Gromit 2.20 with tweaked eval.cfg - Anmon 5.01 = 5.5 - 4.5 (

[Alexandre Côme]

Geschrieben von:/Posted by: Alexandre Côme am 14. November 1999 22:30:19:
.

[Frank Schneider]

Geschrieben von:/Posted by: Frank Schneider am 15. November 1999 21:15:44:
Als Antwort auf:/As an answer to: Gromit 2.20 with tweaked eval.cfg - Anmon 5.01 = 5.5 - 4.5 (5min+3sec)! no text geschrieben von:/posted by: Alexandre Côme am 14. November 1999 22:30:19:

.

A very good result (usually Gromit is rather weak at blitz).
I'm interested in the eval-changes you made: how did you determine what to change and what did you change?

Frank


GromitChess Home Page

[Alexandre Côme]

Geschrieben von:/Posted by: Alexandre Côme am 15. November 1999 23:04:01:
Als Antwort auf:/As an answer to: Re: Gromit 2.20 with tweaked eval.cfg - Anmon 5.01 = 5.5 - 4.5 (5min+3sec)! no text geschrieben von:/posted by: Frank Schneider am 15. November 1999 21:15:44:

.

A very good result (usually Gromit is rather weak at blitz).
I'm interested in the eval-changes you made: how did you determine what to change and what did you change?

Frank

Hello Frank,
I think my tweaked eval.cfg is weaker than the original but I'm not sure!
I have change a lot values of pieces, it's fun but not serious!
I cannot explains this good result against AnMon!
Best Regards
Alexandre Côme

[Andrew Tanner]

Geschrieben von:/Posted by: Andrew Tanner am 16. November 1999 05:08:27:
Als Antwort auf:/As an answer to: Gromit 2.20 with tweaked eval.cfg - Anmon 5.01 = 5.5 - 4.5 (5min+3sec)! no text geschrieben von:/posted by: Alexandre Côme am 14. November 1999 22:30:19:
Hello! Great program!
I use a custom eval.cfg that seems to be significantly stronger at blitz than the default eval.cfg. In short it seems to play more aggressive tactically. This may be due to the high value I assigned to the Bishops. It uses it's bishops in many a game to directly attack the castled kingside. I figured that a higher bishop value might lead the eval to favor more agressive kingside attack type moves. As it turns out this seems to be the case, although I need to do alot more testing.
Out of 10 blitz 3 0 games it won 7 and loss 3 against the default cfg file.
I would like to learn more about what some of the other commands do. I don't see any .doc files that cover this. If you could send/post more info on this to assist in my continuing tests I would be gratefull.

BVALUE = 389 ; value bishop
Giving this a higher value than the default has proven profitable in blitz
PAWNRUN = 0 ; bonus pushing pawns (kingattack)
(I've tried setting this high, and it played a few really powerfull kingside attacks, but overall it lost many games because of premature kingside pawn attacks especially when it castled kingside itself (as expected)
PAWNRUNE = 0 ; pushing pawns in endgame
(why give a value to this? too complex to predetermine with a value..seems it could easily lead to potential disaster)
PAWNSRAM = 0 ; penalty blocked pawns
Need more info..
PAWNSPACE = 0 ; bonus for unblocked pawn on opponentside
PAWNSE5D5 = 6 ; pawns on e5 and d5
suspected that a value of 8 might be too high (6 also might be too high)
PAWNSE4D4 = 6 ; pawns on e4 and d4
PAWNSD4C4 = 2 ; pawns on d4 and c4
These are the first 2 games from the 10.

[Event "Computer chess game"]
[Site "Win98"]
[Date "1999.11.2"]
[Round "1"]
[White "neweval"]
[Black "oldeval"]
[Result "1-0"]
[TimeControl "180"]
1. c4 e5 2. Nc3 Nf6 3. e3 c6 4. Nf3 e4 5. Nd4 d5 6. cxd5 cxd5 7. d3 Bc5 8.
Nb3 Bb4 9. dxe4 Nxe4 10. Qd4 Bxc3+ 11. bxc3 O-O 12. c4 Nc6 13. Qxd5 Qe7 14.
Bd3 Qb4+ 15. Kf1 Nf6 16. Qg5 Be6 17. Qh4 Qc3 18. Be2 Ne5 19. Ba3 Ng6 20.
Qd4 Qxd4 21. Nxd4 Rfd8 22. Nxe6 fxe6 23. Rb1 b6 24. Rd1 Rxd1+ 25. Bxd1 Rd8
26. Ke2 Kf7 27. Bb4 Nh4 28. Rg1 Ng4 29. c5 a5 30. Be1 bxc5 31. f3 Rxd1 32.
fxg4 Ra1 33. Rf1+ Kg6 34. Bxh4 Rxa2+ 35. Kf3 h5 36. gxh5+ Kxh5 37. Bd8 a4
38. Rc1 Kg6 39. Bc7 a3 40. Be5 Rd2 41. Rxc5 a2 42. Rc1 Kh7 43. Bc3 Rd3 44.
Ba1 Rd2 45. h4 Kh6 46. g4 Rh2 47. h5 g6 48. Be5 Rd2 49. Rh1 gxh5 50. e4 Rc2
51. gxh5 Rc6 52. Bf4+ Kh7 53. Ke3 Rc3+ 54. Kd4 Ra3 55. Ra1 Kg7 56. Ke5 Ra6
57. Be3 Kh7 58. Bd4 Kh6 59. Rh1 Ra4 60. Bc3 Rc4 61. Ba1 Ra4 62. Rh3 Ra6 63.
Rh2 Ra4 64. Kf6 Kh7 65. e5 Kh6 66. Kxe6 Ra6+ 67. Kd5 Ra8 68. Kc4 Ra3 69.
Kb4 Ra7 70. Kb3 Ra8 71. Rxa2 Rb8+ 72. Kc4 Kxh5 73. Ra7 Rc8+ 74. Kd5 Rd8+
75. Ke6 Kg6 76. Ke7 Rh8 77. Ra6+ Kh7 78. e6 Rg8 79. Kd7 Rb8 80. e7 Rb7+ 81.
Ke6 Rb8 82. Be5 Re8 83. Kf7 Rg8 84. Rh6+ Kxh6 85. Kxg8 Kh5 86. e8=Q+ Kg5
{Black resigns} 1-0
[Event "Computer chess game"]
[Site "Win98"]
[Date "1999.11.2"]
[Round "2"]
[White "oldeval"]
[Black "neweval"]
[Result "0-1"]
[TimeControl "180"]
1. d4 d5 2. c4 dxc4 3. Nf3 e6 4. Qa4+ Nd7 5. e4 c5 6. Bxc4 cxd4 7. O-O Qb6
8. Na3 Ne7 9. Nb5 Ng6 10. Nfxd4 Bc5 11. Nf3 O-O 12. b4 Be7 13. Bb2 a6 14.
Bd4 Qc6 15. Bd3 Nb6 16. Bxb6 Qxb6 17. Nbd4 Ne5 18. Qd1 Rd8 19. a4 Nxd3 20.
Qxd3 e5 21. a5 Qxb4 22. Rfb1 Qd6 23. Ra4 exd4 24. Rxd4 Qc7 25. Rxd8+ Bxd8
26. Rd1 Bg4 27. Qb3 Qxa5 28. Qxb7 Qa4 29. Qd5 h6 30. Re1 Rc8 31. Ne5 Be6
32. Qd3 Qa2 33. Qd1 Bb6 34. Qf3 Rc2 35. Nd3 Rc3 36. Rd1 Bc4 37. Nb4 Rxf3
38. Nxa2
{White resigns} 0-1