Author: Robert Hyatt
Date: 07:14:29 07/13/99
Go up one level in this thread
On July 13, 1999 at 03:18:51, greg moller wrote:
>On July 13, 1999 at 00:23:03, Robert Hyatt wrote:
>
>>
>>Earlier tonite, Crafty was involved in a match with GM Loek van Wely on
>>ICC. 5 3 time control, and crafty was winning about its normal percentage
>>against him (about 80%, match was 9 wins, 2 losses, 1 draw at the point
>>where the following game was played). In this game, Crafty was
>>in book until move 23. As we passed move 20, I was counting pawns
>>and pieces and concluded white was probably lost. At move 23,
>>Crafty's first search said -4.00, and I thought aha, book learning
>>to the rescue as this is dead lost. But without spoiling anything,
>>this was the result. :)
>
>
>Bob, actually, if the eval said -4.00, how can it be 'book learning to the
>rescue', since the position was (apparently) already lost upon exiting book. I
>don't understand.
>
Book learning does two things.
1. If the score after leaving book is bad, it will adjust book weights so that
opening won't be played again.
2. if the game is lost, it flags the last book move as 'never play this again'
so it won't repeat a losing game multiple times.
In this case it learned the opening was bad, but it actually won...
>regards,
>gm
>>
>>Actually, I will spoil it. At move 40, it announced a mate in 38 moves.
>>
>>:)
>>
>>
>>
>>pgn:
>>
>>[Event "ICC 5 3"]
>>[Site "Internet Chess Club"]
>>[Date "1999.07.13"]
>>[Round "-"]
>>[White "crafty"]
>>[Black "KingLoek"]
>>[Result "1-0"]
>>[WhiteElo "3118"]
>>[BlackElo "2956"]
>>[Opening "Sicilian: Najdorf, Poisoned pawn variation"]
>>[ECO "B97"]
>>[NIC "SI.08"]
>>[Time "00:00:49"]
>>[TimeControl "300+3"]
>>
>>1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Bg5 e6 7. f4 Qb6 8.
>>Qd2 Qxb2 9. Rb1 Qa3 10. f5 Nc6 11. fxe6 fxe6 12. Nxc6 bxc6 13. e5 dxe5 14.
>>Bxf6 gxf6 15. Ne4 Be7 16. Be2 h5 17. Rb3 Qa4 18. Nxf6+ Bxf6 19. c4 Bh4+ 20.
>>g3 Be7 21. O-O h4 22. Bd3 Rg8 23. Qf2 Rg7 24. Bg6+ Rxg6 25. Qf7+ Kd8 26.
>>Rd1+ Bd7 27. Rxd7+ Kxd7 28. Rd3+ Kc7 29. Qxe7+ Kb6 30. Rb3+ Qxb3 31. axb3
>>hxg3 32. c5+ Kb5 33. Qb7+ Kxc5 34. Qxa8 gxh2+ 35. Kxh2 Rg4 36. Qxa6 Rb4 37.
>>Qa5+ Rb5 38. b4+ Kc4 39. Qc7 c5 40. bxc5 Rxc5 41. Qd6 e4 42. Qxe6+ Kd4 43.
>>Qf6+ Kd5 44. Kg3 Rc2 45. Qd8+ Ke5 46. Qh8+ Kd5 47. Qh5+ Kc4 48. Kf4
>>{Black resigns}
>>1-0
>>
>>
>>ICC format with times so that you can see when it went out of book:
>>crafty (3118) vs. KingLoek (2956) --- 1999.07.13 00:00:49
>>Rated blitz match, initial time: 5 minutes, increment: 3 seconds
>>
>> 1. e4 (0:01) c5 (0:02)
>> 2. Nf3 (0:00) d6 (0:01)
>> 3. d4 (0:00) cxd4 (0:01)
>> 4. Nxd4 (0:00) Nf6 (0:00)
>> 5. Nc3 (0:00) a6 (0:02)
>> 6. Bg5 (0:00) e6 (0:01)
>> 7. f4 (0:00) Qb6 (0:01)
>> 8. Qd2 (0:00) Qxb2 (0:01)
>> 9. Rb1 (0:00) Qa3 (0:01)
>> 10. f5 (0:00) Nc6 (0:05)
>> 11. fxe6 (0:00) fxe6 (0:02)
>> 12. Nxc6 (0:00) bxc6 (0:02)
>> 13. e5 (0:00) dxe5 (0:02)
>> 14. Bxf6 (0:00) gxf6 (0:02)
>> 15. Ne4 (0:00) Be7 (0:02)
>> 16. Be2 (0:00) h5 (0:06)
>> 17. Rb3 (0:00) Qa4 (0:08)
>> 18. Nxf6+ (0:00) Bxf6 (0:02)
>> 19. c4 (0:00) Bh4+ (0:18)
>> 20. g3 (0:00) Be7 (0:02)
>> 21. O-O (0:00) h4 (0:19)
>> 22. Bd3 (0:00) Rg8 (1:05)
>> 23. Qf2 (0:17) Rg7 (0:13)
>> 24. Bg6+ (0:16) Rxg6 (0:17)
>> 25. Qf7+ (0:00) Kd8 (0:11)
>> 26. Rd1+ (0:01) Bd7 (0:02)
>> 27. Rxd7+ (0:12) Kxd7 (0:02)
>> 28. Rd3+ (0:11) Kc7 (0:12)
>> 29. Qxe7+ (0:00) Kb6 (0:01)
>> 30. Rb3+ (0:23) Qxb3 (0:03)
>> 31. axb3 (0:09) hxg3 (0:01)
>> 32. c5+ (0:12) Kb5 (0:15)
>> 33. Qb7+ (0:00) Kxc5 (0:01)
>> 34. Qxa8 (0:10) gxh2+ (0:02)
>> 35. Kxh2 (0:09) Rg4 (0:09)
>> 36. Qxa6 (0:11) Rb4 (0:02)
>> 37. Qa5+ (0:11) Rb5 (0:01)
>> 38. b4+ (0:10) Kc4 (0:20)
>> 39. Qc7 (0:11) c5 (0:05)
>> 40. bxc5 (0:11) Rxc5 (0:02)
>> 41. Qd6 (0:00) e4 (0:58)
>> 42. Qxe6+ (0:00) Kd4 (0:01)
>> 43. Qf6+ (0:00) Kd5 (0:03)
>> 44. Kg3 (0:00) Rc2 (0:19)
>> 45. Qd8+ (0:00) Ke5 (0:03)
>> 46. Qh8+ (0:00) Kd5 (0:03)
>> 47. Qh5+ (0:00) Kc4 (0:07)
>> 48. Kf4 (0:00)
>> {Black resigns} 1-0
>>~
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.