Author: Richard Pijl
Date: 05:14:30 12/02/05
Go up one level in this thread
On December 01, 2005 at 22:00:45, Jon Dart wrote:
>A recent offline test game:
>
>[Event "?"]
>[Site "-"]
>[Date "2005.11.29"]
>[Round "?"]
>[White "The Baron 1.7.0"]
>[Black "Arasan 9.0"]
>[Result "1-0"]
>[ECO "B86"]
>[TimeControl "900+10"]
>
>1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e6 6. Bc4 Be7 7. Be3
>O-O 8. Qe2 Nc6 9. O-O-O a6 10. Bb3 Qc7 11. Rhg1 Nd7 12. g4 Nc5
>13. Nf5 b5 14. Bd5 Bb7 15. g5 Rfc8 16. Qh5 b4 17. Bxe6 Nxe6 18. Nd5
>Qd7 19. f4 a5 20. Nf6+ Bxf6 21. gxf6 g6 22. Nxd6 Ncd4 23. Nxc8 Rxc8
>24. Bxd4 Bxe4 25. Rd2 Qa4 26. b3 Qxa2 27. Bb2 Nd4 28. Qh3 Bf5 29. Qh6
>Rxc2+ 30. Rxc2 Nxb3+ 31. Kd1 Qb1+ 32. Ke2 Qxc2+ 33. Ke3 Qc5+ 34. Ke2
>Qc4+ 35. Ke1 Qe4+ 36. Kf2 Qc2+ 37. Kg3 Qd3+ 38. Kh4 Qh3+ 39. Kg5
>Qxh6+ 40. Kxh6 Nc5 41. Rc1 Nd3 42. Ra1 Nf2 43. h3 Bxh3 44. Rxa5 Bg2
>45. Rb5 Ng4+ 46. Kg5 h5 47. f5 Be4 48. fxg6 Bxg6 49. Rxb4 Ne3 50. Bd4
>Nf5 51. Bc5 Kh7 52. Rb3 Nh6 53. Rf3 Ng4 54. Bf8 Be4 55. Rc3 Nf2
>56. Bg7 Bd3 57. Kxh5 Bg6+ 58. Kh4 Bd3 59. Rc1 Be4 60. Rc5 Kg6
>61. Rg5+ Kh7 62. Re5 Bc6 63. Kg3 Nd3 64. Rg5 Be4 65. Bf8 Bg6 66. Rg4
>Bf5 67. Rg7+ Kh8 68. Kf3 Be6 69. Rg5 Kh7 1-0 {Black resigns}
>
>Arasan lost this one but had several chances to win. A few notes:
>
>16. Qh5: Crafty has Rg3 as a book move. It does not like Qh5,
>score is -1.33 after 8 minutes.
The Baron has Qh5, Rg3 and Bxc6 in its book. When left to play on its own, it
would probably play Bxc6, with a drawish score.
>16 .. b4:
>[D] r1r3k1/1bq1bppp/p1npp3/1pnB1NPQ/4P3/2N1B3/PPP2P1P/2KR2R1 b - -
>b4 looks plausible but Crafty considers this a mistake, prefers g6 after a 8
>minute search:
> ({13:+2.32} 16. ... b4 17. Bxc5 g6 18. Qh6 exf5 19. Bxf7+ Kxf7
>20. Nd5 Qd8 21. Qxh7+ Kf8 22. Nf4 Bxg5 23. Bxd6+ Qxd6 24. Rxd6 Bxf4+ $18)
> ({13:-0.42} 16. ... g6 17. Nh6+ Kg7 18. Qh3 b4 19. Bxc5 dxc5
>20. Bxc6 Bxc6 21. Ne2 Bb5 22. c4 Bxc4 23. Qf3 Bxg5+ 24. Rxg5 $15)
> ({13:+0.58} 16. ... Ne5 17. Bxc5 dxc5 18. Bxb7 Qxb7 19. Nxe7+
>Qxe7 20. Rg3 Ng6 21. Rh3 Nf8 22. Rdd3 Rd8 23. Kb1 Rxd3 24. Rxd3 Rd8 $14)
> ({13:+0.98} 16. ... Nb4 17. Nxe7+ Qxe7 18. Bxb7 Qxb7 19. a3
>Nxc2 20. Kxc2 Na4 21. Rd4 Qc6 22. Bd2 Qc5 23. Rd3 Qxf2 24. Rd1 $16)
>
>but I have not done a deep analysis on this.
>
>17. Bxe6:
>This is the position:
>[D] r1r3k1/1bq1bppp/p1npp3/2nB1NPQ/1p2P3/2N1B3/PPP2P1P/2KR2R1 w - -
The Baron plays Bxe6 from the book (last bookmove). When thinking on its own,
the Baron considers a lot of moves, but seems to settle for g6 or Nh6, at least
under normal playing conditions. g6 gets a nice plus score for white after some
time.
I've forced Bxc5, and the Baron did find a similar plus for that move as well.
I guess in these type of positions the score returned by a program depend
heavily on kingsafety and material value tuning.
>Bxc5 is the "only move" here. Arasan gives this analysis at 1 hr./move:
> bm Bxc5
>result: Bxc5 3600 seconds. score: +2.44 1266552516 nodes. ++ corre
>ct
>Bxc5 g6 Qh6 Bf8 Bxd6 exf5 Bxf8 Rxf8 exf5 Ne7 fxg6 hxg6 Bxb7 Qxb7 Na4 Qc6 Rge1
>Qxa4
>result(2): g6 3600 seconds. score: +0.31 1514741473 nodes.
>** error
>g6 hxg6 Rxg6 fxg6 Qxg6 Bg5 Bxg5 Re8 Bf6 Ne5 Bxe5 dxe5 Nd6 Qxd6 Bxb7 Nxb7
>result(3): Bxe6 3600 seconds. score: -0.05 1598238528 nodes.
>** error
>Bxe6 Nxe6 Nd5 Qd7 Rg3 b3 axb3 Bd8 Nxd6 Rcb8 Nf5 g6 Nf6+ Bxf6 Nh6+ Kh8 Rxd7
>
>19. f4: Rg3 is better here:
>[D]r1r3k1/1b1qbppp/p1npn3/3N1NPQ/1p2P3/4B3/PPP2P1P/2KR2R1 w - -
>
The current version of the Baron likes 19.Nf6!? instead (with score +0.64). I'm
not sure whether that is really an improvement over f4, but at least it looks
nice! When given more time, the Baron would switch to Rg3. Analysis output (Xeon
2.8Ghz, single CPU):
ply time nodes score pv
6(14)- 0:01.49 285831 -0.07 Nd5-f6 (Be7xf6) (g5xf6) (g7-g6) (Rd1xd6)
(Qd7-c7) (Be3-d2) (Nc6-d4)
6(15)& 0:01.69 319025 -0.10 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-c7 Qh5-d1 (Rc8-d8)
6(15). 0:01.72 329717 -0.10 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-c7 Qh5-d1 (Rc8-d8)
7(16)& 0:02.80 525533 -0.24 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-c7 Qh5-d1 b4-b3
7(16). 0:07.28 1343975 -0.24 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-c7 Qh5-d1 b4-b3
8(17)& 0:12.03 2197114 -0.33 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-c7 Qh5-d1 b4-b3 Nf5-h6 Kg8-h8 (a2xb3)
8(17). 0:24.36 4513111 -0.33 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-c7 Qh5-d1 b4-b3 Nf5-h6 Kg8-h8 (a2xb3)
9(20)+ 0:28.83 5288621 +0.02 Nd5-f6 (Be7xf6) (g5xf6) (g7-g6) (Rd1xd6)
(Qd7-c7) (Qh5-h6) (Kg8-h8) (Be3-b6)
(Qc7xb6) (Rd6xe6)
9(20)& 0:36.41 6715210 +0.45 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-c7 Qh5-h6 Nc6-e5 Nf5-e7 Kg8-h8 Ne7xc8
9(20). 0:51.73 9724320 +0.45 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-c7 Qh5-h6 Nc6-e5 Nf5-e7 Kg8-h8 Ne7xc8
10(20)& 1:17.89 14583620 +0.62 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-e8 Qh5-h6 Qe8-f8 Rd6-d7 b4-b3 a2xb3
(Qf8-b4)
10(20). 1:53.92 21661406 +0.62 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-e8 Qh5-h6 Qe8-f8 Rd6-d7 b4-b3 a2xb3
(Qf8-b4)
11(21)& 3:04.14 34650040 +0.64 Nd5-f6 Be7xf6 g5xf6 g7-g6 Rd1xd6
Qd7-e8 Qh5-h6 Qe8-f8 Rd6-d7 Ra8-b8 Be3-f4
Qf8-e8 (Rd7-c7)
11(23)& 8:36.95 93107076 +0.93 Rg1-g3 b4-b3 a2xb3 Kg8-h8 Be3-d2
g7-g6 Bd2-c3 f7-f6 Nd5xe7 Nc6xe7 Bc3xf6
Kh8-g8 Nf5xe7 (Qd7xe7)
11(23). 10:50.98 116902013 +0.93 Rg1-g3 b4-b3 a2xb3 Kg8-h8 Be3-d2
g7-g6 Bd2-c3 f7-f6 Nd5xe7 Nc6xe7 Bc3xf6
Kh8-g8 Nf5xe7
> bm Rg3
>result: Rg3 3600 seconds. score: +3.02 1614483454 nodes. ++ corre
>ct
>Rg3 b3 axb3 Qd8 e5 Qf8 exd6 Bd8 Rh3 h6 d7 Rcb8 gxh6 g6 Rg3
>result(2): Rg4 3600 seconds. score: +1.55 1575215789 nodes.
>** error
>Rg4 b3 Nf6+ Bxf6 gxf6 g6 Qh6 Ncd4 Ne7+ Qxe7 fxe7 bxa2 Kd2 Rxc2+ Kd3 a1=Q Rxa1
>Nc5+
>result(3): f4 3600 seconds. score: +0.39 1293350091 nodes.
>** error
>f4 b3 axb3 Kh8 Qxf7 Rf8 Qh5 g6 Qh6 Bd8 Bd4+ Ncxd4 Nxd4 Be7 Nxe6
>
The current version of the Baron plays 22.Rxd6 instead of Nxd6? and displays a
nice score:
11(21)& 2:32.09 29423877 +1.86 Rd1xd6 Qd7-e8 Qh5-h6 Rc8-c7 Be3-b6
Rc7-d7 Nf5-g7 Qe8-c8 Ng7xe6 f7xe6 Rd6xe6
after Nxd6 the score starts to drop a lot, leading to a lost game.
Richard.
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.