Author: Robert Hyatt
Date: 06:35:21 10/06/03
Go up one level in this thread
On October 05, 2003 at 04:55:56, scott farrell wrote:
>Here is a position that I am looking at, it actually arose between crafty
>(khmcrafty on icc) and chompster (chompster is white on the receiving end of
>what is a whooping, but I thought might have been blocked earlier).
>
>this is an interesting position, as it is full of zugs as far as I can see.
>
>[d]8/7p/4k1pP/3p1pP1/1p1B1P2/pP2P3/P2qb3/K6Q w - - 0 84
>
>
>my windows crafty.exe v19.3 (downloaded as compiled from Robert's web site) cant
>solve it, and crafty under linux v19.3 can solve it rather quickly, as does my
>chompster outside the game time controls.
>
>The linux crafty fails high on the right move in 19 seconds, and after 20
>minutes the windows machines cant resolve the wrong fail high move.
This is not unusual. What happens is that there are _many_ wild tactical
wins and losses here, but the aspiration window allows most of them to be
quickly dismissed as fail high or fail low. But once a fail high at the
root happens, the upper (beta) bound is relaxed, and now it has to carefully
follow all those tactical lines to their conclusion.
It would be better were I using an older approach where on a fail high I don't
instantly jump to +infinity, and instead to go +1, then +5, which would let
it see a win of a pawn without having to stomp through all the forced mates.
I did that in Cray Blitz and in early Crafty versions, but for some reason I
took it out years ago. It will be "back in" before too long.
>
>this is from my linux machine, athlon xp3200, 512 Mb DDR400, redhat 9.0, crafty
>compiled linux 686-elf.
>
>EPD Kit revision date: 1996.04.21
>unable to open book file [./book.bin].
>book is disabled
>unable to open book file [./books.bin].
>hash table memory = 192M bytes.
>pawn hash table memory = 6M bytes.
>EGTB cache memory = 256M bytes.
>draw score set to 0.75 pawns.
> game/2 minutes primary time control
>threshold set to 20 pawns.
>choose from book moves randomly (using weights.)
>choose from 3 best moves.
>show book statistics
>display extended statistics at end of each search.
>
>
>Crafty v19.3 (1 cpus)
>White(1): setboard 8/7p/4k1pP/3p1pP1/1p1B1P2/pP2P3/P2qb3/KQ6 b - - 0 1
> puzzling over a move to ponder.
>[snip]
>Black(1): analyze
>Analyze Mode: type "exit" to terminate.
> clearing hash tables
> time limit 3.00 (21.00)
> depth time score variation (1)
> 9 0.07 -0.91 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qh1 Kc6 5. Qg1 Kc7 <HT>
> 9-> 0.07 -0.91 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qh1 Kc6 5. Qg1 Kc7 <HT>
> 10 0.12 -0.92 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bf6 Bc2 7. Be5
> 10-> 0.16 -0.92 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bf6 Bc2 7. Be5
> 11 0.21 -0.89 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Bh8 Kc6 6. Be5 Bc2 <HT>
> 11-> 0.29 -0.89 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Bh8 Kc6 6. Be5 Bc2 <HT>
> 12 0.39 -0.92 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bd4 Kd6 7. Bh8
> Kc6 8. Be5 Bc2 <HT>
>12-> 0.48 -0.92 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bd4 Kd6 7. Bh8
> Kc6 8. Be5 Bc2 <HT>
> 13 0.66 -0.92 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bd4 Kd6 7. Bh8
> Kc6 8. Be5 Bc2 <HT>
> 13-> 0.69 -0.92 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bd4 Kd6 7. Bh8
> Kc6 8. Be5 Bc2 <HT>
> 14 0.89 -0.92 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bd4 Kd6 7. Bh8
> Kc6 <HT>
> 14-> 0.94 -0.92 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bd4 Kd6 7. Bh8
> Kc6 <HT>
> 15 6.16 ++ 1. ... Kd6!!
> 15 14.60 -4.12 1. ... Bd3 2. Qg1 Bc2 3. e4 dxe4 4.
> Be5 e3 5. Qh1 Be4 6. Qg1 Qc2 7. Bf6
> Qf2 8. Qd1 Qxf4
> 15 18.81 -13.82 1. ... Bd1 2. Be5 Bc2 3. Qf1 Kf7 4.
> e4 Bd1 5. Qd3 Qxd3 6. exf5 Qxf5 7.
> Bf6 Qxf4 8. Kb1 Qxf6
> 15-> 18.81 -13.82 1. ... Bd1 2. Be5 Bc2 3. Qf1 Kf7 4.
> e4 Bd1 5. Qd3 Qxd3 6. exf5 Qxf5 7.
> Bf6 Qxf4 8. Kb1 Qxf6
> 16 19.38 ++ 1. ... Bd1!!
> 16 1:23 -Mat08 1. ... Bd1 2. Be5 Bc2 3. Qf1 Kf7 4.
> e4 Bd1 5. Qd3 Qxd3 6. exf5 Qc2 7. fxg6+
> hxg6 8. f5 Qc1#
> 16-> 1:25 -Mat08 1. ... Bd1 2. Be5 Bc2 3. Qf1 Kf7 4.
> e4 Bd1 5. Qd3 Qxd3 6. exf5 Qc2 7. fxg6+
> hxg6 8. f5 Qc1#
> time=1:25 cpu=100% mat=0 n=113660372 fh=98% nps=1335k
> ext-> chk=14178846 cap=389715 pp=944875 1rep=1766151 mate=92760
> predicted=0 nodes=113660372 evals=5129574
> endgame tablebase-> probes done=0 successful=0
> SMP-> split=0 stop=0 data=0/16 cpu=1:25 elap=1:25
>
>
>this is crafty.exe on my celery1200 windows xp 512Mb RAM.
>
>Crafty v19.3 (1 cpus)
>
>White(1): setboard 8/7p/4k1pP/3p1pP1/1p1B1P2/pP2P3/P2qb3/KQ6 b - - 0 1
> puzzling over a move to ponder.
>end-game phase
> clearing hash tables
> depth time score variation (1)
>Black(1): Kd6 [pondering]
> clearing hash tables
> time limit 3.00 (21.00)
> depth time score variation (1)
>analyze
>Black(1): analyze
>Analyze Mode: type "exit" to terminate.
> clearing hash tables
> time limit 3.00 (21.00)
> depth time score variation (1)
> 9 0.15 -1.00 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bf6 Kb5 7. Bd4
> Kc6 8. Qh1 Kd6 <HT>
> 9-> 0.19 -1.00 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc7 5. Be5+ Kc6 6. Bf6 Kb5 7. Bd4
> Kc6 8. Qh1 Kd6 <HT>
> 10 0.28 -1.00 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc6 5. Be5 Bc2 6. Bd4 Bd3
> 10-> 0.31 -1.00 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Bd3 4.
> Qg1 Kc6 5. Be5 Bc2 6. Bd4 Bd3
> 11 0.55 ++ 1. ... Kd6!!
> 11 0.80 -1.81 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Kc6 4.
> Be5 Qxe3 5. Qc2+ Kb5 6. Bf6 Bd3 7.
> Qd1 d4
> 11-> 0.84 -1.81 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Kc6 4.
> Be5 Qxe3 5. Qc2+ Kb5 6. Bf6 Bd3 7.
> Qd1 d4
> 12 1.25 -1.85 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Kc6 4.
> Be5 Qxe3 5. Qc2+ Kb5 6. Bf6 Qg1+ 7.
> Qb1 Qc5 8. Qe1 Bd3
> 12-> 1.40 -1.85 1. ... Kd6 2. Be5+ Kd7 3. Bd4 Kc6 4.
> Be5 Qxe3 5. Qc2+ Kb5 6. Bf6 Qg1+ 7.
> Qb1 Qc5 8. Qe1 Bd3
> 13 2.64 -2.11 1. ... Kd6 2. Be5+ Kc5 3. Bd4+ Kc6
> 4. Be5 Qxe3 5. Qc2+ Kb5 6. Bf6 d4 7.
> Be5 Qg1+ 8. Qb1 Qf2 9. Qc2 d3
> 13-> 2.71 -2.11 1. ... Kd6 2. Be5+ Kc5 3. Bd4+ Kc6
> 4. Be5 Qxe3 5. Qc2+ Kb5 6. Bf6 d4 7.
> Be5 Qg1+ 8. Qb1 Qf2 9. Qc2 d3
> 14 3.63 -1.99 1. ... Kd6 2. Be5+ Kc5 3. Bd4+ Kc6
> 4. Be5 Qxe3 5. Qc2+ Kb5 6. Bg7 Qg1+
> 7. Qb1 Qc5 8. Qe1 Qe3 9. Be5 d4 10.
> 15 4.88 -2.05 1. ... Kd6 2. Be5+ Kc5 3. Bd4+ Kc6
> 4. Be5 Qxe3 5. Qc2+ Kb5 6. Bg7 Kb6
> 7. Be5 Qg1+ 8. Qb1 Qc5 9. Qe1 d4 10.
> Qd2 d3
> 15 8.67 ++ 1. ... Bd3!!
> 15 13:30 1/23? 1. ... Bd3
>
>(that's thirteen minutes, and it cant resolve the fail high, and doesnt see Bd1
>yet.
>
>here is the whole game:
>
>[Event "ICS Rated Standard match"]
>[Site "chessclub.com"]
>[Date "2003.10.05"]
>[Round "-"]
>[White "sfarrellc"]
>[Black "khmcrafty"]
>[Result "0-1"]
>[WhiteElo "2292"]
>[BlackElo "2308"]
>[TimeControl "900+2"]
>
>1. Nf3 Nf6 2. c4 e6 3. Nc3 d5 4. d4 Be7 5. Bf4 O-O 6. e3 c6 7. Qc2 Nbd7 8.
>cxd5 exd5 9. Bd3 Re8 10. h3 Nf8 11. O-O-O Be6 12. g4 Rc8 13. g5 N6d7 14.
>Kb1 c5 15. Ka1 Ng6 16. Bxg6 fxg6 17. Rc1 Qb6 18. h4 cxd4 19. Nxd4 Bf7 20.
>Qe2 a6 21. Qf3 Qc5 22. Nb3 Qc4 23. Rhd1 Nb6 24. Rd2 Qc6 25. Nd4 Qd7 26.
>Rdc2 a5 27. Ncb5 Rxc2 28. Rxc2 Rc8 29. Rc7 Rxc7 30. Bxc7 Nc4 31. Bh2 Qc8
>32. Bf4 Qc5 33. Nc3 a4 34. Qe2 a3 35. b3 Nd6 36. Qc2 Nf5 37. Nxf5 gxf5 38.
>Be5 b5 39. Qd3 b4 40. Ne2 Bg6 41. Qd2 Be8 42. Nf4 Bf7 43. Ne2 Qb5 44. Qe1
>Qe8 45. Nd4 Bf8 46. f4 Qc8 47. Qd2 Bh5 48. Nb5 Bf3 49. Nc7 Qd7 50. Qc1 Be7
>51. Qc2 Bf8 52. Qc1 Be7 53. Na8 Bd8 54. Nc7 Kf7 55. Na6 Be7 56. Nb8 Qe8 57.
>Nc6 Bf8 58. Bc7 Kg8 59. Ba5 Qe4 60. Ne5 Be2 61. Bb6 Bh5 62. Bd4 Be8 63. Qd1
>Bb5 64. Qg1 Bd6 65. h5 Be7 66. Qc1 Be8 67. h6 g6 68. Qd1 Bb5 69. Qg1 Bd6
>70. Qd1 Qg2 71. Qc1 Bf8 72. Kb1 Qe4+ 73. Ka1 Be8 74. Qg1 Bd6 75. Qe1 Qc2
>76. Qh1 Kf8 77. Qe1 Ke7 78. Qf1 Bxe5 79. Bxe5 Qd2 80. Bf6+ Kf7 81. Bd4 Bb5
>82. Qg1 Be2 83. Qh1 Ke6 84. Qb1 Kd6 85. Bf6 Qxe3 86. Be5+ Kc5 87. Qc2+ Kb5
>88. Qb1 Qd2 89. Qg1 Kc6 90. Kb1 Bd1 91. Qg4
>{sfarrellc resigns} 0-1
>
>Scott
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.