White: Anatoly Karpov    0.5



Black: Gata Kamsky       0.5







Played 4 July 96 in Elista, Kalmykia

1. d4

Not much has happened in terms of the score since I stopped annotating the games: Karpov is still ahead, so he sees absolutely no reason to change his first move. Kamsky is once again the one who has to come up with some new way of stirring up trouble against the world champion's solid play. Please notice how throughout the match Karpov kept playing his favorite openings with relative success, while Kamsky was obliged to experience all the time with new systems, some that he might have never played before. This game is no exception from the pattern.

1. ...   Nf6



2. c4    c5!

Maybe Kamsky's best opening choice in this entire match! The Benoni Defense, spelled by many people Ben Oni because these are the names of the two masters who created it, is one of the sharpest openings around. Just that is enough to justify Kamsky's choice! Unfortunately, the Benoni is also considered a very risky defense, and few top players have the courage to play it, especially against someone of Karpov's strength. On the other hand, Kamsky is not known as a great Benoni fan, which means that maybe Karpov has not specifically prepared it for this match. Generally speaking, Karpov has not had too much experience against the Benoni, since not many people had the courage to play it against him. All of the above make the Benoni the perfect choice for Kamsky at this time of the match.

3. d5   e6

For a moment I thought Kamsky would play 3...b5 and enter the sharp Benko Gambit (also called the Volga Gambit). That might have been a little too risky, which is why I consider black's choice 3...e6 and the Modern Benoni that arises a better alternative.

The idea behind this 3...e6 is not very hard to guess. Black is starting to harass white's central pawns, hoping to avoid a complete domination of the centre by the white pieces. He further intends to trade the pawns on d5, thus making the white pawn on c4 move. Once that happens the control of the b5 square is a little weakened so the standard b7-b5 push has more chances of succeeding.

4. Nc3

White's only way to fight for an advantage is to keep a good grip over the central squares and build a fortress in the center capable of resisting black's expected assaults. 3.Nc3 is the perfect move from this point of view. Not only does the knight control the vital e4, d5 and b5 squares, but it also feels much safer now that black has played c7-c5 and is thus unable to play the annoying Bf8-b4 characteristic for the Nimzo-Indian.

4. ...   ed5



5. cd5   d6



6. e4    g6

Maybe black's most important piece in the Benoni, the dark squared bishop finds a beautiful position on g7. Actually, black has this in mind from the moment he plays c7-c5. By making white push his d4 pawn, black insures that the a1-h8 diagonale is open and ready to welcome his bishop. The white pawn on b2 already feels uncomfortable at the thought of the black bishop, but there is not much he can do to stop him, since a maneuver such as b2-b3 and Bb2 would put in danger the white knight, bishop, rook, and king, because of the numerous pins...

7. Nf3   Bg7



8. h3

While this looks like a typical Karpov move, the prophylactic 8.h3 is actually the most popular move in this position. By stopping black's light squared bishop from getting to g4, white makes it hard for black to find any other useful spot for the troubled bishop. The white pawns currently control e6, f5, g4, and h3, so the only square left is d7, which is of course not the greatest square...

Other benefits of 8.h3 include the opening for the king once it castles, the preparation of the h2 square for white's dark squared bishop if it will ever seek refuge from f4, and the prevention of any type of knight maneuvres using the g4 square. All this at the cost of only one tempo, but a fairly important tempo as you will soon realize.

8. ...    0-0



9. Bd3    b5!

Before this move was deeply analyzed several years ago, white was thought to hold an important advantage in this position. 9...b5 however, complicates things enormously. The idea behind b7-b5 is typical for the Benoni: black is trying to get some play on the queen side in exchange for white's center domination. But while usually black has to spend a lot of time preparing this move with maneuvres such as Nb8-a6-c7, Bc8-d7, a7-a6, and Ra8-b8, in this case he decides to sacrifice the pawn based on white's late castling.

By taking the pawn with the knight or bishop, white weakens his e4 pawn and the fact that the king is one the same semi-open file gives black lots of combinative possibilities. Not taking the pawn would mean that black gains several important tempos, so white has to make an important decision, which often depends on the many games that are played all the time in this variation and the novelties they often bring with them.

10. Nb5   Re8



11. 0-0   Ne4



12. Qb3!?

All the moves played up to now have been played many times. 12.Qb3, however, is almost a novelty. I am very curious to know whether Karpov prepared this move or not. The fact that 12.Re1 is the most popular move in this position, and that 12.Qb3 has been played in only one game I could find, makes me believe that there was some sort of home preparation, especially since this one game was played with the white pieces by one of Karpov's former seconds, GM Epishin. On the other hand, Karpov spent more than 20 minutes on this move which is hardly typical for a prepared move. Was he trying to remember his analysis, or was he just checking all the variations again? Very hard to say and I doubt we will find out unless the FIDE world champion decides to tell us himself.

In any case, the main idea of this move is to take Kamsky of the well-trodden paths of 12.Re1 and whatever preparation he might have had for the next few moves. Nonetheless, 12.Qb3 doesn't look like a very useful move in the context of this variation, so I am not sure Karpov would repeat it in the next game with the white pieces. But I guess we'll first have to see if Kamsky will repeat the Benoni...

12. ...   a6



13. Na3

Typical idea for this system, white is moving the knight over to the strongest position available: c4! From there, the knight defends the b2 pawn, attacks black's weak d6 pawn, and is generally annoying black, since he can't use the 'b' pawn to send it away... 12.Qb3 might have been justified if white had played 13.Nc3 here, but after 13...Nc3 14.bc3 Nd7 black's position looks good, having in view the weakness of the c3 pawn and the control of the two open 'e' and 'b' files after the following 15...Rb8.

13. ...   Nd7



14. Nc4

Karpov spent thirty minutes on this very obvious move, and I would be very curious to know what he thought about for so long. This once again raises the question of how effective the home preparation has been, if after such normal moves on black's side like a7-a6 and Nb8-d7, Karpov has spent a precious hour of thinking time. Maybe he found some flaw in the home analysis and was obliged to improvise...

14. ...   Rb8



15. Qc2   f5!

15...f5 is a very risky move in this position since it considerably weakens black's kingside, especially the square e6. But in Kamsky's situation this is the only move. Relinquishing control of the e4 square would seriously hurt any chances for further counterplay and when you're four points behind in a match, counterplay is the only important thing!

16. Bf4   Nb6!

The knight on c4 is by far white's most active and disturbing piece. Kamsky compensates for the absence of the b pawn by using his knight to get rid of white's knight. Once these knights are traded, white will be reduced to the role of defender, especially on the b file, where once again the b2 pawn will bbe harrased not only by black's bishop but also by the battery on the b file.

16...Nb6 has only one problem: it covers the queen's diagonale and thus allows the white knight to jump to a5 and maybe then to c6. Luckily, black has a tactical resource which prevents this move: 17.Na5? Nd5! (now the black queen is once again attacking the white knight) 18.Nc6 Qb6 19.Nb8 Nf4 and even though white is an exchange up he can not pull the knight out of b8...

17. Nb6   Qb6



18. Rb1   Bd7

As weak as the white pawn on b2 might be, taking it right now would be a grave error! After 18...Bb2 19.Bc1! black would have to give up his beatiful bishop or else lose his queen.

19. Nd2!

Karpov realizes that his opening strategy has not exactly been great. After all, any player would be delighted to have such a position with black after the opening: all of black's pieces have beautiful squares and seem to work together perfectly. Nevertheless, Karpov doesn't lose his calm and employs one of the most popular forms of defense: trading the opponent's pieces. The reason this strategy is so successful in this position is that black's superiority only lies in the activity of the pieces and not the pawn structure. Thus, if able to trade most or all of the pieces, white will achieve an equal or even slightly superior endgame, having in view that his pawn configuration is not worse. Kamsky's task, meanwhile, lies in avoiding excessive piece trades and building up the advantage.

19. ...   Nd2



20. Qd2   Bd4?!

I have a feeling that Kamsky took the above mentioned advise a little too literally. In this position there is one black piece that is not doing much: the bishop on d7! At the same time, the white bishop on d3 is doing a great job of defending the white position by keeping an eye on the e4, e2 squares and even the rook on b1. This is definitely a good moment to get rid of the white bishop with 20...Bb5!

Kamsky's move 20...Bd4 is good-looking but little more than that... I suppose he hoped for some sort of action around white's f2 pawn, if the rook moves from f1. Black also has the queen on b6, so after c5-c4 there could be some danger for white. Unfortunately, a short concrete calculation will show that this danger is only apparent. Bd4 doesn't even stop white's rook from leaving f1, so Karpov can further pursue his idea of trading black's active pieces...

21. Rfe1!   Re1



22. Qe1

As much as the white rook would enjoy the 'e' file, the white pawn on b2 is taking its toll. The rook must remain on b1 to defend the pawn, but the queen can also be very dangerous on the e-file, so black will be forced to move his rook off the b file ...

22. ...   Re8



23. Qd2   a5



24. b3    Be5

Kamsky is forced to make this very unpleasant move. I know how much he loved his dark squared bishop, but in this particular position black can't do anything as long as white's bishop is pounding on the weak d6 pawn. At the same time, the black bishop has lost some of its effectiveness, now that the white pawn is gone from b2. So Kamsky didn't have much choice...

25. a3!

Karpov makes a very important move that might appear puzzling. He is weaknening the b3 pawn but for a very good reason! First of all he gains control of the b4 square, which black might have used to propose a queen trade and repair his pawn structure. He also prepares the b3-b4 push which could get him rid of many of his present problems, especially is black had any hopes of pushing his 'a' pawn any further.

25. ...    Bb5

Better later than never...

26. Bb5   Qb5



27. Be5   Re5



28. Re1!

Exceptional defense by the FIDE world champion! His strategy worked perfectly: he managed to trade most of the pieces, and only kept the ones that give the best drawing chances, the queens. Now he is offering black a pawn, that he can not really take: after 28...Qb3 29.Re5 de5 30.d6 white's passed pawn will cause a lot of trouble. 28...Re1 29.Qe1 Qb3 isn't much better either, since after 30.Qe8 white has at least a draw.

28. ...   Kf7?!

Definitely not the best move in this position, but this is a very difficult position to be in when a draw is almost as bad as a loss... Black should not allow white a passed pawn under any circumstances. Unfortunately, that is also the only way black could win this position, if he gave white the pawn and then managed to take it using the king and queen duo. This is exactly what Kamsky is trying to do with 28...Kf7. He could have avoided this by playing 28...Qe8, but after white's quiet 29.Kf1 there would be little left than to repeat moves by checking on b5 and then back to e8....

29. Re5   de5



30. Qh6   Kg8



31. Qd2   Kf7



32. Qc3!

Many people might not understand Karpov's attitude here. After all he got what he wanted: the draw, after the simple repetition with 32.Qh6. Leading 9-5, a draw is almost as good as a win. Nonetheless, he prefers to fight on. The reasons for Karpov's decision to continue are rather simple: first of all a win would almost immediately end the match; secondly, Kamsky was in time trouble; and finally, Karpov calculated that even if black plays the best moves, he still has a way to force the draw a little later. Karpov was also hoping that Kamsky will change his mind and avoid a draw by making a weaker move, which is psychologically understandable...

32. ...   Kf6!

Pushing the pawn to e4 would be a grave error, since it would allow the white queen to enter black's fortress on the a1-h8 diagonale. Black's only hope is to keep the white queen away and get closer the passed pawn on d5.

33. f4   Qe2



34. Qc5

Trading queens in this position is clearly not a good idea since after 34.Qe5 Qe5 35.fe5 Ke5 black not only remains a pawn up but also has a much more active king. Taking first on e5 with the pawn and then on c5 is not great either, because of 34.fe5 Qe5 35.Qc5 Qe1! and after 36.Kh2 Qe5...

34. ...   ef4!

Capturing this pawn is extrememly important for black! Now Kamsky has the threat of perpetual check with Qe1-g3-e1, and he also has a place to go for the king after 35.Qf8 - to g5. Of course, that would mean that the king gets kind of far from white's passed pawn, but having in view the perpetual threat, there is little hope for white to queen the pawn before black obtains the draw.

35. Qc3   Qe5!

Black is still happy to trade queens because his king is so much more active in the resulting pawn endgame. Leaving now to g5 would not be such a good idea, since white's queen is well placed to avoid both the checks on e1 and g3, plus the f4-f3 push.

36. Qc6   Kg5



37. d6

Karpov has exhausted all his possibilities and finally agrees with the draw. There is no good way to stop the perpetual and maintain the advantage of the passed pawn at the same time, so he prefers to create a threat that would force black to go for the perpetual. As much as Kamsky hates this, there is no good way to avoid it...

37. ...   Qe1



38. Kh2   Qg3



39. Kg1   Qe1



40. Kh2   Qg3



41. Kg1   Qe1



42. Kh2   Qg3

And having repeated the position three times, the two players agreed on a draw.

This was a very interesting game in which it seems that Kamsky caught Karpov off guard with his perfect opening choice. After obtaining a slightly better position from the opeing, he didn't fully exploit it, while Karpov defended exceptionally well. I know how hard it must be on Kamsky to calculate all variations and almost always throw out those that end in drawish positions, even when they are the best, but that's life when you are trailing four points. I have to agree that at the present time betting on Kamsky would not be the smartest thing to do, but who knows, maybe we witness a surprise victory in game No. 16 which starts Saturday, July 6th at 8 AM Eastern time. As usual I invite you to join us for the live coverage and then come back a little later for my special post-game analysis.

GM Schwartzman has been playing chess since he was two years old, and started playing tournaments at the age of four. He received his first international invitation by the time he was eight. Since then Gabriel has been invited to events in Germany, Switzerland, Austria, Belgium, the Netherlands, France, Spain, Denmark, Italy, Israel, Canada, as well as the USA.

In 1988 he became vice-world champion of the Under 12 years section, and in 1990 he won the bronze medal at the Youth World championship in Fond du Lac, Wisconsin. Currently 19 years old, Gabriel is the top ranked junior (under the age of 21) player in the United States.

GM Schwartzman is multi-talented, speaking Romanian, English, German and French very well. He also gets by in Russian and Spanish. Gabriel is also an accomplished Internet publisher - his Internet Chess Academy can be found at http://www.yourmove.com

You can send email to Gabriel at [email protected].