Author: Uwe Immel
Date: 12:45:49 02/22/98
Few weeks ago I played the following computermatch with a friend of mine. [Event "10 Min/Zug"] [Site "?"] [Date "1998.??.??"] [Round "?"] [White "Rebel 9 P133"] [Black "Socrates 3.0 Cyrix 166+"] [Result "1-0"] [ECO "A46"] [PlyCount "104"] [EventDate "1998.??.??"] 1. Nf3 Nf6 2. d4 e6 3. Bg5 h6 4. Bxf6 Qxf6 5. e4 Nc6 6. Nc3 Bb4 7. Bc4 Bxc3+ 8. bxc3 d6 9. O-O O-O 10. Rb1 a6 11. Be2 Qf4 12. Nd2 Re8 13. g3 Qg5 14. Rb3 e5 15. f4 exf4 16. Rxf4 Nd8 17. Kg2 b5 18. h3 c5 19. Bh5 Be6 20. d5 Bd7 21. c4 b4 22. Bg4 Bxg4 23. Rxg4 Qe7 24. Rf3 Qe5 25. Qb1 Ra7 26. Rf5 Qc3 27. Qc1 27... Rae7 $6 (27... g6 28. Rf3) 28. e5 Rxe5 29. Rxe5 Qxe5 30. Re4 Qxe4+ 31. Nxe4 Rxe4 32. Qf1 g6 33. c3 a5 34. g4 Nb7 35. Kf3 f5 36. Qa1 Rxc4 37. cxb4 fxg4+ 38. hxg4 g5 39. b5 Nd8 40. Qe1 Nf7 41. Qxa5 Ne5+ 42. Ke2 Re4+ 43. Kd1 Rd4+ 44. Kc2 Rc4+ 45. Kb1 Rb4+ 46. Ka1 Rxg4 47. Qd8+ Kg7 48. b6 Rg1+ 49. Kb2 Rg2+ 50. Kb3 Rg3+ 51. Ka4 Rg4+ 52. Ka5 Rg2 1-0 At the moment I noticed that Socrates is going to exchange his queen for two rooks, I became a misgiving about that. From my experience of playing hundreds of computermatches I remembered, that mostly the side wich remains with the 2 rooks is going to lose. Also in this game my feeling becomes true and Socrates lost. To prove these obvious weakness of chessprograms I did some simple tests by playing some modified positions with several programs in autoplay mode. The results confirmed my suspicion, nearly 80% of the 30 games were won by the side who remains with the queen. So imho I would say, that chessprograms should avoid the exchange of queen for rooks, because they have problems to coordinate the playing with rooks alone vs queen. Surely, this described situation didn't occur very often, but the weakness is real. regards Uwe
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.