In the second game of the 1987 World Championship, Kasparov played an English opening. In the ninth move Karpov played a novelty that he had prepared for his 1981 match with Korchnoi (the usual play is 9. ... exf3). It is claimed that Kasparov's next two moves, 10. d3 and 11. Qb3 completely refute Karpov's surprise move, but in the end Kasparov lost this game (it is the famous one in which he forgot to hit the clock after playing his 26th move).

Then in the fourth game, Kasparov plays an English again and reaches the same position up to move 8! Clearly it was a provocation, but whether he could refute 9. ... e3 or not was not demonstrated, because Karpov avoided it and played the usual 9. ... exf3, only to get trounced (incidentally, this was the 100th championship game between the two).

What is the novelty about, and how do Kasparov moves refute it?

For reference, here are the two games:

[Event "World Championship Game 2"]
[Site "Seville, Spain"]
[Date "1987.10.14"]
[Round "2"]
[Result "0-1"]
[White "Kasparov"]
[Black "Karpov"]
[ECO "A29"]
[FEN ""]

1. c4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. g3 Bb4 5. Bg2 O-O 6. O-O e4 7. Ng5 Bxc3 8. bxc3 Re8 9. f3 e3 10. d3 d5 11. Qb3 Na5 12. Qa3 c6 13. cxd5 cxd5 14. f4 Nc6 15. Rb1 Qc7 16. Bb2 Bg4 17. c4 dxc4 18. Bxf6 gxf6 19. Ne4 Kg7 20. dxc4 Rad8 21. Rb3 Nd4 22. Rxe3 Qxc4 23. Kh1 Nf5 24. Rd3 Bxe2 25. Rxd8 Rxd8 26. Re1 Re8 27. Qa5 b5 28. Nd2 Qd3 29. Nb3 Bf3 30. Bxf3 Qxf3 31. Kg1 Rxe1 32. Qxe1 Ne3 0-1
[Event "World Championship Game 4"]
[Site "Seville, Spain"]
[Date "1987.10.19"]
[Round "4"]
[Result "1-0"]
[White "Kasparov"]
[Black "Karpov"]
[ECO "A29"]
[FEN ""]

1. c4 Nf6 2. Nc3 e5 3. Nf3 Nc6 4. g3 Bb4 5. Bg2 O-O 6. O-O e4 7. Ng5 Bxc3 8. bxc3 Re8 9. f3 exf3 10. Nxf3 Qe7 11. e3 Ne5 12. Nd4 Nd3 13. Qe2 Nxc1 14. Raxc1 d6 15. Rf4 c6 16. Rcf1 Qe5 17. Qd3 Bd7 18. Nf5 Bxf5 19. Rxf5 Qe6 20. Qd4 Re7 21. Qh4 Nd7 22. Bh3 Nf8 23. R5f3 Qe5 24. d4 Qe4 25. Qxe4 Rxe4 26. Rxf7 Rxe3 27. d5 Rae8 28. Rxb7 cxd5 29. cxd5 R3e7 30. Rfb1 h5 31. a4 g5 32. Bf5 Kg7 33. a5 Kf6 34. Bd3 Rxb7 35. Rxb7 Re3 36. Bb5 Rxc3 37. Rxa7 Ng6 38. Rd7 Ne5 39. Rxd6 Kf5 40. a6 Ra3 41. Rd8 Ra2 42. Rf8 Ke4 43. d6 Nf3 44. Rxf3 Kxf3 45. Bc6 1-0
  • 3
    Kasparov annotates these two games at length in the book Garry Kasparov on Modern Chess Part Three: Kasparov vs Karpov 1986-1987. His analysis consists largely of a bunch of concrete lines so I won't try to summarize it here. It's worth noting that while he gives exclamation points to 10.d3 (on which he spent 83 minutes) and 11.Qb3, he does not claim that they "refute" 9...e3. – dfan Oct 2 '13 at 0:36

No, this was unequivocally not refuted. First, I would consider a line "refuted" if it led to a significant advantage, and there was nothing you could do about it...it is forced. That is definitely not the case.

For a very brief moment in time, this appeared to be slightly better for white, but black has since found plenty of resources with Na5, and it has been played many times since, even by some very strong Grandmasters. One of the latest games is Nakamura-Naiditsch from 2018. Another recent game between two highly-rated players is Gelfand-Hammer Palma de Mallorca 2017 (1/2)

 [Event "Olympiad-43"]
 [Site "Batumi"]
 [Date "2018.10.02"]
 [Round "8.3"]
 [White "Nakamura, Hikaru"]
 [Black "Naiditsch, Arkadij"]
 [Result "1/2-1/2"]
 [ECO "A29"]
 [WhiteElo "2763"]
 [BlackElo "2721"]
 [PlyCount "73"]
 [EventDate "2018.09.24"]
 [EventType "team-swiss"]
 [EventRounds "11"]
 [EventCountry "GEO"]
 [FEN ""]

 1. c4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. g3 Bb4 5. Bg2 O-O 6. O-O e4 7. Ng5 Bxc3 8. bxc3 Re8 9. f3 e3 10. d3 d5 11. Qb3 Na5 12. Qa4 c6 13. cxd5 cxd5 14. f4 Bg4 15. Nf3 Nc6 16. h3 Bd7 17. Qa3 Nh5 18. Kh2 Qf6 19. Nd4 Nxd4 20. cxd4 g5 21. fxg5 Qxg5 22. Qd6 Bc6 23. h4 Qg7 24. Bh3 f5 25. Rf3 Rad8 26. Qb4 f4 27. gxf4 Qc7 28. Ba3 Nxf4 29. Qe1 Nxh3+ 30. Qg3+ Qxg3+ 31. Rxg3+ Kh8 32. Kxh3 Re6 33. Rf1 Rg8 34. h5 h6 35. Rxg8+ Kxg8 36. Kh4 Ba4 37. Bc1 1/2-1/2

In that game, black was never worse than the tiniest of white pluses that you see in every sound opening, and by move 19, black was actually better. In fact, per Mega 2020, in the 48 games played in this line since white won 16, black won 20, and 12 draws. The average rating for white was 2386, and for black 2410.

It would seem that this opening holds up quite well for black even to this day, and my modern computer with Stockfish agrees.


I recall these games from his Modern Chess series. Logically, if Kasparov was willing to return to the same position, Karpov believed that it was likely that Kasparov had found a line that resulted in a favorable position for White. Indeed, the moves 10. d3 and 11. Qb3 give White some advantage and Kasparov had a rather pleasant position throughout. Karpov probably didn't want to revisit this line as he knew he got lucky in the first game.

  • True... except Kasparov is known to bluff, plus pushing to repeat the position can also be interpreted simply as a move in the psychological war outside the board. Thus, this is still not an answer. – yrodro Oct 16 '13 at 0:59
  • 3
    Could you please provide a source on Kasparov's bluffing habits? In my experience, nobody in chess bluffs. Each moves is calculated. My answer was given from the perspective of what I would do in Kasparov's shoes. Your comment actually supplements my answer - repeating the position is certainly a psychological move. Kasparov is challenging Karpov to repeat the position, something that he certainly analyzed extensively at least twice before (his preparation for the first game, and his review after the first game). Karpov likely did the same research and realized that the line was faulty. – Andrew Ng Oct 16 '13 at 4:30

Kasparov took 63 minutes over his response to the ....e3 move, that in itself made the novelty a very worthwhile undertaking from Karpov's perspective. Clearly he found a gap in Kasparov's preparation and often that might be all one needs in a contest between two near equals, especially when the psychological factors are complex and significant.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.