Crafty 17.10 - Yace 0.15 G/60 with surprising result

[Dieter Bürßner]

Geschrieben von: / Posted by: Dieter Bürßner at 14 June 2000 13:04:52:
Tonight I had a lucky stroke against crafty. (I should have played lotto;)
Crafty 17.10 vs. Yace 0.15 1.5-8.5 (3 Draws) Game/60
Hash: Crafty hash 6M, hashp 4M, Yace: 10M (I know, that for crafty 8M, 2M
would be better, but it's impossible). Crafty learn off, no endgame
tables (haven't installed them yet). Book: both programs used 2600.pgn
from Dann Corbit's site with maxply 20. AMD K6-2 300 MHz.
I don't have the expertise to judge the appended games. From looking at
the logs, round 3 was interesting. Crafty came out of the book with
+1.11 and saw the mate when it resigned.
I made a few changes to Yace, that may help for longer time controls.
I'll ask Dann to make the new version available soon.
Best wishes,
Dieter "proud" Buerssner
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.13"]
[Round "1"]
[White "wcrafty-17.10.exe"]
[Black "yace.exe"]
[Result "1/2-1/2"]
[TimeControl "3600"]
1. e4 e6 2. d4 d5 3. Nd2 c5 4. exd5 Qxd5 5. Ngf3 cxd4 6. Bc4 Qd6 7. O-O Nf6
8. Nb3 Nc6 9. Nbxd4 Nxd4 10. Nxd4 a6 11. Be3 Qc7 12. Qe2 Bd6 13. h3 O-O 14.
Rad1 Nd5 15. Bxd5 exd5 16. Qf3 Be6 17. Rfe1 Rfe8 18. Re2 Rac8 19. c3 Qc4
20. a3 Bd7 21. Rde1 Re4 22. Nf5 Bxf5 23. Qxf5 Rce8 24. Qd7 Qc6 25. Qxc6
bxc6 26. f3 R4e6 27. Kf2 c5 28. f4 f5 29. Kf3 Re4 30. g3 c4 31. Rd1 Bc5 32.
Rde1 Bxe3 33. Rxe3 h5 34. a4 g6 35. R1e2 R8e7 36. Re1 Kf7 37. R1e2 Ke6 38.
Re1 Kd6 39. h4 Kc6 40. Kf2 Kc5 41. Rxe4 fxe4 42. Ke3 Rf7 43. Rd1 Rb7 44.
Rd2 Rb3 45. Kf2 a5 46. Ke3 Rb8 47. Rg2 Re8 48. g4 hxg4 49. Rxg4 Re6 50. Rg2
Rf6 51. Rg1 Rb6 52. Rg2 Rc6 53. Rg4 Ra6 54. Rg5 Rd6 55. Rg4 Re6 56. Rg5 Rf6
57. Rg1 Rb6 58. Rg2 Rd6 59. Rg5 Re6 60. Rg1 Kd6 61. Rg5 Kc5 62. Rg1 Rd6 63.
Rg5 Rf6 64. Rg1
{Drawn by 3-fold repetition} 1/2-1/2
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.13"]
[Round "2"]
[White "yace.exe"]
[Black "wcrafty-17.10.exe"]
[Result "1-0"]
[TimeControl "3600"]
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 O-O 5. Bg5 c5 6. dxc5 h6 7. Bh4 Na6 8.
O-O-O Nxc5 9. Nb5 Nce4 10. f3 Ng5 11. e4 a6 12. Qa4 Bc5 13. e5 Ne8 14. Nh3
d5 15. Nc3 Qc7 16. cxd5 Qxe5 17. Nxg5 hxg5 18. Bg3 Qe3+ 19. Kb1 exd5 20.
Rxd5 Qe7 21. h4 b5 22. Bxb5 Nc7 23. Bxc7 axb5 24. Qxb5 Qxc7 25. Rxc5 Qa7
26. Rxg5 Be6 27. a4 f6 28. Rc5 Rfb8 29. Qc6 Rb6 30. Qe4 Bb3 31. Nd5 Rb7 32.
Rhc1 Rf8 33. a5 Qa6 34. Ne7+ Kf7 35. Re1 Rd8 36. Nc6 Rh8 37. Rc3 Ra8 38.
Nb4 Rxb4 39. Qxb4 Bd5 40. Rc7+ Kg8 41. Qg4 Bf7 42. Qd7 Rf8 43. b3 f5 44.
Re7 Qf1+ 45. Rc1 Qa6 46. h5 f4 47. Rc7 Qf1+ 48. Kb2 Qxg2+ 49. Ka3 Qg5 50.
Rxf7 Qxa5+ 51. Kb2 Qe5+ 52. Kb1 Qe1+ 53. Rc1 Rxf7 54. Qc8+ Rf8 55. Qc4+ Kh8
56. Rxe1 Kh7 57. Qe4+
{Black resigns} 1-0
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.13"]
[Round "3"]
[White "wcrafty-17.10.exe"]
[Black "yace.exe"]
[Result "0-1"]
[TimeControl "3600"]
1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. Nc3 c6 5. Bg5 dxc4 6. e4 b5 7. e5 h6 8. Bh4
g5 9. Nxg5 hxg5 10. Bxg5 Nbd7 11. Qf3 Bb7 12. exf6 Qa5 13. h4 O-O-O 14. Qd1
c5 15. a4 b4 16. Nb1 Ne5 17. d5 Bxd5 18. Bd2 Qc7 19. f4 Ng6 20. Qc1 Bh6 21.
Bxc4 Nxf4 22. Kf2 Rhg8 23. Rh2 Bxg2 24. Ba6+ Kb8 25. Rxg2 Rxd2+ 26. Qxd2
Nxg2 27. Qd3 Be3+ 28. Ke2 Nf4+ 29. Kxe3 Qe5+ 30. Qe4 Rg3+ 31. Kd2
{White resigns} 0-1
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.13"]
[Round "4"]
[White "yace.exe"]
[Black "wcrafty-17.10.exe"]
[Result "1-0"]
[TimeControl "3600"]
1. e4 c5 2. Nc3 Nc6 3. f4 g6 4. Nf3 Bg7 5. Bb5 Nd4 6. Bd3 e6 7. Nxd4 cxd4
8. Nb5 d6 9. c3 Nf6 10. Nxd4 e5 11. Ne2 O-O 12. O-O Qb6+ 13. Kh1 exf4 14.
Nxf4 Re8 15. Nd5 Nxd5 16. exd5 Qc7 17. Qf3 Re7 18. Bc2 h6 19. d4 b6 20. Bd3
Bb7 21. Bd2 a6 22. h4 Rae8 23. h5 g5 24. Qf5 Kf8 25. Bxg5 hxg5 26. Qh7 Re3
27. h6 Bf6 28. Rxf6 Bxd5 29. Raf1 Rc8 30. Rxf7+ Qxf7 31. Rxf7+ Bxf7 32.
Qh8+ Ke7 33. Qxc8 Re1+ 34. Kh2 Re6
{Black resigns} 1-0
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.14"]
[Round "5"]
[White "wcrafty-17.10.exe"]
[Black "yace.exe"]
[Result "1/2-1/2"]
[TimeControl "3600"]
1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. e3 g6 5. Nf3 Bg7 6. Be2 O-O 7. O-O e6 8.
Qc2 Nbd7 9. Rd1 b6 10. e4 Nxe4 11. Nxe4 dxe4 12. Qxe4 Bb7 13. Bg5 Qb8 14.
h3 Re8 15. Qg4 c5 16. Qh4 Qc7 17. Rd2 e5 18. d5 e4 19. d6 Qc6 20. Ne1 Bd4
21. Rad1 Re6 22. Qg3 Rae8 23. Bg4 e3 24. Rxd4 exf2+ 25. Kxf2 cxd4 26. Bxe6
fxe6 27. Rxd4 e5 28. Rd2 Nc5 29. Kg1 Ne4 30. d7 Nxg3 31. dxe8=Q+ Qxe8 32.
Rd8 Qxd8 33. Bxd8 e4 34. Kf2 Nf5 35. g4 Nd4 36. Ke3 Ne6 37. Bh4 Ba6 38. b3
Bb7 39. Nc2 h5 40. Na3 hxg4 41. hxg4 Nc7 42. Bg3 Na6 43. Nb5 Bc8 44. g5 Bf5
45. Nxa7 Nb4 46. Nb5 Nxa2 47. Bc7 Nc1 48. Bxb6 Nxb3 49. c5 Nc1 50. c6 Nd3
51. c7 Nb4 52. Nc3 Kf7 53. Bc5 Na6 54. Bd6 Bc8 55. Na4 Ke6 56. Bg3 Kf5 57.
Nb6 Bb7 58. c8=B+ Bxc8 59. Nxc8 Kxg5 60. Kxe4 Kf6 61. Kf4 g5+ 62. Kg4 Nb4
63. Nb6 Nd3 64. Nd7+ Ke7 65. Nb6 Kf6 66. Nc4 Kg6 67. Nd6 Nb4 68. Be5 Nd5
69. Bd4 Nf6+ 70. Kg3 Nh5+ 71. Kf3 Nf6 72. Ba7 Kh5 73. Be3 g4+ 74. Kg3 Nd5
75. Bd2 Nc7 76. Bc3 Nd5 77. Ba1 Kg5 78. Bd4 Nf6 79. Be3+ Kh5 80. Bd2 Nd7
81. Ba5 Nf6 82. Bd8 Kg6 83. Kf4 Ng8 84. Nc4 Nh6 85. Bb6 Kf6 86. Bd4+ Ke6
87. Ne3 g3 88. Ba1 Ng8 89. Kxg3 Ne7 90. Kf3 Nf5 91. Ng2 Nd6 92. Ke3 Nf5+
93. Ke4 Nd6+ 94. Kd4 Nf5+ 95. Kc5 Nd6 96. Ne3 Ne4+ 97. Kc6 Nf6 98. Bd4 Nd7
99. Bh8 Nf6 100. Nc4 Kf7 101. Kc5 Ke6 102. Na5 Ne4+ 103. Kd4 Ng5 104. Nb7
Nf3+ 105. Kc5 Ne5 106. Bg7 Nf7 107. Kd4 Ng5 108. Kc5 Ne4+ 109. Kc6 Nf6 110.
Bh8 Nd7 111. Nd8+ Ke7 112. Nb7 Ke6 113. Bc3 Nf6 114. Bb2 Ne4 115. Nd8+ Ke7
116. Nb7 Ke6 117. Bd4 Nf6 118. Ba1 Nd7 119. Na5 Ne5+ 120. Kc5 Nd3+ 121. Kc4
Ne5+ 122. Kc5 Nd3+ 123. Kc4 Ne5+ 124. Kd4 Nf3+ 125. Kc5 Ng5 126. Nb7 Ne4+
127. Kd4 Nd2 128. Bc3 Nf3+ 129. Kc4 Ne5+ 130. Kc5 Nd7+ 131. Kd4 Ne5 132.
Nc5+ Kd6 133. Na6 Ke6 134. Ke4 Nd7 135. Bb2 Nf6+ 136. Kf3 Nd7 137. Ke4 Nf6+
138. Kf3 Nd7 139. Kg4
{Draw by 50 moves rule} 1/2-1/2
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.14"]
[Round "6"]
[White "yace.exe"]
[Black "wcrafty-17.10.exe"]
[Result "1-0"]
[TimeControl "3600"]
1. e4 e5 2. Nf3 f5 3. exf5 d6 4. d4 e4 5. Qe2 Nf6 6. Nc3 d5 7. Ne5 c5 8.
dxc5 Qc7 9. Bf4 Qxc5 10. Bg5 Be7 11. O-O-O O-O 12. Bxf6 Bxf6 13. Rxd5 Qe7
14. Qc4 Kh8 15. Qxe4 Bxf5 16. Qe3 Nd7 17. Nf3 Qxe3+ 18. fxe3 Bxc3 19. bxc3
Nb6 20. Ra5 Be4 21. Ng5 Bc6 22. e4 Rae8 23. Rxa7 h6 24. Nf3 Rxe4 25. Bd3
Ree8 26. Re1 Rxe1+ 27. Nxe1 Kg8 28. Kb2 Rf2 29. h3 Kf7 30. Ra5 Kf6 31. Kb3
g6 32. a4 Rd2 33. Ra7 Nc8 34. Ra8 Nd6 35. c4 Nf7 36. a5 Rd1 37. Nf3 Ra1 38.
c5 Ne5 39. Nxe5 Kxe5 40. Kb2 Ra4 41. g3 g5 42. a6 bxa6 43. Rh8 Ra5 44. Rxh6
Rxc5 45. Bxa6 Kd4 46. Rd6+ Ke3 47. Rf6 Bg2 48. Bf1 Rc6 49. Rf8 Bxf1 50.
Rxf1 Rb6+ 51. Kc3 Rc6+ 52. Kb3 Rb6+ 53. Ka4 Rb2 54. Rf5 g4 55. h4 Rxc2 56.
Rf4 Rc6 57. Rxg4 Re6 58. h5 Rh6 59. Rg5 Ke4 60. g4 Rh7 61. Rg6 Ra7+ 62. Kb5
Rb7+ 63. Ka5 Rb2 64. h6 Ra2+ 65. Kb4 Rb2+ 66. Kc3 Rh2 67. g5 Rh3+ 68. Kc4
Rh2 69. Rg7 Kf5 70. h7 Rxh7 71. Rxh7 Kg6 72. Re7 Kxg5 73. Re5+ Kg4 74. Kd3
Kf4 75. Rb5 Kf3 76. Rb4 Kf2 77. Rf4+ Kg2 78. Ke3 Kg3 79. Re4 Kg2 80. Rg4+
Kh3 81. Kf3 Kh2 82. Kf2 Kh3 83. Re4
{Black resigns} 1-0
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.14"]
[Round "7"]
[White "wcrafty-17.10.exe"]
[Black "yace.exe"]
[Result "1/2-1/2"]
[TimeControl "3600"]
1. e4 e5 2. Bc4 Nf6 3. d3 Nc6 4. Nc3 Bc5 5. Bg5 h6 6. Bh4 Na5 7. Bb3 Nxb3
8. axb3 a6 9. Nf3 g5 10. Bg3 d6 11. d4 exd4 12. Nxd4 O-O 13. O-O Bd7 14.
Nf5 Bxf5 15. exf5 b5 16. Re1 Qd7 17. Nd5 Nxd5 18. Qxd5 Rfe8 19. Rxe8+ Rxe8
20. c3 Re2 21. b4 Qe8 22. Kf1 Bb6 23. Rb1 Kg7 24. Qf3 Rd2 25. f6+ Kg6 26.
h4 Qe6 27. h5+ Kh7 28. Re1 Qc4+ 29. Kg1 Qd3 30. Qxd3+ Rxd3 31. Kh2 Rd5 32.
Re7 g4 33. Rxf7+ Kh8 34. Bh4 Rxh5 35. Kg3 Bxf2+ 36. Kxf2 Rxh4 37. Kg3 Rh5
38. Kf4 Kg8 39. Rxc7 Rd5 40. Rg7+ Kf8 41. Kxg4 h5+ 42. Kf4 Rd2 43. Kf5 Rf2+
44. Ke6 Re2+ 45. Kxd6 Rxb2 46. g3 Rf2 47. Ke5 Rf3 48. Rg5 Rxc3 49. Rxh5
Rxg3 50. Ke6 Re3+ 51. Re5 Rf3 52. Rd5 Re3+ 53. Kf5 Rf3+ 54. Kg6 Rg3+ 55.
Rg5 Rb3 56. Rg4 Rb1 57. Rd4 Rg1+ 58. Kf5 Rf1+ 59. Ke5 Re1+ 60. Re4 Rc1 61.
Rh4 Re1+ 62. Re4 Rc1 63. Ke6 Rf1 64. Rh4 Re1+ 65. Kf5 Rf1+ 66. Kg5 Rg1+ 67.
Rg4 Rb1 68. Kg6 Re1 69. Rg2 Ra1 70. Rg3 Ra2 71. Rg1 Ra3 72. Re1 Rg3+ 73.
Kf5 Rf3+ 74. Ke6 Rc3 75. Re4 Rc1 76. Re5 Rc4 77. Rd5 Re4+ 78. Kf5 Rxb4 79.
Rd8+ Kf7 80. Rd7+ Ke8 81. Re7+ Kf8 82. Ra7 Ra4 83. Ke6 Re4+ 84. Kf5 Ra4 85.
Ke6 Re4+ 86. Kd5 Ra4 87. Ke6
{Drawn by 3-fold repetition} 1/2-1/2
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.14"]
[Round "8"]
[White "yace.exe"]
[Black "wcrafty-17.10.exe"]
[Result "1-0"]
[TimeControl "3600"]
1. d4 Nf6 2. Nf3 d5 3. c4 e6 4. Nc3 Be7 5. Bf4 O-O 6. e3 c5 7. dxc5 Bxc5 8.
Qc2 Nc6 9. Rd1 Qa5 10. a3 Be7 11. Ne5 Nxe5 12. Bxe5 Rd8 13. cxd5 Nxd5 14.
e4 Nb6 15. Rxd8+ Bxd8 16. f4 Nd7 17. Bd6 Qh5 18. Nb5 Bh4+ 19. Kd2 Qg4 20.
Kc1 Bd8 21. Nxa7 Nb6 22. Nb5 Bd7 23. Kb1 Rc8 24. Nc3 Na4 25. Qd2 Nxc3+ 26.
bxc3 f6 27. Ka1 Qh5 28. e5 Be8 29. Be2 Qf7 30. Bf3 f5 31. Qd4 Ba5 32. Bb4
Qc7 33. Bxa5 Qxa5 34. Qb4 Qc7 35. Rc1 Ra8 36. Ka2 Ra4 37. Qxb7 Qc5 38. Qb3
Rxf4 39. Qb8 Kf7 40. Rb1 Qc4+ 41. Kb2 Qc5 42. Qd8 Qc4 43. Qb6 Rh4 44. h3
Bb5 45. Rd1 Ba4 46. Rd6 h6 47. Qb7+ Kg8 48. Qb4 Qxb4+ 49. axb4 Kf7 50. Rb6
Rc4 51. Rb7+ Kf8 52. b5 Rc5 53. b6 Bc6 54. Rb8+ Ke7 55. Bxc6 Rxc6 56. c4
Rxc4 57. b7 Rc6 58. Re8+ Kxe8 59. b8=Q+ Kd7 60. Qf8 Rb6+ 61. Kc3 h5 62.
Qxg7+ Ke8 63. Qc7 Rb1 64. Qc6+ Ke7 65. Qd6+ Kf7 66. Qd7+ Kf8 67. Qxe6 Ra1
68. Qxf5+ Ke7 69. Qf6+ Ke8 70. e6 Ra3+ 71. Kd4 Ra4+ 72. Kc5 Ra7 73. Kb6 Rh7
{Black resigns} 1-0
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.14"]
[Round "9"]
[White "wcrafty-17.10.exe"]
[Black "yace.exe"]
[Result "0-1"]
[TimeControl "3600"]
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. e3 c5 5. Bd3 d5 6. Nf3 O-O 7. O-O dxc4 8.
Bxc4 cxd4 9. exd4 b6 10. Bg5 Bb7 11. a3 Bxc3 12. bxc3 Qc7 13. Bxf6 Qxc4 14.
Ne5 Qd5 15. Qg4 Qxg2+ 16. Qxg2 Bxg2 17. Bxg7 Kxg7 18. Kxg2 f6 19. Nd3 Nd7
20. Rfe1 Kf7 21. Rab1 Rfc8 22. Rb3 Rc6 23. Re3 Rac8 24. Nb4 R6c7 25. f4 f5
26. Na6 Rc4 27. Nb4 a5 28. Na2 R8c6 29. Rh3 Nf6 30. Kg3 Kg6 31. Kf2 Nd5 32.
Rb1 h5 33. Rg1+ Kh6 34. Rf3 Ra4 35. Rb1 Rxa3 36. Rb2 Rc4 37. Re2 Kg7 38.
Rc2 b5 39. Rd2 b4 40. Rg3+ Kf6 41. cxb4 Rxg3 42. hxg3 axb4 43. Rb2 b3 44.
Rxb3 Rc2+ 45. Kf1 Rxa2 46. Rb8 h4 47. gxh4 Rd2 48. h5 Rxd4 49. Kf2 Rxf4+
50. Kg3 Kg5 51. h6 Kxh6 52. Re8 Re4 53. Rh8+ Kg7 54. Ra8 e5 55. Re8 Kf7 56.
Rd8 Rd4 57. Rh8 e4 58. Rh2 f4+ 59. Kg4 e3 60. Kf3 Nc3 61. Rc2 Rd3 62. Kxf4
e2 63. Rxe2 Nxe2+ 64. Ke4 Rg3 65. Kd5 Re3 66. Kc5 Ke6 67. Kc4 Rc3+ 68. Kb4
Kd5 69. Kb5 Nd4+ 70. Kb4 Ne2 71. Kb5 Nd4+ 72. Kb6 Rb3+ 73. Ka5 Kc6
{White resigns} 0-1
[Event "Computer chess game"]
[Site "HUGO"]
[Date "2000.06.14"]
[Round "10"]
[White "yace.exe"]
[Black "wcrafty-17.10.exe"]
[Result "1-0"]
[TimeControl "3600"]
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. e3 c5 5. Bd3 d5 6. Nf3 O-O 7. O-O dxc4 8.
Bxc4 Nbd7 9. Qe2 a6 10. a3 cxd4 11. axb4 dxc3 12. bxc3 Qc7 13. Bb2 Nb6 14.
Bb3 e5 15. Ra5 e4 16. Ne5 Be6 17. Bxe6 fxe6 18. c4 Qe7 19. b5 axb5 20. Rxb5
Nc8 21. Bd4 Nd6 22. Rb4 Rfc8 23. Rfb1 Rc7 24. Qb2 Nf5 25. g3 Nxd4 26. exd4
Ra7 27. Re1 Qe8 28. Rb6 Qc8 29. Qb3 Ra8 30. c5 Nd5 31. Rd6 Re7 32. Rxe4 Nc7
33. Re1 Ra7 34. Rb6 Nd5 35. Nc6 bxc6 36. Rb8 Rac7 37. Rxc8+ Rxc8 38. Qa4 g6
39. Qa6 Rcc7 40. Qa8+ Kg7 41. f4 h6 42. Kf1 Nc3 43. Qa1 Nd5 44. Qa6 Nc3 45.
Qc4 Nd5 46. Qe2 Rb7 47. Qe5+ Kf7 48. Qh8 h5 49. Qh7+ Kf6 50. Qg8 Rbd7 51.
Qf8+ Rf7 52. Qh8+ Rg7 53. h3 Nc7 54. g4 hxg4 55. hxg4 Kf7 56. g5 Nd5 57. f5
Rb7 58. fxe6+ Ke7 59. Qxg7+ Ke8 60. Qxb7 Nb6 61. e7 Nd7 62. Qc8+ Kf7
{Black resigns} 1-0

[Mogens Larsen]

Geschrieben von: / Posted by: Mogens Larsen at 14 June 2000 14:00:24:
Als Antwort auf: / As an answer to: Crafty 17.10 - Yace 0.15 G/60 with surprising result geschrieben von: / posted by: Dieter Bürßner at 14 June 2000 13:04:52:

Tonight I had a lucky stroke against crafty. (I should have played lotto;)
Crafty 17.10 vs. Yace 0.15 1.5-8.5 (3 Draws) Game/60
Hash: Crafty hash 6M, hashp 4M, Yace: 10M (I know, that for crafty 8M, 2M
would be better, but it's impossible). Crafty learn off, no endgame
tables (haven't installed them yet). Book: both programs used 2600.pgn
from Dann Corbit's site with maxply 20. AMD K6-2 300 MHz.
I don't have the expertise to judge the appended games. From looking at
the logs, round 3 was interesting. Crafty came out of the book with
+1.11 and saw the mate when it resigned.
I made a few changes to Yace, that may help for longer time controls.
I'll ask Dann to make the new version available soon.
Best wishes,
Dieter "proud" Buerssner

Congratulations, the result suggest that you're doing something right ). Even though Crafty does perform better with its own book, it's a very impressive nonetheless. There's not even any duplicate games. Looking forward to a release of Yace 0.15 as soon as possible.
BTW, what do you plan to implement in the nearby future? That would be interesting to hear if it's not to technical or top secret ).
Best wishes...
Mogens

[Aaron]

Geschrieben von: / Posted by: Aaron at 14 June 2000 14:08:29:
Als Antwort auf: / As an answer to: Re: Crafty 17.10 - Yace 0.15 G/60 with surprising result geschrieben von: / posted by: Mogens Larsen at 14 June 2000 14:00:24:

Tonight I had a lucky stroke against crafty. (I should have played lotto;)
Crafty 17.10 vs. Yace 0.15 1.5-8.5 (3 Draws) Game/60
Hash: Crafty hash 6M, hashp 4M, Yace: 10M (I know, that for crafty 8M, 2M
would be better, but it's impossible).

Congratulations, the result suggest that you're doing something right ).
Even though Crafty does perform better with its own book, it's a very >impressive nonetheless. There's not even any duplicate games. Looking forward >to a release of Yace 0.15 as soon as possible.
BTW, what do you plan to implement in the nearby future? That would be >interesting to hear if it's not to technical or top secret ).

Yes.. I use exactly the same configurations for Blitz..when using 10 Mb..

Definitely..Best new winboard engine for the year..Unless Gully can upstage Yace..
Yes.. I'm curious as well...Features to support Analysis mode?, Book Learning? etc??

[Gábor Szõts]

Geschrieben von: / Posted by: Gábor Szõts at 14 June 2000 14:51:16:
Als Antwort auf: / As an answer to: Re: Crafty 17.10 - Yace 0.15 G/60 with surprising result geschrieben von: / posted by: Mogens Larsen at 14 June 2000 14:00:24:

BTW, what do you plan to implement in the nearby future? That would be interesting to hear if it's not to technical or top secret ).

I'd suggest to implement resigning... although you may not need it in the future!!?
Best wishes,
Gábor

[Mogens Larsen]

Geschrieben von: / Posted by: Mogens Larsen at 14 June 2000 14:55:29:
Als Antwort auf: / As an answer to: Re: Crafty 17.10 - Yace 0.15 G/60 with surprising result geschrieben von: / posted by: Gábor Szõts at 14 June 2000 14:51:16:

BTW, what do you plan to implement in the nearby future? That would be interesting to hear if it's not to technical or top secret ).

I'd suggest to implement resigning... although you may not need it in the future!!?

I would like to second that. It makes testing much more efficient.
Best wishes...
Mogens

[Dieter Buerssner]

Geschrieben von: / Posted by: Dieter Buerssner at 14 June 2000 16:40:21:
Als Antwort auf: / As an answer to: Re: Crafty 17.10 - Yace 0.15 G/60 with surprising result geschrieben von: / posted by: Gábor Szõts at 14 June 2000 14:51:16:

I'd suggest to implement resigning

Experimental in 0.15. Put a line
resign on
into yace.ini.
Sent it to Frank and Dann. I assume, it will be available soon.
Regards,
Dieter

[Dieter Bürßner]

Geschrieben von: / Posted by: Dieter Bürßner at 14 June 2000 16:51:00:
Als Antwort auf: / As an answer to: Re: Crafty 17.10 - Yace 0.15 G/60 with surprising result geschrieben von: / posted by: Mogens Larsen at 14 June 2000 14:00:24:

BTW, what do you plan to implement in the nearby future? That would be
interesting to hear if it's not to technical or top secret ).

Nothing secret. After the first point, no particular order. Everything
will take it's time, though.
1. Fix two unsolved problems (Unfortunately, I cannot reproduce them yet)
a) Yace doesn't use it's book sometimes (Never happens here)
b) Yace doesn't move when it's black (Never happens here either)
c) Draw Problems with crafty.
2. Download xboard and try xboard mode (should be easy, Yace works under linux already)
3. Don't draw won endgames anymore (especially rook pawn with wrong bishop)
4. Move now
5. Pondering
6. Analyze (should be easy, when I understood fully what is needed)
7. Endgame tables
8. Learning
And, because testing was only done against tactical test positions, there
should be many possibilities for tweaking it's strength for real games.
Regards,
Dieter

[Mogens Larsen]

Geschrieben von: / Posted by: Mogens Larsen at 14 June 2000 18:25:52:
Als Antwort auf: / As an answer to: Yace features geschrieben von: / posted by: Dieter Buerssner at 14 June 2000 16:40:21:

I'd suggest to implement resigning

Experimental in 0.15. Put a line
resign on
into yace.ini.
Sent it to Frank and Dann. I assume, it will be available soon.
Regards,
Dieter

When is Yace designed to resign?
Best wishes
Mogens

[Aaron]

Geschrieben von: / Posted by: Aaron at 14 June 2000 18:55:33:
Als Antwort auf: / As an answer to: Yace features geschrieben von: / posted by: Dieter Bürßner at 14 June 2000 16:51:00:

BTW, what do you plan to implement in the nearby future? That would be
interesting to hear if it's not to technical or top secret ).

And, because testing was only done against tactical test positions, there
should be many possibilities for tweaking it's strength for real games.

Here;s the opinion from someone who doesn;t know much chess or programming..Perhaps you should work on adding knowledge? (pawn structures,king safety,endgame) etc..
Your search looks great, and it's tactical abilities are first rate

I would like to add to the already long list of features, how about adding a list of evaluation terms such as king safty,passed pawns, development,search depth,extensions in a txt file so that even a computer idiot like me can play around with..
See Gormit2.2 or Exchess

[Dieter Buerssner]

Geschrieben von: / Posted by: Dieter Buerssner at 15 June 2000 11:17:09:
Als Antwort auf: / As an answer to: Re: Yace features geschrieben von: / posted by: Mogens Larsen at 14 June 2000 18:25:52:

When is Yace designed to resign?

It is very primitive. When it found for 6 moves a score smaller -8.0.
Time on the clocks is not considered. Feel free to suggest a better approach.
Regards,
Dieter

[Mogens Larsen]

Geschrieben von: / Posted by: Mogens Larsen at 15 June 2000 11:58:39:
Als Antwort auf: / As an answer to: Re: Yace features geschrieben von: / posted by: Dieter Buerssner at 15 June 2000 11:17:09:

When is Yace designed to resign?

It is very primitive. When it found for 6 moves a score smaller -8.0.
Time on the clocks is not considered. Feel free to suggest a better approach.
Regards,
Dieter

The Crafty approach is three times under -x, where x is determined by the user. I believe that is the best way to do it. Six moves under -8 is too much IMHO. Something like 3 moves with a score smaller than -6.0 is sufficient. The time factor is best suited against human players I think, but definately an interesting factor to include later on.
Best wishes...
Mogens

[Dieter Buerssner]

Geschrieben von: / Posted by: Dieter Buerssner at 15 June 2000 18:10:37:
Als Antwort auf: / As an answer to: Re: Yace features geschrieben von: / posted by: Mogens Larsen at 15 June 2000 11:58:39:

The Crafty approach is three times under -x, where x is determined by the user. I believe that is the best way to do it. Six moves under -8 is too much IMHO. Something like 3 moves with a score smaller than -6.0 is sufficient. The time factor is best suited against human players I think, but definately an interesting factor to include later on.

Thanks, for your suggestion.
I will make the score and the number of moves settable by the user. (It's only
a few lines of code). But right now, Yace is too stupid for "aggressive
resigning". You may have recognized, that it doesn't understand a draw
with the rook pawn and the wrong bishop. It's score is far off. Also,
it sometimes is very lucky, and finds an "infinete chess" (don't know
the correct English term) from a very low score.
Regards,
Dieter

[Dann Corbit]

Geschrieben von: / Posted by: Dann Corbit at 15 June 2000 18:50:31:
Als Antwort auf: / As an answer to: Re: Yace features geschrieben von: / posted by: Dieter Buerssner at 15 June 2000 18:10:37:

The Crafty approach is three times under -x, where x is determined by the user. I believe that is the best way to do it. Six moves under -8 is too much IMHO. Something like 3 moves with a score smaller than -6.0 is sufficient. The time factor is best suited against human players I think, but definately an interesting factor to include later on.

Thanks, for your suggestion.
I will make the score and the number of moves settable by the user. (It's only
a few lines of code). But right now, Yace is too stupid for "aggressive
resigning". You may have recognized, that it doesn't understand a draw
with the rook pawn and the wrong bishop. It's score is far off. Also,
it sometimes is very lucky, and finds an "infinete chess" (don't know
the correct English term) from a very low score.

I think the resignation mode stated is a good one for calibration runs, etc.
But in some contests, you will want to keep trying until you are dead. So it should be configurable, ideally.


My FTP site

[Aaron]

Geschrieben von: / Posted by: Aaron at 18 June 2000 12:42:04:
Als Antwort auf: / As an answer to: Re: Yace features geschrieben von: / posted by: Dann Corbit at 15 June 2000 18:50:31:

The Crafty approach is three times under -x, where x is determined by the >>>user

The time factor is best suited against human players I think, but definately >>an interesting factor to include later on.
Thanks, for your suggestion.
I will make the score and the number of moves settable by the user. (It's >>only a few lines of code). But right now, Yace is too stupid for "aggressive
resigning".
You may have recognized, that it doesn't understand a draw
with the rook pawn and the wrong bishop. It's score is far off. Also,
it sometimes is very lucky, and finds an "infinete chess" (don't know
the correct English term) from a very low score.

I think the resignation mode stated is a good one for calibration runs, etc.
But in some contests, you will want to keep trying until you are dead. So it >should be configurable, ideally.

Yes for swindling..

I think it's great that Yace allows resigning to be so customisable..Even more so than Crafty!

Yes,maybe you could plug in TBs? But that helps a little only..
Perpertual check?

Definitely more options is better..
One question, though, I never see programs offer draws? Why?

[Dieter Buerssner]

Geschrieben von: / Posted by: Dieter Buerssner at 19 June 2000 14:13:39:
Als Antwort auf: / As an answer to: Re: Yace features geschrieben von: / posted by: Aaron at 18 June 2000 12:42:04:

You may have recognized, that it doesn't understand a draw
with the rook pawn and the wrong bishop. It's score is far off. Also,
it sometimes is very lucky, and finds an "infinete chess" (don't know
the correct English term) from a very low score.

Yes,maybe you could plug in TBs? But that helps a little only..
Perpertual check?
One question, though, I never see programs offer draws? Why?

This is planned. And it would of course help in this situation.
Unfortunately, I can't get the tbgen.exe from Robert Hyatt's site
running for four men table bases. I get GPF's under windows, and
also errors under Linux, or with the DJGPP compiled version under
DOS:
*** Unable to probe minor -- something wrong
White pieces: 02 70 16
Black pieces: 70
En passant 7F
The values shown are always different. I suspect, it has something to do with
my system.
Thanks. (I wanted to write "infinete check", check and chess have the
same German translation ...)
One reason might be, that it isn't easy for a chess program, to judge,
whether a position is a draw.
One could try to start with something like: The score didn't change much
for n moves. But what is a good choice for n. How do you quantify: score
doesn't change much? Do these choices depend on the material evaluation?
Do these choices depend on the absolute value of the score? (Say I am ahead
by one pawn, but can't advance). A human player may be able to make such an
decision correctly after just looking for a few seconds at the board. It
seems difficult do do this for a chess program, that usually doesn't
understand the position. The chess programm just sees how the position
developes. It might be that one side has an easy time, while the other side
must play very accurate. These matters are complicated, when clock time
is considered as well.
Please suggest a detailed approach. Implementing it shouldn't be too difficult.
Regards,
Dieter