Author: Vasik Rajlich
Date: 09:22:16 01/05/06
Go up one level in this thread
On January 05, 2006 at 07:18:40, Arnold Gove wrote:
>On January 05, 2006 at 07:03:01, Eduard Nemeth wrote:
>
>>Thanks Vasik!
>>
>>Cool! Nice tool this Engine Rybka WinFinder 1.0!
>>
>>Rybka WinFinder 1.0 finds a mate (mate in 11) on my "only P3 600 MHz" in
>>7 seconds here.
>>
>>[D]rnb2r1k/qp1nb1pB/4p2p/1P1p4/p7/P4N2/1BQN1PPP/4RRK1 w - - 0 1
>>
>>Analysis by Rybka WinFinder 1.0 32-bit:
>>
>>1.Txe6
>> +- (1.72) Tiefe: 3 00:00:00
>>1.Txe6
>> +- (1.84) Tiefe: 4 00:00:00
>>1.Lxg7+ Kxg7 2.Txe6
>> +- (2.54) Tiefe: 4 00:00:01
>>1.Lxg7+ Kxg7 2.Dg6+ Kh8 3.Dxh6 Tf7
>> +- (7.04) Tiefe: 5 00:00:03
>>1.Lxg7+ Kxg7 2.Dg6+ Kh8 3.Dxh6 Tf7 4.Lg6+ Kg8
>> +- (10.33) Tiefe: 6 00:00:04 16kN
>>1.Lxg7+ Kxg7 2.Dg6+ Kh8 3.Dxh6 Tf7 4.Lg6+ Kg8 5.Lxf7+
>> +- (320.00) Tiefe: 7 00:00:07 22kN
>>
>>(Nemeth, Privat 05.01.2006)
>>
>>On playchess _only_ one experimantal game:
>>
>>Toga II on AMD 2000 MHz.
>>Rybka WinFinder 1.0 on P3 600 MHz.
>>
>>[Event "Wertungspartie, 5m + 0s"]
>>[Site "Maschinenraum"]
>>[Date "2006.01.05"]
>>[Round "?"]
>>[White "Polezhaev, Toga II 1.0 Beta2"]
>>[Black "Solist, Rybka WinFinder 1.0"]
>>[Result "0-1"]
>>[ECO "A29"]
>>[PlyCount "110"]
>>[EventDate "2006.01.05"]
>>[TimeControl "300"]
>>
>>1. c4 {0} e5 {0} 2. Nc3 {0} Nf6 {0} 3. Nf3 {0} Nc6 {0} 4. g3 {0} Bb4 {0} 5. Bg2
>>{0} O-O {0} 6. O-O {10} Re8 {-0.15/12 0} 7. Nd5 {0} Bf8 {0} 8. d3 {0} Nxd5 {0}
>>9. cxd5 {0} Nd4 {0} 10. Be3 {0} Bc5 {0.29/10 12} 11. Nxd4 {-0.05/13 12} exd4 {
>>0.21/9 3} 12. Bd2 {-0.03/14 8} d6 {0.14/18 1} 13. a4 {0} a5 {-0.07/10 5} 14.
>>Re1 {0} Bf5 {-0.08/12 5} 15. Bf3 {-0.08/13 11} h6 {-0.07/12 5} 16. Rc1 {
>>-0.07/13 6} Ra6 {-0.08/10 5} 17. Rc4 {-0.04/13 7} c6 {-0.03/12 1} 18. Qb3 {
>>0.07/12 17} Qc7 {0.03/10 9} 19. Rec1 {0.15/12 9} Qe7 {0.00/10 9} 20. R4c2 {
>>0.23/12 9} Rea8 {0.00/9 6} 21. Bh5 {0.10/12 8} Rb6 {0.00/9 6} 22. Qc4 {
>>0.01/13 0} Ra7 {0.00/10 5} 23. Bf3 {0.13/11 5} Ra8 {0.00/14 0} 24. Qa2 {
>>Solist,Rybka WinFinder 1.0 bietet Remis 0.09/12 15} Qf6 {0.00/9 5} 25. Bf4 {
>>0.10/12 0} g5 {0.00/9 6} 26. Rxc5 {0.25/12 5} gxf4 {-0.08/10 1} 27. R5c4 {
>>0.26/12 9} c5 {-0.09/10 6} 28. Rf1 {0.26/12 8} fxg3 {-0.40/12 4} 29. fxg3 {
>>0.11/12 4} Qg5 {-0.42/12 0} 30. b3 {0.16/12 3} Bh3 {-0.54/10 3} 31. Re1 {
>>0.00/12 0} Re8 {-0.74/10 7} 32. Kh1 {0.00/12 4} Bg4 {-0.84/12 0} 33. Bg2 {
>>-0.88/13 7} Qe3 {-1.66/12 0} 34. Rc2 {-1.53/13 13} Re5 {-1.73/15 0} 35. Rf1 {
>>-1.77/12 5} Rh5 {-1.98/13 0} 36. Be4 {-1.75/13 8} Qxg3 {-2.27/10 6} 37. e3 {
>>-1.86/13 0} Qxe3 {-2.10/10 8} 38. Rcf2 {-1.66/13 0} Bh3 {-1.81/9 5} 39. Rd1 {
>>-1.51/13 2} Rxb3 {-1.22/10 7} 40. Rf3 {-1.74/11 3} Qxe4 {-1.44/10 0} 41. dxe4 {
>>-2.65/13 2} Rxf3 {-1.42/10 0} 42. Qe2 {-2.75/14 7} Rg3 {-1.47/14 0} 43. Qxh5 {
>>-3.18/14 4} Bg4 {-1.98/16 0} 44. Rd3 {-3.11/16 2} Bxh5 {-2.18/15 1} 45. Rxg3+ {
>>-3.73/16 1} Kf8 {-2.23/16 2} 46. Kg2 {-3.76/16 1} c4 {-2.36/16 4} 47. Kf2 {
>>-3.80/17 0} Bd1 {-2.55/16 10} 48. Ke1 {-4.83/18 9} Bc2 {-2.78/18 0} 49. Kd2 {
>>-5.03/16 8} Bxe4 {-3.64/18 3} 50. Ke2 {-5.38/15 1} c3 {-3.77/14 2} 51. Rh3 {
>>-5.80/15 0} Kg7 {-3.97/13 3} 52. Kd1 {-6.22/15 0} f5 {-4.29/13 1} 53. Rg3+ {
>>-9.80/18 15} Kf6 {-4.96/10 1} 54. Rh3 {-9.01/15 0} Kg5 {-6.03/12 1} 55. Rg3+ {
>>-11.02/15 2} Kf4 {Polezhaev,Toga II 1.0 Beta2 resigns -7.00/22 0} 0-1
>>
>>After move 42:
>>
>>Polezhaev,T - Solist,R, Wertungspartie, 5m + 0s 2006
>>
>>[D]6k1/1p3p2/3p3p/p1pP3r/P2pP3/6rb/4Q2P/3R3K w - - 0 1
>>
>>Black is winning.
>
>Perhaps this new engine needs explaining. It will only perform its real purpose
>at long time controls. The mate it found is also found rapidly by Rybka Preview;
> the game you quote has a time control that permits only the "Brief
>stabilization stage" hardly the special "Win-finding stage" that is the whole
>point of the engine. If Winfinder also performs well enough as an ordinary
>engine, then I suppose that is interesting, although Vasik himself thinks that
>"the results will be poor."
>
Looking through the log of the game, you will see that many of black's moves
were played after a search of >10 ply.
These additional plies either found something really strong (unlikely) or were
completely wasted (very likely).
IMHO this type of search only makes sense if there is a human there to say "look
now".
Vas
>AG
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.