The Longest Game: The Five Kasparov/Karpov Matches for the World Chess Championship

2018

Introduction

In chess, a match is the purest trial of strength. There are no third parties, and the amount of white and black games is always equal. Therefore, it is not surprising that chess has a long tradition of matches for the highest crown. In the beginning, there wasn’t an official world title at stake. In the first half of the 19th century, the Frenchman Louis-Charles Mahé de la Bourdonnais and the Irishman Alexander McDonnell fought each other in a contest held over six matches at the Westminster Chess Club in London; they were considered the strongest players of their time. All these matches were played at short intervals in 1834. De la Bourdonnais won four of them, McDonnell one, and the final match was never finished. In total, they played 85 games against each other.

It was only in 1886 that the first official World Championship match was played. Wilhelm Steinitz defeated Johannes Zukertort in a 24-game contest that was held in New York, Saint Louis and New Orleans. This 24-game format was to become the standard much later, after the Second World War. At the end of the 19th century and in the first half of the 20th century, a different format was often chosen. There was no coordinating organization like FIDE back then, which meant that the world champions and their challengers had a free hand to arrange their matches and choose the format.

Steinitz-Lasker in 1894 was the first match where a player had to win 10 games to gain the title. It was precisely this format that Bobby Fischer wanted to use roughly 80 years later in his match against Anatoly Karpov. Fischer argued that this would prevent the eventuality that one of the players would sustain his lead in the match by playing for draws. The obvious drawback of Fischer’s proposal was that matches could be drawn out for many months. In the days of Steinitz and Lasker, this danger didn’t exist. Emanuel Lasker defeated one rival after the other in less than 20 games, in matches where the victor had to win 8 or 10 games. Remarkably, the Capablanca-Alekhine match was scheduled for 6 won games – and it was exactly this match that took much longer than Lasker’s matches. This format was employed three times in the 1970s and 1980s, as a direct consequence of Fischer’s proposal. In fact, FIDE was responsive to the argument that a player who was leading in the match would start playing too cautiously. World championship matches can have so much impact that they sometimes mark the end of an era. José Raul Capablanca took Lasker’s place, and in turn he was dethroned by Alexander Alekhine; three chess Caesars, each with their own era. In recent times, the same has happened: Magnus Carlsen put an end to the Anand era. But things can go differently, too. Fischer managed to put an end to the era of Boris Spassky, but at the same time he also put an end to his own era, since he stopped playing after winning the title. The case of Mikhail Botvinnik was more complicated. He remained in the saddle for fifteen years, playing six matches with four different opponents. Tigran Petrosian dethroned him for good, but David Bronstein and Mikhail Tal didn’t manage this. And neither did Vasily Smyslov, actually, but I hesitate to mention his name in this context.

Botvinnik and Smyslov played three World Championship matches in the 1950s. The first ended drawn (12-12), the next was won by Smyslov, and in the third match, Botvinnik recaptured his title. In fact, we could say that there were two chess Caesars ruling in those years. They played 69 games for the highest honour in four years; Smyslov won 18 of them and Botvinnik 17. The other 34 games ended in draws. It didn’t seem probable that such a protracted titanic struggle would ever be repeated in chess history. And yet, this is exactly what happened three decades later.

The road to the top

Neither Anatoly Karpov nor Garry Kasparov experienced any kind of setbacks on their road to the top. Both of them succeeded in their first attempt to cover the long road through the Interzonal tournament and the Candidates Matches. Before, only Tal had shown such a meteoric rise. Petrosian and Spassky had risen to the top level slowly. Fischer belonged to the world’s top at a very young age, but he only became Champion ten years after his appearance in the Candidates tournament at Curaçao. Karpov was 20 when he gained access to the world elite. In 1971, the year when Fischer celebrated his great triumphs in the Candidates Matches, Karpov won his first top tournament. In the Alekhine Memorial in Moscow, he came shared first with Leonid Stein, ahead of four former World Champions: Smyslov, Petrosian, Tal and Spassky. At the time of his World Championship match with Spassky, Fischer must have sensed that Karpov could be his challenger in the next cycle.

Karpov made no mistakes in the Interzonal tournament in Leningrad 1973. He raged through the event with an iron hand, winning when necessary and making draws if there was nothing more to be gained. In the Candidates Matches, Karpov had one difficult moment: he lost the first game with white against Spassky. But he recovered, and eventually won the match convincingly. Now the Final match against Kortchnoi was scheduled. Karpov said later that at that point, he was already gearing up for the World Championship match against Fischer; Kortchnoi would be an easy hurdle to take. However, the match against Fischer never took place, and this was a huge disappointment for Karpov. He had wanted to learn from Fischer, just as Kasparov would later learn valuable lessons from him. For seven years, Karpov was condemned to playing matches with Kortchnoi. After he had disposed of the latter once and for all in 1981, it was clear that he would be facing a younger challenger in the next cycle.

For a short time, I hoped to be that challenger myself. I had won great tournaments ahead of Karpov and Kortchnoi, and had reached second place on the Elo rating list, ahead of Kasparov and Kortchnoi. But in 1982, many things went wrong for me. Tortured by sleeping problems, I failed miserably at the Interzonal tournament in Las Palmas. On the other hand, 1982 was the year of Kasparov’s breakthrough. He won the traditional top tournament in Bugojno overwhelmingly. The Interzonal tournament in Moscow went less smoothly for him. FIDE had determined that there would be three Interzonal tournaments with two qualifying places each, instead of two tournaments with three places each. This put extra pressure on the favourites. Kasparov started with 5½ points out of 8 games. In the 9th round, he was paired with black against Ulf Andersson, who had the same score. It transpired that 1982 was an excellent year for the Swedish strategist as well. He managed to outplay Kasparov completely in their game. The latter saved himself by a draw offer, inspired by desperation. In itself, there was no reason at all for Andersson to even consider accepting this offer. He was running no risk whatsoever, and he had different ways to liquidate into a technically winning endgame. However, the tension must have been too much for him, and he accepted the offer. After that, Kasparov made no mistakes in the concluding phase of the tournament.

In hindsight, Kasparov didn’t seem to experience any difficulties in the Candidates matches. Yet, there were several awkward moments that had nothing to do with his chess-technical qualities. Due to a conflict between FIDE and the Soviet chess federation, Kasparov ran the risk of losing his match with Kortchnoi by forfeit. FIDE chairman Florencio Campomanes had allotted the match to Pasadena, which was the place where Fischer was living as a recluse at the time. The Soviets insisted that the match be played in Rotterdam. Kasparov didn’t show up, and initially Kortchnoi was declared the winner. Eventually, the match did take place – in London, several months later. It turned out not to be a walkover for Kasparov. He lost the first game with white. There are parallels with Karpov’s opening loss against Spassky, but the difference was that Kasparov didn’t manage to recover as fast as Karpov had. He failed to get the match in his grip, until in the 6th game Kortchnoi grossly overplayed his hand in a superior position. Now, the deadlock was overcome, and Kasparov convincingly took control and won the match.

On the January 1984 rating list, Kasparov was leading with 2710 Elo, and Karpov had 10 points less. In this Orwell and Murakami year, these two top players started a series of matches that would grow to be even more epic than the struggle between Botvinnik and Smyslov. In a six-year time span, they played five matches, totalling 144 games – almost doubling that earlier clash of titans (and also much more than the total number of games between De la Bourdonnais and McDonnell). And here also, the score remained almost equal: 21 victories for Kasparov, 19 for Karpov. Before the fifth and last match, in 1990, Kasparov announced that he would destroy his rival; he wanted to end the Karpov era once and for all. He didn’t succeed – in the end he only managed to win by a one-point margin.

Curiously, it was Nigel Short who ended the Karpov era. In 1992, the Englishman defeated the Russian in the semifinal of the Candidates matches. Thus, Karpov was eliminated and couldn’t play a sixth match against Kasparov. But he could still be regarded as the second player in the world. We can say that the absolute hegemony of the two K’s lasted more than a decade; they were the two strongest players in the world from 1983 to 1994. One year later, this second position was taken over by Viswanathan Anand.

Botvinnik’s comment on the eve of the first Karpov-Kasparov match is interesting. In an interview with Alexander Münninghoff for New In Chess, he said the following: ‘The match between Karpov and Kasparov will distinguish itself from the usual title fights. I think that, from a creative point of view, this will be the third top match in this century. The first was Alekhine-Capablanca in 1927. The second was the two matches between Tal and me in 1960/61. And Karpov-Kasparov will be the third high point – since it will not actually be about the question who is the strongest. This match will be creative especially because it will be a clash between two fundamentally different approaches to chess.’ Next, Botvinnik explains that Karpov is the practician, and Kasparov the researcher. It is curious that he doesn’t mention the greatest match of the previous century: Spassky-Fischer. However, he wasn’t talking about aspects like the Cold War on the chessboard, or the lonely warrior fighting against an entire realm. Botvinnik was only talking about the clash of styles: the strategist versus the attacker, the practical player versus the more scientifically oriented player. And here we see something curious: he – Botvinnik – was the only strategist who was a researcher at the same time. Alekhine and Kasparov were attackers and researchers, while Capablanca and Karpov were strategists and practical players.

Naturally, Botvinnik could not have foreseen that four more matches would follow. And in fact, the first match wasn’t the most interesting one. The first part of this match was rather one-sided – Karpov was vastly superior. Kasparov was receiving some instructive chess lessons from Karpov. This is how many experts summarized the events, and Karpov agreed. The second part of the match was characterized by a huge number of bloodless draws. For Karpov, this was the third match that was played according to the Capablanca-Alekhine format, i.e. for 6 won games. In Baguio City 1978, he had needed 32 games to defeat Kortchnoi – two less than Alekhine in his match with Capablanca. In the second match, in Merano 1981, Karpov had followed Lasker’s example; he had finished the job in under 20 games. The general expectation was that the first match between Karpov and Kasparov in Moscow would be a long one. But when Karpov was already leading 4-0 after nine games, the end seemed to be near. However, at that moment, an effect occurred that Fischer had pointed out: playing with the aim of consolidating a lead will not bring you success. Nevertheless, this was just what Karpov did, and the result was that after five months – with 48 games played – the match was stopped with the score being 5-3 in Karpov’s favour. The last two games had been won by Kasparov, but even in those, there was no question of a clash between styles; Karpov was just exhausted, and he played without his customary accuracy. FIDE decided to avoid the disaster scenario of endless matches in the future, and the next four matches were traditionally played over 24 games.

The second match in Moscow, and the third in London and Leningrad, did follow Botvinnik’s criteria. The number of sharp struggles, both in a strategic and a combinatory sense, was enormous. In an interview before the second match, Spassky said that Karpov should actually be paying a thousand dollars for every post-mortem with Kasparov. This was clearly a reaction to the general notion that Karpov was supposed to be the tutor. By the way, Spassky certainly didn’t mean that he considered Kasparov to be the better player. On the contrary, he thought that Karpov was stronger. But his words followed naturally from Botvinnik’s theory: Karpov, the strategist and practical player, could learn a lot from the dynamic style and the systematic preparation of his 12-year-younger rival.

The fourth match, in Seville 1987, was somewhat disappointing from a creative point of view. A lot of energy had been demanded from the two giants: four long matches in as many years. It was actually miraculous that the match could take place at all, due to an administrative aberration by the FIDE board. In hindsight, we can be grateful for this – with this match, the epic clash of titans was stretched out even further.

The fifth and final match in New York and Lyon in 1990 was probably the best of them all. Even though Karpov was already approaching forty, he had not lost much of his strength, while Kasparov was at the height of his powers. Especially the first half of the match in New York showed an explosion of new ideas. This fifth match was much more interesting than Kasparov’s later matches against Anand and Kramnik, and it also outshines Carlsen’s (shorter) World Championship matches against Anand and Sergey Karjakin in quality.

CHAPTER 1

Moscow 1984/85

Chief arbiter Svetozar Gligoric inspects Karpov’s chair in the Hall of Columns.

In the Soviet Union, there was a tradition to hold World Championship matches in the Estrada Theatre. This venue is situated opposite the Cathedral of Christ the Saviour, the colossal edifice which was destroyed in the Stalin era but was carefully rebuilt after the fall of communism, on the other bank of the Moskva River.

Here, Botvinnik reconquered his title against Tal in 1961, and eight years later, the Estrada Theatre was the venue where Spassky dethroned Petrosian. Its concert hall is one of the largest in the world, with more than 1,700 seats. Many of those seats used to be taken up by a knowledgeable audience that regularly gave vent to their emotions; the hall would be buzzing, and suppressed cries could be heard when apparent mistakes were made. On the stage, a solemn kind of pantomime was shown. In the middle were the players, sitting motionlessly or pacing up and down the stage. In the back, on the right, was the arbiters’ table. They too would often be watching the battle scene motionlessly; now and then, they would indulge in a muffled conversation. In the back, on the left, hung the wooden demonstration board. The wooden pieces were wielded by hand. Grooves had been carved out at the bottom of the ranks. The board boys were very skilled at moving the pieces, using hooks to make the large wooden figurines glide to their destination squares. The demonstration boards were large, but they still seemed tiny due to the high ceiling.

The first match between Karpov and Kasparov was played at an even more prestigious place: the Hall of Columns in the House of the Unions. This underlined the prominent position of Karpov – and of chess in general – in the former Soviet Union. The Hall of Columns, which features 28 pillars of wood and marble in Corinthian style, is slightly smaller than the concert hall of the Estrada Theatre. There are over 1,000 seats. Here, Pushkin, Lermontov and Tolstoy visited the balls, and Chaikovsky and Rakhmaninov gave concerts. But there was also a dark side to this venue. It was used not only for cultural events, but for political ones as well. Stalin’s infamous show trials had been held here. Later, Stalin himself was laid in state in the Hall of Columns, as were all the leaders and a number of dignitaries of the former Soviet Union.

Not everybody was happy with this eminent playing location. Alexander Nikitin, who had been Kasparov’s coach for nine years already, turned out to be vehemently opposed to it. In his book, Mit Kasparov zum Schachgipfel, he explained that, in his opinion, the Hall of Columns was too pompous. The implication was that Kasparov would feel less at ease in such an environment than Karpov did. This view was probably justified. Karpov was used to moving in higher circles, while the 21-year-old Kasparov was a newcomer to this world. There was yet another problem with the location: the interest in the match was so overwhelming that the hall got filled up very quickly. Every day, standing on the street in front of the building, there were throngs of people who had not managed to secure a ticket.

Nikitin also stressed that Kasparov had been able to engage not more than a small group of seconds. There were many top players who didn’t want to risk falling into disfavour with the Soviet authorities by working for Karpov’s opponent. Still, Kasparov had a tightly-knit team that concentrated on preparations five months before the match: besides Nikitin, it consisted of Alexander Shakarov, Evgeny Vladimirov and Georgy Timoshenko. Shortly before the match, they were joined by the strongest player of the team: Josif Dorfman.

A foreign grandmaster was also involved in the preparation: Andras Adorjan. This was an unusual step by Kasparov. The World Championship matches in the Soviet Union had always been entirely internal affairs. Interestingly, Nikitin was not impressed by the Hungarian grandmaster. ‘He was not a great help to us’, he wrote. Nevertheless, Adorjan was known for his great theoretical knowledge, and for having original ideas in the opening. Kasparov himself had a high opinion of Adorjan’s ideas. He thought that Karpov would start opening with 1.♘f3 instead of 1.d4 at a certain point, out of fear for Adorjan’s preparation in the Grünfeld Indian.

Kasparov could also count on the support of Botvinnik, who had been his mentor since his youth. The former World Champion was not present during the analysis sessions, but Kasparov could always telephone him for advice after the game. Another former World Champion had also offered his services: Tigran Petrosian. His relations with Karpov were downright bad. However, Petrosian’s assistance didn’t last long. He was already seriously ill in the preparation phase of the match, and died a month before its start.

Karpov had an impressive team of seconds. His official seconds were Yuri Balashov and Igor Zaitsev. He could no longer count on the support of Mikhail Tal, who had been Karpov’s second during the matches in Baguio and Merano. Tal had withdrawn for this match. This was doubtlessly connected with his sympathy for Kasparov; both had a highly dynamic playing style, and it is no wonder that Tal would later collaborate with Kasparov. However, Karpov didn’t need Tal. He could also count on the support of Efim Geller, Lev Polugaevsky, Sergei Makarychev, Rafael Vaganian, Tamas Georgadze, Mikhail Podgaets, Adrian Mikhalchishin and Elizbar Ubilava. Some of these helpers were registered as journalists. Mikhalchishin, for example, was officially registered as a correspondent of a Lviv newspaper. In the Soviet Union, it was common usage to use journalism as a cover in this way.

Curiously, real Soviet journalists thought that top players elsewhere in the world made the same arrangements. During my match with Valery Salov in St. John 1988, one of them wrote that, apart from Andersson and Hort, I also had the ‘so-called journalists Ligterink and Ree’ in my seconds’ team. In reality, these two were writing extensive daily reports for national Dutch newspapers. I only spoke to them after the games to give comments.

Mikhalchishin once recalled that the unofficial seconds drove with Karpov in the car to the playing hall, but did not get out at the same time as he did; they would ride around the block one more time, so that nobody could see them. Curious manoeuvrings!

It probably didn’t give Kasparov any sleepless nights that Mikhalchishin and Ubilava helped his opponent. Geller and Polugaevsky were the theoretical pillars Karpov was relying on. Karpov once told me that his cooperation with Geller had been very useful; for him it had opened new worlds in opening study. Polugaevsky was also theoretically excellently-versed, but he often called in sick, so he was of less use to Karpov. Probably, Polugaevsky was worrying about the future. He was a bit fearful. What if Kasparov won? What would his further career look like then?

The match was played according to the rules of the day: 40 moves in 2½ hours; when the playing time of 5 hours was completed, the game would be adjourned. This was a ritual that illustrated the romanticism of chess. The arbiter would come up with an envelope, in which the player whose turn it was had to put his scoresheet after writing down the move he intended to play. That was the ‘sealed move’. The opponent’s scoresheet was also enclosed in the envelope. In most tournaments, there was an evening session of 16 moves per hour and the game would be resumed on a free day. If, after another six hours of play on that day – so, after 104 more moves – the game still wasn’t over, the resumption had to take place on the next free day. However, this occurred very rarely. These days, we regularly see games of more than 100 moves. This is because you can bleed drawn positions completely dry when your opponent is chased by the clock. If your opponent repeatedly gets another hour’s thinking time, playing on doesn’t make much sense.

In World Championship matches, there were no evening sessions. The game would be resumed on the next day, and could be adjourned again after 88 moves. In fact this happened only once in all five matches between Karpov and Kasparov. So, it was quite common for a game to spread over two days. This was one of the reasons why World Championship matches took much longer in those days than they do today. And then there was another thing: the players had a right to take time-outs. In 24-game matches, both sides were allowed to claim three time-outs without a doctor’s certificate. Now that the match was to be decided after six won games, they were entitled to three time-outs during the first 24 games, and then one time-out during every following eight games. The game was adjourned until the next playing day (i.e., not until a special day reserved for adjourned games), and so every time-out resulted in a two-day delay. You can imagine the horror of present-day organizers being stuck with unsold entrance tickets while the match keeps expanding in time. In the early 1990s, the time-outs were abolished. But they were still in effect during all five matches between Karpov and Kasparov.

Notably, Karpov wasn’t regarded as the clear favourite before the match. In his short career, Kasparov had already achieved such impressive results that he was seen as a worthy opponent, in spite of his lack of experience in match play.

More than 300 journalists from inside the Soviet Union and abroad were accredited. A press conference was held on the opening day. ‘Do you think that there will be other players who will come into the struggle for the highest crown in the foreseeable future?’ Karpov was asked. ‘Not any time soon’, the World Champion replied, and he continued: ‘I certainly do not exclude the possibility that we will face each other again, the way Botvinnik and Smyslov continued their man-to-man fight for years on end.’ Initially, it didn’t look as if this prediction would come true. Something happened that nobody had foreseen: Karpov quickly built up a huge lead.

The first game was played on 9 September. As had been expected, Karpov opened with the e-pawn, and Kasparov didn’t have any trouble staying afloat against the Keres Attack in the Sicilian. The second game was very sharp.

Queen’s Indian Defence

Garry Kasparov

2715

Anatoly Karpov

2705

Moscow Wch m 1984 (2)

1.d4 ♘f6 2.c4 e6 3.♘f3 b6 4.g3 ♗b7 5.♗g2 ♗e7 6.0-0 0-0 7.d5

This move was all the rage at the time. Of course, Karpov was prepared for it, since Kasparov had played the pawn sacrifice before.

7…exd5

8.♘h4

The critical move, which had been introduced into practice by Polugaevsky in 1980. Before that, 8.♘d4 had been the usual move. However, in that case, Black can return the pawn by 8…♘c6 (Kortchnoi’s 8…♗c6 is also possible, intending 9.♘xc6 dxc6) 9.cxd5 ♘xd4 10.♕xd4 c5 11.♕d2 d6, with a playable Benoni position.

8…c6 9.cxd5 ♘xd5 10.♘f5 ♘c7

This is how Karpov always played; it is also Black’s most reliable method of defence. Black prepares the push of his d-pawn, while he may be able to put the knight on e6. After Portisch’s 10…♗f6 or Padevsky’s 10…♘f6, White has better chances to get an opening advantage.

11.♘c3 d5 12.e4 ♗f6 13.♗f4

Against Marjanovic in La Valetta 1980, Kasparov had played 13.exd5 and won convincingly. Later, however, a reliable method of defence had been found for Black. It is better for White to keep the tension in the centre.

13…♗c8!

A strange move at first sight. Instead of further developing his queenside, Black withdraws his bishop. However, this retreat is entirely logical. Black wants to exchange White’s most important attacking piece: the knight on f5. In Kasparov on Modern Chess, part two, Kasparov calls this move a novelty. But it had been played one year earlier in the Lloyds Bank Open, in the game Rogers-Levitt. It is possible that Karpov knew this game. Of course there were no databases in those days, but there was a series of books titled Tournament Chess, in which all the games of all the important tournaments were given without comments. The Lloyds Bank games were in there.

14.g4

‘Not a particularly pretty move’, Jon Speelman remarked in the match book Moscow Marathon that he wrote with Jon Tisdall. However, it is understandable that White wants to support his knight, to keep the pressure on the black centre. The alternative was 14.♕a4, to increase the pressure on the enemy queenside. An interesting variation is 14…♘ca6 15.♖ad1 ♗xf5 16.exf5 ♘c5 17.♕c2 a5 18.♘xd5 cxd5 19.♗xd5 ♘bd7 20.♗xa8 ♕xa8 21.♖fe1, and chances are roughly equal.

Less good is Rogers’ choice of 14.e5. Now, Black can play his bishop to e7 or g5, after which White has insufficient compensation for the pawn.

14…♘ba6 15.♖c1 ♗d7

Black hesitates, allowing White to take the initiative. A few weeks after this game, in Tilburg, Tukmakov played 15…♗xf5 16.gxf5 ♗g5 against Sosonko. Here, White has some trouble maintaining the balance.

16.♕d2

Now, Kasparov was very satisfied with his position.

16…♘c5

17.e5

Kasparov provides this thrust with a ‘?’. However, it is fully in accordance with Kasparov’s sharp playing insights in his younger years. True, objectively 17.♗xc7 is better, as White wins the pawn back with advantage. But we can imagine Kasparov’s reluctance to trade his strong attacking bishop for a knight. After 17…♕xc7 18.exd5 ♖ad8 (Speelman’s recommendation 18…a5 is dubious, in view of 19.g5. Now, Black has to go for 19…♗d8, but it doesn’t look convincing) 19.b4 ♗xf5 20.gxf5 ♘a6, there is a critical moment. At the board, Kasparov didn’t see how to continue. Later, he found the powerful move 21.♕e2!, with the point 21…♘xb4 22.♕c4, and Black is in trouble. He can hardly prevent the creation of a strong passed pawn on c6 for White. Perhaps then, 22…♗g5 is Black’s best chance. The game could continue 23.d6 (after 23.♖cd1 ♕f4, Black manages to exchange the queens) 23…♕xd6 24.♘e4 ♕h6 25.♘xg5 c5 26.♘e4 ♖d4, and this position is not easy to win for White.

17…♗e7 18.♘xe7+

Afterwards, Kasparov was dissatisfied with this trade, because it eases Black’s defensive task. Objectively, it is probably White’s best option. After both 18.♘e2 a5 19.♘eg3 ♖e8 20.♘h5 ♗xf5 21.gxf5 g6 and 18.♗g3 a5 19.f4 ♗xf5 20.gxf5 f6, Black has the advantage.

18…♕xe7 19.♗g5 ♕e6 20.h3

This small defensive move was also criticized by Kasparov. However, it is again White’s best option. After 20.b4 h6! 21.♗e3 ♘5a6 22.b5 ♘c5, White has insufficient compensation for the pawn.

20…♕g6 21.f4

Again criticized by Kasparov, and again, undeservedly so. After the alternative 21.♗e7 ♖fe8 22.♗d6 ♘7e6 23.f4 ♕h6 24.♖cd1 ♕h4!, White’s offensive is brought to a standstill. The white bishop is a bit misplaced on d6.

21…f6!

The right reaction. Black has to fight for the initiative in the sharpest possible way. On 21…♕d3, White would have withdrawn his queen from the exchange by 22.♕f2.

22.exf6 gxf6 23.♗h4 f5

Again, the sharpest move. After 23…♕h6 24.♗f2 ♖ad8 25.♗xc5 bxc5 26.b3, White would win the pawn on c5, with approximately equal chances.

24.b4

Dorfman recommended 24.gxf5 ♗xf5 25.♖f3 here, to play along the open g-file. White gets enough counterplay in that case: 25…♕h6 26.♖g3+ ♗g6 27.♗g5 ♕g7 28.b4 ♘5e6 29.♘e2 c5 30.bxc5 bxc5 31.♘c3, winning back the pawn. The text move is still possible, but White will have to play very inventively in the following phase.

24…fxg4!

The point of the previous move. After 24…♘5e6 25.♖f3, White would have gotten his way.

25.hxg4?

But this is way off the mark. The text move amounts to an exchange sacrifice for which White gets insufficient compensation. Better was 25.bxc5 gxh3 26.♖f2. Now, the play becomes very sharp. The main line runs as follows: 26…♖f7 (26…hxg2 27.♖xg2 ♗g4 28.♘e2! is also playable for White) 27.♔h2 ♕h5 28.♖g1! ♖g7 29.♗xd5+! ♘xd5 30.♖xg7+ ♔xg7 31.♕d4+ ♔g8 32.♖f1 ♗g4 33.f5!, with a dynamic equilibrium.

25…♘d3 26.♖f3

Also insufficient was 26.f5 ♕xg4 27.♕xd3 ♕xh4 28.b5, and now 28…c5! is sharpest. After 29.♘xd5 ♗xb5 30.♕b3 ♘xd5 31.♗xd5+ ♔g7 32.f6+ ♔h8 33.♗xa8 ♗xf1 34.♖xf1 ♖xa8 35.f7 ♖f8, Black eventually reaches a technically won endgame.

26…♘xc1 27.f5 ♕g7 28.♕xc1 ♖ae8 29.♕d2

29…d4

The most direct continuation, which, however, gives White a chance to save himself. Black could also have opted for cautious manoeuvring with 29…♗c8 30.♗f2 ♖f7, when White’s initiative is toothless. Black can, if necessary, transfer his bishop to a6.

30.♘e2

The players have entered the time-trouble phase, and Kasparov keeps playing as sharply as possible, spurning the push 30.f6!. After 30…dxc3 31.fxg7 cxd2 32.gxf8♕+ ♖xf8 33.♖d3, White has good chances to hold the endgame.

30…♘d5

Understandably, Black wants to centralize his knight, but this will be at the cost of an important pawn. 30…c5 was winning. It is quite possible that Karpov was worrying about the continuation 31.bxc5 bxc5 32.♘f4. Indeed, this looks menacing for Black. However, he has a winning combination here: 32…♗c6! 33.♘h5 ♗xf3 34.♘xg7 ♖e2!, and White is helpless.

31.♘xd4 ♔h8

Kasparov called this ‘unnecessary prophylaxis’. This is one of many examples of wrong usage of Aron Nimzowitsch’s terminology. The king move only serves to create a threat: to take on g4. However, since White wants to push his g-pawn anyway, the move is rather pointless. Stronger was the direct 31…♖e4!, and Black keeps his advantage.

32.g5 ♖e4 33.♗f2 ♕e5

34.♖g3

The time scramble is well under way, and Kasparov increases the pressure. 34.g6 was too sharp, in view of 34…♖g4! 35.♘b3 ♖xf5 36.♗d4 ♖xd4 37.♘xd4 ♖f4!, with advantage to Black. But 34.♗g3 ♕xd4+ 35.♕xd4+ ♖xd4 36.♗e5+ ♔g8 37.♗xd4 ♗xf5 38.a3 was possible, when Black’s extra pawn has no value.

34…♖f4 35.f6 ♗e8

An unnecessary precautionary measure, after which the initiative passes to White. Black didn’t have to fear the march of the g-pawn. On the contrary, he could have become active on the king’s wing with 35…h6. The game could continue: 36.♗xd5 (the direct 36.g6 is no good, on account of 36…♖8xf6) 36…cxd5 37.♘f3 ♕e4 38.♗d4 ♖g4 39.♖xg4 ♕xg4+ 40.♔f2, and White has sufficient compensation for the exchange.

36.b5!

Again, forcing Black to play accurately.

36…c5?

Now, White gets the c6-square for his knight. Necessary was 36…cxb5. After 37.♘e2 ♕a1+ 38.♗e1 ♖f5!, Black can narrowly hold.

37.♘c6 ♕a1+

An understandable move in time trouble: each check brings you closer to the time control. 37…♗xc6 38.bxc6 ♖f5 was also insufficient, as White has a hidden way to win: 39.♖d3 ♘f4 40.♖e3 ♕d4 41.♖e8!, winning the queen.

38.♗f1?

A more or less automatic move; White arranges his bishops in such a way that his king remains optimally protected. But the other bishop move, 38.♗e1!, would have been much stronger. White keeps the black knight under cover, and on 38…♗xc6, he has the surprising blow 39.g6!! (not 39.bxc6, in view of 39…♕d4+). After 39…hxg6 (or 39…♖d8 40.bxc6 ♕d4+ 41.♕xd4 cxd4 42.♖g5 ♘xf6 43.c7 ♖c8 44.♗g3 h6 45.g7+ ♔g8 46.♖g6, and wins) 40.♖h3+, Black gets mated.

38…♖f5!

Now, Black controls the position again.

39.g6

Not right now! White should have settled for 39.♕e1 ♕xe1 40.♗xe1 h6 41.♗d2, with good chances to hold.

39…♗xg6 40.♖xg6 ♖5xf6

The notorious 40th move, just before the time control. With his flag about to fall, Karpov fails to spot that 40…♘xf6 wins immediately. If White moves his rook, the black knight has a devastating sortie to e4.

41.♖xf6

This was the move Kasparov had sealed in his envelope. Karpov and his seconds had the whole night and the following morning to analyse the position.

41…♕xf6

Not the best way to recapture. Stronger was 41…♘xf6, although White can narrowly hold the position with 42.♕e1 ♖g8+ (42…♕xa2 43.♗g3) 43.♗g2 ♕b2 44.a4 ♖g7 45.♗g3 c4 46.♘e5 ♕d4+ 47.♗f2 ♕f4 48.♔f1.

After the game, Karpov was highly dissatisfied with the work of his seconds’ team. Sometimes a small, tightly-knit team can achieve more than a large group of grandmasters who do not know each other very well.

42.♕e1!

A very accurate defensive move.

42…♖g8+

On 42…♕g5+, 43.♗g3 was sufficient.

43.♔h2 ♕f4+ 44.♗g3 ♖xg3 45.♕xg3 ♕xf1 46.♕b8+ ♔g7 47.♕g3+

Draw. Black cannot evade perpetual check.

A marvellous struggle that revealed both the strong and the weak sides of the two players. The computer mercilessly exposes the mistakes, but it would be unfair to condemn the players’ performance for this reason. Both combatants had to solve extremely complex problems, and succeeded in many instances. Lesser gods would have lost the thread of the game much sooner. Curiously, Karpov had found himself in serious time trouble. Until then, he had been known for his fast and efficient play. However, it was mainly Kasparov who was shaken by the game; he claimed his first time-out, with a slight flu as the official reason. He wasn’t very ill, though. English grandmaster and chess writer Raymond Keene describes in his book, The Moscow Challenge, how the American journalist Eric Schiller visited Kasparov in the Rossiya Hotel: ‘He is jolly, comfortable and in complete control of himself. The only time a smile leaves his face is when his mother Klara Kasparova insists on administering traditional Azerbaijani anti-cold remedies, prepared by her using a special oil obtained from her sister.’ Kasparov and Schiller discussed the length of the match. Kasparov thought that it might go on until the end of November. Schiller thought this was too optimistic: it might take until the end of December. Both predictions would prove to be too optimistic.

In the Russian daily newspaper Izvestia, Bronstein criticized Karpov’s opening choice in the first two games. He regarded Karpov’s choice of the Queen’s Indian as a provocation, motivated by the fact that Kasparov’s preparation was superior. With white, Karpov would derive more benefit from 1.d4 as the first move; Kasparov’s repertoire – the Tarrasch Defence and the King’s Indian – gave Karpov positions that he commanded like no other player. Therefore, Bronstein thought that Karpov would switch to 1.d4 in the critical phase of the match. And he was proved right. Karpov mostly played 1.e4, but already in his match with Spassky, he had varied with 1.d4. And in tournament games preceding