Interesting Endgame Blunder

[Colin Frayn]

Geschrieben von: / Posted by: Colin Frayn at 10 May 2000 13:20:40:
Here's a game I just played between ColChess v6.2.1 and Sjeng on 10 minute blitz level. Sjeng gains an advantage in a favourable exchange in the middle game. I don't think that Colchess saw the future advantages of the sacrifice, but that doesn't matter as it managed to cunningly manipulate the game into one that it knew that it would have a good chance of winning. The knight and pawn endgame function gives quite a large benefit for defendable passed pawns, especially those on the far side of the board away from the enemy king.
Sjeng was looking more than 2 ply further ahead for most of the game, but I think that it does this at the sacrifice of accuracy and complicated analysis.
Well at least I'm satisfied that my static endgame evaluation routines are working OK now. Sjeng probably shouldn't have exchanged queens, but I'm not an expert so I wouldn't really know.
Cheers,
Colin
[Event "Computer chess game"]
[Site "cmf29.robinson.cam.ac.uk"]
[Date "2000.05.10"]
[Round "6"]
[White "sjeng"]
[Black "colchess"]
[Result "0-1"]
[TimeControl "600"]
1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. Bg5 Be7 5. e3 O-O 6. Bd3 Nbd7 7. Nf3 b6 8.
cxd5 exd5 9. O-O Bb7 10. Qe2 c5 11. Rac1 c4 12. Bf5 g6 13. Bc2 Bd6 14. b3
Ba3 15. Rb1 Bb4 16. Na4 cxb3 17. Bxb3 Qe7 18. Rfc1 Bd6 19. Nc3 Qe6 20. Nb5
Ba6 21. a4 Rfe8 22. Qa2 Bxb5 23. axb5 h5 24. Rc6 Nb8 25. Rxd6 Qxd6 26. Bxf6
Qxf6 27. Bxd5 Nd7 28. Bxa8 Rxa8 29. Qd5 Rd8 30. Ra1 Nf8 31. Qe5 Qxe5 32.
Nxe5 Rd5 33. Nc6 Rxb5 34. Nxa7 Ra5 35. Rxa5 bxa5 36. d5 Kg7 37. d6 Kf6 38.
Nb5 Ke6 39. h3 Kd5 40. Na3 a4 41. Nb5 Kc6 42. Nc3 a3 43. h4 Ne6 44. Kf1 f6
45. f3 Ng7 46. g3 Nf5 47. Kf2 Nxd6 48. Kg2 Nb5 49. Na2 Kd5 50. Kg1 Kc4 51.
Kf2 Kb3 52. Nc1+ Kc2 53. Na2 Nc3 54. Nb4+ Kb3 55. Nd3 Nd1+ 56. Ke2 Kc3 57.
Nc1 Kc2 58. Na2 Nc3+ 59. Nxc3 Kxc3 60. Kf2 a2 61. f4 a1=Q 62. e4 Kd3 63.
Kg2 Qb2+ 64. Kh3 Kxe4 65. g4 Kf3 66. gxh5 Qg2#
{Black mates} 0-1


ColChess Homepage