Author: Joachim Rang
Date: 04:06:42 09/27/05
Go up one level in this thread
On September 27, 2005 at 06:56:04, Uri Blass wrote:
>I downloaded the games of Fritz9 to see how it beat FruitWCCC
>
>Here is one of them
>
>[Event "20'/40"]
>[Site "?"]
>[Date "2005.09.23"]
>[Round "6"]
>[White "Fritz 9"]
>[Black "Fruit WCCC'05"]
>[Result "1-0"]
>[ECO "A90"]
>[Annotator "0.19;-0.01"]
>[PlyCount "87"]
>[EventDate "2005.??.??"]
>[TimeControl "40/1200:0"]
>
>{GenuineIntel 1598 MHz W=13.6 ply; 1.022kN/s B=14.1 ply; 546kN/s} 1. d4 f5 2.
>g3 Nf6 3. Bg2 e6 4. c4 c6 5. Nf3 d5 6. O-O Bd6 7. b3 Qe7 8. Bb2 O-O {
>Beide letzter Buchzug} 9. Ne5 {0.19/12 64} Nbd7 {(b6) -0.01/13 36} 10. Qd3 {
>(f4) 0.29/12 66} Ne4 {(a5) -0.07/13 45} 11. f3 {(f4) 0.26/13 52} Nec5 {
>(Nxe5) -0.14/14 35} 12. Qe3 {0.31/14 75} Nxe5 {-0.13/14 22} 13. dxe5 {
>0.38/13 30} Bc7 {-0.16/14 39} 14. h3 {(Kh1) 0.16/12 39} dxc4 {-0.65/12 22} 15.
>Nd2 {0.22/14 40} b5 {(f4) -0.81/14 69} 16. Bd4 {0.35/12 24} Na6 {-0.75/14 30}
>17. bxc4 {(f4) 0.29/13 51} Rd8 {-0.73/15 60} 18. c5 {0.13/14 38} Nb4 {
>-0.66/15 56} 19. Qc3 {(Rfc1) -0.04/14 35} Nd5 {-0.89/15 34} 20. Qd3 {
>(Qb2) 0.00/14 41} f4 {(Nb4) -1.23/14 54} 21. Ne4 {-0.07/13 36} a5 {
>(b4) -1.23/13 26} 22. Qd2 {(Kh2) -0.10/13 44} fxg3 {(Ne3) -1.53/14 40} 23. f4 {
>(Nxg3) 0.49/14 27} a4 {(Qh4) -0.74/13 49} 24. e3 {0.95/13 38} Kh8 {
>(a3) -0.65/13 47} 25. Rf3 {(Nxg3) 1.41/13 24} Ba5 {-0.46/13 27} 26. Qc2 {
>1.50/13 32} Ra7 {(Ba6) -0.23/13 48} 27. Rxg3 {2.04/13 33} Qh4 {0.14/13 63} 28.
>Rg4 {(Kh2) 2.02/14 33} Nb4 {(Qh6) 0.13/13 19} 29. Qd2 {(Qe2) 2.32/14 58} Qe7 {
>0.16/15 30} 30. Nd6 {2.34/13 25} Nd5 {0.36/14 43} 31. Qf2 {(Qc2) 2.29/12 37}
>Ba6 {0.40/12 23} 32. Kh2 {(a3) 2.55/13 100} Qf8 {(b4) 0.67/12 29} 33. e4 {
>(Rg1) 2.70/12 26} Nc7 {(Ne7) 1.07/13 26} 34. a3 {(Qh4) 2.57/12 18} Rd7 {
>(Ne8) 1.11/13 21} 35. f5 {(Qh4) 3.38/13 24} Ra8 {(Qg8) 1.15/14 28} 36. f6 {
>(Qh4) 3.50/12 16} g6 {(Rb8) 2.19/14 55} 37. Qh4 {3.96/13 18} Bd2 {2.34/14 29}
>38. Rxg6 {4.55/13 30} b4 {3.45/14 45} 39. Bf1 {(Rd1) 4.35/11 14} Ra7 {
>3.76/11 28} 40. Bxa6 {5.64/10 5} Nxa6 {5.68/11 20} 41. Rf1 {(Rag1) 8.71/13 39}
>Rf7 {6.52/12 27} 42. Nxf7+ {12.23/15 42} Rxf7 {10.12/15 37} 43. Rg7 {
>15.29/15 25} h6 {11.16/16 25} 44. Rfg1 {(Qh5) 15.88/15 22} 1-0
>
>I wonder how fruit WCCC played fxg3 with score of 1.53 pawns for black when
>Fruit2.1 that is supposed to be weaker does not like that move and never get
>scores that are close to 1.53(I do not think that a small bonus for the side to
>move can be the reason for it)
>
>Note that the time control suggest that the game was played on hardware that is
>twice faster than 2 ghz so it also seems that fruit WCCC did not search very
>deep because Fruit2.1 with clean hash could get depth 14 in 28 seconds on A3000
>when in the pgn it needed 40 seconds when I expected Fruit WCCC to do it faster
>thanks to pruning that was added.
>
>
>
>Fritz 9 - Fruit WCCC'05, 20'/40 2005
>r1br2k1/2b1q1pp/2p1p3/ppPnP3/3BNp2/5PPP/P2QP1B1/R4RK1 b - - 0 1
>
>Analysis by Fruit 2.1:
>
>22...fxg3 23.Nxg3
> µ (-1.06) Depth: 1/2 00:00:00
>22...fxg3 23.Nxg3 Bb7
> µ (-1.22) Depth: 2/9 00:00:00
>22...fxg3 23.Nxg3 Bb7 24.Ne4
> µ (-0.90) Depth: 3/9 00:00:00
>22...Ne3 23.gxf4 Nxf1 24.Kxf1
> µ (-1.22) Depth: 3/9 00:00:00
>22...Ne3 23.gxf4 Nxf1 24.Kxf1 Kh8
> µ (-1.39) Depth: 4/12 00:00:00
>22...Ne3 23.gxf4 Nxf1 24.Kxf1 Kh8 25.Rd1
> µ (-1.08) Depth: 5/13 00:00:00
>22...Ne3 23.gxf4 Nxf1 24.Kxf1 Ba6 25.Rd1 b4
> µ (-1.32) Depth: 6/15 00:00:00 35kN
>22...Ne3 23.gxf4 Nxf1 24.Kxf1 Rf8 25.Rd1 Qh4 26.e3
> µ (-1.16) Depth: 7/20 00:00:00 76kN
>22...Ne3 23.gxf4 Nxf1 24.Kxf1 Ba6 25.Nd6 Rf8 26.Rd1 b4
> µ (-1.28) Depth: 8/20 00:00:01 157kN
>22...Ne3 23.gxf4 Nxf1 24.Kxf1 b4 25.Nd6 Rf8 26.a3 b3 27.Rd1
> µ (-1.22) Depth: 9/23 00:00:01 483kN
>22...Ne3 23.gxf4 Nxf1 24.Rxf1 b4 25.Nd6 Ba6 26.Rd1 Rd7 27.Kh2 Rf8
> µ (-1.23) Depth: 10/23 00:00:01 883kN
>22...Ne3 23.gxf4 Nxf1 24.Rxf1 b4 25.Nd6 Ba6 26.Rd1 Rf8 27.Kh2 Kh8 28.e4
> µ (-1.19) Depth: 11/24 00:00:02 1932kN
>22...Ne3 23.gxf4 Nxf1 24.Rxf1 b4 25.Rd1 Ba6 26.Nd6 Rf8 27.e4 Qh4 28.f5 exf5
>29.Nxf5
> µ (-1.28) Depth: 12/31 00:00:05 4124kN
>22...Ne3 23.gxf4 Nxf1 24.Rxf1 b4 25.Rd1 Kh8 26.Nd6 g6 27.Kh2 Ba6 28.e4 Rd7
>29.Qc2
> µ (-1.15) Depth: 13/31 00:00:11 9055kN
>22...Ne3 23.gxf4 Nxf1 24.Rxf1 b4 25.Rd1 Kh8 26.Nd6 g6 27.Kh2 Ba6 28.e4 Rd7 29.f5
>exf5 30.exf5 gxf5
> µ (-1.17) Depth: 14/34 00:00:28 22738kN
>22...Ne3 23.gxf4 Nxf1 24.Bxf1 Kh8 25.Nd6 g6 26.e3 Ba6 27.Rd1 b4 28.Bxa6 Rxa6
>29.a3 Rb8 30.Qe2 Ra7
> µ (-0.99) Depth: 15/37 00:01:14 61421kN
>22...Ne3 23.gxf4 Nxf1 24.Bxf1 Kh8 25.Nd6 Rf8 26.e3 g6 27.Bd3 b4 28.Kh1 Ba6
>29.Be4 Qd7 30.Rg1 Rad8
> µ (-1.01) Depth: 16/39 00:02:52 137394kN
>22...Ne3 23.gxf4 Nxf1 24.Bxf1 Kh8 25.Nd6 Rf8 26.Kh2 g6 27.Rd1 a4 28.e3 Ba5
>29.Qc2 Bb7 30.Bd3 Rad8 31.Bc3 Bxc3 32.Qxc3
> µ (-0.86) Depth: 17/40 00:06:06 294079kN
>
>(, 27.09.2005)
I wonder too. On my nachine WCCC'05 rejects fxg3 on depth 14. However contrary
to 2.1 it considers fxg3 with high scores for black until depth 14. Maybe some
kind of pruning effect...
Joachim
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.