Author: Robert Hyatt
Date: 22:14:24 06/14/05
Go up one level in this thread
On June 14, 2005 at 23:24:35, Robert Hyatt wrote:
>On June 14, 2005 at 23:20:36, Robert Hyatt wrote:
>
>>On June 14, 2005 at 21:53:11, Pablo Ignacio Restrepo wrote:
>>
>>>Doctor
>>>Robert Hyatt
>>>
>>>I wonder.
>>>Why Crafty made a draw ?
>>>Best Regard,
>>>Pablo Restrepo
>>
>>
>>Hard to say...
>>
>>if this was really a guest vs crafty game, Crafty plays a bit differently since
>>guests have no "rating" and Crafty has no idea how good or bad they are as a
>>chess player. If you want to play from a real account, I can certainly set it
>>up so that this will be nearly impossible to do...
>>
>>
>>
>
>Looked further. Classic case of a 3-fold repetition hidden by a hash table hit,
>that wrecked the search. I'll look to see if there is a way to recognize this
>and prevent this particular case from haunting me. It's a known problem for any
>program that uses a hash table however... not just for Crafty.
>
>
After a couple of hours of debugging, this is not a hash problem. It appears to
be a very subtle bug in 3-fold repetition detection in Crafty. I have not yet
found it. Crafty is one of the rare programs that requires 3-fold repetitions
at plies 1-2 in the search to trigger a draw, but it does like most everyone
else and requires just one repetition deeper in the tree to trigger a draw.
Somehow it couldn't find the 3-fold rep. Whether it is the SMP search stuff, or
just a repetition bug is unknown as I can't reproduce this in a test, yet. More
to come...
>
>
>
>>>
>>>[D]5r1k/4b1p1/4p1r1/p2p1q1p/1ppPpP1P/P1P1P1P1/1PR2B1N/Q6K b - - 0 142
>>>
>>>
>>>[Event "ICC 5 0 u"]
>>>[Site "Internet Chess Club"]
>>>[Date "2005.06.14"]
>>>[Round "?"]
>>>[White "guest5918"]
>>>[Black "crafty"]
>>>[Result "1/2-1/2"]
>>>[ECO "A45"]
>>>[BlackElo "2868"]
>>>[PlyCount "283"]
>>>[EventDate "2005.??.??"]
>>>[TimeControl "300"]
>>>
>>>1. d4 Nf6 2. e3 e6 3. c3 d5 4. Bb5+ Nbd7 5. f4 Bd6 6. Nf3 O-O 7. O-O c5 8. Bd3
>>>Re8 9. Re1 Nb6 10. g3 Bd7 11. Qe2 c4 12. Bc2 Qe7 13. Nbd2 Bc6 14. Nf1 Ba4 15.
>>>Qd1 Bxc2 16. Qxc2 Ne4 17. N3d2 f5 18. Re2 Rf8 19. Rg2 Rf6 20. Nf3 Raf8 21. Qe2
>>>R6f7 22. Bd2 Qf6 23. Be1 Qh6 24. Rc1 Rf6 25. Rc2 Na4 26. Rc1 b5 27. a3 Qh3 28.
>>>Ra1 Rg6 29. Ra2 Rf7 30. Ra1 a5 31. Ra2 Rg4 32. Ra1 Rf8 33. Ra2 Rf6 34. Ra1 Rgg6
>>>35. Ra2 Rf7 36. Ra1 Nb6 37. Ra2 Nd7 38. Ra1 Ndf6 39. Ra2 Rf8 40. Ra1 Ng4 41.
>>>Ra2 Bc7 42. Ra1 Nh6 43. Ra2 Ng4 44. Ra1 Rf7 45. Ra2 Bb6 46. Ra1 Bd8 47. Ra2 Bb6
>>>48. Ra1 Bd8 49. Ra2 Bf6 50. Ra1 Ra7 51. Ra2 Rb7 52. Ra1 Rb8 53. Ra2 Rb7 54. Ra1
>>>Rb8 55. Ra2 Rh6 56. Ra1 Kh8 57. Ra2 Rg6 58. Ra1 Rf8 59. Ra2 Bd8 60. Ra1 Kg8 61.
>>>Ra2 Bb6 62. Ra1 Bc7 63. Ra2 Kh8 64. Ra1 Kg8 65. Ra2 Rh6 66. Ra1 Bd6 67. Ra2 Rf7
>>>68. Ra1 Rg6 69. Ra2 Ngf6 70. Ra1 Rg4 71. Ra2 Bc7 72. Ra1 Rf8 73. Ra2 Bd6 74.
>>>Ra1 Rf7 75. Ra2 Rf8 76. Ra1 Bc7 77. Ra2 Re8 78. Ra1 Bd6 79. Ra2 Rd8 80. Ra1 h6
>>>81. Ra2 Rf8 82. Ra1 Nd7 83. Rc1 Be7 84. Rc2 Bd8 85. N3d2 Bc7 86. Nb1 Bd6 87.
>>>Rc1 Rg6 88. Rc2 Ndf6 89. Rc1 Ng4 90. Rc2 Be7 91. Rc1 Kh7 92. Rc2 Kh8 93. Rc1
>>>Kg8 94. Rc2 Bd6 95. Rc1 Rf7 96. Rc2 Bc7 97. Rc1 Rf8 98. Rc2 Rd8 99. Rc1 Rf8
>>>100. Rc2 Bd6 101. Rc1 h5 102. Nbd2 Rf7 103. Nxe4 fxe4 104. Rc2 Rgf6 105. Rf2
>>>Nxf2 106. Qxf2 Rg6 107. Qg2 Qf5 108. Kh1 Be7 109. Qe2 Rf8 110. Bf2 Rg4 111. Qe1
>>>Bd6 112. h4 Be7 113. Nh2 Rg6 114. Re2 Bd6 115. Rc2 Rb8 116. Rc1 Rf8 117. Rc2
>>>Be7 118. Re2 Qf6 119. Rc2 Qf7 120. Re2 Bd8 121. Rc2 Qf5 122. Re2 Rf7 123. Rc2
>>>Be7 124. Re2 Bd6 125. Rc2 Kh8 126. Re2 Kg8 127. Rc2 Kf8 128. Re2 Bc7 129. Rc2
>>>Kg8 130. Re2 Bd8 131. Rc2 Be7 132. Re2 Rff6 133. Rc2 Kh8 134. Re2 Kg8 135. Rc2
>>>Kh8 136. Re2 Rf8 137. Rc2 b4 138. Qa1 Kg8 139. Re2 Kh8 140. Rc2 Bd6 141. Re2
>>>Be7 142. Rc2 {Game drawn by repetition} 1/2-1/2
>>>
>>>
>>>
>>>Statistics for crafty(C) On for: 50 Idle: 2
>>>
>>> rating [need] win loss draw total best
>>>Wild 2033 [6] 343 108 11 462 2204 (03-Oct-2000)
>>>Bullet 3241 [8] 6675 1468 1088 9231 3286 (27-Dec-2002)
>>>Blitz 2868 61054 18860 15044 94958 3388 (09-Jun-2000)
>>>Standard 2595 5533 3040 2694 11267 2792 (25-Oct-2000)
>>>
>>> 1: Crafty v19.20 (2 cpus)
>>> 2: crafty uses all 3/4/5/6 piece endgame databases, over 150 gigs so far.
>>> 3: Dell Poweredge 2600, 2 x 2.8ghz xeon
>>> 4: hours online: 70,519; percent of life online: 78
>>> 5: hours online: 63,273
>>> 6: anonymous ftp to ftp.cis.uab.edu/pub/hyatt
>>> 7: note that I +noplay computers without notes giving their
>>> hardware/software, or computers that don't play an even number of games.
>>> This also includes anyone matching me while having !computer in their
>>> formula.
>>> 8: computer policy: no more than 4 blitz or 2 standard games by any one
>>> program in Crafty's 20-game history here. No crafty clones whatsoever
>>> (ZIPPYNOPLAYCRAFTY 1 is used all the time)
>>> 9:
>>>10: (2) do not interrupt human games under any circumstances. I am mainly
>>> interested in playing strong humans, but will allow computers to play so long
>>> as they don't monopolize crafty's time. Beware.
>>>
>>>history crafty
>>>6 = 2868 B ? guest5918 5 0 [Blitz u] A45 Rep 2005.06.14 21:35:18
>>>5 - 2868 W 2948 BountyHunter 5 3 [Blitz r] A03 Res 2005.06.14 18:42:01
>>>4 a 2882 W 2608 RingForce 5 0 [Blitz r] A40 Sho 2005.06.14 17:05:12
>>>3 = 2595 W 2681 Somnus 30 30 [Standard r] B89 50 2005.06.14 03:32:35
>>>2 + 2882 B 2860 Excalibar 5 3 [Blitz r] A30 Mat 2005.06.14 03:07:53
>>>1 = 2866 W 2891 Excalibar 5 3 [Blitz r] B18 50 2005.06.14 02:28:11
>>>0 = 2865 B 2892 Excalibar 5 3 [Blitz r] B20 50 2005.06.14 01:58:55
>>>99 + 2864 W 2719 ClubberLang 3 0 [Blitz r] A02 Res 2005.06.14 01:15:44
>>>98 + 2854 W 2704 IMOrlov 5 0 [Blitz r] A02 Res 2005.06.14 00:57:06
>>>97 + 2844 B 2667 ClubberLang 3 0 [Blitz r] A00 Res 2005.06.13 19:54:30
>>>96 + 2835 W 2659 ClubberLang 3 0 [Blitz r] C01 Res 2005.06.13 19:36:21
>>>95 - 2826 B 2994 ajop 5 3 [Blitz r] C41 Res 2005.06.13 18:07:24
>>>94 - 2836 W 2984 ajop 5 3 [Blitz r] B52 Res 2005.06.13 17:55:10
>>>93 + 2846 B 2958 BountyHunter 5 3 [Blitz r] D47 Res 2005.06.13 14:50:43
>>>92 + 2823 W 2474 Crasharin 5 3 [Blitz r] C10 Res 2005.06.13 10:32:52
>>>91 - 2819 B 2986 BountyHunter 5 3 [Blitz r] E97 Res 2005.06.12 22:12:57
>>>90 = 2829 B 2674 SpiderChessX 5 3 [Blitz r] D45 Agr 2005.06.12 20:19:12
>>>89 + 2836 W 2667 SpiderChessX 5 3 [Blitz r] B67 Res 2005.06.12 20:02:52
>>>88 a 2826 B 2507 ELLI 5 3 [Blitz r] A00 Sho 2005.06.12 14:12:06
>>>87 a 2826 B 2484 Lashkhor 5 0 [Blitz r] A00 Sho 2005.06.12 08:01:36
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.