Author: Robert Hyatt
Date: 00:58:40 07/28/98
Go up one level in this thread
On July 27, 1998 at 23:27:24, blass uri wrote:
>
>On July 27, 1998 at 19:07:42, Robert Hyatt wrote:
>
>>On July 27, 1998 at 18:19:07, Mark Young wrote:
>>
>>>Crafty 15.18 running on 4 Ppro 200 512mb ram.
>>>
>>>Fritz 5 running on a P II 300 64 mb ram.
>>>
>>>Time control for game 1 was 120 min. with 180 sec./incs.
>>>
>>
>>
>>this was an interesting game. I didn't get to watch much of it, but I
>>did get to see a few key points. Around move 40, crafty was at +1.0,
>>fritz was at -.75 (it thought white was better) yet material was even.
>>
>>I looked at this myself and said "uh-oh" because I think both programs
>>were miscalculating. I think the advantage for white was based on black
>>having 3 isolated pawns, while white's pawns were at a2/b2/c2. And black's
>>king was exposed.
>>
>>however, from a human's perspective, I told Mark I liked black, even though
>>both program thought white was much better. I based this on several things:
>>
>>black had the bishop pair, and they made it very difficult for white to develop
>>any sort of attack. Black had a passer at h7 and e5, and I thought that the h
>>pawn (distant passer with kings on the queen-side) might be an endgame problem.
>>
>>It seems my judgement was better, because Crafty tried quite dilligently to
>>whip up an attack, but the two bishops could just cover all the key squares >and prevent it, and in the end, the passed h pawn told the tale.
>
>I think the losing mistake of crafty was 54.Qxf5.
>Junior5 and fritz5 want to play 54.Qxf5 but reject it after a short time.
>They prefer 54.Qc6.
>The evaluation of Qc6 after 10 minutes is 0.00 by both programs.
>
If that is true, this might have been caused by me. Crafty was playing on
my ALR, but only using 2 cpus most of the day, as I was working on a kernel
problem and doing a lot of compiles. A couple of times I totally forgot about
the game in progress and would build a new kernel telling the make to use all
4 cpus. That would slow Crafty to a crawl for several minutes. I'll look at
the log file to see if I caused it to overlook this. It was spending about 10
minutes per move normally, and in most cases searched at least 12 plies every
time, often more...
unless I happened to hit on the machine while it was on a key move...
>>
>>Interesting game, however, and it would be interesting to see how other >programs evaluate things around move 36 or so.
>
>Junior5 evaluated the position at move 36 as 0.89 pawns advantage for white
>at depth 16 and wanted to play 36.Bd3
>
>Uri
>> Both thought white was better, but it had
>>the look of Rebel-Anand game 8 to me... the two bishops looked very strong. And
>>I have code for two bishops and bishop mobility so Crafty isn't exactly a dummy
>>here... :)
>>
>>
>>
>>>
>>>Game 1
>>>[Event "?"]
>>>[Site "?"]
>>>[Date "????.??.??"]
>>>[Round "?"]
>>>[White "Crafty 15.18 4xPro200 512m ra"]
>>>[Black "Fritz 5 P II 300 64 mb ram."]
>>>[Result "0-1"]
>>>[ECO "B99"]
>>>[Annotator "Young,M"]
>>>[PlyCount "168"]
>>>
>>>{46144kB, f5book.ctg
>>>} 1. e4 {0} 1... c5 {0} 2. Nf3 {12} 2... d6 {0} 3. d4 {8}
>>>3... cxd4 {0} 4. Nxd4 {10} 4... Nf6 {0} 5. Nc3 {7} 5... a6 {0} 6. Bg5 {6} 6...
>>>e6 {0} 7. f4 {7} 7... Be7 {0} 8. Qf3 {7} 8... Qc7 {0} 9. O-O-O {9} 9... Nbd7 {0
>>>} 10. g4 {7} 10... b5 {0} 11. Bxf6 {7} 11... Nxf6 {0} 12. g5 {6} 12... Nd7 {0}
>>>13. f5 {6} 13... Nc5 {0} 14. f6 {7} 14... gxf6 {0} 15. gxf6 {7} 15... Bf8 {0}
>>>16. Qh5 {8} 16... Bd7 {0} 17. Bh3 {6} 17... b4 {0} 18. Nce2 {6} 18... O-O-O {0}
>>>19. Qxf7 {8} 19... Bh6+ {0} 20. Kb1 {6} 20... Rdf8 {0} 21. Qh5 {7} 21... Rxf6 {
>>>0} 22. Rhf1 {9} 22... Rhf8 {0} 23. Rxf6 {7} 23... Rxf6 {0.16/12 233} 24. Qh4 {9
>>>} 24... Rg6 {0.34/13 809} 25. Nf3 {11} 25... Bf8 {0.47/12 991} 26. e5 {799}
>>>26... dxe5 {0.25/13 1312} 27. Qxb4 {9} 27... e4 {0.44/12 771} 28. Nd2 {9} 28...
>>>Qxh2 {0.59/12 1054} 29. Bf1 {9} 29... Qc7 {0.28/12 862} 30. Nb3 {126} 30... e5
>>>{0.31/13 546} 31. Nxc5 {17} 31... Bxc5 {0.28/13 516} 32. Qxe4 {57} 32... Kb8 {
>>>0.50/12 711} 33. Nc1 {14} 33... Bc6 {0.47/12 598} 34. Qf5 {607} 34... a5 {
>>>0.63/12 43} 35. Nb3 {589} 35... Bb4 {0.44/13 693} 36. c3 {1353} 36... Bd6 {
>>>0.69/12 387} 37. Nd2 {661} 37... Qg7 {0.41/11 404} 38. Qf2 {1588} 38... Bc7 {
>>>0.34/13 594} 39. Bd3 {689} 39... Rd6 {0.28/12 1} 40. Rg1 {510} 40... Qh6 {
>>>0.47/12 594} 41. Qe2 {0} 41... Rd7 {0.31/11 469} 42. Nc4 {663} 42... Rg7 {
>>>0.19/12 305} 43. Rf1 {814} 43... Rg2 {0.22/12 357} 44. Qd1 {57} 44... e4 {
>>>0.38/12 484} 45. Bc2 {12} 45... Rh2 {0.41/11 232} 46. Ba4 {219} 46... Bb7 {
>>>0.44/12 227} 47. Qd7 {636} 47... Rh1 {0.25/13 362} 48. Rxh1 {89} 48... Qxh1+ {
>>>0.06/13 79} 49. Kc2 {344} 49... Qg2+ {0.00/13 280} 50. Kb3 {136} 50... Qg6 {
>>>0.13/14 274} 51. Bb5 {445} 51... Bc8 {0.16/13 89} 52. Qe7 {1079} 52... Qe6 {
>>>0.13/13 243} 53. Qc5 {480} 53... Qf5 {0.06/15 1} 54. Qxf5 {430} 54... Bxf5 {
>>>-0.63/15 1} 55. Ne3 {459} 55... Bg6 {-0.41/13 30} 56. Bc4 {404} 56... h5 {
>>>-0.63/14 725} 57. Ng2 {11} 57... Bg3 {-0.91/13 321} 58. Ka4 {393} 58... h4 {
>>>-0.81/13 268} 59. Be6 {69} 59... Bh5 {-1.00/13 294} 60. Kxa5 {72} 60... Bf3 {
>>>-1.84/13 305} 61. Ne3 {381} 61... Bf2 {-2.13/13 42} 62. Nc2 {958} 62... e3 {
>>>-2.13/13 326} 63. Nd4 {348} 63... Bg2 {-2.72/13 1} 64. a4 {324} 64... h3 {
>>>-3.09/14 103} 65. Bxh3 {220} 65... Bxh3 {-3.50/13 1} 66. Kb4 {321} 66... Bf1 {
>>>-3.66/13 242} 67. a5 {333} 67... e2 {-4.00/14 1} 68. Nxe2 {375} 68... Bxe2 {
>>>-4.25/14 0} 69. Ka4 {791} 69... Kb7 {-4.75/14 231} 70. b4 {869} 70... Kc6 {
>>>-6.22/16 1} 71. Kb3 {487} 71... Kb5 {-5.16/13 59} 72. Kc2 {246} 72... Be1 {
>>>-6.22/15 160} 73. Kb3 {93} 73... Bc4+ {-6.34/14 68} 74. Kc2 {275} 74... Ka4 {
>>>-6.31/15 1} 75. a6 {256} 75... Bxa6 {-6.50/14 82} 76. Kb2 {175} 76... Bd3 {
>>>-6.59/15 108} 77. b5 {142} 77... Bxb5 {-#9/14 29} 78. Kc2 {221} 78... Ka3 {
>>>-#8/14 0} 79. c4 {254} 79... Bxc4 {-#8/12 20} 80. Kd1 {10} 80... Bc3 {-#6/10 0}
>>>81. Kc2 {14} 81... Kb4 {-#5/9 0} 82. Kd1 {10} 82... Kb3 {-#6/8 2} 83. Kc1 {9}
>>>83... Be2 {-#5/6 0} 84. Kb1 {9} 84... Bd4 {-#4/4 0} 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.