Author: Uri Blass
Date: 08:35:57 05/02/02
Go up one level in this thread
On May 02, 2002 at 09:43:33, Vine Smith wrote: >On May 02, 2002 at 06:41:58, Steve Maughan wrote: > >>A business trip freed up my computer for 36 hours - so I decided to play a match >>between Shredder 6.02 and Hiarcs 8 (game in 30 min, ponder off, 1.5 GHz P4). So >>far Hiarcs has won, by a slim margin, all the matches I have played (Gambit >>Tiger 2 and Fritz 7). This was the first one it lost! Shredder manager to beat >>the new kid on the block by +13, -9, =17. The games are given below. >> >>A couple of comments. Firstly, this number of games does not prove that >>Shredder is better than Hiarcs (even at the 90% confidence level). Secondly >>there are quite a few draws - 17! I don't know for certain but I imagine that >>Shredder and Hiarcs probably are close cousins in terms of playing strengths / >>style which may be a factor in a higher draw count. I haven't had a chance to >>go through the games so I can't comment on them but I post them for your >>enjoyment! >> >>Regards, >> >>Steve Maughan >> >>========================================================================== >[snip] >>[Event "Durham Riding, Blitz:30'"] >>[Site "Durham Riding"] >>[Date "2002.04.30"] >>[Round "12"] >>[White "Shredder 6.02"] >>[Black "Hiarcs 8"] >>[Result "1-0"] >>[ECO "C00"] >>[TimeControl "1800"] >>[PlyCount "143"] >> >>1. e4 e6 2. Qe2 Nc6 3. Nf3 Be7 4. d4 d5 5. e5 f6 6. Be3 fxe5 7. dxe5 Nh6 8. >>Bxh6 gxh6 9. Nc3 O-O 10. O-O-O Bd7 11. Kb1 Be8 12. h4 Bb4 13. Qe3 Qe7 14. >>h5 Rf5 15. Bd3 Rxh5 16. Rxh5 Bxh5 17. Qxh6 Bxf3 18. gxf3 Rf8 19. Ne2 Rf7 >>20. c3 Bc5 21. f4 Bxf2 22. Rh1 Rg7 23. Rh5 a5 24. f5 exf5 25. e6 Ne5 26. >>Bxf5 Ng6 27. Qd2 c6 28. Qh6 Nf8 29. Rh1 Rg5 30. Bd3 Be3 31. Qh3 Re5 32. Bf5 >>h6 33. Qg3+ Qg7 34. Qh3 c5 35. Rf1 b5 36. Bg4 d4 37. Ka1 b4 38. cxd4 cxd4 >>39. Ng3 Qc7 40. Nf5 Rxe6 41. Nxe3 Rxe3 42. Qxh6 Qe7 43. Be2 Nh7 44. Rg1+ >>Kh8 45. Qg6 Nf6 46. Bd1 Rh3 47. Qg2 Rh7 48. Bc2 Rh4 49. Qg3 Rh6 50. Qb8+ >>Qe8 51. Qd6 Qf7 52. Qxd4 Qf8 53. Bb3 Rh5 54. Qf4 Nd7 55. Bf7 Rh6 56. Kb1 >>Nf6 57. Bc4 a4 58. Rg8+ Qxg8 59. Bxg8 Nxg8 60. Qxb4 Kg7 61. a3 Re6 62. Qxa4 >>Nf6 63. Qa7+ Kg6 64. a4 Kf5 65. b4 Nd5 66. a5 Ke5 67. a6 Nb6 68. Qb8+ Kd4 >>69. a7 Re1+ 70. Kc2 Na8 71. Qxa8 Re2+ 72. Kc1 1-0 >[snip] > >One has to wonder exactly what kind of positional knowledge has been included in >Hiarcs 8 for it to play the above game, which was the 12th in your series. >First, its book gives out at move 3, when 3...e5, leaving White with a misplaced >queen in a king's pawn game, is normal and instead gets a bad French with >3...Be7. Then it "solves" its development difficulties with 7...Nh6? and castles >into the busted up kingside with 9...O-O? It's something of a miracle that it >survived as long as it did -- too bad Smirin couldn't have studied this one >before his match. I suspect that it could not help smirin because hiarcs used a special book against smirin and not the commercial book. I also doubt if Hiarcs8 has better evaluation than hiarcs7.32 I read that it can search clearly deeper than hiarcs7.32 and it can explain a big improvement. If you assume that it also know much more about evaluation then you can expect it to be clearly the best program. The fact that it did not happen suggest that the improvement relative to Hiarcs7.32 is only or almost only in search rules. Uri
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.