Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: here my guess...

Author: Robert Hyatt

Date: 11:31:59 08/12/05

Go up one level in this thread


On August 12, 2005 at 12:37:34, Uri Blass wrote:

>On August 12, 2005 at 11:52:26, Darrel Briley wrote:
>>but I used Fruit 2.1 against 8x Crafty in one game.  Score follows.
>>
>>
>>[Event "ICC 5 3"]
>>[Site "Internet Chess Club"]
>>[Date "2005.08.10"]
>>[Round "-"]
>>[White "Frutit 2.1"]
>>[Black "crafty"]
>>[Result "0-1"]
>>[ICCResult "Crafty"]
>>[WhiteElo "3020"]
>>[BlackElo "2813"]
>>[Opening "Sicilian: Richter-Rauzer, Rauzer attack, 7...a6"]
>>[ECO "B66"]
>>[NIC "SI.29"]
>>[Time "23:22:40"]
>>[TimeControl "300+3"]
>>
>>1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 d6 6. Bg5 e6 7. Qd2 a6 8.
>>O-O-O h6 9. Be3 Be7 10. f4 Nxd4 11. Bxd4 b5 12. Qe3 Bb7 13. Bxf6 gxf6 14.
>>Bd3 b4 15. Ne2 Qa5 16. Kb1 Rc8 17. Rhe1 Qc5 18. Qg3 Bf8 19. Nd4 Qa5 20. f5
>>e5 21. Nb3 Qb6 22. Nd2 Qc5 23. h3 Qc7 24. Bc4 a5 25. Bb5+ Bc6 26. Bxc6+ Qxc6
>>27. Rc1 a4 28. c3 b3 29. axb3 axb3 30. Nxb3 Rb8 31. Nd2 Qb5 32. Rc2 h5 33.
>>c4 Qa5 34. Qc3 Qa6 35. Rd1 Ra8 36. b3 Bh6 37. Nf3 O-O 38. Qb2 Rfb8 39. Rc3
>>Rb6 40. g4 Qa7 41. g5 fxg5 42. h4 g4 43. Ng5 Ra6 44. Rcd3 Ra1+ 45. Kc2 Ra2
>>46. Rxd6 Bf8 47. R6d5 Bb4 48. Rd8+ Kg7 49. f6+ Kg6 50. Rg8+ Rxg8 51. Nh3
>>Rxb2+ 52. Kxb2 Ra8 53. Nf4+ exf4 54. Rd5 Qa1+ 55. Kc2 Ra2+ 56. Kd3 Rd2#
>>{White checkmated}
>>0-1
>>
>>Fruit showed a pretty good advantage in the eval for most of the game, that is
>>until those Q-side pawns came crashing down.
>>
>>It's only a blitz game, but as a result I've revised my thinking on Fruit's
>>chances.  Lack of multi-processor support and using the hardware provided
>>locally (PIV 3Ghz) will keep Fruit from being a serious contender.
>>
>>                                   DB
>
>You forget that you used the public version of fruit when bob used better
>version of Crafty.
>
>I believe that fruit WCCC is better than fruit that played in this game.
>It may be interesting to know at what move Crafty started to see an advantage
>for itself.

Found the log.  Crafty came out of book at +1.0 (good for white, crafty was
playing black).  I did not have a decent book working at that point and was
using something very small just to get the thing to playing so I could work on
the other more important issues...

at move 38 for white, Crafty expected Rb2 and white played Qb2.  Score went from
well over +1 to near zero at that point.  After white's 40. g4, the score went
to -1.0 (good for crafty now) and continued to go south for white from that
point forward.

Here's the relevant part of the log after g4:

               11     0.80  -0.64   40. ... Qa7 41. Rcd3 Ra6 42. g5 Bxg5
                                    43. h4 Ra1+ 44. Kc2 Ra2 45. hxg5 Rxb2+
                                    46. Kxb2 fxg5 47. Rxd6
               11->   1.01  -0.64   40. ... Qa7 41. Rcd3 Ra6 42. g5 Bxg5
                                    43. h4 Ra1+ 44. Kc2 Ra2 45. hxg5 Rxb2+
                                    46. Kxb2 fxg5 47. Rxd6
               12     2.43     -1   40. ... Qa7!!
               12     4.32  -1.06   40. ... Qa7 41. g5 fxg5 42. Rcd3 Qb7
                                    43. c5 dxc5 44. Nxe5 Bg7 45. Rd8+ Bf8
                                    46. R8d7 Qxe4+ 47. Kc1
               12->   4.74  -1.06   40. ... Qa7 41. g5 fxg5 42. Rcd3 Qb7
                                    43. c5 dxc5 44. Nxe5 Bg7 45. Rd8+ Bf8
                                    46. R8d7 Qxe4+ 47. Kc1
               13     6.92  -1.35   40. ... Qa7 41. g5 fxg5 42. Rcd3 Ra6
                                    43. Rxd6 Ra2 44. Qxe5 Ra1+ 45. Kb2
                                    Qf2+ 46. R1d2 R8a2+ 47. Kc3 Qxf3+ 48.
                                    Kb4 Bf8 49. Rxa2 Bxd6+
              time=8.08  mat=1  n=111695644  fh=91%  nps=13.82M

after the expected g5 move from white:

               12     1.23     -1   41. ... fxg5!!
               12     2.70  -2.26   41. ... fxg5 42. Rcd3 Ra6 43. Rxd6
                                    Ra1+ 44. Kc2 Ra2 45. Nxe5 Qe3 46. Rd8+
                                    Kh7 47. R8d4 Rxb2+ 48. Kxb2 Bg7 49.
                                    R1d3
               12->   2.72  -2.26   41. ... fxg5 42. Rcd3 Ra6 43. Rxd6
                                    Ra1+ 44. Kc2 Ra2 45. Nxe5 Qe3 46. Rd8+
                                    Kh7 47. R8d4 Rxb2+ 48. Kxb2 Bg7 49.
                                    R1d3
               13     4.95     -1   41. ... fxg5!!
               13     8.25  -2.54   41. ... fxg5 42. c5 Ra6 43. Rxd6 Ra1+
                                    44. Kc2 Ra2 45. Qxa2 Qxa2+ 46. Kd3
                                    Qg2 47. Kc4 Bf8 48. Rd7 Qe2+ 49. Kd5
               13->   8.31  -2.54   41. ... fxg5 42. c5 Ra6 43. Rxd6 Ra1+
                                    44. Kc2 Ra2 45. Qxa2 Qxa2+ 46. Kd3
                                    Qg2 47. Kc4 Bf8 48. Rd7 Qe2+ 49. Kd5
               14    11.66     -1   41. ... fxg5!!


Couple of moves later I saw this:

               14->   3.23  -5.70   43. ... Ra6 44. Rcd3 Ra1+ 45. Kc2 Ra2
                                    46. Rxd6 Bf8 47. R6d5 Bb4 48. Rd8+
                                    Kg7 49. f6+ Kg6 50. Rxa8 Qf2+ 51. Kd3
                                    Qg3+ 52. Ke2 Rxb2+ 53. Kf1 Qxh4
               15     5.63     -1   43. ... Ra6!!
               15    12.07  -6.90   43. ... Ra6 44. Rcd3 Ra1+ 45. Kc2 Ra2
                                    46. R3d2 Rxb2+ 47. Kxb2 Bxg5 48. hxg5
                                    g3 49. Rxd6 g2 50. g6 Qa3+ 51. Kc3
                                    g1=Q 52. gxf7+ Kxf7 53. Rd7+ Ke8 54.
                                    Re7+ Kxe7 55. Rxg1
               15->  12.38  -6.90   43. ... Ra6 44. Rcd3 Ra1+ 45. Kc2 Ra2
                                    46. R3d2 Rxb2+ 47. Kxb2 Bxg5 48. hxg5
                                    g3 49. Rxd6 g2 50. g6 Qa3+ 51. Kc3
                                    g1=Q 52. gxf7+ Kxf7 53. Rd7+ Ke8 54.
                                    Re7+ Kxe7 55. Rxg1
               16    19.46  -6.82   43. ... Ra6 44. Rcd3 Ra1+ 45. Kc2 Ra2
                                    46. R3d2 Rxb2+ 47. Kxb2 Bxg5 48. hxg5
                                    g3 49. Rxd6 g2 50. g6 Qa3+ 51. Kc3
                                    g1=Q 52. gxf7+ Kxf7 53. Rxg1 Qxd6 54.
                                    b4
               16->  20.07  -6.82   43. ... Ra6 44. Rcd3 Ra1+ 45. Kc2 Ra2
                                    46. R3d2 Rxb2+ 47. Kxb2 Bxg5 48. hxg5
                                    g3 49. Rxd6 g2 50. g6 Qa3+ 51. Kc3
                                    g1=Q 52. gxf7+ Kxf7 53. Rxg1 Qxd6 54.
                                    b4
               17    40.19     -1   43. ... Ra6!!
               17    54.17  -7.24   43. ... Ra6 44. Rcd3 Ra1+ 45. Kc2 Ra2
                                    46. R3d2 Rxb2+ 47. Kxb2 Bxg5 48. hxg5
                                    g3 49. Rxd6 g2 50. g6 g1=Q 51. Rxg1
                                    Qxg1 52. gxf7+ Kxf7 53. Rd7+ Ke8 54.
                                    Rh7 Qd4+ 55. Kc1 Qxe4 56. Rxh5
               17->  55.79  -7.24   43. ... Ra6 44. Rcd3 Ra1+ 45. Kc2 Ra2
                                    46. R3d2 Rxb2+ 47. Kxb2 Bxg5 48. hxg5
                                    g3 49. Rxd6 g2 50. g6 g1=Q 51. Rxg1
                                    Qxg1 52. gxf7+ Kxf7 53. Rd7+ Ke8 54.
                                    Rh7 Qd4+ 55. Kc1 Qxe4 56. Rxh5

>
>Maybe some improvement in the king safety evaluation could help fruit.
>
>Fruit is free source code so I can know that public fruit does not evaluate
>attackers that do not attack directly squares near the king so at move 40 it did
>not evaluate the rook a8 as additional attacker(it gave it small bonus for other
>reasons like open file or being at distance of one file from the opponent king
>but it was not enough) and I think that other evaluation terms to evaluate
>better king safety are missing from public fruit.
>
>I think that fabien needs some code to tell fruit that lack of pawns near the
>king is more important when the opponent attacks the squares near the king
>because if I understand fruit's code correctly lack of pawn shield and king
>attack are different terms in fruit's code.
>
>
>Public Fruit needs too much time to see that black is better at move 40 of the
>game and I think that in the type of position that happened fruit's evaluation
>is worse than Crafty and other programs.
>
>Crafty's evaluation is not symmetric but even from black side it can see that
>white is not a pawn up at move 39 when fruit is uncapable of doing it.
>
>Here is some analysis of both fruit and old Crafty on the same old
>hardware(A1000)
>
>Frutit 2.1 - crafty, ICC 5 3 Internet Chess Club 2005
>rr4k1/5p2/q2p1p1b/4pP1p/2P1P3/1PR2N1P/1Q4P1/1K1R4 b - - 0 1
>
>Analysis by Fruit 2.1:
>
>39...h4 40.Nxh4
>  +-  (1.75)   Depth: 1/2   00:00:00
>39...Qb6
>  ±  (1.26)   Depth: 1/8   00:00:00
>39...Qc6
>  ±  (1.18)   Depth: 1/8   00:00:00
>39...Bf4
>  ±  (1.13)   Depth: 1/8   00:00:00
>39...Ra7
>  ±  (1.11)   Depth: 1/8   00:00:00
>39...Kh7
>  ±  (0.98)   Depth: 1/8   00:00:00
>39...Kh7 40.Rcd3
>  ±  (1.09)   Depth: 2/8   00:00:00
>39...Kh7 40.Rcd3 Qc6
>  ±  (1.04)   Depth: 3/8   00:00:00
>39...Qc6 40.Qe2 Kh7
>  ±  (0.97)   Depth: 3/8   00:00:00
>39...Qc6 40.Rd5 Kh7 41.Rcd3
>  ±  (1.03)   Depth: 4/10   00:00:00
>39...Qc6 40.Rd5 Kh7 41.g4 hxg4 42.hxg4
>  ±  (0.99)   Depth: 5/18   00:00:00
>39...Qc6 40.Rd5 Kh7 41.Rcd3 Ra7 42.Rd1
>  ±  (1.02)   Depth: 6/18   00:00:00  26kN
>39...Qc6 40.Rd5 Kh7 41.Rcd3 Ra7 42.Nd2 Bxd2 43.Qxd2
>  ±  (1.05)   Depth: 7/22   00:00:00  89kN
>39...Kh7 40.Rcd3 Qc6 41.Rd5 Be3 42.R1d3 Bf4 43.Rxd6 Qxe4
>  ±  (1.00)   Depth: 7/24   00:00:00  128kN
>39...Rb7 40.Rd5 Kh7 41.Rcd3 Rab8 42.Qc2 Qa4
>  ±  (0.93)   Depth: 7/24   00:00:01  179kN
>39...Rb7 40.Rd5 Kh7 41.g4 hxg4 42.hxg4 Qc6 43.Rcd3 Bf4 44.Rxd6 Qxe4
>  ±  (0.95)   Depth: 8/24   00:00:01  277kN
>39...Ra7 40.Rd5 Kh7 41.g4 hxg4 42.hxg4 Qc6 43.Rcd3 Bf4 44.Rxd6 Qxe4
>  ±  (0.93)   Depth: 8/24   00:00:02  481kN
>39...Ra7 40.Rd5 Kh7 41.g4 hxg4 42.hxg4 Qc6 43.Rcd3 Bf4 44.Rd1
>  ±  (1.00)   Depth: 9/24   00:00:03  669kN
>39...Kh7 40.Rcd3 Qb7 41.Re1 Qc6 42.g4 hxg4 43.hxg4 Ra7 44.Rd5
>  ±  (0.95)   Depth: 9/25   00:00:04  892kN
>39...Kh7 40.Rd5 Rb4 41.g4 hxg4 42.hxg4 Qb7 43.Rcd3 Rb8 44.Rxd6 Rxb3 45.Rxb3 Qxb3
>  ±  (0.85)   Depth: 10/32   00:00:07  1708kN
>39...Kh7 40.Rd5 Rb4 41.g4 hxg4 42.hxg4 Rab8 43.Rdd3 Qc6 44.Qe2 Qa4 45.Qb2
>  ±  (0.81)   Depth: 11/32   00:00:13  3570kN
>39...Kh7 40.Rd5 Rb4 41.g4 hxg4 42.hxg4 Rab8 43.Nd2 Qb6 44.Kc2 Qg1 45.Rxd6 Qxg4
>46.Rxf6
>  ±  (0.76)   Depth: 12/35   00:00:24  6954kN
>39...Kh7 40.Rd5 Rb4 41.g4 hxg4 42.hxg4 Rab8 43.Nd2 Qb6 44.Kc2 Qg1 45.Rxd6 Qxg4
>46.Rxf6
>  ±  (0.76)   Depth: 13/37   00:00:49  14560kN
>39...Kh7 40.Rd5 Rb4 41.g3 Rab8 42.g4 hxg4 43.hxg4 Qb7 44.Kc2 Rg8 45.Rcd3 Rxg4
>46.Kb1 Qa8 47.Rxd6 Qxe4
>  ²  (0.62)   Depth: 14/38   00:03:39  67630kN
>
>(Blass, Tel-Aviv 12.08.2005)
>
>Frutit 2.1 - crafty, ICC 5 3 Internet Chess Club 2005
>rr4k1/5p2/q2p1p1b/4pP1p/2P1P3/1PR2N1P/1Q4P1/1K1R4 b - - 0 1
>
>Analysis by Crafty 19.13:
>
>39...Rb4
>  ³  (-0.31)   Depth: 1/9   00:00:00
>39...Rb4 40.Kc2
>  ²  (0.41)   Depth: 2/9   00:00:00
>39...Bf4 40.Kc2
>  ²  (0.31)   Depth: 2/9   00:00:00
>39...Qc6 40.Qe2
>  =  (0.23)   Depth: 2/9   00:00:00
>39...Qc6 40.Qe2 Rb4
>  =  (-0.09)   Depth: 3/9   00:00:00
>39...Qc6 40.Qe2 Ra3 41.Kc2
>  =  (0.17)   Depth: 4/11   00:00:00
>39...Qc6 40.Rd5 Rb4 41.g4 hxg4 42.hxg4
>  =  (0.02)   Depth: 5/15   00:00:00
>39...Qc6 40.Qe2 Rb4 41.Qd3 Ra6 42.Kc2
>  =  (0.16)   Depth: 6/18   00:00:00  84kN
>39...Rb4 40.g4 Qc6 41.Qe2
>  =  (0.09)   Depth: 6/18   00:00:00  113kN
>39...Rb4 40.g4 hxg4 41.Rg1 Kf8 42.Rxg4 Bf4
>  ²  (0.47)   Depth: 7/18   00:00:01  208kN
>39...Qc6 40.Rd5 Bf4 41.g3 Bh6 42.g4 hxg4 43.hxg4
>  ²  (0.34)   Depth: 7/18   00:00:02  390kN
>39...Qb7 40.Rd5 Qb4 41.Rcd3 Rb6 42.Qc3 Qa3
>  ²  (0.27)   Depth: 7/18   00:00:02  616kN
>39...Rb6 40.g4 Qb7 41.Qc2 Rba6 42.Qd3 Ra1+ 43.Kc2
>  =  (0.17)   Depth: 7/18   00:00:02  674kN
>39...Rb6 40.Rd5 Qb7 41.Rcd3 Qc7 42.Qf2 Bf4 43.Qh4
>  =  (0.20)   Depth: 8/20   00:00:04  1013kN
>39...Rb6 40.Rd5 Qb7 41.Rcd3 Qc7 42.g4 hxg4 43.hxg4 Bf4 44.Qc2
>  ²  (0.30)   Depth: 9/22   00:00:06  2097kN
>39...Rb6 40.Kc2 Qa7 41.Nd2 Ra6 42.Rg3+ Bg7 43.Qc3 Ra2+ 44.Kd3 Qf2 45.Re3 Qxg2
>  =  (0.13)   Depth: 10/24   00:00:18  7193kN
>39...Rb6 40.g4 Qb7 41.Rd5 hxg4 42.Nh2 gxh3 43.Rxh3 Bg5 44.Ng4 Rb8 45.Kc2
>  ²  (0.36)   Depth: 11/27   00:00:29  11951kN
>39...Rb4 40.g4 Qb7 41.Rxd6 Qxe4+ 42.Rdd3 Qb7 43.Rd5 e4 44.Nd4 hxg4 45.hxg4 e3
>  ²  (0.26)   Depth: 11/27   00:00:48  20685kN
>39...Rb4 40.Rd5 Rb6 41.Kc2 Qa7 42.Qb1 Qb7 43.g4 hxg4 44.Qg1 Ra2+ 45.Kb1 Rxb3+
>46.Kxa2 Rxc3 47.Qxg4+
>  ²  (0.47)   Depth: 12/30   00:03:01  82136kN
>39...Rb6 40.Kc2 Qa7 41.Kd3 Ra6 42.Qe2 Ra2 43.Qe1 Rb2 44.g4 hxg4 45.hxg4 Qc5
>  =  (-0.02)   Depth: 12/30   00:04:27  122440kN
>39...Rb6 40.Kc2 Qa7 41.Kd3 Ra6 42.Qe2 Ra2 43.Qe1 Rb2 44.Qg1 Qb7 45.g4 Rxb3
>46.gxh5+
>  =  (0.13)   Depth: 13/33   00:05:18  146094kN
>
>(Blass, Tel-Aviv 12.08.2005)
>
>Uri



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.