Author: Mark Young
Date: 05:59:52 06/20/01
Go up one level in this thread
On June 20, 2001 at 06:14:41, odell hall wrote: >On June 20, 2001 at 05:19:01, Otello Gnaramori wrote: > >>Hi Odell, >>What do you exactly mean for understanding ? >>Do you mean the evaluation function of the program ? >>In the final count I think that is the performance that counts as many others >>pointed out. > > > Well I just mean the knowledge that the programmers incorparate into their >programs, such as, knowledge of bishops of opposite colours, strategy, pawn >structure, ofcourse they really have no "real' understanding at all, but they >have enough artificial knowledge to allow them to play chess strategically at >the 2100 level, but that does not mean they can not perform at the grandmaster >level. as i pointed out earlier they have special talent's that humans lack, >they don't get tired, they don't get intimadated, they calculate accurately, >ect, ect, this is what makes them perform at the grandmaster level, but as far >as their understanding of Chess, they are in no way grandmasters, i think >everyone can agree with this. But if we speak of Results, i think they are >grandmaster, as deep jr, did perform at the 2600 level performance rating Dr. >hyatt and fide requires, along ofcourse with Fritz. The main reason computers can show understanding in chess that produces moves of a high positional nature that beats many grandmasters is that chess is 100% tactical and this lets them see with their search many so called positional moves. The other main reason why computes are so strong is their ability to defend, or the ability to counter direct threats. We offen forget how important defence is in chess, and computers can defend some positions better then any human could ever hope to. If you want to understand how a great defense can lead to winning games or in this case beating a 2700+ Grandmasters, check out the below game GM Leko Vs Deep Junior. Showing that if you can hold on, any human player is bound to crack. I know some will take this as a negetive and not a plus for the computers, but this is the nature of chess. Humans have weaknesses too that computers can exploit. [Event "Chess Meeting 2000 Super"] [Site "Dortmund"] [Date "2000.07.??"] [Round "9"] [White "Leko, Peter"] [Black "Junior 6"] [Result "0-1"] [ECO "C48"] [WhiteElo "2743"] [PlyCount "120"] [EventDate "2000.07.07"] 1. e4 e5 2. Nf3 Nc6 3. Nc3 Nf6 4. Bb5 Nd4 5. Bc4 Bc5 6. d3 c6 7. Nxd4 Bxd4 8. Qf3 h5 9. h3 d5 10. exd5 Bxc3+ 11. bxc3 cxd5 12. Bb3 Qc7 13. O-O O-O 14. Qg3 a5 15. a4 Be6 16. Bh6 Ne8 17. Bd2 f6 18. f4 exf4 19. Qxf4 Nd6 20. Rae1 Bf7 21. Qd4 Qc6 22. Re7 Rfe8 23. Rxf7 Nxf7 24. Bxd5 Qd7 25. Rb1 Rad8 26. c4 Qxa4 27. Rxb7 Rd7 28. Rxd7 Qxd7 29. Bxa5 Re2 30. Bb6 Qe8 31. Bc7 Rxc2 32. c5 Kh8 33. Bg3 Qd8 34. Bf2 Rc1+ 35. Kh2 Qb8+ 36. g3 Nh6 37. Kg2 Rc2 38. Be4 h4 39. Qe3 f5 40. Bd5 hxg3 41. Qxg3 Qb5 42. Qe3 Qb4 43. d4 Qb2 44. c6 Kh7 45. Qf4 Qa3 46. Be4 Rxf2+ 47. Kxf2 fxe4 48. c7 Qa8 49. d5 Qa7+ 50. Kg2 Qa2+ 51. Qf2 Qc4 52. d6 Qc6 53. Qf8 e3+ 54. Qf3 Qc2+ 55. Kf1 Qc4+ 56. Kg2 Qe6 57. Qc6 Qa2+ 58. Kh1 Nf5 59. Qe4 Qa1+ 60. Kh2 Qb2+ 0-1
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.