illusion, as the king isn’t nearly as safe as it seems. In the following example, Petrosian badly mishandles the middle game and was lucky to come out with a full point. It was perhaps experiences like this that led Petrosian towards the defensive style that made him almost unbeatable in his prime.

Petrosian – Kluger Budapest 1955

Having achieved a pleasant opening advantage, White now treats his opponent too lightly. Or does he? White’s next move is very hard to assess, but later White’s central king position could have come back to haunt him. But since Black fails to exploit these mistakes… 15.Kd2!? g5 16.h3 White should be consistent and continue with 16.Kc2. 16…h5 17.Qg1!? a6 18.g4! fxg3 This is only natural, as otherwise Black will just suffer on the queenside. 19.Qxg3? 19.Nxg3! was the right follow up. 19…Nh7 20.Be3 h4 21.Qg2 Kh8 22.Rhf1?! Bh6 23.c5 Bd7 24.a4?! Nef6 25.Ng1? Nh5 26.Rfb1 Nf4 27.Qh2?

White has played inexactly and has landed in trouble. 27…Nf6? Missing a pretty win with 27…Nxh3! 28.Nxh3 Rxf3 29.Nf2 Rxe3! 30.Kxe3 g4+ 31.Ke2 g3 – a thematic King’s Indian Defense combination. 28.Bf1 N6h5? White’s 28.Bf1 covered the threat against his h3-pawn, but also removed a defender from the vicinity of his king. For the next few moves, Black overlooks the very strong idea of …c7-c6!, opening up the center. 29.Nd1 Qe8 30.a5 Ng3 30…c6 31.Nc3 Qc8? Black also had 31…Nxf1 32.Rxf1 c6! 32.Rc1 Qe8? 33.Bc4 Bg7?! Now 33…Bb5 was right. Black just doesn’t realize that he should open lines to White’s king. White may well have realized the danger, but there wasn’t a lot he could do about it other than hope.

34.Kc2 Rf6 35.Kb3

Black has missed his best opportunities, but still stands better. 35…Bf8 36.Qc2?! Rg6 37.Bxf4 gxf4 38.c6 One theory of chess is “play like you’re winning.” Your opponent may believe you. Here it works. 38…bxc6 39.dxc6 Bxc6 40.Bd5 Rg7? 40…Bxd5! 41.Nxd5 c5! would have refuted White’s idea. White’s exposed king is more important than positional considerations such as White’s seemingly strong d5-knight. 41.Bxc6 Qxc6 42.Nd5 Qb7? Black had to play 42…Qxc3+ 43.Rxc2 c5, with equality. 43.Qc6 Qb8 44.b5? 44.Nxc7! 44…axb5? 45.Nxc7 Ra7 46.Nd5 Raf7 47.a6 Nh5 48.Qb6 Qxb6 49.Nxb6 Ra7 50.Rc8 Kg8 51.Ne2 Nf6 52.Nc3 Rg3 53.Nxb5 Rxf3+ 54.Kc4 1-0

Now White’s king is well placed! And, of course, there are the desperation kingwalks that don’t help. To dispel any illusions that every Petrosian kingwalk was a magic bullet, we give the following example.

Simagin – Petrosian Moscow 1956

Black’s position is in ruins and the ensuing kingwalk is pure desperation. 42…Kc7 43.Rd2 Qxb4 44.Qg8 Rd7 45.Rc2+ Bc5 46.Qa8 Kd6 47.Rd2+ Bd4 48.Be3 Ke6 49.Qe8+ Kf5 50.g4+ Ke4

Black’s king can go all over the board – there’s no escaping the murderous crossfire from White’s pieces. 51.Qa8+ Rd5 52.Rxd4+ 1-0 Here is another example of an attempt to confuse matters with a kingwalk, with the attempt failing badly.

Averbakh – Petrosian Leningrad 1960

Having fallen into a passive position out of the opening, Petrosian has struggled to escape, but to no avail. In time trouble, he tries a kingwalk to confuse matters, but Averbakh takes it all in stride. 36…Kh6!? 37.Re4! Qf3 38.Rh4+ Kg5 Going back doesn’t help: 38…Kg7 39.Re2!, with the idea of 40.Re7+. 39.Qf8

39…f5 39…h5 is also met by 40.Rf4, threatening 41.h4 mate.

40.Rf4 Qc3 41.h4+ Kh5 42.Rxf5+ Kg4 42…gxf5 43.Qxf5+ Kh6 44.Qg5 is mate. 43.f3+ 1-0

Black resigned, because both 43…Kxg3 and 43…Qxf3 are met by 44.Rg5+ Kxh4 (44…Kh3 45.Rh2 mate) 45.Qh6 mate. As we have pointed out, Petrosian’s kings had nerves of steel. In the following game, Petrosian knew that his king could always escape White’s counterattack.

Kupper – Petrosian Tel Aviv 1964

Having outplayed his opponent from an inferior position, Petrosian now invites complications. 35…Nf8!? 36.Ne3 Bf3 37.b6+ Kb8 The start of a kingwalk. 38.Qc1 Bxg2 39.Qa3 Qh7 Here Black could have changed gears with 39…Nd7! 40.Qd6+ Ka8!, and Black’s attack is faster. It’s understandable that a human would be more cautious than the fearless engine. 40.Qd6+ Kc8 41.Qc5+ Kd8

42.Bxf5 White has nothing to lose. 42…Rxh2 43.Qd6+ Qd7 44.Qb8+ Ke7

Black’s king was on a7 ten moves ago. 45.Bc2 Bf3 46.Rg3 Rxc2 47.Nxc2 Be4 48.Ne3 Rh1+ 49.Kf2 Qxa4 0-1 White threw in the towel, because he soon runs out of checks. In the next example, White walks his king up the board, to prepare for a breakthrough on the kingside, then walks his king again, to prepare for a second breakthrough, on the queenside.

Petrosian – Bertok Zagreb 1965

White has a small, but real, advantage. He has more space and a kingside pawn majority, but is hampered by weaknesses on the queenside. White will soon try something on the kingside, but first he brings his king forward. 31.Kg2 Re8 32.Kh3 Bf8

33.e5!? fxe5 34.f5 Bxf3 35.Qxf3 Bg7 36.h5 g5?! It is more blessed to give than to receive. 36…e4! gave Black a slight edge. 37.Ne4

White now has ample compensation for his sacrificed pawn. Breaking through, however, is an entirely different question. 37…Bf6 38.Bd2 Kg7 39.Rf1 Rff8 40.Raa1 Rf7 41.Rfc1 Rff8 42.Be1 Nc6 43.Rab1 Nb4 44.Rb2 Rf7 45.Bg3 Rfe7 46.Re1 Kf7 47.Rbe2 Kg7

White now undertakes a second kingwalk, transferring his king to the queenside, where it can take over some defensive duties. 48.Kg2 Qc6 49.Ra1 Qc7 50.Kf1 Kf7 51.Ke1 Kg7 52.Kd1 Kf7 53.Be1 Rb8

54.Bxb4 White’s thinking is that there is no way to make progress without making this trade. 54…axb4 55.Rea2 Rb7?

55…Kg7! 56.a5 b3 57.a6! bxa2? 58.axb7 Kg7 If 58…Qxb7, 58.Nd6+ wins. 59.Qg2 59.Nxf6! Kxf6 60.b8Q! Qxb8 61.Qc6+ Kf7 62.Qg6+ wins, but only an engine would give up White’s powerful e4-knight, then White’s b7-pawn, in quick succession. 59…Qxb7 60.Qxa2 b5 61.Qa6 1-0 After some of the heavy duty examples we’ve looked at, here’s an easy one.

Petrosian – Lutikov Moscow 1966

14.Ke2?! Clever, but 14.axb4 Qxb4+ 15.Nd2 would win a piece, although Black has some compensation. 14…Qa6! 15.Nxd4 Be7 16.Nxc6 Qxc6 17.Rxd8+ Bxd8 18.Kd1

White can’t hold his pawn, so the game peters out quickly. 18…b6 19.Nc3 Ba6 20.Qa4 Qxa4+ 21.Nxa4 Rc8 22.Nc3 ½-½ In the next example, Petrosian’s king again proves to be a slippery target.

Petrosian – Yudovich Moscow 1967

White has a slight advantage, because of his pressure on Black’s d5-pawn. It is difficult to convert this advantage, however, if Black plays precisely. 28.Kh2!? Qb1 29.Kg3

29.e4! was strong, as 29.dxe4? 30.Bxe6 fxe6 31.Qh6 wins, but Petrosian prefers to wait. 29…Rc5 30.a4 a5 31.e4! Qg1 32.h4 To prevent 32…h4+, although 32.exd5 would work as well (32…h4+ 33.Rxh4).

32…Kh7 33.Bxd5 Bxd5 34.Rxd5 Rc6?! 35.Rd7 Rf6 36.Qd4 1-0 Here is another kingwalk in the opening, but a misguided one.

Kupper – Petrosian Venice 1967

19…Kf8!? We can say with certainty that there are very few reigning world champions who would make this move! One way of looking at it is to term 19…Kf8!? a type of “waiting move,” as it gives White absolute freedom to play whatever move he likes. 20.Rb1?! h4 21.b4 axb4 22.axb4 Bxe3 23.Nxe3 g6 24.Ra1 h3 25.c3? hxg2+ 26.Kxg2 Kg7

It seems as though White and Black have been singing from the same hymn book. Things couldn’t have worked out better for Black. White’s position is

full of weaknesses, and Black has finished castling by hand, with both his rooks being “developed” on their original squares. 27.b5?! Qc7 27…Rxa1 28.Rxa1 Rh4 was more incisive. 28.Ra6 Rac8 29.Rf3 Qb7 30.Qd4 Rc5 31.Qa4? d4! 32.Ra7 Rxh2+ 33.Kxh2 Qxf3 34.Qxd4 Nd5 34…Rxb5! was faster, with the idea of invading White’s first or second rank. 35.f5 Nxe3

36.Rxf7+ Kh6! Of course Black saw this second, minor kingwalk. 36…Kxf7? 37.Qd7+ is a draw. 37.Qh4+ Qh5 38.Qxh5+ gxh5 39.Rf6+ Kg5

It’s over. 40.Rxe6 Rxb5 41.f6 Ng4+ 42.Kg3 Nxe5 43.Kh2 Ng4+ 0-1 Here is another example of a kingwalk to prepare a breakthrough, then to elude the opponent’s counterattack.

Paoli – Petrosian Venice 1967

Black has outplayed his opponent and now must think of how to convert his advantage. The answer is to play …c6-c5, opening the queenside. To prepare for this, Petrosian first carries out a short kingwalk, to increase the defense of his d5-pawn. 34…Kf8 35.Ree1 Rc7 36.Qd2 Kf7 37.Rh1 Ke6

38.a3 Rhh7 39.Rh2 c5 40.h5?! White loses patience. 40…Rxh5 41.Rxh5 gxh5 42.Rh1 Rh7 43.Qd1 h4 44.Rh3 c4 45.Qa4 Qd7 46.Qd1 hxg3 47.Qh1 Rxh3 48.Qxh3

Black’s advantage has increased. He now moves his king to the queenside, both to escape checks from White’s queen and to go after White’s b2-pawn. 48…Kd6 49.Qh8 Kc6 50.Qf6+ Kb5 51.Qf8 Qe6 0-1 In this example, Petrosian’s king again escapes a counterattack by a kingwalk along the light squares. This allowed Petrosian to pursue his own attack.

Visier – Petrosian Palma de Mallorca 1968

Material is equal, but White’s king has issues on the light squares. The problem, as we’ve seen many times, is that if Black brings his queen into the attack, he has to be prepared for a kingwalk to escape White’s counterattack. 36…Qe4!? 37.Rb8+ Kf7 38.Qb7+ Qe7 To gain time, Black repeats moves, although White should take advantage of this by exchanging queens on e7. 39.Qb3 Qe4 40.Qb7+ Kg6! 41.Rg8+ Kh5 42.Qf7+ Kg4

As planned. Black now threatens mate on g2, so White must give another check.

43.Qd7+ The alternative was 43.Qe6+. Both endings are difficult for White. 43…Qf5 44.Qxf5+ Kxf5 45.g4+? Panic. White should hang in there with 45.Be1. 45…Kxg4 46.Bg3 Rxc3 47.Bxd6 Kf3 47…Rc2! was more precise. 48.Rf8+ Ke4 49.Rd8 Rc2 50.Bf8 0-1 This allows Black’s king to join in the attack without White’s rook being able to check it away: 50…Kf3 now mates, so Black resigned. Another easy kingwalk to the queenside, followed by a kingside breakthrough.

Bednarski – Petrosian Lugano 1968

White has overreacted to his opponent’s positional approach to the opening and it is Black who has the advantage. By this point, Petrosian’s next move will seem completely natural. 17…Kd7! 18.Qc2 Qf8 One of the points of Black’s last move. His major pieces can now operate freely. 19.Rag1 Ne8 20.R4g2 Ng7 21.Ng3 Qf7 22.h5 White tries to break Black’s hammerlock on the f5-square, but to no avail. 22…Raf8 23.hxg6 hxg6+ 24.Rh2 g5! 25.Rxh8 Rxh8+ 26.Nh2 gxf4 27.Rf1 Bg5 28.Qf2 Nf5!

Which is safer: White’s castled king or Black’s uncastled king? 29.Nxf5 fxe3 30.Qg2 e2 0-1 In the next example, the purpose of the kingwalk is a bit different – to allow White’s rooks to take control of a file that Black had opened earlier in the game.

Petrosian – Vasiukov Moscow 1969

White stands better because he has a nice blockade, while his knight is better than Black’s light-square bishop. As compensation, Black has at least opened the h-file against White’s king. That’s true, but it’s not Black who can exploit it. 33.Kf1! Bg7 34.Ke2 Rf8 35.Rh1

White’s king is perfectly safe on e2, because only White can open lines (with f2-f3). This is what happens, but only after White first improves the position of his rooks. 35…Be8 36.Rh5 Qe6 37.d5 Qd7 38.Rg5 Rf6 39.Rd1 Bg6 40.Rh1 Ref7

41.f3! exf3+ 42.gxf3 Qe8 43.Kf2 c6 44.Qd1 cxd5 45.Qxd5 Qc6 46.Nh5 Bxh5 47.Rhxh5 Qxd5 48.cxd5 Kf8 49.fxg4 fxg4 50.Rxg4 1-0 (78)

White went on to win the endgame. The following example is a classic Petrosian kingwalk: the position is closed, Black’s king wants to relocate to the queenside, and ultimately Black breaks through on the kingside once his king has reached its destination.

Díez del Corral – Petrosian Palma de Mallorca 1969

The position is equal, with White’s doubled d-pawns having pluses and minuses. White is preparing to attack on the kingside, so Black immediately carries out a kingwalk to safety to the other flank. 22…Kf8!? 23.Qg5?! White should operate on the queenside, with 23.Bd2 or 23.a4, so as to deny Black’s king a safe haven. 23…Ke8 24.Rac1 Kd7 25.h5?! gxh5 26.Rxh5 Rg8 27.Rh7 Kc8 28.Qh4?! Qg6 29.Rh8 Rxh8 30.Qxh8+ Kb7

Black’s king has completed its journey and the position has been transformed as a result. Black’s advantage is by no means decisive, as White’s queen is active, but Black has no structural weaknesses and White must be careful. 31.Qf8 Rc8 32.Qd6 Qe8 33.a4 Rd8 34.Qa3 Qe7 35.Qc3 Exchanging queens was the lesser evil. 35…Rc8 36.Bd2 g5! 37.Qc2 f4! 38.gxf4 gxf4 39.Bxf4 Rg8+ 40.Bg3 Nxd4 41.Qc3 Ne2 42.Qc6+ Kb8 43.Re1 Nf4+ 44.Kf1 Nxd3 45.Rb1 Qf7

Black’s pieces are starting to close in on White’s king. 46.Qd6+ This makes things worse, but 46.Rxb6+ fails to 46…axb6 47.Qxb6+ Qb7 48.Qd6+ Ka7. 46…Kb7 47.Ke2 Rc8 48.a5 If 48.Kxd3, then 48…Qh7+! 49.Kd2 Rc2+ mates – Black needn’t bother taking White’s b1-rook. 48…Rc2+ 49.Kf1 Nxf2 50.Rxb6+ axb6 51.Qxb6+ Kc8 52.Qa6+ Kb8 53.Qb6+ Qb7 54.Qd6+ Qc7 0-1 As discussed in Chapter 9, kingwalks can take place in the endgame. The following game is a good example of this.

Petrosian – Fuller Siegen 1970

White has a large advantage, but has to figure out how to make it count. Petrosian now makes a very instructive decision. 28.Qg5! Qxg5 29.fxg5 Normally one would hesitate before going into an endgame with bishops of opposite colors, which tend to be drawish, but there are a number of important factors which put this position in a different category: (1) There are also rooks on the board, and White controls the open d-file; (2) White’s bishop is much more active than Black’s; (3) White has a mobile queenside pawn majority, while Black’s extra pawn on the kingside (on e6) is worse than useless; and (4) White’s king can invade Black’s position via the dark squares. 29…Bc6 30.Kf2! Kf7 31.Kg3! Ke8 32.Kf4 Bd7 33.Bd6 a6 34.Ke5

Black, likely in time trouble, now breaks on the queenside, but all this really does is give White a passed pawn. 34…b5 35.axb5 axb5 36.c5 Ra8 37.Kf6 A bit of a bluff, since White doesn’t intend to chase after Black’s h7-pawn with his king, although you never know. 37…Ra7 38.g3 Ra1 39.Be5 Ra3 40.Rb2 40.Rd6 was winning, but on the last move of the time control, White understandably deferred this sacrifice. 40…Ra7 41.Bb8 Rb7 42.Bd6 Ra7 43.h4 Bc6 44.Kxe6 b4 45.Rd2 Bd7+ 46.Kd5 Kf7 47.c6 The winning idea. 47…Ra5+ 48.Bc5 Be6+ 49.Kd6 Bxb3 50.Rb2 Bc4 51.Rxb4 Ba6

52.Rb7+ Bxb7 53.cxb7 Rb5 54.Kc6 Rb1 55.Bb6 Rc1+ 56.Kd6 Rd1+ 57.Ke5 1-0 It was games like this that gave Petrosian the reputation of being a boa constrictor. An earlier example saw Petrosian constricting Pavey to the point of zugzwang, but this could happen to the strongest players as well.

Petrosian – Mecking Wijk aan Zee 1971

We start this example early, because here Black castled by hand. Consider this an appetizer for the real kingwalk that occurs later on. 19…Kf8 20.Nf3 Kg7 21.Be3 Nc5 22.Nd4 Ra6 23.Bc1 Raa8 24.g3 b6 25.h4 h5 26.Qf3 Qd8 27.Bd2 Qe8 28.Bg5 White doesn’t have to make this exchange, but his long-term plan is to create dark-square weaknesses in Black’s position. 28…Qd8 29.Qf4 Rc8 30.Re3 Bxg5 31.hxg5 Ra8 32.Qf6+ Qxf6?! An understandable response, but Black should have avoided giving White access to the e5-square. 33.exf6+ Kh7

White’s advantage is now more pronounced, and the next step is for White to blockade the d4- and e5-squares. 34.Kg2 Rae8 35.f4 Rb8 36.Ree1 Nb7 37.Rh1 Kg8 38.Nf3 Nd6 39.Ne5 Be8 40.Bd3 Rc8

There is so much to learn from this game! White’s e5-knight dominates Black’s e8-bishop, and White’s d3-bishop dominates Black’s d6-knight in much the same way. The engine, of course, recommends the crude 41.g4, breaking through on the kingside. But what then of Petrosian’s reputation? With Black’s h5-pawn also restrained, White can now undertake a kingwalk to blockade Black’s d5pawn. 41.Kf3! Bc6 42.Rh2 Be8 43.Ke3 Rc7 44.Kd4

Now White gradually opens the position on the queenside, since Black will be playing without his h8-rook. 44…Nb7 45.b4 Nd8 46.Rh4 Nb7 47.Ra2 Nd6 48.Rh1 Nb7 49.b5 Nc5 50.Bc2 Nd7 51.Ra3 Nc5 52.c4 Nd7 53.Rc3 Nxe5 54.Kxe5 The engine considers 54.fxe5 to be the better recapture, but it makes more sense to a human to retain a route into Black’s position for White’s king. 54…dxc4 55.Be4 Rc8 56.Kd6 Rc5 57.Rhc1

57…h4 1-0 This break is too late to matter. After 58.Rxc4 Rxc4 59.Rxc4 hxg3 60.Ke7, Black loses his e8-bishop, so Black resigned without waiting for White’s reply.

The next example features another seemingly easy escape by Petrosian’s king.

Tukmakov – Petrosian Moscow 1973

White is a pawn down, and rather than suffer a slow death, he gambles on a kingside attack. 32.f5?! exf5 33.Rh4 Rxe5 The engine finds an elegant line for Black: 33…Kg8!? 34.Qh6 f6!, defending Black’s h7-pawn from the side. Petrosian seeks clarity by taking White’s e5pawn, and going for a kingwalk to escape the attack. 34.Qh6+ Kf6 35.Rf1 Qb6 36.Rhf4 Ke7 37.g4 Qe3 38.Qh4+ Kd6

39.R4f3 Qd2 40.gxf5 Kc5!?

The engine favors 40…g5!? 41.Qh6+ Kc7. 41.fxg6 hxg6 42.c4? Rh5! 43.Qg4? Rg5 0-1 One of the all-time great kingwalks was seen in Petrosian-Unzicker, 1960, as described in Chapter 1. Fourteen years later, Petrosian carried out a similar maneuver against an even stronger opponent.

Petrosian – Ljubojević Manila 1974

White has eliminated any possibility of active play by his opponent, which must have greatly irritated Ljubojević, who always favored open combat. The next step for Petrosian is a kingwalk to the queenside, to prepare for the opening of the kingside. 50.Kf1! Rf8 51.Ke1 Qc7 52.Kd1 Rf6 53.Kc2 Qd8 54.f3 Rff7 55.Qg1 Qc7 56.Kb2

White has completed his “classic” kingwalk – as we saw in Chapter 1, all Black could do was (literally) mark time and move his pieces back and forth. 56…Rd8 57.Qd1 Rfd7 58.g4!

The logical culmination of White’s strategy. 58…Kh8 59.Qg1 Rg7 60.Qe3 Kh7 61.Qg5 Be8 62.Qf6 Re7 63.gxh5 gxh5 64.Rg2 Rdd7

65.Nb5 A strong, completely human move. But the engine instantly comes up with 65.Nd5! exd5 66.Qf5+ Kh6 67.Qf8+ Kh7 68.Rg8! Rg7 69.Rh8+ Kg6 70.Qf5 mate. 65…Qc5 66.Rxd6 Qe3 67.Na3?! Again, a natural move, but now the engine unleashes 67.Qf8! Bg6 68.Nc7! Rxc7 69.Rd8. It is very difficult to switch from grinding an opponent and winning pawns to giving away pieces for a mating attack! 67…Qf4 68.Qxf4 exf4 69.Rxd7 White could have played 69.Rxb6. 69…Bxd7 70.c5! bxc5 71.Nc4 Bc8 72.Rg5 Kh6 73.Rxc5 1-0 (100)

And, after some imprecise moves, White won.. A kingwalk by one of Petrosian’s toughest opponents – the great Korchnoi.

Korchnoi – Petrosian Moscow 1975

This time it’s Petrosian who has an inferior position without counterplay. Korchnoi is not too proud to borrow from his opponent’s playbook. 48.Kf2 Ne6 49.Ke2 Ng7 50.Re3 Ne6 51.Kd2 Qd7 52.Rc3 Qe8 53.Bc2 Qf8 54.Bb1 Qe8 55.Rf3 Qf8 56.Rf1 Qe8 57.Rf2 Qf8 58.Rf3 Qg7 59.Rc3 Qf8 60.Kc1 Qe8 61.Kb2

White’s king has reached the safest square it can find, so now White can play for a breakthrough on the kingside. 61…Qa8 62.Qe7 Qb8 63.Rf3 Qf8 64.Qf6 Qe8 65.Rc3 Qa8 66.g4!? Qe8?! 66…Qb8!?

67.Qe5 Qd8

68.Bxg6! hxg6 69.Rh3 Nd3+ 69…Ng7 is met by 70.Qh2!, and the h5-square is covered by White’s g4pawn. That was the idea behind 66.g4!? 70.Rxd3 Nxg5 71.Rc3 Ne4 72.Rh3 f6 73.Qh2 Kf7 74.Rh8 1-0 To paraphrase Tarrasch, kings walks, like love, like music, have the power to confuse. A timely kingwalk can throw off an opponent and salvage a lost position. Sometimes – but not always!

Hübner – Petrosian Biel 1976

Black is in trouble and his best chance is to go into an inferior endgame by exchanging queens. He would have some compensation for his material deficit, but two pawns are two pawns… Instead Petrosian keeps the queens on, in the hope of swindling his opponent in time trouble. Black’s impending kingwalk is a key part of this plan. 33…Qf8 34.f4! Nd3 35.Qb5 The first of several missed opportunities. 35.Re7! would have tied down Black’s queen. 35…Bd4 36.Nh1! Qd6!? The position is very sharp, and in a time scramble anything can happen. 37.g3? Missing a forced mate with 37.Qe8+ Kg7 38.Re7+ Kh6 39.Qf8+ Bg7 40.Qxg7+ Kh5 41.Qxh7 mate. 37…Nxf4!? The engine assesses 37…Kg7! as equal.

38.Qe8+ Kg7

39.Re7+? After 39.gxf4, Black can either play 39…Kh6 or 39…Qxf4+ 50.Ng3 Rg1, with compensation. But 39…Qxd5? would lose to 40.Re7+ Kh6 41.Qf8+ Kh5 42.Rxh7 mate. The move played loses. 39…Kh6 40.Nf2 Bxf2 41.Rxh7+ Kg5! 0-1

Black would lose after 41…Kxh7? 42.Qg8+ Kh6 43.Qh8+ Kg5 44.Qh4 mate. But after 41…Kg5! 42.h4+ Kg4 43.Bf3+ Kxf3 44.Qa8+ Nd5, White is out of useful checks.

Position after 44…Nd5 (analysis) The next example is almost trivial, and it is actually White’s rook, rather than his king, that gives it interest.

Petrosian – Geller Moscow 1976

The position is equal, and remains so. White now “castles by hand,” but in an unusual way – he uses his queen’s rook to complete the castling, since his king’s rook is somewhat indisposed on h4. 21.Kf2!? a5 22.Kg1 Rc7 23.Rf1

A puzzle theme – which rook is on f1 and which on h4? 23…Bc8 24.Bh3 Bxh3 25.Rxh3 Qc8 26.Kg2 Rd7 27.Rh4 Qa6 ½-½ Black’s kingwalk in the following game is instructive, but so is the way the stronger player somehow gets a decisive advantage from a seemingly dead

position. Under closer examination, it’s apparent that the starting position is not so dead after all, and also that Petrosian got some help from his opponent.

Bukić – Petrosian Banja Luka 1979

This is the kind of position where annotators used say “Black stands better, because he has an outpost on d5.” But the engine knows better – White has more space and his bishop is not so bad. In short, while Black wins this position, in doing so he has to “make something out of nothing.” A big part of this involves Black’s kingwalk. 26…Kf8 27.Rfc1 Ke8 28.Bh4 Nf5 29.Bf2 Kd7 30.g4 Nfe7 31.Bg3 Rh8 32.h4 h5!

33.g5?

A terrible positional move which wrecks White’s kingside by giving Black permanent control of f5. 33…Rhd8 34.Bf2 Ke8 35.Kh2 Rd7 36.Rc4 Kd8 37.Ng3 g6 38.Ne4 Kc8 39.Nd6+ Kb8 40.Ne4 Kb7

Black’s king is now on an ideal square, and Black can begin to maneuver his pieces. 41.R4c2 Nf5 42.Kh1 Rad8 43.Rd2 Rc8 44.Kh2 Nde7 45.Nf6 Rdd8 46.Ne4 Nc6 47.Rcd1 Na5 48.Rc2 Nb3 49.Rc4

49…c5! At just the right moment.

50.Rd3 Nbxd4 51.b4 Nc6 52.Rxd8 Rxd8 53.bxc5 Nxe5 54.Rc3 Rd3 55.Rxd3 Nxd3 56.cxb6 Nxf2 57.Nxf2 Kxb6 58.Kh3 Kb5 This hardly qualifies as a kingwalk and isn’t worth a diagram. 59.Nd3 Ka4 60.Ne5 Kxa3 61.Nxf7 a5 62.Ne5 a4 63.Nxg6 Kb3 0-1 Not every kingwalk is a quiet affair. In the following game, Black decides not to go quietly into the night. White’s king gets a good workout before the curtain comes down.

Petrosian – Popović Vrbas 1980

Facing a slow death, Black decides to stake everything on a counterattack. Petrosian’s king is quite ready to be “driven out of the corner” for a good cause. 29…Nd4?! 30.Nxc7 Qc6 31.R1b6 Nxe2+ This is what Black had planned – otherwise he has to give up his queen. 32.Kf1 Nxg3+ 32…Qf3 33.Qe4 leads to a queen exchange. 33.hxg3

33…Qh1+? Paradoxically, Black’s best chance was not to drive White’s king into the center, but instead to play 33…Qf3! 34.Ke2 Rd4 35.Qxf7 e4 36.Qxg7+ This “sacrifice” eliminates most of Black’s remaining pieces. 36…Kxg7 37.Ne6+ Kg8 38.Nxd4 exd3+ 39.Kxd3 Rd8 40.Bb2 Qd1+ 41.Kc4 Qf1+ 42.Kc3 Qe1+ 43.Kb3 Qd1+ 44.Nc2

44…Kf8 If Black’s rook leaves the back rank, he is quickly mated: 44…Rd3+ 45.Ka2 Qxc2 46.Rg7+ Kf8 47.Rb8+. 45.Bg7+ Kg8 46.Bf6 Rd7 47.c6 1-0

Chapter 15 Duncan Suttles While Petrosian can be considered to be one of the main “interpreters” of Nimzowitsch, Canadian grandmaster Duncan Suttles may also be placed in this category. Their styles differ, however, in that Suttles was much more inclined to take risks than Petrosian, so while Petrosian used his positional understanding to avoid danger, Suttles often used his to seek it out – to see just how close to the line he could go. Needless to say, he sometimes went over the line, but what was lost in ELO points was gained by the spectators. Almost all of Suttles’ games are annotated in the three-volume collection of his games by the authors of this book – Chess on the Edge. We will look at Suttles’ kingwalks in this chapter, with abbreviated notes. In our first example, Suttles is actually the victim of a kingwalk by his opponent, grandmaster William Lombardy, as his counterattack fails when Black’s king slips up the board to safety.

Suttles – Lombardy Seattle 1966

White is on top and should continue with 28.Rxc8+ Qxc8 29.a4! 28.Qc3?! Rxc5 29.Qxc5 Qc8 30.Qd6?! 30.Qa7! 30…Kf7! 31.Bd2 And here White should have played 31.Bd3!, keeping Black’s queen out of c2. 31…Qc2 32.Bb4? It’s hard to know quite what White was trying to do, since Black’s g6-knight covers e7 and f8. White must have underestimated the effectiveness of Black’s impending kingwalk. Perhaps time trouble played a role. A draw would have resulted from 32.Qd8! Qxd2 33.Qf6+ Ke8 34.Qxe6+ Ne7 35.Qxa6, and Black must give perpetual check. 32…Qc1 Now Black is winning.

33.Bc5 Qe3+ 34.Kh1 Qxf4

All White can do is give some checks, but we have seen this kind of position before. 35.Qc7+ Kg8 36.Qd8+ Kg7 37.Kg1 Qe3+ 38.Kh1 Qf2 39.Qf6+ Kh6 40.Bd3 Kh5

Yet another sad example of an advanced king being much safer than a king in the corner. 41.Be7 Qe1+ 42.Kg2 Nf4 mate 0-1 The next kingwalk may not have been the right one, because White could have defended, but psychologically it’s often easier to find such defenses after the game is over. The provocation engendered by a kingwalk seems to bring out the worst in people at times, as we shall see later on in this chapter.

Romani- Suttles Havana 1966

Black is winning. He is a pawn up and his a2-rook and e5-knight are both strongly placed. White’s hope lies in his advanced b7-pawn, which is tying down Black’s other rook. Suttles converts his advantage with a kingwalk. 41…Kh5!? The engine recommends 41…Kf6!, followed by …Ne5-d3, opening a path for Black’s king by vacating the e5-square. There is sometimes more than one road to Rome. 42.Bc8 Kh4 43.Rf2

This move is necessary, since 43.Ra1? loses to 43…Nf3+.

43…Ra3 44.Re2 Nf3+? 44…f3 was winning. 45.Kf1?! The natural 45.Kf2 would have eliminated even a hint of a Black advantage. One strongly suspects that the time control in this game wasn’t reached at move 40. 45…Nd4 46.Re8? This is a serious blunder that opens the floodgates for Black’s pieces. 46.Rg2 was equal. 46…Kg3 46…Ra2 comes out to more or less the same thing. 47.Bd7 Ra2!

A completely sound rook sacrifice, because after 48.Rxb8? Nf3, Black wins with 49…Rf2 mate. 48.Ke1 f3! 49.Bc8 Very sad, but Black was threatening to take White’s b7-pawn, to lure White’s b1-rook off the first rank.

49…Rh2 50.Rg8 f2+ 0-1 Black gets a queen, because if 51.Kf1 Rh1 mate. In the next example, what starts as a defensive kingwalk morphs into an attacking one, forcing Black to go into a lost endgame.

Bouaziz – Suttles Sousse 1967

27.Rae1 In Chess on the Edge, this move is criticized, but the 2018 engine also doesn’t think much of the recommended 27.Rad1. 27…Rxa2 28.Qe5+ Qf6 29.Qc7+ This is the move the engine questions, recommending instead 29.Qe8. 29…Kh6

30.Re7?

This counterattack will fail, although going into an endgame with 30.Qe7 holds little joy. 30…Rxf2 31.Rxh7+ Kg5 32.h4+ Kg4 33.Qd7+ Kxg3

Certainly photogenic, although the engine’s recommendation of 33…Rf5! is even better. 34.Qd3+ Rf3 35.Rxf3+ Qxf3 36.Qxg6+ Kh3

37.Qe6+ White must force the exchange of queens. 37…Qg4+ 38.Qxg4+ Kxg4 39.h5 Kg5 Now Black’s king goes back.

40.h6 c5 41.Kf1 Kg6 42.Rh8 Rc2 0-1 The following game is like an Infinity Stone – no matter how much one of the authors wishes it would disappear, he keeps writing books in which it surfaces. From the standpoint of kingwalks, White’s is a complete success. Objectively White was lost, but subjectively he was completely winning.

Suttles – Harper Vancouver 1968

Black has just played 18…f6, and the position is balanced. But now White comes up with an idea – to drive Black’s knight to h4, then go after the knight with his king. The idea is atrociously bad. 19.g4? Nh4 20.Kh2? f5! Chess on the Edge gives this move a “!?,” but the 2018 engine applauds it. 21.gxf5?! exf5 22.Nd2 Re6? 22…Nxe5! is just crushing. Even now, it’s hard to understand how Black could miss this move. Compare Black’s piece development with White’s. 23.Kg3 Nxe5 24.Nde4 Neg6 And here 24…Nef3 is completely winning. 25.hxg6 Rxg6+ 26.Kxh4 fxe4 27.Nxe4 Be7+ 28.Ng5

28…Re8? Now 28…Rh6+ forces White’s king into a pin: 29.Kg3 Bxg5, regaining the sacrificed piece. 29.Kh5? 1-0 Perhaps recognizing that 29.Be6+ isn’t good enough. At this point, completely thrown off by White’s kingwalk, Black resigned, which is the very last of many missed opportunities. Considerably better was: 29…Rxg5+ 30.fxg5 Qg3, forcing checkmate.

Williams – Suttles Vancouver 1968

White’s positional pawn sacrifice in the opening has rebounded, and Black has heavy pressure on White’s queenside. Rather than suffer through a passive defense, White tries to create counterplay on the kingside. 20.h4?! h5 21.g4 hxg4 Here Black could have exploited the drawbacks to White’s kingside pawn advances with 21…Qd8!, and if 22.Bg5, 22…Bf6! 22.fxg4

22…Rxb2?

This sacrifice, which is given an exclamation mark in Chess on the Edge, is both dubious and unnecessary. 23.Rfxb2 Rxb2 24.Rxb2 Qxc3

25.Rb8+? A blunder. White had a perfectly good defensive move: 25.Qc1!, which protects both his b2-rook and his e3-bishop. Black doesn’t have a knight fork on e2, because White’s b2-rook covers that square. Grabbing a pawn by 25… Qxc1? 26.Bxc1 Nxe4 gives Black impressively centralized knights, but in fact it’s White’s rook and bishops that would be impressive after 26.Kg2, as Black’s pieces actually don’t coordinate well. After 25.Rb8+?, on the other hand, White’s position quickly collapses. 25…Kh7 26.Bf2 Qh3! 27.Bxd4 Qg3+ 28.Kf1 exd4 29.h5 Qf4+ 29…d3 would have ended matters. 30.Kg1 Qe3+ 31.Kf1 Nxe4

This at least has the merit of allowing Black to finish the game with a kingwalk. 32.hxg6+ Kxg6 33.Qc2 d3 34.Bxf7+ Kg5 0-1

Like everyone else, Suttles had his share of unsuccessful kingwalks.

Zuk – Suttles Seattle 1969

Black has been outplayed. In a desperate attempt to salvage the game, Black counterattacks, with a kingwalk being part of the plan. The problem, however, is that the light-square path that Black’s king intends to follow is by no means safe – White is not weak on the light squares and Black’s king has no realistic hope of reaching safety. 38…Ne5 39.Qc8+ Kg7 40.Qxb7+ Kg6 41.Qc8 The greedy 41.Qxb5 was also good. 41…b4 42.Qg8+ Kh5

43.Bf3+!

Simple and strong. The first point is that Black has to capture White’s f3-bishop, because 43…g4 44.Bxg4+ Nxg4 45.Rf5 is mate. 43…Rxf3 43…Nxf3 also loses to 44.Qf7+ Qg6 (45…Kg4 46.Rxf3) 45.g4+ (or simply 45.Qxg6+ Kxg6 46.Rxf3). 44.Rxf3 Nxf3 45.Qf7+ Qg6 46.g4+ Kxg4 47.Qxf3+ 1-0 White doesn’t even have to take Black’s queen – after 47…Kh4, 48.Qg3+ Kh5 49.Qh3 is mate. Some kingwalks are less purposeful than others. In the following game, White’s king wanders around the kingside somewhat aimlessly – or at least the reasons for White’s king moves are too subtle for detection. One term for this sort of maneuvering is “tacking,” and what appear to be aimless moves might, at least at times, have deep points to them. Whether or not that is so, if the opponent weakens his or her position while you’re treading water, you will make progress. That seems to be what happened in the following example.

Suttles – Taimanov Sukhumi 1972

We might call this a “typical Suttles game,” in that it is very difficult to understand. Keep in mind that Black was Mark Taimanov, who had been in the candidates matches just a year earlier (from his point of view, the less said about that the better). 24.Kf2 White’s king starts to wander. The engine gives Black a slight advantage after this move, but it’s difficult to explain why. 24…Ba4 25.Kg2 Bxd1 26.Qxd1 Rb2 Here the engine recommends 26…g5, which is easier to explain – Black would secure the e5-square for his pieces. 27.g5! Qa5 28.Be1 Qa3 29.Kg3 Nb6 30.e5 Na4 31.Be4 e6 32.Bc6?! 32.Rf2! 32…Nb6 33.Kg4

What exactly is White’s king doing on g4, you might ask? That’s a good question, and since it doesn’t stay there, the answer is probably “nothing.” 33…Nc8 34.Kh3 Ne7 35.Be4 Nf5 While White was wasting time, Black has improved his position by bringing his knight to f5. 36.Qc1 Qa4 37.Ng3 Rxh2+ 38.Kxh2 h6!? 39.gxh6 Bxh6 40.Bd2 Nd7?! 40…Qe8!, repositioning Black’s queen, gave Black an edge. 41.Bf3 Kf7 42.Nxf5 gxf5 43.Kg3 Off again. 43…Ke7 44.Kf2 Rh8?

45.Bd1?! The engine’s suggestion 45.Rb1! is very strong – Black shouldn’t have moved his rook from the b-file. 45…Qc6 ½-½ Here a draw was agreed, presumably in view of a repetition after 46.Bf3 Qa4 47.Bd1. But White still has a big advantage, and should have played on. In the next game, we again see a king escape up the board to safety in response to an enemy attack along the back rank.

Rubin – Suttles Montreal 1973

White is down two pawns, and while he has some compensation, it is insufficient. As the time control approaches, White commits to a desperate counterattack along the eighth rank, which Black meets with the tried and true method of a kingwalk up the board. 37.Qd8!? g5! 38.Qh8+ Kg6 39.Bd3+ Ne4 40.Rg8+?! As is so often the case, each check actually improves Black’s position. 40…Kf5

The engine considers 40…Kh5! to be even better.

41.Bxe4+ Kxe4

Nimzowitsch would have wholeheartedly endorsed Black’s centralization, which contrasts sharply with White’s decentralized pieces. White now withdraws his queen to prevent Black’s rook from getting to the first rank. 42.Qa1 Rc6 The second rank will do as well. 43.c3 Rb6 44.Qd1 Rb2 45.Qg4+ Qf4+ 46.Qxf4+ gxf4 The endgame is an easy win. 47.Ra8 f3 48.Kg1 f2+ 0-1 The last two examples in this chapter are both from the 1973 U.S. Open in Chicago, which, thanks to the “Fischer boom,” attracted nearly 800 players. In the first game, Suttles saved a lost position against the eventual tournament winner, Norman Weinstein, with one of the longer kings walks in this book. In the second game, he beat Walter Browne and denied him first place in the event, with the shortest kingwalk in this book.

Weinstein – Suttles Chicago 1973

By playing simple, principled chess against a dubious variation of the Pirc Defense, White has achieved a completely winning position. He has a material advantage (which is about to increase), pressure against Black’s king, and Black’s a5-knight is badly placed. Clearly there is only one chance for Black… 32…g5 33.Qxh7+ Ke6 34.Rd8 Nb7 35.Rc8 Kd6

White’s play has been quite logical up to this point, and with the simple 36.h3 he would remove any possibility of a back rank mate, while keeping an eye on his c2-pawn, which would still be defended by his queen.

Instead White attacks Black’s kingside pawns, and immediately gives up his advantage. We must assume all of this happened in a time scramble. 36.Qg7? Qxc2! 37.Qf8+ Kd5 This kingwalk is very strong – the farther forward Black’s king goes, the more his position improves. At some point, of course, White has to take time out to prevent mate. 38.Qg8+ Kd4 39.g3 Nd6! Black could now give perpetual check, but naturally he plays for more. 40.Rd8? 40.Qd8! was still equal, with Black’s best option being to force the draw. 40…Kc3!? A wonderful time trouble move, if that’s what this was. Objectively Black should first play 40…Qb1+ 41.Kg2 Qe4+, and then 42…Kd3. 41.Qd5 Kb2

42.Ra8? Since Black’s kingwalk began, White has become completely disoriented. He was still all right after 42.Qf3 or 42.b4.

42…Ne4! An unexpected attack on White’s f2-pawn, against which there is no defense. 43.Kg2 Qxf2+ 43…g4!, ensnaring the king, forced mate in a maximum of ten moves. 44.Kh3 f5 45.Qxe5+ Kb1 0-1

A spectacular – as well as a spectacularly successful – kingwalk.

Browne – Suttles Chicago 1973

To fully appreciate Black’s next move, it’s important to understand the context. This was a key last-round game in the U.S. Open, with White having only to draw to win the tournament. The game had been complicated and confusing, but White had let his advantage slip. Browne, who was notorious for getting into – and getting out of – time trouble, had literally seconds left on his clock and had not yet reached the fifty-move time control. The importance of the game, together with White’s extreme shortage of time, made for an electrifying level of tension. Added to this, a crowd of hundreds of spectators were watching, as most of the games in the round had finished. As Browne analyzed every variation he could think of, Black played… 43…Kh7!! This move completely threw White off his game, as it was one of the few moves he hadn’t considered. In fact, according to the engine, this move doesn’t significantly change the evaluation of the position, nor is there any other move for Black which significantly improves his position. The engine therefore confirms Suttles’ conclusion that the position was just a position, and Black had a wide range of choices as to what he played. As he explained

afterwards, his idea was to give White as many options as possible, while he was short of time and facing a mass of potential threats. For obvious reasons, in Chess on the Edge this game was placed in the chapter on “Psychology,” but 43…Kh7!! is also a kingwalk – even if it is a short one. 44.f4 g4 45.Rc6 Bb5 46.Rd6? And here is the time trouble blunder.

46…Bf8? Either Black was now playing a very dangerous psychological game (46… Rxd6 47.exd6 Nd1+ won immediately), or he overlooked this possibility. 47.Re1 Bxd6 48.Rxd1 Nxd1+ 49.Ke1 Bxb4 50.cxb4 Ne3 0-1

Chapter 16 Yasser Seirawan At least one of the authors of this book comes by his knowledge of kingwalks honestly. We have already seen some Seirawan games in the previous chapters, but now we will see that kingwalks are common in his games. Whether they are always justified and successful is, of course, a different issue… In our first example, it is actually Seirawan’s opponent who embarks on an unnecessary and dangerous kingwalk, which ends in disaster. As the reader now knows, kingwalks have to have a sound tactical and strategic basis, as otherwise the king may be simply walking the plank.

Farooqui – Seirawan Columbus 1977

The position is equal. White has more space, and this balances whatever structural weaknesses he might be said to have. But White now comes up with a dangerous idea – to use his king actively in the center. White must not have realized that so many pieces remained on the board. 22.Kd3?! Ndb8 23.Ne2 Na7 24.c4? dxc4+ 25.Nxc4 Nbc6 26.Nb6 Nb5 27.Rc1 0-0 28.Ke4? Black now has a big advantage. What is White’s king doing on e4, other than being a target? 28…Bd8 The engine suggests 28…f5+ as more incisive, but it is very human to assume that White will continue on his chosen path and make things worse. 29.f4 Re7?

But this really is an error. As before, I should have played 29…f6, opening the position. 30.d5? White continues to be helpful. He missed a golden opportunity to equalize with 30.Rxc6! bxc6 31.Bb4. 30…exd5+ 31.Nxd5 Rd7 32.Be1?

32…f5+! Now Black strikes, releasing the bottled up energy of his rooks. 33.exf6 Re8+ 34.Ne7+ Bxe7 35.fxe7 Rdxe7+ 36.Kf3 Rxe2 37.Bf2 Nbd4+ 38.Bxd4 Nxd4+ 0-1

White’s kingwalk ended in complete disaster – he is actually getting mated after 39.Kg3 R8e3+. A kingwalk can take the steam out of an attack, but the very absence of the king can increase the positional effects of a sacrifice. That is the case in the following game, where Black moves his king out of the threatened sector, but White fails to realize that the bishop sacrifice he had in mind as part of a direct attack on his opponent’s king would still give him significant compensation, in the form of a dangerous passed pawn.

Bergmans – Seirawan Graz 1978

Black stands better, because his queenside pawn majority is worth more than White’s advanced e5-pawn. To make the most of this, White has transferred his queen to the kingside, and now threatens 19.Bxh6. Black must move his king, but in which direction? 18…Kf8!? It shouldn’t come as a surprise that the engine recommends 18…Kh8!? 19.Qh5 Renewing the threat of 20.Bxh6. 19…Ke8!?

Black continues his kingwalk, hoping to neutralize White’s attack and get back to positional chess. 20.Nd2? White blinks. In a surprise to one of your authors, the engine assesses the position as equal after 20.Bxh6! gxh6 21.Qxh6, as White’s passed h-pawn gives him enough compensation for his sacrificed piece. This continuation is actually very logical, because otherwise White’s c1-bishop isn’t very useful. 20…Bf8 21.e4 Nb6 22.f4 White’s idea is to break through with his pawns, but there’s no chance this will work, in view of Black’s superb piece placement. 22…Rad8 23.Qe2 Rd3 24.Nf1 Nc4 24…c4! was even better. 25.Rb1 Qd7 26.b3 Na5 27.Qc2 c4 28.b4 Nb3 29.Be3 Nd4 30.Qa2 a5 31.Red1?

Now Black crashes through down the d-file, while his king serenely observes from its original square. 31…Rxd1 32.Rxd1 Ne2+ 0-1 As we have stressed throughout this book, not every kingwalk is justified, and an ill-advised kingwalk can ruin a position just as quickly as any other strategic error. Players who undertake kingwalks frequently will inevitably be harshly reminded of this truth.

Fries Nielsen – Seirawan Graz 1978

After blundering two rooks for a queen, Black has fought back and, while there is no draw in sight, there is no clear win for White either. With a few moves required to reach the time control, White naturally checks. 39.Bf7+ Kf5? A very ill-advised kingwalk. After the modest 39…Kg7, discovered checks by White do nothing (other than gain time on the clock), because every black piece and pawn is on a dark square. 40.Rb5+ Kg4

41.Be6+

41.f3+! Kh4 42.g3+ finished things off right away, because after 42…Kh3 43.Be6+ g4, White has a pretty mate with 43.Bxg4+ Nxg4 44.Rh5 mate and a beautiful one with 43.Rh5+ Nxh5 44.Bxg4 mate. 41…Kh4 42.g3+ Kh5 43.Bf7+ Kg4

44.f3+ 44.Kg2!, threatening both 45.f3 mate and 45.Nh2 mate, won immediately. Taking the opposition can be important in the endgame. 44…Kxf3 45.Rf5+ Ke2 46.Rxf6 The butcher’s method, rather than the artist’s, but equally effective in the end. 46…Qb4 47.Rf2+ Ke1 48.Bh5 Qe7 49.Rd2 1-0 A bad kingwalk can ruin a good position, but that doesn’t mean a good kingwalk will save a bad position. What such kingwalks can do, though, is to make the finish of a game more entertaining, even when the psychological effect of the kingwalk isn’t enough to change the objective reality of the game.

Seirawan – Evans Greenville 1980

Black has been outplayed and has a lost position. He is a pawn down, and is burdened by a very unpleasant double-pin along the seventh rank. Black concludes his best chance is to go forward with his king, which at least partly solves the problem of the pin on his c7-knight. And maybe Black’s kingwalk will improve the safety of his king? Maybe not. 34…Kh6!? 35.Nc3! 35.Bd6 runs into 35…Qd5, with some resistance. Now Black’s c2-rook no longer defends his c7-knight. 35…Bd8 36.Qxe5 White’s predilection for taking pawns is well known, but here White’s capture also brings his queen closer to Black’s king. 36…Qd7

37.Bf8+ Black’s king must now continue his walk, but things won’t end well. 37…Kh5 38.g4+! Kxg4 39.Qf4+ Kh5 40.Qh6+ Kg4

41.Rb4+ 1-0 Both 41.f3+ and 41.h3+ mated in two, but 41.Rb4+ is even better, because now White gets to capture Black’s queen as well, after 41…Qd4. But Black resigned instead. A kingwalk in the opening can be justified in several ways – one being to generate a rich, complicated middle game. To those who adhere to the belief that the goal of the opening in chess is to get a playable middle game, kingwalks are therefore a valuable weapon. This assumes that “playable” means “interesting and complex, where the players can make errors.”

In the following example, White accomplishes this, as his small opening kingwalk leads to a difficult position for both sides.

Seirawan – Korchnoi Wijk aan Zee 1980

11.Kf1!? We have seen this move before, in Petrosian-Karaklajic, Yugoslavia 1954. The idea is the same: to keep the position more complex, by avoiding exchanges, and to throw the opponent off with an unexpected move. Against Korchnoi, the former goal is immediately reached (but be careful what you wish for) and the latter goal is more difficult. Nevertheless, Black’s reply is not the best. 11…Nc6?! A natural but poorly timed developing move. My prep focused on 11…Bc3! 12.Rb1 Nc6!, only now. Capturing the pawn by 12…Bxd4?! 13.Nxd4 Qxd4 14.Bb2 allows a skewer, as White favorably regains the g7-pawn. I had intended 13.d5 Nd4 14.Ba3, with a small advantage. 12.Bb2 I had reached this position before in my tournament practice. Better still, I had analyzed this exact position deeply at home. My conclusion was emphatic – Black is just busted. With natural play by Qd1-e2, Ra1-d1 and d4d5, with Qe2-e4 to follow, Black will face insurmountable threats. Happy dreams of an early scoresheet signature filled my head. Korchnoi now

thought for over an hour and played a move I had never considered at all, even after many hours of study. 12…Bc5 How to even annotate such a move? Recognizing the dangers he is facing, Korchnoi tries to throw a monkey-wrench into White’s plans. Now it was my turn to think. 13.Bc2 Just like Korchnoi’s eleventh move, this natural move wasn’t best. My struggle was where to put this bishop in order to protect my d4-pawn. I really liked 13.Bb1!, which stops kingside castling cold: 13…0-0? 14.Qc2, threatening checkmate while defending the b2-bishop. Black loses his c5bishop. But what if Black decides to try to go queenside, with 13…Bd7! 14.Qd2 Bb4 15.Qe2 Qe7? I began feeling sorry for the entombed a1-rook. Crossing up Black’s queenside castling plans with 13.Be4! felt right, but that allows Black to go kingside: 13…0-0 14.Qd2 Bb4 15.Qc2, with an altogether wrong battery. Annoyed with the defects of each choice, I settled on the text. At least the patterns would be familiar. Quantum chess would allow White’s bishop to be on all these squares at once. 13…0-0 14.Qd3 Rd8 Funnily enough, I have the bishop and queen battery I wanted – just on the wrong squares. In my prep my queen was always on the e4-square, making my d-pawn a free radical. The pin along the d-file is annoying. 15.Rd1! Kf8! Preparing to flee the jurisdiction in which I intended to prosecute. The battery that I had hoped would be a decisive factor is, surprisingly, not so terrifying after all.

16.Qe4 Annoyed by a potential …Nxd4, followed by, …e6-e5, I decided to simply get out of the pin, while also stopping Black’s best scheme of development, which is fianchettoing on the queenside. 16…Bd6

Despite my not having landed an early knockout blow, my opening has been a success. White has more space and active pieces which are ideally situated to attack Black’s king, with the lone exception of the h1-rook. 17.h4! This is one way to solve the problem of the h1-rook. A more conventional approach was 17.g3, followed by 18.Kg2 and 19.Rhe1, developing the rook on a central file. I thought the text was a clever prelude to a rook lift to the ffile. Fanciful ideas like Rh1-h3, Qe4-h7, Nf3-e5 and Rh3-f3 filled my head. The engine isn’t concerned about White’s h1-rook at all, recommending: 17.Qh7!, a move I considered but wrongly rejected because of 17…g5, with counterplay against the d4-pawn. This assessment isn’t even close, as after the very powerful 18.Rb1!, Black is in big trouble. 17…Qf5! Simply put, White’s queen is far too strong on the e4-square. 18.Qe2

Thanks to my previous move I wanted to attack, and therefore I retained the queens. Had I played 17.g3, I suspect I would have just continued with 18.Kg2, accepting the queen trade. Attacking moves can have a narcotic-like effect. Come hell or high water, I’m in attack mode. The engine is too, as it still gives White a big advantage after 18.Qe2, 18…Qa5! Annoyingly, Korchnoi avoids a very fine trap. 18…Qh5? at first blush appears to be an obstinate way to prevent White’s kingside play. Actually, Black’s queen instantly becomes a target as after 19.d5! Nb4 (19…exd5 20.Rxd5! Qg4 21.Rxd6! Rxd6 22.Ba3) 20.Bxg7+! is a nasty surprise. If Black captures with 20…Kxg7, 21.g4! Qxg4 22.Rg1 is a nice way to involve the recalcitrant White rook in the game. 19.Bb3?! A natural move with a dual purpose: to protect the a2-pawn while preparing the launch of White’s d4-pawn to d5. By this time, I had given up on any glories to be had on the b1-h7 diagonal and had set my sights on Black’s f7pawn. Ideas such as c4-c5 and d4-d5 were dancing in my mind. The merciless engine, however, points out the ultra-aggressive 19.Ng5! as best. The knight leap is based on an attacking line far beyond our capabilities: 19…hxg5 20.hxg5 Qxg5 21.d5 Na5 22.Rh5 Qf4 23.g3 Qxc4. Even if I had gotten this far, I would have stopped and given up on this ghost of an attack. But the engine continues with 24.Bd3 Qa4 25.Re1!, and White’s position is absolutely crushing. Damn! 19…Ne7! Another fine defensive move. Black removes the sting of a d-pawn advance while improving the prospects of his c8-bishop. Finally, if White comes back to a line like Qe2-e4-h7, Black’s e7-knight might prove to be a useful defender. 20.h5

Enthused with the spirit of youth I press forward. I was determined to make a rook lift work. 20…b6?! Once more, the most natural move in the position turns about to be wrong. Having almost equalized, Black should have played either 20…Bd7, connecting his rooks at once, or 20…Bc7, increasing his control of the d5square. Now my attack springs to life. 21.Qe4! Ba6 Due to the hanging a8-rook, Black has no time to play 21…Qf5, which was a key defensive reaction. 22.d5 A move of which I was unjustly proud. In my mind the raking white bishops would rule supreme. Sadly, the engine gives the mundane 22.Qh7!, with a straightforward winning advantage for White. I had wanted to close the fifth rank and prevent 22…Qf5, as a possible defense. It wasn’t necessary. 22…exd5?! Black’s only chance was 22…Ba3, seeking trades. 23.Qh7 f6 24.Kg1

Tucking the king away. We can think of this as the completion of the kingwalk started with 11.Kf1. 24…Bxc4 I expected 24…dxc4 25.Bc2 Bc5 26.Re1!, with a decisive attack. The text allows me my rook lift. 25.Rh4! 1-0 (39) At last, everything has come marvelously together. White’s pieces are all working in harmony. Korchnoi was under his usual time pressure duress, and I won just before the time control in a tense finish. The traditional kingwalk to safety before undertaking an attack on the opposite flank will always be part of chess. Here it is successful, although only with some help from my illustrious opponent.

Gligorić – Seirawan Baden bei Wien 1980

Black has a strong grip on the position because of his pressure on White’s f3pawn. The potentially weak white c3- and c4-pawns are money in the bank, although it will be a long time before Black can take advantage of them. Black’s kingwalk is easy to see – his king goes to the queenside before he opens up the kingside. 34…Kf8! 35.Rf2 Ke8 36.Rf1 Kd7 37.Rf2 Kc8 38.Rf1 Once again, we see one player literally moving a piece back and forth while the other completes a kingwalk. Mikhail Suba referred to this as “one player doing while the other one is watching.” Okay, I’m the one doing but now what do I do? 38…Qh7 39.Bd2 Rf8 40.Qe1 Qf7 41.Qd1 Kc7 42.Be3 Qg7 43.Kh2 Qh7 44.Kg2

Now Black must find a way through, which is by no means an easy task. 44…Rg8 45.Bd2 Sidestepping a possible …Ng6-f4+ leap. 45…Nf8 46.Qb1 Nd7 47.Ree1? Nb6! 48.Qe4 Nxa4 49.Ra1 Nb6 50.Rxa7 Rf7! Exchanging queens will remove White’s most important defensive piece. 51.Re1?! Qxe4 52.Rxe4 Rgf8

Success! Black’s king defends his b7-pawn, and White’s pieces have lost their coordination, so White can no longer defend both his f3- and c4-pawns. The “two weakness theory” is confirmed once more!

53.f4 gxf4 54.gxf4 Nd7 55.Kg3 Rg8+ 56.Kf2 Rfg7 57.Ke2 Rg2+ 58.Kd1 Rh2 59.Re2 Rg1+ 60.Be1 Rhh1 61.Kd2 exf4 62.Bh4 0-1 Every now and then it’s necessary to dispel the illusion that all kingwalks are carefully calculated, strategically sound maneuvers which accomplish a clear positional or tactical goal. These exist, but so do examples where the kingwalk more resembles a drunken Viking chieftain stumbling out of a tavern in mid-winter into a midnight snowstorm. Sometimes, against all odds, the king makes it back in one piece. Seirawan – Bellon Las Palmas 1981 1.c4 e6 2.Nc3 d5 3.d4 c6 4.e3 Nf6 5.Nf3 Nbd7 6.Qc2 Bd6 7.e4? This central break is too simplistic. 7…dxe4 8.Nxe4 Nxe4 9.Qxe4 e5! 10.dxe5 Qe7?! 10…0-0! was better. 11.Bf4 Bb4+ 12.Bd2 Bxd2+ 13.Kxd2?!

White is nothing if not ambitious! I have my precious pawn, and giving it back with the superior 13.Nxd2 just wasn’t going to happen. 13…Nc5

Even stronger was 13…Qb4+!, followed by 14…Nc5. 14.Qh4? After 14.Qf4!, the position is equal. Once the implications of the lines starting with 13…Qb4+ had well and truly sunk in, I was anxious to trade queens. I had made a petite oversight, though. 14…g5!

Duh! And this is it. In an instant my position is in ruins. 15.Qh6 Black’s g5-pawn is of course protected by the knight fork on e4. 15…Bf5 Developing pieces is always good. In this instance, 15…Ne4+ was better. 16.Re1 Hoping to escape to the queenside. 16…Qd8+ 17.Qd6 Qa5+ 18.Ke3

When playing 13.Kxd2?!, I definitely did not want my king to end up here! 18…Ne4! 19.Qa3 Qb6+ 20.c5 Nxc5 21.Rd1

The engine assesses Black’s advantage as being worth more than a rook. That’s about how it felt at the time. While the engines have taught us much, they have not affected the truth of the adage that “no one ever won a game by resigning!” 21…Ne6+? There was more than one way to win, but the simplest was 21…Nd3+! 22.Nd4 0-0-0! 23.Qc3 Nxb2, simply grabbing a pawn with 24…Rfe8 to follow. White can do nothing against these natural moves. 22.Ke2 g4 23.Nh4 Bc2 24.Rd6 Rd8 25.Ke1

25.Kd2 is better, objectively speaking, but it’s understandable that White wants to bring his king back to its original square and try to start over. 25…Rxd6 26.exd6 Kd7 Cheeky. Black shows when it is proper and fitting to make a kingwalk. 27.Be2 Nd4? A simple developing move was all that was required. 27…Re8! 28.Bxg4 h5! gives White’s g4-bishop a terrible choice about which diagonal to give up. Unpinning Black’s e6-knight allows a discovered check on the e-file, while giving up control of d1 allows the deadly 29…Qd4!. Remember, White can’t castle… 28.Bxg4+ f5 29.Bh5! Still hanging in the game. Now the e8-square is covered by White’s h5bishop. 29…Be4 30.Qc3 Rf8 31.Kd2 Rf6 32.Rd1 Rh6? The immediate 32…Rxd6 was right – why spend a move to let White bring his h4-knight back into the game? 33.Nf3! Rxd6

White has fought his way back into the game, and after 34.Nxd4 c5 35.Ke3! the position would be dead equal. Instead, I gamble on trying for a win in my opponent’s time trouble. Adrenalin induced overload can cause us to make wacky decisions. 34.Ne5+? Ke6 Now it’s Black’s turn for a kingwalk! 35.Kc1 Kxe5 36.f3 Rd5 37.b4

37…Qa6? A terrible time trouble blunder. There was no reason for Black’s queen to abandon the d4-knight. After 37…Bc2, Black is winning. 38.fxe4 Qxa2? 39.exd5 1-0

At this point, looking aghast in total horror, Black resigned. Is it right to use the term “kingwalk” when the king’s perambulation is caused by the opponent’s attack? We think so, and the terminology shouldn’t depend on the intention of the defender. In other words, accidental kingwalks are still kingwalks – not that such an expanded definition is required in the next example.

Kogan – Seirawan South Bend 1981

Black has won a pawn by exploiting a pin on the f1-a6 diagonal. Passive defense holds little hope, so White decides to sacrifice a piece for attacking chances. A kingwalk to safety ends White’s hopes. 22.Ne4 Rxd5 23.Nxf6+!? gxf6 24.Qxf6 Rf5! 25.Qh8+ Kf7 26.Qxh7+ Ke8 27.Qh6 Bd3 28.Ra1 Qe7 29.Qg6+ Qf7 30.Qg3 Kd7

Mission accomplished, although Black’s king can keep going if it has to. Generally speaking, in this sort of position, White only gets enough for the piece if Black’s king continues to be open to attack, or if White’s h-pawn also is a threat, as in the Capablanca-Schroeder game in Chapter 11.

Neither of these factors applies in this position, and Black soon takes the initiative. 31.Bd4 Rg8 32.Qe1 Rxg2+! 33.Kxg2 Rg5+ 34.Qg3 Rxg3+ 35.hxg3 Nb3 Or 35…Be2. 36.Rc3 Nxa1 37.Rxd3 Nc2 38.Bxb6+ Kc6 39.Bxa5 Kb5 40.Bb4 e5 41.e4 Nd4 42.g4 Ka4

Black’s king is completely safe on a4 and Black’s queen and knight run rings around White’s depleted defenders. 43.Be1 Qc4 44.Rc3 Qe2+ 45.Bf2 Ne6 0-1 A kingwalk may be an integral part of a sound defensive plan, may be improvised, or may be something in between – as when it turns out that the kingwalk fails, or could have failed, as in the next game.

Llanos – Seirawan Morón 1981

White has sinned – he advanced his e4-pawn to e5, opening the b1-h7 diagonal for Black’s h7-bishop. Faced with a queenside collapse, White sacrifices a piece. 19.Bxh6 gxh6 20.Qxh6 Kh8 For the moment Black’s king goes to the corner. 21.Ng5?! Bxg5 22.hxg5 Rg8 23.f4 Ncd4 24.Rac1 bxa3 25.bxa3 Nc3 Uncharacteristically, Black doesn’t take White’s a3-pawn. 26.Kf2 Rg6 27.Qh5

27…Rb8? An interesting moment. The engine gives Black a big plus after either 27…c4 or 27…Qb6, exploiting the position of White’s king. After 27…Rb8?, in contrast, the engine assesses the position as slightly better for White! 28.Ng4! Kg7 Now Black resorts to a kingwalk, as the danger to his king is real. 29.Nf6 Bg8 30.g4 Kf8 31.f5 Rg7 32.Qh6 Rb2

The position I had been playing for. I thought that all my bases were covered and I would crash through against the c2-pawn 33.Nh7+? A move I was prepared for. Until now, White has made the most of his opportunities, and with 33.g6! he would have crashed through first. The deadly threat is 34.Nh7+ Bxh7 35.gxh7, getting a second queen. 33…Bxh7 34.f6 Ke8 35.Qxg7? Short of time, White makes it easy for me to keep my kingside intact. 35…Bg6 36.Rh1 Kd7

The pendulum has swung back completely in the last four moves. White’s attack has ground to a sickening halt and, as a result, White’s queen is totally out of play. 37.Bf1 Rb8 Preventing 38.Rh8 is a fine idea, but 37…Qb8!, attacking White’s e5-pawn, would lead to a checkmate. 38.Qh6 Qa5 39.Qh2 Ne4+ 40.dxe4 Qd2+ 41.Kg1 Nf3 mate 0-1 Here is another example of a short kingwalk in the opening, to avoid unwanted simplification. Seirawan – Christiansen Lone Pine 1981 1.c4 e6 2.Nc3 d5 3.d4 Nf6 4.Nf3 Be7 5.Bf4 0-0 6.e3 b6 7.cxd5 Nxd5 8.Nxd5 exd5 9.Bd3 Bb4+

10.Ke2!? It’s interesting that of the three possible ways that White can reply to 9.Bb4+ (10.Nd2, 10.Kf1 and 10.Ke2), the engine dislikes 10.Ke2 the most! 10…Bd6 11.Bxd6 Qxd6 1-0 (44) Black has comfortably equalized, but White went on to win. Covering a weakness prior to starting action in another part of the board is a strategy that can apply in the endgame, as well as in the middle game.

Seirawan – Jigjidsuren Lucerne 1982

This pawn structure often offers Black dynamic chances in the middle game, because White’s b2-pawn is just as much of a target as is Black’s d5-pawn. But we have reached the endgame, and Black still has his worst minor piece. White therefore has a permanent advantage, and now brings his king to the queenside, to defend his b2-pawn, before initiating active operations. 38.Kf1!? Kf8 39.Rfd4 Rc5 40.Ke1 Ke7 41.Kd1 Rb7 42.Kc2 Rd7

With his kingwalk completed, White now plays to create dark-square weaknesses in his opponent’s position by fixing Black’s kingside pawns on light squares. This involves White advancing his kingside pawns, which makes White’s repositioning of his king doubly useful.

43.Rh4! h5 44.g4 hxg4 45.hxg4 Rd8 46.g5 Rb8 47.Rhd4 Ke6 48.Rf4 Rh8 49.Rf6+ Ke7 50.Rd4 Rh5 51.Rdf4 Be8 52.Rxa6 1-0 On time? White still has an edge. The reader may be asking “but do I really have to understand kingwalks to play good chess?” Our answer would be “maybe not, but you might.” Feeling comfortable with kingwalks means that the idea is ready to be used in key games, such as when you are playing the world champion. This doesn’t guarantee success, however, as the following tragedy illustrates.

Karpov – Seirawan Mar del Plata 1982

Black has achieved an ideal position out of the …b6 variation of the Winawer French, thanks to a miscalculation by the reigning world champion. White’s f6-bishop is as good as a bad bishop can be, but, while it controls a number of important dark squares, it can’t do anything about the light squares, especially c4, from which Black’s knight dominates the queenside. White will try to create counterplay on the kingside, but in fact it is the queenside where all the action will occur, and with this in mind Black begins a kingwalk. Chess Duels puts it this way: “Black’s position on the queenside is so nice that his king decides to stroll over and have a closer look.” 23…Kd7 24.g4 Kc7 25.Kf2 Kb7 26.Bh4

26…Ka6?! An interesting example of rushing to complete a kingwalk. As Larsen pointed out after the game, Black should first play 26…a5!, and only then 27…Ka6. Black would then have a very clear – and unstoppable – winning plan. After suitable preparation, he would play …b6-b5, opening the b-file in order to exchange one or both rooks, while at the same time creating a powerful passed a-pawn. 27.Ke2 c5 ½-½ (81)

After many misadventures and more than 50 additional moves, Black managed to throw away numerous winning positions and the game ended in a painful draw. Here is another example of the kingwalk as a defensive resource, nullifying the opponent’s attack.

Böhm – Seirawan OHRA Open 1983

Black has lots of material for his queen, and has just captured a white pawn on b2, giving him a very dangerous passed b3-pawn. In short, from a material point of view White is lost. His last chance is a desperate counterattack against Black’s king, trying for a perpetual check. 30.Rxg7+!? Kxg7 31.Qg5+ Kf7 32.Qg6+ Ke7 33.Qg7+ Kd6 34.Qxf6

Black’s king is safe, surrounded by friendly pieces. My goal is now to advance Black’s b3-pawn and tie down White’s queen, so that White’s counterattack can’t flare up again. 34…Rc2 35.Qf8+ Kc7 36.Qb4 b2 37.Kh2 Rg7! 38.Kh3 Rgxg2 39.Qe7+ Kc6 40.Qe8+ Kc5 0-1

Sometimes you can run and hide. In this case, the a1-square will make for a great hideout. Escaping from checks along a pathway of dark or light squares is one of the most common kingwalks, and you will find yourself on both sides of this technique.

Miles – Seirawan Linares 1983

After a doubtful opening and an ill-conceived attempt to gain compensating activity, Black has arrived in a completely lost position. All that is required of White is the realization that his king can take care of itself with a short kingwalk. 27.Bh6! Re2+ 28.Kg3 Rg2+ 29.Kh4

And that’s that. White’s king is completely safe on h4, while the same cannot be said of Black’s king. 29…Qd8+ 30.Bg5 Qb8 31.Rf3 Qe5 32.Re3 Rxg4+ 33.hxg4 1-0

The following example features a short kingwalk by White in the opening, in a known variation, then a second kingwalk by White in the middle game, then a third, and unsuccessful, kingwalk by Black. Seirawan – Böhm Lugano 1983 1.c4 c5 2.Nf3 Nf6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nb4 6.Bc4 A well-known opening variation, in which White allows his king to be kicked around in return for a lead in development. 6…Nd3+ 7.Ke2 Nf4+ 8.Kf1

A small kingwalk in the opening, but that alone doesn’t justify the inclusion of this game in a book on kingwalks. 8…Ne6 9.b4 cxb4 10.Nd5 g6 11.Bb2 Bg7 12.Bxg7 Nxg7 13.Nxb4 0-0 14.h3 Qd6 15.Rb1 White’s lead in development has dissipated and his position remains a little loose. Not surprisingly, the engine gives Black a small edge. But in the real world, this is just a position. White rolls up his sleeves and starts to outplay his opponent. 15…Nc6 16.Nxc6 bxc6 17.Qc2 Be6

18.Ke2!? An interesting, and challenging, continuation. White could also play 18.g3 and 19.Kg2, but he decides to risk using his king in the center, mainly so it can help defend White’s d-pawn. 18…c5 19.Rb7 Rfb8 20.Rhb1 Bxc4+ 21.Qxc4 Ne6 22.g3 Kf8 23.d3 Rc8

24.Ke3 The engine, which is not always fond of kingwalks, dislikes this move. But it turns out just fine. The game now enters a maneuvering phase, in which White gains the advantage by exploiting the weaknesses on Black’s queenside. The game doesn’t end with White promoting his a-pawn, however.

24…f6 25.R1b3 Rd8 26.Ra3 a5 27.Rab3 Ra6 28.Rb8 Kg7 29.Rxd8 Nxd8 30.Rb5 Ne6 31.a4 Qc7 32.Qc3 Qd8 33.Nd2 h5 34.h4

White is on top, as soon Black will lose a pawn. Black decides the best chance is a kingwalk of his own, to try to generate counterplay on the kingside, but he’s wrong. 34…Kh6?! 35.Nb3 Qd7 36.Qc4 Rd6 37.Rxa5 Nd4? 38.Qg8! It turns out that Black’s king isn’t all that well placed on h6. 38…g5 39.Ra8

There won’t be any “kingwalk up the board to safety” this game. 39…Nc2+ 40.Kd2 1-0

And now another example of a kingwalk in a bad position that churns up enough dust to obscure what is going on. White not only got out of trouble – in time trouble he could even have won.

Seirawan – Alburt Estes Park 1986

White has a miserable position after a failed attempt to refute the Benko gambit. 19…g5 20.a4 f6 21.a5 Kf7 22.Ra3 Ng6

Black’s advantage has grown, so White resorts to a proven strategy – a kingwalk, to give the position a makeover. 23.Kf1!? Qa6 24.Ke1 Rhh8 25.Kd1 Rab8 26.Nd4 Nhf4 27.Kc2 Ne5 28.Kb1

Surprisingly, White’s kingwalk has worked, up to a point. Of course Black is still better. 28…Rhc8 29.Rc1 c3 30.Rcxc3 Rxc3 31.Rxc3 Nc4 32.Rxc4 Qxc4 33.Nc6 Rb3 34.Qc2 Qf1+ 35.Ka2 Rxf3

It looks as though White’s position is finally collapsing, but White’s king is safe, and that allows him to launch a counterattack against Black’s king, which is in more danger than it seems. White’s next move shows us why. 36.e5! Ng6 A risky winning try 37.Nd8+ Kg7? 38.Ne6+ Kh7

39.Nf8+? Anything can happen in time trouble. White had a surprising win with 39.exf6! Rxf6 (39…Rxf2 40.Nf8+ Kh8 41.Nxg6+ Kg8 42.Nxe7+ Kf8 43.Qc8+ Kf7 44.Qe6+ Ke8 45.Qg8+ Kd7 46.Qc8 mate) 40.Ne4. In these variations, where else could White’s king be to keep out of trouble? 39…Kg8 40.Qxg6+ Kxf8 41.Qh6+ Kg8 42.Qg6+ Kf8 43.Qh6+ Kg8 ½-½ Black can’t avoid perpetual check with 43…Ke8, because of 44.Qh5+ and 45.Qxf3. If anyone needed reminding, the next example again demonstrates that a kingwalk in a bad position doesn’t always help.

Miles – Seirawan Sarajevo 1987

Black, having overlooked a surprising opportunity several moves earlier, has a difficult position. Lines are starting to open on the kingside, and things will only get worse. Black therefore tries a dangerous kingwalk, hoping to at least get his king out of danger by transferring it to the queenside. Unfortunately, this takes time, and the position is not static. 23…Kf7!? 24.d5!? 24.Rh3! is even stronger. 24…exd5 25.Bxf5 Ne5 26.cxd5 Bxd5 27.Bxf6 Kxf6 Why not? The opening of the position has clearly increased White’s advantage, so Black tries to complicate matters. 28.Rxh5? The confusion often brought on by a kingwalk kicks in. 28.f4! was right. 28…Rh8 29.Bh7? Another inexact move. Black is back in the game.

29…Bf7?! The position is difficult for both sides, and Black unnecessarily allows White to unpin his h7-bishop. The immediate 29…Ke7 was better. 30.Rh3 Ke7 31.Bf5 Kd8 32.f4 Nc6 33.Ne4 Kc7

Black has completed his kingwalk, and his position is considerably better than it might have been. White has an extra pawn, though, which gives him a comfortable edge – and since Black’s king is not completely safe on the queenside, any Black efforts at counterplay will expose Black’s king to renewed dangers. 34.Qd2 Rd8 35.Rxh8 Qxh8 36.g4 Na5 37.Rh2 Qg8 38.Rh7 Kb8 39.Nf6 Qf8 40.Nd7+ White cashes in, although 40.g5! was strong as well. 40…Rxd7 41.Bxd7 Nc4 One well-placed piece isn’t enough. 42.Qh2 Qe7 43.Qe2 d5 44.Ba4 Nd6 45.g5 Ne4 46.Qa6 Qe6 47.Rh8+ Bg8 48.Bd7 1-0 The kingwalk in the next example is elegant and has two functions – White’s king finds safety, and also clears the way for a surprising, and very strong, attack by his major pieces.

Seirawan – I. Sokolov Sarajevo 1987 1.c4 Nf6 2.Nc3 d5 3.cxd5 Nxd5 4.g3 g6 5.Bg2 Nb6 6.d3 Bg7 7.Be3 0-0 8.Qc1

White’s attempt to trade dark-square bishops is entirely justified, especially since Black takes the time to avoid it. This gives White the opportunity to start a direct attack on Black’s king by opening the h-file. 8…Re8?! 9.Bh6 Bh8?! 10.h4! Nc6 11.h5 Nd4 12.Nf3 Nxf3+ 13.Bxf3 c6 14.hxg6 hxg6

15.Kf1! The exclamation mark is for the strength of this move, not the difficulty in finding it. By walking his king to g2, White keeps his h1-rook on the h-file,

while at the same time putting his king on a safe square (g2) and allowing his major pieces to operate along the first rank. The importance of this last point becomes apparent later on. 15…Nd5 16.Kg2 Nxc3 17.bxc3 Qa5 Black needs counterplay, but there is little to be found. 18.Bd2 Bg7 Black can’t allow 19.Rxh8+. 19.Rb1 Be6 20.c4 Qc7 21.Bh6 Bf6

22.Qg1! See the note to 15.Kf1! This pretty queen maneuver decisively increases the pressure on Black’s kingside. 22…Bf5 23.Qh2 e5 24.Bg5 Bg7 25.Qh4 f6 26.Bxf6 Rab8 27.g4 1-0 Sometimes chess is a pleasure to play, at least for one side. Often a king escapes from a counterattack by walking to safety – we have seen many examples. It is less common to induce an enemy attack in order to execute a kingwalk that has an ulterior motive.

T. Brown – Seirawan Los Angeles 1987

Black got a big advantage out of his Sicilian Defense, but a few inaccurate moves let White back into the game. White should now play calmly with 33.h3. but instead sees an opportunity to drive Black’s king up the board. What to White is an attack is to Black a winning kingwalk, thanks to White’s weak queenside pawns. 33.Nxb7? Kxb7 34.Qe1 Kb6!? 35.Kh1 Ka5

36.b3? In a different universe, Black’s king might be uncomfortable on a5, but here White’s b3-pawn is weak, and Black can destroy White’s queenside pawn

chain. 36…Rc3 37.Re2 Rxb3 38.Rxe5+ Kxa4 39.Re2 Ra3 40.Qd1+ Ka5 41.Re1 0-1 Some test readers were found to have developed an unhealthy affection for kingwalks. To counter this, it is necessary to look at examples where a king walks into mate. Here is an elegant one.

Seirawan – Andrijević Lugano 1988

Material is equal, but Black’s a5-knight is so far removed from the kingside that White is, in effect, a rook up. Black tries to run away with his king, but White, having done so himself on many occasions, knows how to weave a mating net. 44…Kh6 45.Rd4! Qg5+ 46.Kh1 Qc1+ 47.Kh2 Qc5

48.Qf4+ Qg5 If Black’s king goes back with 48…Kg7, White has 49.Rd7+, while 48…Kh5 runs into 49.Qh4 mate. 49.Qf8+ Kh5 50.Qg7!

This maneuver is so common that it is easy to gloss over. However, it is important enough that it deserves a nice label like a “fitting.” This is for a straightjacket! By inducing Black’s h7-pawn forward, Black’s king’s retreat is cut off, while his kingside structure is weakened. Look for this “fitting” maneuver in your king hunts. 50…h6 51.Rg4 Qf5 52.Rh4+! 1-0 Black resigned, because he is mated after 52…Kxh4 53.Qxh6+ Qh5 54.Qf4+ Qg4 55.Qxg4 mate. The following example is a pretty one, which shows that king safety can be an optical illusion. Black’s king, which ends up on the fourth rank, is safe, while White’s castled king, abandoned by his attacking pieces, is the one that is in trouble.

Gavrikov – Seirawan Lugano 1988

Having striven for over 20 moves to demonstrate that a knight is better than a good bishop, Black has no advantage – other than the paradoxical fact that his king seems less secure than his white counterpart. After White’s next move, Black decides to offer an exchange of minor pieces. 46.Qb2?! Nc4 47.Bxc4 dxc4 48.Rb8?! White sees the chance of an attack against Black’s seemingly exposed king, but in fact White’s king is more vulnerable. 48…Rxa6 49.h3? Qf5! 50.Qb4 Qe4+ 51.Kh2 Qf3 52.Qf8+ Kg6 53.Qg8+ Kh5 0-1

Since 54.g4+ Kh4 only makes things worse, and White has no way of either attacking or defending, he resigned. We have seen many kingwalks that are objectively bad. Sometimes they work, and sometimes they don’t. But even when a kingwalk works, the king may find that the security systems in his new home aren’t up to code, and accidents may happen. This is what occurs in the next example, but White is not to be faulted, as his kingwalk was a winning attempt.

Seirawan – Ljubojević Barcelona 1989

In this complicated position, Black sacrifices his e6-bishop. The sacrifice is only good for a draw, but any other move leaves him with a worse position. 22…Bxh3! 23.gxh3 Qxh3 24.Rh2 The tournament bulletin queried this move, but it doesn’t change anything. 24…Qg4+ 25.Kf2 Qf5+

26.Kxe3 The tournament bulletin treats this as a blunder, but it’s actually just a risky attempt to win. If White retreats his king with 26.Kg1, the game ends in a

draw by perpetual check. 26…Qxe5+ 27.Kd2 Qxh2 28.Qxc4 Rae8 29.Rd1 Re5 30.Kc2 Qh7+ 31.Kb3 Qh3 32.Nd4 Rc8 33.Qb5 Re3

White’s kingwalk has neither succeeded nor failed – the position is still equal, although the equilibrium is certainly dynamic. Add time trouble to the mix and the difference between an “equal” and a “drawn” position becomes apparent. 34.Ka3? I had been playing for this mistake from afar for quite some time. I reasoned that I’d play Nd4-b3, when my king would be super-safe and my d-pawn would just rush up the board to win me the game. It turns out that now White’s king is actually on a much worse square! 34…Qg4! 35.Rd2 Qf4 36.Rd1? Qd6+! 37.Kb3 a6!

Now White’s d5-pawn falls, as 38.Qa5 Rh3, followed by 39…b5. cuts off White’s queen. 38.Qb4 Qxd5+ 39.Ka3 Rc4 0-1 Sacrificing the kingside in the opening in order to knock out the opponent’s center, with a kingwalk as an essential element of the sacrifice. What could be more natural? Timman – Seirawan Hilversum 1990 1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nc3 e6 5.g4 Bg6 6.Nge2 c5 7.h4

7…h5!?

As Daniel King points out, at the time it was thought necessary for Black to interject 7…cxd4 8.Nxd4, so that White can’t respond to …h7-h5 with Nf4. Black takes a different approach, which involves a short kingwalk. 8.Nf4 Nc6 9.Nxg6 fxg6

10.Qd3? White can resist everything but temptation. He sacrifices his d4-pawn in order to immediately exploit the weaknesses on Black’s kingside. There are several problems with this idea. One is that central pawns are generally worth more than wing pawns. The second is that there are only so many black kingside pawns that White can take. Finally, White gets a check, but no lasting attack, so it is actually White’s king that ends up in danger. 10…cxd4! 11.Nb5 hxg4! Not just grabbing a pawn, although there is that. The text wins some squares on the kingside, while also developing Black’s h8-rook. Grabbing the central pawn with 11…Nxe5? would be a mistake, because of 12.Qe2. White’s e5pawn actually helps Black defend his center. 12.Qxg6+ Kd7 13.Qxg4 Qb6 0-1 (29)

Black has a clear advantage. His king remained on d7 for another eight moves, retreating to c8 only after he had an opportunity to play …Ra8-e8. The next example is very nice, with the kingwalk forming only part of a greater strategic plan. But the example is also instructive, as things are not as they seem (or at least as they were later represented to be).

Seirawan – Kozul Wijk aan Zee 1991

In his magisterial, iconoclastic work, Secrets of Modern Chess Strategy, John Watson includes this game in his chapter on “Exchange Sacrifices,” and is rapturous about White’s next move, which involves a deep positional exchange sacrifice. 19.Qg4!? It is difficult to know how to assess a move that embodies a beautiful idea and ultimately works, but objectively isn’t best. Historically, commentators haven’t worried about such objective assessments, which is why there are countless examples of smooth, instructive wins when, in reality, the loser had all sorts of inconvenient chances. A good example is Richter-Tarrasch, Nuremburg, 1888 – the immortal “Constriction Game,” in which Tarrasch’s two bishops famously dominated his opponent’s hapless knights. The following position was reached after White’s 28th move:

Black is creating the text book example of how two bishops can defeat two knights. He aims to take more and more space, to deny White’s knights squares. However, knights are tricky and Black has to be careful. 28…f5?! Taking away the e4-square, but now White’s e2-rook has a square. 29.Re5! Bd6?

Either 29…h4 or 29…g4 was better. 30.R5e2? White signs on as co-author of the text book example, and in so doing achieves chess immortality – at the cost of the game. After 30.Rd5! Rg6

31.g4!, White wins the battle for the e4-square, and has at least equal chances. An engine spots this possibility instantly, as would a strong player seeking counterplay for White, but the flow of the game after 30.R5e2? was so compelling that White’s resource was either overlooked or ignored for a century. For example, Fred Reinfeld, in Tarrasch’s Best Games of Chess, presents this game as an unbroken demonstration of how to win with two bishops against two knights. Tarrasch himself saw the game this way as well. Check out the position less than ten moves later and you will see why. 30…Ra8 31.Na5 Rab8 32.Nab3 h4 33.Kh1 Rg6 34.Kg1 Be6 35.Rf2 Ra8 36.Rfe2 a5 37.Nb1 a4 38.N3d2

Black broke through with 38…c4! and won convincingly in another eight moves. Returning to the position in our game, before White played 19.Qg4!?:

In this position, the engine instantly gives 19.Qe1!, assessing the position as clearly winning for White. After 19…Qxe1 20.Raxe1, White’s a1-rook is no longer on the a1-h8 diagonal, which takes the sting out of Black’s …e5-e4. 19…Qd4 20.Be2 is no better, as White threatens both 21.f6 and 21.Rad1, chasing Black’s queen away from d4. The conclusion is obvious – don’t believe everything you read, because often things are too good to be true. But let’s ignore this missed possibility and go along for the ride. 19…Qxg4 20.hxg4 e4 21.Be2 Bxa1 22.Rxa1 Rab8

White has sacrificed an entire exchange, although Black’s e-pawns are living on borrowed time, and White’s a4-knight ties down Black’s b8-rook. But does White actually have any advantage? The engine doesn’t think so, whereas at the time I thought my position almost winning.

23.Kh2 h5! 24.gxh5 Bxf5

25.Kg3 After the game, White suggested an improvement: 25.Rf1 Bh7 (?) 26.Rf6 Rfd8 27.g4 Kf8 28.Kg3 Ke7 29.Rh6 Rh8 30.g5.

Position after 30.g5 (analysis) and Black is in a complete bind. White’s king can calmly take both of Black’s e-pawns, after which g5-g6 is decisive. This beautiful variation, suggested by White after the game, is given in Secrets. But Black can resist more strongly with 25…Bd7! 26.Rf6 Bxb5! 27.cxb5 Rfc8, and suddenly Black’s pieces are far more harmonious and his passed e-pawns are a real threat.

Is humanity better off with truth or illusion? Time will tell. 25…Kg7 26.Kf4 Bh7 27.g4

Here is the kind of position White was aiming for with his exchange sacrifice. His kingwalk has given him an active king which participates in the blockade of Black’s position, but should this lead to anything concrete? Let’s see. 27…f6 28.Nc3 Kh6 29.Nd1 Rbe8 30.Nxe3 Re5 31.Rf1 Rg8 32.Bd1 f5? Black cracks under the pressure. After 32…Rge8 or 32…Rf8, the game goes on. 33.Be2 A good time trouble move, although the immediate 33.c5! was strong. 33…Rg5?

34.c5! bxc5 35.b6 Re8 36.gxf5 Rb8 37.Rb1 Rgg8 38.b7 Kg7 39.Rb6 Kf7 40.Ba6 Ke7 41.Rc6 Kd7 42.Rc8 1-0 The kingwalk in the next example is really just funny, as kingwalks sometimes (often?) are. Black could have played differently, but then his king wouldn’t have had a story to tell. Hübner – Seirawan Wijk aan Zee 1992 1.d4 d5 2.c4 dxc4 3.Nf3 c5 4.e4 cxd4 5.Qxd4 Qxd4 6.Nxd4 Nf6 7.Nb5 Kd8 8.Bxc4 a6 9.Nd4 e6 10.f3 Bd7 11.Bf4 Nc6 12.Nxc6+ Bxc6 13.Nd2 Rc8 14.Ke2 Bb4 15.Rac1

The position is equal, but Black has to be careful – not just because of who he’s playing, but also because he has to resolve the issue of his king position.

15…Ke7 16.a3 Ba5 17.Nb3 Bb6 18.Bd2 Nd7!? More interesting, although not any better, than the “natural” 18…Rhd8. 19.Bb4+ Kf6

A moment’s reflection reveals Black’s thinking. The queens have been exchanged and all of White’s minor pieces are on the queenside, so Black’s king is actually quite safe in front of his kingside pawns – at least for the moment. Black doesn’t press his luck, and soon his king reaches its destination. 20.Bc3+ Kg6 21.h4 h5 22.Bd3 Kh7

An unusual way to “castle by hand.” 23.Nd4 Nc5 24.Nxc6 ½-½

The next example is a bit uneventful – White’s kingwalk could have been more dramatic on the kingside, and the pending kingwalk to the queenside was cut short by Black’s resignation.

Seirawan – Soylu Moscow 1994

White has a very pleasant position – he has the bishop pair, Black’s queenside pawns are split, and Black has weak squares on the kingside. White has many ways to proceed. He chooses a kingwalk. 34.Kf1 a5 35.Qc4 Qd6 36.Ke2 Ne8 37.Qb5 Nc7 38.Qd7 Qa6+ A depressing exchange of queens was better. 39.Kf3

39…Qb6 40.Qe7

Now Black’s e5-pawn is attacked. Black saves it, but ruins what’s left of his position by doing so. 40…Ne6 41.Bxe6 Qxe6 42.Qxe6 fxe6

43.Ke2 1-0 White’s next kingwalk will gobble up the c5-pawn, because after 43…Kf7 44.Kd3 Ke7 45.Kc4 Kd6, White wins with 46.Bxg5! Chess made easy.. No kingwalk ever made would have been enough to save Black in the next example.

Christiansen – Seirawan Chandler 1997

Things have gone badly wrong for Black, with the only bright spot in his position being his b3-knight, which has just moved to that square, attacking White’s a1-rook. 21.Ra2? White underestimates just how well he is doing. 21.Nxf7! Kxf7 22.Qf3+ Ke8 23.Bxe6 would destroy Black’s position immediately, and this game would never have made it into this book, although it might have been seen in a different publication. 21…Rf8 22.Qc2 22.Nxf7! was still right. 22…Bxg5 Black removes the danger, although his position is still bad. 23.Bxg5 Qc7 24.h5 gxh5 25.Rxh5

25…Kd7 Black’s king takes a walk. The problem is that Black’s position isn’t bad because his king is on e8 – it’s bad because it’s bad all over. 26.Rh7 Rg8 27.f4 Kc6 28.f5 exf5 29.Qxf5 Kb5

Black’s king is actually beautiful, whereas his b3-knight only looks beautiful. This allows Black’s king to dazzle by comparison, while the rest of his position doesn’t even look good. We’ll go on for a few moves, just to prove it. 30.Rf2 Raf8 31.Bf4 31.Qf3!, followed by 32.Bh5, won right away.

31…Qe7 32.Ra2 Na8 33.e6 Nb6 34.Re2 fxe6 35.Qxe6 Qxe6 36.Bxe6 Re8 37.Re5 Rgf8 38.Rxg7 Kc6 39.Rg6 Kb5 40.Bxd5 Rxe5 41.Bc6+ Ka5 42.dxe5 1-0 (49)

With the time control reached, we can leave. Black’s kingwalk wasn’t enough to even begin to solve his positional problems. Kingwalks are not for the faint of heart. Next we see a kingwalk conducted against a renowned attacking player, as though it was the most natural thing in the world.

Shabalov – Seirawan Chandler 1997

Black has gone to a lot of trouble to achieve and maintain his d3-knight, c4pawn configuration. On top of everything else, this involves a kingwalk. 19.Bxd7+?! Kxd7 20.d5! Quiet play with 20.Qc3 gives Black a comfortable advantage. 20…exd5 21.Nd4 Kc8 Black had several other possible moves, but his top priority is to try to avoid tactical threats against his king. 22.f4?! Too slow. 22…Rb8 23.Rab1 Rb6 24.Kh1 Bc5 25.N2f3 Kb7 26.e6 Ka8 0-1 (47)

With his king ultra-safe on a8, Black has a durable advantage and went on to win in another 20 moves.

Chapter 17 Recent Examples As should have been obvious from the previous chapters, kingwalks not only have a long history, but are still an important part of the game today. Despite the pervasive influence of computers on modern chess, psychology continues to play an important role, and when players are confronted with unusual and surprising developments at the board, all-too-human errors often result. Whether kingwalks are “mere technical maneuvers” that have been internalized by today’s top players, or are creative and inspired achievements, there is no doubt that kingwalks are alive and well – more than ever. In this final chapter, we give a selection of kingwalks from recent games, and invite the reader to compare them to some of the classic kingwalks that appear earlier in the book. Ultimately, we think the maestros of the past would be pleased. Kingwalks to prepare an attack In our first example, White could well have been Tigran Petrosian himself. At least for the kingwalk part of the game.

Tomashevsky – Ponomariov Baku 2016

White has a nice edge. He has more space, and Black’s e7-bishop is somewhat “bad,” as it is on a poor diagonal. The normal approach to this position would be to advance on the queenside, but White comes up with a different idea. He decides to try to take advantage of Black’s h5-pawn by attacking on the kingside. To prepare for this, White undertakes a classic kingwalk. 22.Rcg1!? Kg7 23.Kf1 Rh8 24.Ke1 Qd7 25.Kd1 Kg8 26.f3 Bf8 27.g4 Rh7 28.Kc2 Be7 29.Rg2 Qd8 30.Kb1 0-1 (58)

A typical end position of a kingwalk of this type. If you didn’t know better, you would think that White had castled on the queenside, brought his queen’s

rook to the kingside, and started a pawn advance on that side of the board. White now has many ways to build up the pressure on the kingside, but instead he sacrificed a pawn and ultimately lost. Given the players in the next example, it might be better to term Black’s kingwalk a “kingwalk to prepare a counterattack.” Black transfers his king to the queenside not so much for safety reasons, but rather to facilitate his response to White’s potential attack on the kingside. Subtle stuff, but this is a world championship match we’re talking about.

Carlsen – Karjakin New York 2016

The position is equal. Karjakin decides to try to infuse some life into the position with a kingwalk, anticipating future activity on the king side. 32…Kf8!? 33.Qc3 Ke8 34.Rb4 g5 35.Rb2 Kd8 36.Rf2 Kc8 ½-½ (51)

Mission completed. The position is still equal, but later White advanced his g2-pawn to g4 and Black did indeed get chances for a serious advantage. Carlsen held the draw, however, and the match headed into a rapid chess tiebreak, which he won, retaining his title.

The next example is very hard to categorize. In part this is because it is from a blitz game (albeit a “serious” one), and in part because White’s kingwalks may or may not have had an actual purpose. But since White was eventually planning to break into Black’s position, there presumably were variations where the position of White’s king would have mattered. In one sense this game may be the best example in this entire book, since it features the most kingwalks. It could have been placed in the “double kingwalk” category, but really this example mainly shows how irritating and wearing kingwalks can be for an opponent consigned to passive defense.

Nakamura – Caruana St. Louis 2017 (blitz)

White has a winning material advantage, but what is the best way of converting it? The engine favors the direct approach – crashing through with f4-f5 as soon as possible. A human being is inclined to take a more subtle approach, wearing down the opponent’s resistance by maneuvering – especially in a blitz game. 38.g4 Qc8 39.Kg2 White rejects the forcing approach, in favor of what can be termed “tacking.” This sailing term isn’t wholly accurate, though, as it refers to the series of oblique maneuvers required to sail into the wind. Here the wind is at White’s back, but he still defers a breakthrough until Black’s resistance has been worn down. So perhaps “preliminary bombardment” might be a better term. 39…Kh7 40.Kf3 Qd7 41.Ke3 Bh4 42.Kd2 Be7 43.Kc2 Qc8

White has brought his king to the queenside, and now explores the possibility of invading down the a-file. 44.Ra1 Qd7 45.Ra4 Rb7 46.Rba3 Bf8 47.Ra8 Bg7 48.R3a6 Rc7 49.h4 Qe7 50.g5 h5

51.Rb8 51.Qa3! was cleaner, followed by a rook trade. 51…Qd7 52.Rbb6 Qc8 53.Kd3 d4 54.Qc2 Bf8 55.Qg2 Be7 56.Qa2 And here 56.Qe4, followed by 57.Rxc6, did the job. 56…Qd7 57.Rb8 Rb7 58.Rxb7 Qxb7 59.Rb6 Qd7 60.Rb8 Bd8 61.Ra8 Bc7 62.Ra7 Kg7 63.Qc4 Qd8 64.Qxd4 Qb8 65.Ra3 Bd8 66.Qc4 Qb7 67.Kc2

Be7 68.Rd3 Qa7 69.Kb2 Qb7 70.Qd4 Bf8 71.Qc4 Be7 72.Ra3 Qd7 73.Kc2 Bf8 74.Qd3 Qb7 75.Ra4 Be7

76.Kd1 Now White decides that his king should return home. 76…Kf8 77.Ke1 Kg7 78.Kf2 Kf8 79.Kg3 Kg7 80.Kf3 Kf8 81.Ke2 Kg7 82.Kd2 No surprise. After all, everyone knows the saying “you can’t go home again.” 82…Kf8 83.Kc3 Kg7 Black’s only hope is that White’s kingwalks lead to an accidental draw by repetition. 84.Ra1 Kf8 85.Ra2 Kg7 86.Qa6 Qd7 87.Rd2 Qc7 88.Qa8 Kh7

It took a while, but Black’s position is near collapse. Black can’t move his queen or bishop, so he is reduced to shuffling his king between g7 and h7. White’s king now stretches his legs one more time, for no particular reason other than to give Black one more opportunity to err. 89.Qe8 wins immediately, but the opportunity won’t go away. 89.Kb3 Kg7 90.Kc2 Kh7 91.Qe8 1-0 The classic kingwalk to prepare for action on one side of the board still has a role to play, although getting a position which is under sufficient control to allow time for such a maneuver is no easier now that it ever has been.

Mamedov – Shankland Batumi 2018

Black’s advantage is indisputable – White’s d3-pawn is a target and White’s kingside is weak because of his advanced g4-pawn. Before trying to exploit his advantage by opening lines on the kingside, Black conducts a classic kingwalk, bringing his king from g8 to b7. 40…Kf7!? 41.Kf1 Ke7 42.Ke2 White conducts a short kingwalk of his own, to free up his pieces by overprotecting his d3-pawn, but White’s king is by no means safe on e2. 42…Qc6 43.Qc3 Kd7 44.Qc2 Kc8 45.Kf1 A change of plan! 45…Kb7

Mission accomplished. Now Black can stir up trouble on the kingside. Note that if queens are traded, Black’s king will have a quick path to the b4square. We will follow the game a little bit longer. 46.Rg3 Rf4 47.Kg1 g5 48.Re1 Qe8 49.Kg2 h5! 50.Re4 Qc6 51.Rf3 Qd7 52.Rg3 Qh7 53.gxh5 Qxh5 54.c5 White seeks counterplay, but it’s not easy to attack Black’s king. 54…Rd4 55.b4 Rxb4 56.Rxb4 Rxb4 57.Rf3 Qf7 58.Kg1 Qe6 59.cxb6 Qxb6 0-1 (96)

Black went on to win in another 30 moves or so. In the following blitz game, Black’s kingwalk succeeds against none other than the world champion.

Carlsen – Ding Kolkata 2019 (blitz)

This position, which arose out of a Deferred Exchange variation of the Ruy Lopez, is more tense than it looks, especially in a blitz game. Both players would like to open up the position on their own terms – White by breaking with f3-f4, when he’s ready, and Black by advancing his kingside pawns. These similar plans are based on the position of White’s b2-bishop and its counterpart on b7. With these considerations in mind, Ding now muddied the waters by starting a kingwalk. 19…Kf7!?

Black’s idea is simple, at least on the surface. He pictures his king safely on c8 or b8, after which the kingside opens and his king is on the right side of the board and White’s isn’t. If the position remains unchanged until Black’s king completes its journey, Black will have created chances for an advantage. White, who knows a few things about chess, realizes all this in a flash and now tries to disrupt Black’s plans. 20.Qf2?! The main recommendation of the engine is 20.g3, but close behind is 20.Kf2!?, starting a kingwalk of his own. 20…Ke8 21.Qh4 Kd7

22.f4!? This sacrifice is what Carlsen had in mind, and it is only on his next move that White goes wrong. Interestingly, the engine disagrees strongly with the assessment on the ChessBase website that 22.Nd5 Qf7 is equal, considering that Black has a big advantage if White plays quietly. 22…Nxf4

23.Rxf4 Understandable in a blitz game, but White actually had adequate compensation after the simpler 23.g3! Ne6 24.Rf2, because Black’s f6-pawn has been weakened by Black fifteenth move (15…g6). 23…exf4 24.Ng4 g5 25.Qh6

25…Qe6?! It does seem wrong to spend more time commenting on a blitz game than the players were given to play it, but objectively 25…Qe6 is a mistake. After the human move 25…f5!?, Black has an advantage, and if the engine is to be believed, after the computer move 25…c4!, Black’s advantage is even greater (26.Bxf6 Qc5+ 27.d4 Qb4!, hitting White’s e1-rook). 26.Nxf6+?!

We will pass on the engine’s opinion one more time – after 26.h3!, preparing to take on f6 with the b2-bishop, White maintains the balance. 26…Kc8

Black has completed his kingwalk, and now everything goes according to plan. 27.Qxg5 f3!? A good move, but even better was first 27…Rd6!, driving away White’s f6knight, as 28.e5 fails to 28…Rxf6!, and 29.exf6 is impossible because White’s e5-pawn is pinned to White’s e1-rook, while 29.Qxf6 is met by the crushing 29…Qg4!, threatening mate on g2, while also uncovering an attack on White’s queen. 28.Rf1 Rg8! 29.Nxg8 Rxg8 30.Qe5 White’s last chance was 30.Qf6, but Black is still winning after exchanging queens and invading White’s second rank with his rook. 30…Rxg2+ 31.Kh1 Qg6 32.h4 Bc6

This position might be said to be the realization of the idea behind Black’s kingwalk. His king is completely safe, while White’s king is fatally exposed. 32….f2 was winning, but the move played, which makes b7 available as the ultimate haven for Black’s king, is a triumph of humanity. There was another even more incredible winning move, which combines the best of man and machine: 32…Bxe4! 33.dxe4 (33.Qxe4 Rh2+! 34.Kxh2 Qg2 mate) 33…Kb7!, completing the kingwalk in style. White is helpless against the threat of 34…Qg4, as 34.Qf4 is met by 34…Rg4. Black can be forgiven for not playing 32…Bxe4! near the end of a blitz game… 33.Qf4 Rxc2 34.Rg1 Rg2 35.Qe3 Qg3 0-1

Black’s king is safe, White’s is not. Could there be a better illustration of the continuing power of the kingwalk? Mating attacks This one is simple, but elegant. The short kingwalk is not a lost art.

Nakamura – So Las Vegas 2015 (rapid)

White has an extremely strong attack and it’s not surprising that a decisive sacrifice is available. It’s not the sacrifice, which is actually pretty obvious (although the moves leading up to it weren’t), but rather the follow up that interests us. 31.Ng6+! hxg6 32.hxg6 Ra7

33.Kf2! White’s king participates in the attack – by getting out of the way. The point of 31.Ng6+ was to open the h-file, and White’s f1-rook will go to that file with fatal effect.

33…g3+ 34.Kf3!

34.Kxg3? is weaker, because it allows 34…Bd6+. Pawns don’t matter in this position. 34…Rxg6 35.fxg6 Qxg6 36.Qxc8+ Kh7 37.Rh1 1-0 Here is an example, taken from Jonathan Speelman’s popular “Agony” column on the ChessBase website, of a kingwalk as part of a mating attack. And, not for the first or last time, the king gets help along the way from the opponent.

Samaniego – Gacosta Philippines 2016

21…e4? Black has a big advantage, and after 21…exf4 he keeps it. 21…e4? allows a pseudo-sacrifice which lets White eliminate most of his material deficit, and at the same time start an attack. 22.Nxg6! Kf7 Of course 22…hxg6 23.Qxg6 mate won’t do. 23.Nxf8 Qxf8 24.Be5? There was no reason to exchange bishops, as this just reduces White’s attacking chances against Black’s king. This can be considered the moment where White’s king hunt turns into a black kingwalk, with surprising results. 24…Bxe5 25.fxe5 Ke6 26.Qc3? After 26.d3!, White still had an edge. 26…Rd8

27.g4? With the same idea as the missed 26.d3!, but there’s a big difference. 27.g4? exposes White’s king, and Black still has three pieces that can attack it. Yes, three, not two. 27…f4 28.Re1 Rd4 29.Qb3+ Kxe5 29…c4!, with the idea of 30…Qc5, was stronger, but taking White’s e5-pawn makes sense. 30.Qxb7 f3 31.Qc7+ Kd5?! The natural 31…Qd6 was better. 32.Qd7+? White returns the favor, most likely in time trouble. After the engine recommendation 32.Kf1!, things aren’t so easy for Black. 32…Qd6 33.Qb7+ Ke5

34.Qg7+? Suicide. White forces Black’s king into the heart of his position. 34…Kf4! 35.Qf7+ Kg3 36.Kf1 Kh2

Once again we see the apparent paradox – Black’s advanced king is much safer than White’s conventionally placed king. 37.Qb3 Qg3 38.Qe3 Qg2 mate 0-1 As one might expect, kingwalks to prepare an attack at the highest level can be very dynamic and integrated into a broader attacking scheme. Such was the case in the next example, where Aronian conjured up an attack from seemingly nothing.

Aronian – Grischuk St. Louis 2018

18.Rxf7!? One unkind, engine-addicted commentator gave this sacrifice a question mark. If that’s the right way to annotate a game, why not just ban Levon Aronian and his inspired chess from tournaments altogether? Objectively, 18.Rxf7!? is unsound, in that Black can successfully defend – if he finds the right moves. And it is in that one word “if” that all the hopes for chess rest in the face of superior machine abilities. White is playing a person and that person is not aware of the engine evaluation. We will add that Aronian’s chess experience told him there was something fishy about his opponent’s position, because Black hasn’t developed his queenside, and he felt obliged to try to prove that his intuition was correct. 18…Kxf7 19.Rf1+ Bf5! This is right, because Black has to connect his rooks in order to adequately defend f8. 20.g4 g6 21.Qc1 Kg7?! Getting out of the pin is a knee-jerk, human response. Black would have been winning had he brought his e8-rook into the defense with 21…Re6!, covering

some key squares and activating his pieces. 22.gxf5 gxf5 23.Bxe4 fxe4 24.Qf4 h6! 25.Qc7+

25…Kh8?! Here the bold 25…Kg6 retained Black’s advantage 26.Bd6! Rg8+ 27.Kf2!! Simply brilliant, especially since White must have had this kingwalk in mind much earlier! White’s king is headed for e3, where it will be safe and out of the way of White’s f1-rook. 27…Rg6! 28.Be5+ Kg8 29.Ke3

A key moment.

29…Rd8? Both 29…b5 and 29…Re8 are assessed as equal by the engine. 30.Qe7! Suddenly Black is paralyzed, as his major pieces have lost their coordination. White breaks through by bringing in his only reserve unit – his h-pawn. In the meantime, Black has no counterplay against White’s king, which is completely safe on the e3-square. 30…b5 31.h4! a5 31…h5 is met by 32.Rf5. 32.h5 Rg5 33.Rf6 Rxe5 34.Rg6+ 1-0 Escaping to safety across the board The next example is short and sweet – and nothing we haven’t seen before. Some kingwalks have become “standard techniques” in the hands of today’s grandmasters.

Vachier Lagrave – Naiditsch Germany 2017

After an intense opening, White has pressure, although he is two pawns down. Naturally the engine finds defensive resources for Black, but the human response is a short kingwalk, eliminating any concerns about the safety of Black’s king. 22…Kd7!? 23.Rxg7+ Kc8 0-1 (43)

White has compensation, but he played imprecisely and Black went on to win. The next example is a simple illustration of a kingwalk across the board to escape from danger, but with a slight twist that should remind the reader that

kingwalks are rarely a panacea. Even after a king reaches its “safe haven,” problems may yet arise.

Rodshtein – Artemiev Skopje 2019

White has just played 28.h4, with the obvious, but dangerous, idea of 29.h5, opening the h-file. Because of the reduced material, White’s kingside attack should not be decisive, but Black decides to ensure his king’s safety with a kingwalk. 28…Kf8!? 29.h5 Ke7 30.Qd3 Rc7 Opening a route for his king. 31.Rh2 Kd7 32.hxg6 hxg6 33.Kb1 Kc8 34.Ka2 Kb7 35.Rh3 Ka7

Black hasn’t solved all his problems, but he’s solved one of them – his king is as far from the kingside as possible.

36.Rh2 Rd7 37.Bd1 Rb7 38.Ba4 Rd7 39.Rh3 Rb7 40.Qc2 Rb8 41.Qg2 Rc8 42.Qd2

42…Qa6 A slip which gives White an unexpected chance to complicate. 43.Qd3? With 43.Rh7! Rc7 (43…Qxc4+? 44.Bb3) 44.f5!?, White could create favorable chaos, but now the game peters out to a draw. 43…Rb8 44.Rh2 Rb7 45.Kb1 Qb6 46.Kc1 Qa6 `-`

Dubov – Giri Moscow 2019

After a very sharp opening, a complex position has arisen. Objectively White is slightly better, and the engine recommends 19.Nd2, with an edge. Instead Dubov stuns his opponent with a shocking move. 19.0-0-0!?

Is there another example in chess history where a player castled on the side where there were no friendly pawns at all? We can’t term castling a “kingwalk,” but don’t worry – White’s king is just getting started. White has unpinned his c3-knight, so he threatens 20.Nxb5, as well as 20.Qd8 mate. But Black has a logical reply that ruins White’s dream.

19…Qa5?! After 19…Qb6!, Black defends against mate and forces a queen trade, because 20.Nb5? fails to 20…Qc6+! 21.Kb1 Na6. 20.Nb5 Na6 21.Qd7+ Here is the difference – Black’s queen doesn’t cover the d7-square, as it would after 19…Qb6! 20.Nb5? Qc6+!. 21…Kf8 22.Kb1 Ba3?

This move is the real culprit. After 22…Bc5!, the outcome of the game would be very much in doubt. Dubov now embarks on an inspired kingwalk to safety. Having brought his king to b1, it now runs to the kingside. 23.Rd3! Qb4+ 24.Kc2 Qa4+ 25.Kd2 Bb4+ 26.Ke2

Now Black has to worry about his own king’s safety, as White threatens 27.Qd8+! Rxd8 28.Rxd8 mate. 26…Kg8 27.Ne5 27.Rhd1 was a calmer way to proceed, but White must have seen everything to the end at this point. The threat is now 28.Qxf7 mate. 27…Qc2+ 28.Kf3

Only six short moves ago, White’s king was on b1! 28…Rf8 29.Rhd1 h5 30.Qd4 Rh7 31.Qf4

Black’s problem is that his pieces are completely uncoordinated, while White’s army is an efficient mating machine. White’s advantage is absolutely decisive, and Dubov finishes his masterpiece in style. 31…Bc5 32.Nd4 Qa2 33.R1d2 Qd5+ 34.Ke2 Bb4 35.Ndc6 Qc5 36.Ne7+ Kh8 1-0 Here Black lost, perhaps on time, but White has a lovely mate with 37.Nxf7+! Rhxf7 38.Qh6+ Rh7 39.Qxf8 mate, or 37…Rfxf7 38.Rd8+ Rf8 39.Rxf8 mate. Escaping to safety up the board It should be comforting to the reader to see examples of kingwalks where even elite players lose their balance. The next game could easily have ended up in the “fails” category, but in the end Black’s kingwalk succeeded, after several near-death experiences for His Majesty.

Wei – Mamedyarov Zhejiang 2016

26.g5! The start of a strong attack. 26…Nxe4 27.Nxe4 dxe4 28.Bxf7+ Kg7 29.Bxe8 Rxe8 30.Nd2 This move isn’t really a mistake, but 30.Nh2!? was better, both because White’s knight strengthens his king’s defenses, and because White can follow up with 31.Rcd1 and 32.Bc1, getting his a3-bishop back in the game. 30…Bxg5 31.Rxc6 Qxc6 32.Nxe4 Bf4 33.Bb2 Nf6 34.Nxf6 Kxf6

Black’s king starts its journey. White still has an edge, but positions like this are almost always dangerous for both sides. 35.Qb3 Kg5!? 36.Qf7?! 36.Qd1 was more circumspect. 36…h5 37.a6 Kh4 Black plays to win! As is so often the case, the players must have both been short of time at this point. Kingwalks and time trouble go together like peanut butter and jam. 38.a7 Kxh3 39.Qb3+ Without this resource, White would be lost. 39…Kh4?

The wrong retreat. 39…Kg4! would have given Black’s king the option of retreating to f5. 40.Qd3? The accursed 40th move! After the computer move 40.Bc1!, White is suddenly winning, as 40…Bxc1 fails to 41.Qg3 mate (this is why 39…Kh4 was wrong).

Black therefore has make a move like 40…Rc8, allowing 41.Bxf4 exf4 42.Kh2!, and White’s advanced a7-pawn, Black’s weak pawns on f4 and g6, and the ever-present threats of checkmate all combine to give White a won position. This is hard for a human to understand, and even more difficult to see on the last move of the time control. However, it is easy to see that Black’s f4bishop is a tenacious defender. 40…e4 41.Qc3?! Qd5?! 42.Qf6+? 42.Bc1! would again be decisive, but this idea is not at all obvious. 42…g5

43.Bc1? In his wonderful Endgame Manual, Mark Dvoretsky inserts negative examples which he terms “tragicomedies.” They are the best part of the book, because they appeal to that shameful, but near universal, human trait which is best expressed in German – schadenfreude (“pleasure derived by someone from another person’s misfortune”). If you have enjoyed this book, you are guilty of it as well… White finally plays Bb2-c1, but whereas it would have won earlier, it now loses immediately. The difference? Black can avoid the exchange of bishops

by advancing his e4-pawn, so his king is completely safe and White’s king is ripped to shreds. Chess can be a very cruel game. That’s why we play it. Getting back to the game, after 43.Qg6! Qd7! 44.Rxe4 Qg4+, the game ends in a draw by perpetual check. 43…e3! 44.fxe3 Qf3 45.Qb2 Qg3+ 46.Kf1 Be5 0-1

We’ve seen this type of position before. In the next example, it was impossible not to give the entire game, in part because it would be too difficult to explain how the position after 15 moves came about, and in part because it shows that not every Petroff has to be dull. Instead we see a no-holds-barred knife fight, to the credit of both players. In the end, White’s attack is defeated by – you guessed it – a kingwalk. And this time Wei is on the winning side of the kingwalk. Rapport – Wei Wijk aan Zee 2017 1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nc4 Nxe4 5.Qe2 Qe7 6.Ne3 Be6 7.Nc3 Nf6 8.g3 Nc6 9.Bg2 Qd7 10.b4 d5 11.b5 Nd4 12.Qd3 c5 13.0-0 0-0-0 14.Re1 h5 15.h4

15…g5?! Whatever else you can say about this move, accusing Black of playing for a solid draw isn’t one of them. 16.b6!? a6 17.hxg5 Ng4 18.Ncxd5 White is on top, as he has won an important central pawn. All he has to do now is avoid getting mated. 18…h4

19.Nc7?! The way through the labyrinth was 19.Nxg4! Bxd5 20.Nf6 Bxg2 21.Nxd7 Nf3+ 22.Qxf3 Bxf3 23.Bb2 Rh5 24.Nxf8 hxg3 25.fxg3 Rxf8 26.Bf6, but it is

easy to understand White’s fear that a minotaur might devour him at any moment. With 19.Nc7?!, White, true to his style, plays for mate himself. 19…h3

20.Bxb7+? This must have been the result of a miscalculation, as after 20.Nxg4, there is nothing clear for either side. White’s counterattack looks frightening, but Black stays calm and his king goes for a walk. 20…Kxb7 21.Qxa6+ Kc6 22.Nxg4 Bxg4

Now Black has a textbook attack against White’s weakened king position. White can only keep attacking, but he doesn’t have enough pieces for it to work.

23.b7+ Kxc7 24.Qa5+ Kb8 25.Qa8+ Kc7 26.Qa5+ Kd6 27.Qb6+ Kd5

Black’s king is completely safe, and now it’s Black’s turn to attack. 28.Re3 h2+ 29.Kh1 Qc6 30.Rb1 Qxb6 31.Rxb6 Bd6 32.Bb2 Rde8 0-1

This position should be on the cover of this book. Whose king is safer? Just when Wei seemed to be getting the hang of kingwalks, we come to our next example, where he again comes out on the wrong side. To be fair to the brilliant Chinese star, his idea was sound, but mistimed by a move. In a complex position, with the board full of pieces, the slightest detail can make a big difference to the outcome. Ivanchuk – Wei Wijk aan Zee 2017

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Nf3 d5 5.cxd5 exd5 6.Bg5 h6 7.Bh4 0-0 8.e3 Bf5 9.Rc1 Nbd7 10.Qb3 Bxc3+ 11.Rxc3 c6 12.Nd2 Qa5 13.f3 Rfe8 14.Kf2 Rac8 15.Be2 c5 16.dxc5 Rxc5 17.Rhc1

17…g5 This gives White a slight pull. Black had a sharp alternative: 17…Rxe3!? 18.Kxe3?! d4+ 19.Kxd4? Rc4+! with mate next move. Of course, White does not have to be so cooperative. Unlike checkers, in chess captures are not compulsory. 17…Rxe3!? 18.Bxf6! Rxe2+ 19.Kxe2 Qa6+ 20.Kf2 Rb5 leads to a balanced position, where Black has compensation for his slight material deficit. Even 18.Kxe3?! d4+ 19.Kf2 is playable for White. 18.Bg3

18…Rxe3?! One move makes all the difference between a king hunt and a kingwalk that ends with White’s king safe in the middle of the board. Sadly, the enterprising text is just unsound. 19.Kxe3 d4+ 20.Kxd4! Re5 21.Bd3! White finds the most accurate defense. It might look like Black should mate, but in fact he only has one check, and that helps White escape (21…Rd5+ 22.Ke3). 21…Be6 22.Qxb7 Kg7 23.Ne4! Now White’s king can retreat if it has to. But White’s king is just safe on d4, period. 23…Nd5 24.Bxe5+ Nxe5 25.Rc5! 1-0

Black resigned before White could fully implement his next defensive idea, which was to capture all of Black’s pieces. A good general should not always lead his troops into battle. As we have seen, kingwalks can be startlingly simple or mind-bogglingly complicated. Our next example falls into the second category, in part because of the unusual material balance in the starting position, where “any of three results is possible” as they like to say.

Andreikin – Eljanov Wijk aan Zee 2017

35…Rd8? Black errs in an exceptionally tense and complicated position. The win was there with 35…Reh8! 36.f4 (Black threatened mate, so White creates an escape route for his king and covers h1 with his queen) 36…Qf7! (36…g4?, hemming in White’s king, fails to 37.Nd4, because White can answer 37… Rh1+ with 38.Qxh1) 37.Qxf7+ Kxf7 38.g4 (38.c7 is too slow, because after 38…g4 Black mates) 38…Ke6!, and Black’s king will go to c7, stopping White’s passed pawns. The idea of 35…Rd8 is to force White’s queen to a different square, so that after White plays f3-f4, it no longer controls the h1-square. This is all pretty difficult stuff to find in time trouble. 36.Qe6 Rdh8 37.f4 Rh1+ This sends White’s king on its way to safety, but 37…Qf7 was no longer as effective, because White can reply 38.Nd4!, and he recaptures on e6 with check. 38.Kg2 R1h2+ 39.Kf3 Rf2+ 40.Kxe3 Rhh2 41.Bd1

A wise person might ask “what on earth is going on?” The engine gives a simple reply: “the position is completely equal.” 41…gxf4+ 42.Nxf4 Qb6+ 43.Ke4 Qb1+ 44.Kd5 Qxd1+ 45.Kc5

And so it is. One side or the other will give perpetual check, and neither White’s fearsome queenside pawns nor Black’s menacing rooks can do anything about it. 45…Rxf4 46.Qe7+ Kg6 47.Qe8+ Kg7 48.Qe7+ Kh6 49.Qf8+ Kh7 50.Qf7+ Kh6 51.Qf8+ Kh7 52.Qf7+ Kh6 53.Qf8+ Kh7 54.Qf7+ ½-½ In the following game, White’s king boldly embarks on a walk up the board in order to escape attack from his opponent’s queen. White’s kingwalk both succeeded (his king made it to safety) and failed (in that in the end the game was drawn), but at the outset Black missed a beautiful opportunity to prevent

the kingwalk, leaving White’s king fatally exposed in the middle of the board.

Nyzhnyk – Shimanov St. Louis 2017

After some twists and turns, a complicated position has arisen in which neither king is completely safe. White has just played 35.Ra8, to trade a pair of rooks. Black now takes White’s b2-pawn, and White has to make a choice, likely in mutual time trouble, as is so often the case. 35…Qxb2+ 36.Kd3?!

The exclamation mark is added only as a nod to White’s courage – 36.Kd3 is objectively bad, but the refutation is by no means obvious, or even easy to explain. 36…Bf1+?

Black’s h3-bishop goes to the wrong diagonal. The winning move, believe or it or not, was 36…Kh7!, when Black’s king is safe and White’s is not. Presumably it’s a coincidence that with this move Black’s king takes the diagonal opposition with respect to White’s king. Black’s c8-rook isn’t a problem, because after 37.Rxc8 Qxa1, White’s back rank attack is too slow, as 38.Qb8 fails to 38…Bf5+!, which stops White’s kingwalk in its tracks. If necessary, Black has a kingwalk of his own to escape White’s attack. 37.Ke4 Rxa8 38.Rxa8+ Kh7 39.Qb8 Another importance difference between this position and the variation that could have arisen after 36…Kh7! 37.Rxc8 Qxa1 38.Qb8 is that now White’s queen, rather than his rook, will be checking on h8. 39…Qb1+ 40.Ke5 Qf5+ 41.Kd6 Qd5+ 42.Kc7

White’s kingwalk has succeeded, although in this particular position it is only good for a draw. 42…Kg6 43.Qh8 f6+ 44.Kb6 b4 45.Qe8+ Kh6 46.Qh8+ Kg6 47.Qe8+ ½-½ Chess has changed a lot since the olden days, but people haven’t. The urge to punish a kingwalk seems to be a universal human trait, and it’s likely we will see it as long as chess is played.

Anon – Caruana ICC 2017 (one minute)

This is a bullet game, which we join with White having 31 seconds left, while Black has 19 seconds. White has also outplayed his renowned opponent, as Black’s king is in greater danger than its white counterpart. Most of all, White has a significant advantage, in part because two rooks are worth more than three minor pieces, but mainly because White’s a4-pawn is very dangerous in any kind of endgame. Since writing this introductory note took considerably longer than it took to play the entire game, we will now sit back and enjoy some bullet chess, only pausing now and then to reflect on the fact that kingwalks have their role to play in this form of chess as well. Since White threatens 31.Rb7, Black’s king must head up the board. 30…Kg6 31.Rb6+ Checks are very tempting in bullet. 31…Kg5 There’s no turning back. 32.Qf8

It’s not fair to give anything but a blunder a question mark when the players only have a minute each for the entire game. It may be fair to point out that White’s approach to the position is wrong, He is trying to checkmate his opponent, when he should be content with exchanging queens, especially when he is ahead on time. 32.Qe8! was the right continuation, when it’s hard to see how Black can continue the game. 32…Qc7

Black threatens mate, hoping his king can escape. It can, but as pointed out in the previous note, this only matters if White chases Black’s king. Now White can win safely with 33.Qe7+, trading queens. But such is the mesmerizing power of a kingwalk that White instinctively rejects such a tepid approach to the position and keeps trying for a mate that isn’t there. 33.h4+ Kxh4 34.Rh6+ If we were going to give out any question marks, it would be for this move, since 34.Qe7+ was still winning for White (if it’s possible for anything but checkmate on the board to be “winning” in a bullet game). 34…gxh6 35.Qxh6+

35…Nh5 This leads to an amusing finish, but denies us the more thematic 35…Kg3 36.Qg5+ Kh2 37.Qh4+ Bh3

Position after 37…Bh3 (analysis) and Black’s king has completed its escape. 36.Re2 Qc1+ 37.Qxc1 Ng3+! 0-1

Ideally, kingwalks to safety involve calculation – it is clearly better to work out that your king can escape than to leave matters to chance. In practice, however, is it not always possible to see things through to the end, either because of time trouble or fatigue (our most charitable explanations) or because calculating long and complex variations is just not easy for most players (the real explanation in many cases). Fortunately the variations sometimes aren’t that difficult, and one salutary effect of this book might be to help players develop an intuition as to when kings can escape safely and when they can’t. In the following example, there is more than one road to safety, because White’s king has many squares open to it, and it only has to contend with Black’s queen.

Mazur – Keymer Balatonszárszó 2017

38.Rxe5 38.Nf6+ was faster. If 38…Bxf6, then 39.Qxd6 Rxd6 40.Re8 mates. 38… Kh8 doesn’t help either, because 39.Nxd7 wins a rook, as White’s queen is defended by his h6-bishop (White might have missed this, because backwards diagonal moves are one of the hardest things to see). That leaves 38…Qxf6, but then 39.Qxd7 wins right away, because of the mate threats on the back rank. 38.Qxd6 is a reasonable alternative to 38.Nf6+!, as it gives White a large advantage after 38…Rxd6 (38…Bxd6? 39.Re8+, mating) 39.Rxe5 Bc6, but there is not yet a clear win. 38…Qxe5 39.Qxd7

39…Bxg2+ The best try – Black goes for a perpetual. However, White’s king can escape up the board, using two instructive techniques. As White’s king moves into the center, it triangulates, forcing Black’s queen to check from different colored squares. This allows White’s king to reach the queenside safely. Once Black runs out of checks, White threatens mate, and his king escapes by running back and hiding in front of Black’s queenside pawns. 40.Kxg2 Qe2+ 41.Kg3 Qe1+ 42.Kf3 Qf1+ 43.Ke4 Qe2+ 44.Kf4 Qf2+ 45.Ke5 Qxb2+ 46.Kd5 Qg2+ 47.Kc5 Qf2+ 48.Kc6 Qf3+ 49.Kb6 Qa8

Stage one is complete. 50.Nf6+ Kh8 51.Qxf7 Qb8+ 52.Kc5 Qc8+ 53.Kb4 1-0

And now stage two is over, and with it, the game. Every strong player is now familiar with the idea of escaping an attack by moving his or her king up the board to safety. Having this technique in your toolbox allows you to see the resource long before it is needed. In the next example, we give the entire game, to give White’s winning kingwalk better context. Esipenko – Shirov Sochi 2018 1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.Nc3 e6 5.e3 Nbd7 6.Bd3 dxc4 7.Bxc4 b5 8.Bd3 Bb7 9.0-0 a6 10.e4 c5 11.d5 Qc7 12.dxe6 fxe6 13.Ne2 Bd6 14.a4

White offers a pawn, which Black unwisely takes. 14…Nxe4?! 15.Bxe4 Bxe4 16.Nc3 Bb7 17.Re1 0-0 18.Rxe6

White has regained his pawn, and is attacking Black’s d6-bishop. Shirov tries to make something of his temporary lead in development, but White’s king slips away from Black’s attack. 18…Ne5 19.Qxd6 Nxf3+ 20.gxf3 Qf7 21.Re3 b4 22.Ne4 Rad8 23.Qg3 Rd1+ 24.Kg2 b3 25.Rc3 c4 26.Nd6 Qe7 27.Nxb7

27…Qe2 Since 27…Qxb7 28.Rxc4 is hopeless (White’s c1-bishop is pinned, but is also defended by White’s c4-rook), Black tries a desperate counterattack. White’s answer is easy to see. 28.Kh3! Rf5 29.Re3 Qf1+ 30.Kh4

White’s king is completely safe. 30…Rxc1 31.Re8+ 1-0

Artemiev – Caruana Wijk aan Zee 2020

This position arose in the final game of the final round of the grueling 2020 Tata Steel Masters tournament. Caruana was on a tear, and had already locked up first place. Under pressure for most of the game, White now errs. With 44.Qe1!, he could continue to resist, although it wouldn’t have been much fun. 44.Rd1? Qc4 45.Qc3 Qa2+ 46.Nd2

46…Qc2! American grandmaster Sam Shankland, annotating this game, gives 46…Qc2 an exclamation mark, and comments: “Well calculated. White can stick his queen in the center and give a lot of checks, but there is a clear escape plan.”

Black’s escape plan might have been clear to the number two player in the world, and is also clear with the benefit of hindsight, but in reality some impressive calculation was required. 47.Qe5 Bxd2

48.Qxe6+ And here Shankland writes: “Now all Black has to do is bring his king to b2. Once it sits there, any queen check can be blocked with the bishop and a countercheck. Black wins easily.” This is true, although it isn’t trivial to calculate that Black’s king can find a path to safety. But what is really interesting, apart from the fact that in 2020 kingwalks are still very much a part of high-level chess, is how mundane Shankland considers it to be for Black to walk his king from one corner of the board to the other. We still don’t find such kingwalks to be routine. 48…Kg7 49.Qe7 Rf7 50.Qe5+ Kf8 51.Qb8+ Ke7 52.Qe5+ Kd8 53.Qb8+ Kd7 54.Qb7+ Kd6

Here we see White’s problem. He can never take Black’s f7-rook, because Black will respond with …Bf4+ and White will be mated. 55.Qb6+ Ke5 56.Qb5+ Kd4 57.Qb6+ Kc4 58.Qe6+ Kc3 59.Qe5+ Kb3 60.Qd5+ Kb2

We have seen this triangulation technique before as well. Black’s king has now reached its destination, as White is out of useful checks. 61.Qb5+ Bb4+ 62.Kg3 Qxd1 62…Rxf3+ also won. 63.Qxb4+ White resumes checking, but not for long.

63…Qb3 64.Qd2+ Kb1 65.Qe1+ Kc2 66.Qf2+ d2 0-1 After 67.Qc5+ Kb1, White has again run out of checks. Kingwalks in the opening A kingwalk can be part of opening preparation, but rarely to the extent we see in our next example. White’s idea was objectively good only for a draw, but Black had to spend so much time and effort working through the myriad of variations that in the end he faltered.

Navara – Wojtaszek Biel 2015

After a ridiculously complicated opening, the players have arrived at a balanced position. Actually it’s a bit misleading to say “after,” because the players are still in the opening – specifically, what follows was apparently all part of White’s preparation. 18.Re1 Bh4 19.Rb1 White doesn’t intend to repeat moves, but his rook is better placed on b1, should he capture Black’s c4-bishop with his b3-pawn. 19…Bg5 20.Kf2 Bh4+ 21.Kf3!? And so begins what is sure to become a modern-day classic kingwalk. To this point, White had taken less than 30 seconds per move. 21…e4+! Black, in contrast, was thinking long and hard at this point. 22.Kf4 Knight captures allow perpetual check: 22.N5xe4 Ne5+ 23.Kf4 Ng6+, with a draw.

22…g5+ 22…Rd6 and 22…Rd2 were options. 23.Kf5

The engine assessment: dead even. 23…Rhe8 24.Rhd1 Re5+ 25.Kf6

Ever forward! 25…Rg8 25…Rxc5 was a reasonable alternative. 26.bxc4 Rg6+ 27.Kxf7 Re7+ 28.Kf8

An astounding trek by White’s king, and it’s not over yet. Black still has ways to give perpetual check, but there is no mate. Presumably Black went wrong looking for what wasn’t there. 28…Rf6+? After 28…Reg7!, Black sets up a perpetual check. 29.Kg8 Rg6+ 30.Kh8!

How can this not justify another diagram? 30…Rf6 31.Rf1! With a big time advantage, White avoids the repetition after 31.Kg8. 31…Bf2 32.Rxf2! Rxf2 33.Rf1 Rxg2?

The losing move. Black had to throw in 33…Re8+. 34.Rf8+ Kc7 35.Nd5+? 35.N5xe4! was right, taking the d6-square away from Black’s king. 35…Kd6 36.Nxe7 Kxc5 37.Rf5+ Kxc4 38.Nxc6 bxc6 39.Rxg5 1-0 (48)

The excitement is over, and a difficult rook and pawn endgame has been reached. After some mutual errors, White came out on top, justifying his fantastic opening preparation. Kingwalks can be a messy business. In the next example, White tries something we’ve seen before – he answers a bishop check on b4 by moving his king. This ambitious idea succeeds, to the extent that it leads to a complicated position, but otherwise it lands in him hot water. There follows an up-and-down slugfest, where the evaluation of the position changes on almost every move. In the end, White loses, but because of something much simpler than his kingwalk. He just doesn’t manage to develop two of his pieces. Stupak – Bok Baku 2016 1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 c6 5.Bg5 h6 6.Bxf6 Qxf6 7.e3 g6 8.e4?!

This is premature, but tricky. White reasons that he might be able to exploit the weakened f6-square. 8…dxe4 9.Nxe4 Bb4+

10.Ke2? The big idea! Three cheers for White, but in this position this move doesn’t work very well. 10…Qg7 11.c5 The logical follow up, but it will come as no surprise that the tactics favor Black. Maybe because White’s king is on e2? 11…0-0 12.Qa4 Na6

13.Ke3 This may be where the fail-safe switch failed. The desirable 13.a3 is met by 13…Nxc5, because if White captures on c5 with his d4-pawn, Black has … Qxb2+ and …Qxa1. So White’s king bravely marches on. The immediate threat is 14.Bxa6. 13…b5 Stopping White’s threat of 14.Bxa6, but the immediate 13…f5 was even stronger. 14.cxb6 axb6 15.Qxc6 f5?! Now the timing isn’t as good. This is the type of move that White is hoping to provoke. After the simple 15…Rb8!, Black is just better. 16.Qxa8 Nc7 17.Nf6+! Qxf6

18.Qc6 After 18.Qa4, White is surviving – provided he plays like an engine. In real life, not so much. 18…Nd5+ 19.Ke2 Rd8 20.g3? White errs in a complicated position, and this time it should be fatal.

Interestingly, 20.a3! Bc5! 21.Rd1 was right, for reasons that will become apparent. 20…Bc5! 21.Qa4

Now 21.Rd1 fails to 21…Ba6+ 22.Ke2 Bb4+, which wouldn’t be possible if White’s pawn was on a3. 21…Nb4? Missing 21…Bxd4 22.Qxd4 Nf4+ 23.Qxf4 Qxb2+, and White’s position collapses. 22.Rd1 Now the position is equal, but there are more adventures ahead. 22…Bb7 23.Rd2?! Things really would have worked out better had White developed the rest of his pieces. 23.Bg2 was right. 23…Ra8 24.Qd1 Bf8? Now Black blinks. He should have played 24…Rxa2, not fearing 25.dxc5 Rxb2. 25.a3 e5?

25…g5! was the best chance. 26.axb4 exd4 27.Qb3+ Kh8

28.Rd1? Kingwalks are one thing, but kingwalks without developing your pieces are something else! Why not 28.Bg2!, when White would have every chance of consolidating completely? 28…Re8+? 28…Qe7+ was the right way to check. 29.Kd2 Bxf3

30.Qxf3?

White misses his last chance to develop his f1-bishop. 30.Bb5!, attacking Black’s e8-rook, would lead to something completely unexpected – a level position with equal material. 30…d3 White has kept his extra rook, but what use is it on h1? 31.Rb1? Now it’s over. 31.Kxd3 had to be played. 31…Bxb4+ 32.Kxd3 Qe6!

This powerful quiet move, which threatens 33…Qb3+, ends matters. 33.Kc2 Rc8+ 34.Qc3+ Bxc3 35.bxc3 Qc6 0-1 One reason that kingwalks early in the game (or any time in the game, for that matter) attract our attention is that they at least appear to be dangerous. Appearances can be deceiving, but not always. Sometimes a kingwalk is every bit as risky as it looks. The next game, a brilliancy by China’s leading player, illustrates this in classic fashion. It could also have been categorized as a “fail,” but that would have been unfair, as it is quite a spectacular game. Perhaps a “spectacular fail?”

Bai – Ding Zhejiang 2017 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Nf3 0-0 5.Bg5 c5 6.e3 cxd4 7.Qxd4 Nc6 8.Qd3 h6 9.Bh4 d5 10.Rd1 g5 11.Bg3 Ne4!? This initiates complex play, including a positional queen sacrifice. 12.Nd2 Nc5 13.Qc2 d4 14.Nf3 e5 15.Nxe5

15…dxc3!? A bold choice, since Black had simpler alternatives. 16.Rxd8 cxb2+ 17.Ke2?

White is not to be outdone – his king goes for a walk, hoping to take advantage of Black’s supposedly exposed pieces by avoiding exchanges. But Black’s pieces are also active. 17…Rxd8

18.Qxb2 Na4! 19.Qc2 Nc3+ 20.Kf3 We have seen other examples of a king moving to f3 in the opening, but here White’s king is swarmed by Black’s entire army. 20…Rd4! The most spectacular option, although there were others. 21.h3 h5 22.Bh2 Giving back the queen with 22.Qxc3 is no different from resigning, as Black’s attack would continue and inevitably White would lose material. 22…g4+ 23.Kg3

23…Rd2! 24.Qb3 Ne4+ 25.Kh4 Be7+ 26.Kxh5

This is not a “kingwalk up the board to safety.” Rather this is walking the plank! 26…Kg7 27.Bf4 Bf5 28.Bh6+ Kh7 29.Qxb7 Rxf2 30.Bg5

30…Rh8 31.Nxf7 Bg6+ 32.Kxg4 Ne5+ 0-1 Kingwalks in the endgame The following light-hearted example, from a blitz game, could have been categorized as a kingwalk to help deliver mate, but it certainly occurred in an endgame, so here it is. As we have seen, even the world champion can fall victim to a kingwalk – at least in blitz.

Aronian – Carlsen Paris 2016 (Blitz)

The position is equal, but this is a blitz game and Aronian is Aronian, so White decides to play his final card – he starts a kingwalk to bring his king into the attack. Nothing can really go wrong, as White always has a perpetual check with Nf8+ and Ng6+. 43.Kf4!? Rf2+ 44.Ke5 Rf3 45.Ke6 Rxe3+ 46.Kf7

White has achieved everything he could have hoped to achieve in what is still a completely equal position – he has managed to force his opponent to find the only move that holds the draw. 46…Rb3?

Even the world champion can err in the face of a kingwalk! 46…Rd3! was the move to hold, covering the d7-square. White then has nothing better than the perpetual. Carlsen must have expected that he would have an easy draw after the text, but he was rudely awakened. 47.Nf8+ Kh8 48.Ng6+ Kh7 49.Nf8+ Kh8 50.Nd7!

The winning move, as White’s d7-knight shields his king from checks along the seventh rank. Incidentally, our suspicion is that the repetition of moves must have been a way for White to gain a few seconds on the clock, rather than to tease Black into thinking he would escape with a draw. 50…Rb7 51.Rxg8+ Kh7 52.Rxg7+ Kh8 53.Kg6 a3 54.Rh7+ 1-0

After 54…Kg8 55.Nf6+ Kf8 56.Rxb7, Black loses his rook, then is mated with 57.Rf7.

Double kingwalks In games where both sides undertake kingwalks, it can happen that one succeeds and the other fails. As the next example makes clear, which is which is not always a foregone conclusion.

Chandra – So St. Louis 2016

White has played with commendable violence in an effort to avoid falling victim to his elite opponent’s superior technique. It looks as though his efforts are on the verge of success, as Black’s material advantage will count for little if his king succumbs to an attack. However, Black has a beautiful positional sacrifice which solves all his problems. 23…Rxg5+! 24.fxg5 0-0-0! This was Black’s idea. By eliminating White’s g5-bishop, he was able to castle and bring his king to safety, at least for the moment. Material is equal, but Black’s pieces are better coordinated, which gives him a significant advantage. 25.Nb3 Nxb3 26.Qxb3 Qa1+ 27.Kg2 Rd1 28.Qc2

28…Kd8?! As full of admiration as we are at So’s idea of sacrificing to castle, then walking his king back to where it came from only a few moves later, it must be said that this maneuver is poorly timed. Fischer (Johannes, not Robert) recommends 28…Be5, but the engine favors the simple 28…Kb8, which not even we would term a kingwalk. 29.Kf3 Ke8 30.Rb4 Rc1 31.Qe2 b5 32.Rb3 Rh1?

A blunder that gives White an unexpected opportunity. 33.Kg4? Giving this move two question marks seems too cruel, but here it is tempting. The first question mark is justified by the fact that 33.Rxe6+! wins on the

spot, as Black’s king is mated after 33…fxe6 34.Qxe6+ Kf8 35.Rd3!, preventing 35…Qd1+, while threatening 36.Rd8 mate. A second question mark might be added because 33.Kg4? also loses. White had other moves – for example, 33.Rxb5!? leads to a perpetual after 33… cxb5 34.Qxb5+. An additional reason to refrain from being too judgmental (apart from the obvious one – that we’ve all made worse moves) is that both players must have been short of time. 33…Kf8 Black’s king passes out of the danger zone, which is more than can be said for its white counterpart. 34.Rg3 Kg8

Black’s king is safe and White’s king isn’t. Chess can be a simple game. 35.Ree3 Qd4+ 36.Re4 f5+ 37.gxf6 Qxf6 38.f4 Qxh4+ 39.Kf3 Rh2 40.Qd3 Qh5+ 0-1 The next example, from a blitz game, features a second, symmetrical, kingwalk that didn’t happen. The opportunity was fleeting, but was no less real for being ephemeral. And, like a glorious sunrise or a rose in bloom, beauty is sometimes enhanced by its brevity.

Nakamura – So Belgium 2018 (blitz)

After winning a pawn in an opening trap that claimed three victims in the same event, White has been on top for most of the game, but he has lost control of the position, and now makes a mistake that could have been fatal. 48.Qc4? The only move was 48.Qc8+ Kg7 49.h4, although this allows 49…Bxe5!, and Black restores material equality. White prevents this by protecting his d5-bishop, but this allows Black a surprising opportunity. 48…h5? By covering g4, Black now threatens mate with 49…Qg1+ 50.Kf3 Qf2, but this gives White time for 49.h4, giving his king a safe haven on h3. Black should have played 48…Qg1+ 49.Kf3 Qe3+ 50.Kg4 h5+ 51.Kh4

Position after 51. Kh4 (analysis) Both players must have seen this, despite it being a blitz game, and thought that White’s king was safe, while White is breaking through on f7. But Black can meet White’s kingwalk to what appears to be safety with a kingwalk of his own: 51…Kg7!! 52.Bxf7 Kh6!

Position after 52…Kh6! (analysis) Now Black’s king covers the crucial g5-square, and after 53.Bxg6, 53… Qxf4+! 54.gxf4 Bf2 is mate!

Position after 54…Bf2 mate (analysis) White can only prevent this beautiful mate by giving up his queen for Black’s bishop, which is hopeless. 49.h4!

Now everything is fine for White, as Black’s attack grinds to a halt. 49…Qg1+ 50.Kh3 Bf2 51.Bg2 Qd1 52.Qc8+ Kg7 53.Qc3 Kg8 54.Qf3 Qd4 55.Qxb7

White’s king is safe in its shelter on h3, and the rest of the game doesn’t contain any surprises. 55…Qd3 56.Qf3 Qf5+ 57.Kh2 Bc5 58.Qc3 Bf2 59.Qd2 Bb6 60.d7 Bd8 61.Qd6 a5 62.Qc6 Kg7 63.Qc8 Bxh4 64.d8Q 1-0 Sometimes, as they say, the most beautiful part of a game is in the notes. Unsuccessful Kingwalks In the next example, White gains material and walks his king from e1 to a1, just where a king should be. But after a few inexact moves, Black breaks through and rips White’s king position to shreds. Sometimes you can run, but you can’t hide. Nguyen – Jones Baku 2016 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 0-0 5.Nf3 d6 6.h3 e5 7.d5 Na6 8.Be3 Nc5 9.Nd2 Nh5 10.b4 Na6 11.a3 Qe8 12.c5 f5 13.cxd6 cxd6 14.Nb5

Objectively, White is better, but often that is what is found on White’s tombstone after Black wins in a King’s Indian Defense. Here Black decides on an imaginative, and not completely correct, queen sacrifice. White counters with a kingwalk to the safety of the queenside, but things don’t turn out well. 14…f4!? 15.Nxd6 White can’t resist, although he had the option of playing 15.Bxa7!? 15…fxe3 16.Nxe8 exf2+ 17.Ke2 Ng3+ 18.Kd3 Rxe8 What’s going on? Black has only two pieces for his queen, but White’s h1rook is hanging and White’s king is in the middle of the board. Black’s real problem isn’t so much material, as that his pieces lack coordination. But it’s not easy for White to take advantage of this, because his pieces have the same problem… 19.Kc2 Bd7 20.Bc4 Ba4+ 21.Bb3 Bb5 22.Kb2 Be2 23.Qb1 Bh6 24.Nf3 Kh8

25.Ka2? White completes his kingwalk, but here we must withhold our applause, as the ruthless 25.Re1! was better. The engine evaluates the resulting position as equal. 25…Rac8 Black plays for the brilliancy prize, but 25…Nxh1 was safer. 26.Qb2? Now Black crashes through, proving that White’s kingwalk didn’t solve his king safety problems. The very annoying 26.d6!, threatening a pawn fork with 27.d7, would have distracted Black just long enough for White to have defended. 26…Nxe4! 27.Raf1 Bxf1 28.Rxf1 Nc3+ 29.Ka1

29…e4! “Black opens the a1-h8 diagonal, with decisive effect.” That comment has probably been made thousands of times in the annotations to King’s Indian games. It’s as true here as it was all the other times. 30.Qxf2 exf3 31.Qxf3 Bg7 32.d6 Re2 0-1

We keep telling you – kings just aren’t safe in the corner of the board. No matter how lovely a kingwalk to safety looks after the game, in the heat of battle the difference between success and failure can turn on a single move. Add in time shortages, and chess indeed can be a game chock full of mistakes.

Pelletier – Gelfand Zürich 2017 (Blitz)

17…Bxh2+? This was a blitz game, so nothing really needs an explanation. Black’s sacrifice is attractive, but quite unsound. Care is required by White, however. 18.Kxh2 Ng4+ 19.Kg3 19.Kg1 also works, provided White can find two strong moves: 19…Qh4 20.Bb5! Qh2+ 21.Kf1 Qh1+ 22.Ke2 Qxg2 23.Nxe4!, and White’s knight is immune from capture because of the weakness of Black’s first rank (23… Rxe4 24.Rxd8+). 19…Qe5+ 20.f4 Qh5 21.Rxd8? White had to take Black’s e4-pawn, with either his knight or his queen. Then Black’s attack peters out. 21…Qh2+ 22.Kxg4

22…h5+? 22…f5+ led to mate in a few moves, as Black’s queen can check on h5 or h6, as required. 23.Kg5? It is Black who is mated after 23.Kf5 g6+ 24.Kg5 Qg3+ 25.Kh6 Rxd8 26.Nxe4, when White’s king also earns an assist.

Position after 26.Nxe4 (analysis) 23…Qg3+ 24.Kxh5 g6+ 25.Kh6 Qh4 mate

The absence of Black’s h5-pawn makes all the difference. It’s appropriate that we end this book with a modern fail. Kingwalks are always exciting, certainly provocative, can be beautiful, and often work. But we would be remiss not to follow the advice of legal counsel to again remind the reader that checkmate ends the game, and a king that wanders too far away from its army can indeed be checkmated.

Harikrishna – Adhiban Wijk aan Zee 2017

34.h4! The best chance to make something happen in a level position. 34…gxh4 35.Nxh4+ Kg5 Black’s king goes for a walk, as 35…Kf7? 36.Nf5 gives White the advantage. 36.Nf3+ Kxg4 Once again, Black’s king has to advance, as 36…Kg6 allows either 37.g5 or 37.Nf4+. 37.Rh1

Black’s king is not yet in any actual danger, but it must not have seemed that way to the king himself. He has led a charge against the enemy, winning a pawn in the process, but when he looked back, no one had followed him… So it is time to withdraw. 37…Kf5 38.Rh5+ Ke6 38…Kg6? 39.Nf4+ is much worse. 39.Rh6

39…Kf5? An invitation to White to repeat moves, but 39…Kf7! was the way to keep the balance.

40.Nh4+! Kg5 After 40…Ke4 41.Rc3, Black’s king is stuck, and 40…Ke6 41.Nf4+ Kd6 42.Rxf6+ Kc7 43.Nxd5+ isn’t any better. 41.Rg6+ Kxh4 42.Rh1+ Bh3 43.Rg3 1-0

Black is mated.

Bibliography Books Chernev, Irving, Combinations: The Heart of Chess (Dover 1960) Harper, Bruce & Seirawan, Yasser, Chess on the Edge, Volumes 1-3 (Chess’n Math Association 2008) Kasparov, Garry, My Great Predecessors, Volumes 1-5 (Everyman 20032006) Mednis, Edmar, The King in the Middlegame (Chess Enterprises 1999) Tolkien, J.R.R., The Hobbit (George Allen & Unwin 1937) Tolkien, J.R.R., Lord of the Rings (George Allen & Unwin 1954) Databases & Software ChessBase 12 MegaBase Websites Chessbase.com Chessgames.com Engine Stockfish 10