Stockfish 1.7

[Uri Blass]

What I noticed most is that the evaluation of SF 1.7 is very close to R3 ... usually within .10 cp.

I do not see it
Here are examples:

I see significant higher evaluations by stockfish(see moves near move 40 in the first game)

I see strong disagreement in evaluation in the second game that I post
in moves 20-30 of the second game stockfish evaluates itself as better when rybka evaluates the position as close to equal.

[Event "URI-AMD, Blitz:5'+5""]
[Site "URI-AMD"]
[Date "2010.04.09"]
[Round "2"]
[White "Stockfish 1.7 JA"]
[Black "Rybka 3 1-cpu 32-bit"]
[Result "1-0"]
[ECO "A30"]
[Annotator "0.32;0.22"]
[PlyCount "181"]
[TimeControl "300+5"]

{W=17.3 ply; 539kN/s B=11.5 ply; 23kN/s} 1. c4 c5 2. Nf3 Nf6 3. Nc3 e6 4. g3
b6 5. Bg2 Bb7 6. O-O Be7 7. d4 cxd4 8. Qxd4 d6 9. Rd1 a6 10. b3 Nbd7 11. e4 {
Both last book move} Qc8 {0.22/12 25} 12. Ba3 {0.32/16 16} Nc5 {0.10/12 8} 13.
Qe3 {(e4-e5) 0.40/17 11} O-O {-0.05/11 13} 14. Nd4 {0.28/17 16} Qc7 {
-0.02/12 32} 15. f3 {0.40/16 10} h6 {(Nf6-d7) -0.01/12 39} 16. Rd2 {
(Ba3-b2) 0.40/16 13} Rac8 {(Rf8-c8) 0.00/12 31} 17. Rad1 {0.48/17 17} Rfe8 {
0.02/11 11} 18. Nc2 {(Qe3-f2) 0.52/16 11} Rb8 {(Nf6-h5) 0.03/10 12} 19. Qe2 {
0.40/17 17} Rbc8 {(Re8-c8) 0.02/11 11} 20. Bb2 {0.48/18 14} Ncd7 {
(Bb7-c6) 0.03/11 14} 21. Ne3 {0.44/17 9} Red8 {(Qc7-b8) 0.04/12 52} 22. Ba3 {
(f3-f4) 0.48/17 11} Nc5 {0.00/11 5} 23. Qf2 {(Ba3-b2) 0.44/17 12} Bc6 {
0.00/12 26} 24. Rc1 {(Qf2-e2) 0.44/17 11} Qb8 {-0.05/11 11} 25. Nc2 {
(Ba3-b2) 0.40/16 13} Qc7 {(Bc6-d7) 0.00/12 29} 26. Nd4 {(Nc2-e3) 0.44/16 11}
Ba8 {(Bc6-b7) 0.00/11 6} 27. Bh3 {(Ba3-b2) 0.44/16 12} Kh7 {(Ba8-b7) 0.00/10 10
} 28. b4 {(Ba3-b2) 0.64/16 12} Ncd7 {0.32/11 7} 29. Bxe6 {0.72/17 13} fxe6 {
0.32/10 0} 30. Nxe6 {0.68/18 11} Qa7 {(Qc7-b8) 0.23/11 9} 31. Nxd8 {0.72/15 17}
Rxd8 {0.22/12 11} 32. Nd5 {0.60/15 10} Rf8 {0.31/12 29} 33. Qe3 {
(Rd2-d1) 0.56/16 8} Nxd5 {0.37/10 14} 34. exd5 {0.44/16 22} Bg5 {
(Be7-f6) 0.45/10 6} 35. f4 {0.60/15 7} Bf6 {0.38/11 8} 36. g4 {
(Qe3-d3+) 0.84/16 13} g6 {0.34/9 11} 37. Rf1 {(Rd2-e2) 1.37/15 10} Rf7 {
(Qa7-c7) 0.58/8 13} 38. Qe6 {(g4-g5) 1.49/16 14} Re7 {0.88/9 8} 39. Qxd6 {
1.61/12 1} Re4 {0.48/10 7} 40. f5 {1.25/16 10} gxf5 {0.48/11 12} 41. Rxf5 {
(g4xf5) 1.53/15 7} Rxc4 {0.54/11 10} 42. h3 {1.49/16 10} b5+ {0.67/11 8} 43.
Kg2 {(Rd2-f2) 1.53/17 9} Qc7 {0.49/10 7} 44. Rxf6 {1.41/18 10} Nxf6 {0.77/11 6}
45. Qxf6 {1.33/10 1} Bxd5+ {0.77/11 7} 46. Rxd5 {1.33/13 1} Rc2+ {0.77/9 0} 47.
Kf3 {1.41/17 7} Rc3+ {0.71/11 8} 48. Ke4 {1.65/18 9} Rxa3 {0.71/11 3} 49. Qf5+
{1.61/18 9} Kh8 {0.71/10 0} 50. Rd7 {1.61/17 7} Qc6+ {0.71/11 9} 51. Kd4 {
1.85/19 10} Qc3+ {(Qc6-c4+) 1.03/12 8} 52. Kd5 {1.89/16 0} Qc4+ {0.92/12 7} 53.
Ke5 {1.85/20 8} Re3+ {0.92/13 6} 54. Kf6 {1.85/21 6} Qc6+ {0.91/13 4} 55. Kf7 {
1.85/17 1} Qc4+ {0.91/12 0} 56. Rd5 {1.81/23 10} Qc7+ {0.91/12 0} 57. Qd7 {
1.61/22 7} Qb8 {0.91/12 0} 58. Qd8+ {1.57/21 8} Qxd8 {0.91/12 0} 59. Rxd8+ {
1.73/16 0} Kh7 {0.91/17 12} 60. Rd6 {1.49/22 8} Rf3+ {1.11/16 7} 61. Rf6 {
1.69/22 8} Rxh3 {1.11/17 2} 62. Rxa6 {1.77/21 9} Rg3 {(Rh3-f3+) 1.11/18 10} 63.
Rb6 {2.54/18 22} Rxg4 {(Rg3-f3+) 1.11/17 3} 64. Rxb5 {2.34/17 6} Rg7+ {
(Rg4-f4+) 1.11/17 6} 65. Ke6 {3.23/17 5} Rg6+ {1.11/17 6} 66. Kd7 {3.23/19 8}
Rg7+ {1.11/17 9} 67. Kd6 {(Kd7-c6) 3.31/17 8} Rg6+ {1.11/14 8} 68. Kc7 {
3.11/19 8} Rg7+ {1.11/13 7} 69. Kb6 {3.67/18 8} Rg6+ {2.26/12 6} 70. Ka5 {
3.59/18 5} Rg8 {(Rg6-g7) 2.26/14 6} 71. Rb6 {(Rb5-d5) 4.12/17 6} h5 {0.75/13 6}
72. b5 {4.04/16 7} Rg2 {(h5-h4) 1.23/12 6} 73. a4 {(Rb6-c6) 6.22/16 6} h4 {
(Rg2-g8) 3.32/10 5} 74. Rc6 {(Rb6-d6) 7.95/17 5} Rg5 {(h4-h3) 3.32/11 5} 75.
Rc1 {10.14/18 8} h3 {(Rg5-e5) 3.32/10 5} 76. Rh1 {11.91/19 8} Kg6 {4.91/10 5}
77. Ka6 {(Rh1xh3) 13.37/17 6} Re5 {(Rg5-c5) 5.11/8 5} 78. b6 {
(Rh1xh3) 19.71/17 10} Re6 {(Re5-e4) 5.58/8 5} 79. Rxh3 {77.09/17 6} Kg5 {
(Re6-d6) 6.75/9 5} 80. Ka7 {(a4-a5) 82.87/22 7} Kg4 {6.92/7 4} 81. Rh1 {
(b6-b7) 82.92/23 6} Kf5 {(Kg4-f3) 10.29/9 5} 82. b7 {(Rh1-f1+) 95.45/16 7} Re7
{10.07/9 5} 83. a5 {(Rh1-f1+) 101.67/15 11} Rd7 {(Re7-c7) 11.80/8 5} 84. Ka8 {
(a5-a6) 95.81/17 7} Rd3 {(Rd7xb7) 17.31/6 5} 85. b8=Q {#11/11 2} Rd2 {
(Rd3-e3) 18.12/7 5} 86. Rf1+ {(Qb8-b5+) #8/10 1} Rf2 {(Kf5-e6) #11/8 4} 87.
Rxf2+ {#6/6 0} Ke6 {(Kf5-e4) #11/7 2} 88. Re2+ {(a5-a6) #5/6 0} Kd5 {
(Ke6-f5) #6/3 0} 89. Qb4 {#3/6 0} Kc6 {#3/3 0} 90. Rd2 {(Qb4-c4+) #2/6 0} Kc7 {
#2/3 0} 91. Qc3# {#1/6 0} 1-0


[Event "URI-AMD, Blitz:1'+1""]
[Site "URI-AMD"]
[Date "2010.04.08"]
[Round "11"]
[White "Rybka 3 1-cpu 32-bit"]
[Black "Stockfish 1.7 JA"]
[Result "0-1"]
[ECO "B25"]
[Annotator "0.03;0.04"]
[PlyCount "190"]
[TimeControl "60+1"]

{W=8.8 ply; 23kN/s B=13.3 ply; 426kN/s} 1. e4 c5 2. Nc3 Nc6 3. g3 g6 4. Bg2
Bg7 5. d3 d6 6. f4 e6 7. Nf3 Nge7 8. O-O O-O 9. Be3 Nd4 {Both last book move}
10. Rb1 {0.03/9 7} Bd7 {(Qd8-a5) 0.04/13 2} 11. Qd2 {(Nc3-e2) 0.08/10 4} Rc8 {
(Qd8-c7) -0.16/13 4} 12. Ne2 {0.05/9 2} Nxf3+ {(Ne7-c6) 0.00/12 4} 13. Bxf3 {
0.06/10 3} d5 {(b7-b6) -0.20/15 3} 14. b4 {(Rf1-d1) 0.09/10 5} cxb4 {-0.12/13 4
} 15. Rxb4 {(Be3xa7) 0.08/11 7} b6 {(Qd8-c7) -0.24/14 3} 16. e5 {
(Be3-d4) -0.07/9 4} f6 {(Ne7-f5) -0.52/13 2} 17. exf6 {0.02/9 1} Bxf6 {
-0.52/14 2} 18. Bd4 {(Be3-f2) 0.00/10 2} Nf5 {-0.52/14 3} 19. Bxf6 {0.00/11 2}
Qxf6 {-0.76/14 3} 20. c3 {(c2-c4) -0.03/11 2} Qe7 {-0.64/13 2} 21. Re1 {
(g3-g4) -0.04/11 4} Qc5+ {-0.68/14 2} 22. d4 {-0.05/12 2} Qc7 {-0.84/16 4} 23.
Qd3 {(Rb4-b3) -0.04/11 3} Kg7 {(Nf5-d6) -0.72/15 2} 24. Bg2 {
(Rb4-b3) -0.04/10 3} h6 {(Nf5-d6) -0.76/16 3} 25. Rbb1 {(Rb4-b3) -0.04/10 3}
Nd6 {-0.84/14 3} 26. Rb4 {-0.08/9 3} Rf5 {(Nd6-c4) -0.88/14 7} 27. Qa6 {
(Bg2-f3) -0.05/9 2} Bc6 {(Rf5-f7) -0.80/12 2} 28. Qd3 {-0.05/9 3} Bb7 {
-0.88/13 2} 29. Rb3 {(Bg2-h3) -0.03/9 2} Rf6 {(Qc7-d7) -0.80/12 1} 30. Nc1 {
(Bg2-f3) -0.04/9 2} Ne4 {-0.80/13 1} 31. Ne2 {-0.08/9 2} Bc6 {
(Rf6-f7) -0.92/14 7} 32. Rbb1 {(Bg2xe4) -0.01/9 3} Nd6 {(Rf6-f8) -1.05/12 5}
33. Nc1 {(Rb1-b3) 0.00/10 2} Ba4 {(Bc6-b7) -1.05/13 1} 34. Ne2 {
(g3-g4) -0.16/10 4} Qc4 {(Qc7-d7) -1.13/14 2} 35. Qd2 {(Qd3xc4) -0.24/9 1} Bb5
{-1.33/13 1} 36. Qe3 {(Rb1-b4) -0.34/9 3} Qd3 {-1.37/12 1} 37. Qf2 {
(Qe3xd3) -0.36/9 1} Ba4 {(Qd3-f5) -1.73/12 2} 38. Ra1 {(Re1-c1) -0.44/9 3} Nc4
{(Rf6-f8) -1.81/12 1} 39. Rab1 {(a2-a3) -0.57/9 4} Ne3 {(Qd3-e3) -2.26/13 2}
40. Qf3 {(Re1-c1) -0.63/8 3} Qd2 {-2.98/12 2} 41. Ra1 {(Re1-c1) -1.72/7 3} Bc2
{-3.55/14 1} 42. Bh3 {-2.19/8 1} Nd1 {-3.23/15 2} 43. Rf1 {(Kg1-f1) -2.28/9 1}
Be4 {-3.83/15 1} 44. Raxd1 {-2.28/10 1} Qxd1 {-3.87/13 1} 45. Qe3 {
(Rf1xd1) -2.28/11 1} Qc2 {(Qd1-d3) -3.91/14 1} 46. a3 {-2.24/10 1} Qa2 {
-3.75/15 1} 47. Bg2 {(g3-g4) -2.25/10 1} Bxg2 {-3.67/15 4} 48. Kxg2 {-2.25/9 0}
Qxa3 {-3.39/14 3} 49. Qe5 {-2.23/10 2} Kf7 {-3.39/13 1} 50. g4 {-2.13/10 2} Qe7
{-3.35/12 1} 51. h4 {(Rf1-f3) -2.24/9 1} Qd7 {(a7-a5) -3.47/12 3} 52. Ng3 {
(Qe5-e3) -1.99/8 2} Ke7 {-3.35/11 1} 53. Qe3 {-2.09/7 0} Qc6 {-3.27/12 2} 54.
Rf3 {(Rf1-e1) -2.13/7 2} a5 {-3.63/11 2} 55. h5 {(f4-f5) -2.26/8 1} Rg8 {
(Ke7-d7) -3.87/11 2} 56. Kh3 {-2.38/7 1} g5 {-3.79/12 1} 57. f5 {-2.56/8 1} Re8
{(a5-a4) -3.95/13 1} 58. Qc1 {-2.50/7 1} a4 {-4.16/12 1} 59. Rf2 {
(Qc1-a3+) -2.50/8 1} Rc8 {-3.91/11 1} 60. Qa3+ {(Rf2-f3) -2.47/8 1} Qd6 {
-4.32/11 1} 61. Qb2 {(Qa3-c1) -2.47/8 1} Kd8 {-4.08/12 2} 62. Qa1 {
(Rf2-e2) -2.63/7 1} b5 {-4.24/11 1} 63. Kg2 {(Qa1-e1) -2.66/7 1} Kd7 {
(a4-a3) -4.44/10 1} 64. Qb2 {(Rf2-f3) -2.44/7 1} Kc6 {-4.16/11 1} 65. Rf1 {
-2.65/8 1} a3 {-4.12/14 1} 66. Qb3 {-2.58/9 1} Ra8 {-4.00/14 1} 67. Rb1 {
(Rf1-a1) -2.58/9 1} Ra5 {-4.00/12 1} 68. Ra1 {-2.58/10 1} exf5 {
(Kc6-b6) -4.88/13 1} 69. Nxf5 {-2.58/8 1} Qf4 {-4.84/11 1} 70. Ne7+ {
(Qb3-d1) -2.58/8 1} Kd7 {-3.91/12 1} 71. Qxd5+ {-2.38/9 1} Ke8 {
(Kd7xe7) -3.67/12 1} 72. Qg8+ {(Ne7-c6) -3.15/7 1} Rf8 {-4.96/11 1} 73. Qe6 {
-3.15/9 1} Qf2+ {-6.02/12 1} 74. Kh1 {-3.15/8 0} Rf6 {-5.53/13 1} 75. Qc8+ {
-3.15/9 1} Kxe7 {-9.57/12 1} 76. Qc7+ {-3.15/9 1} Kf8 {-8.00/14 1} 77. Qxa5 {
-3.15/10 1} Qh4+ {-16.12/14 1} 78. Kg1 {-3.15/8 0} Qxg4+ {-17.05/14 1} 79. Kh1
{-3.15/7 0} Qxh5+ {-19.91/14 1} 80. Kg2 {(Kh1-g1) -3.15/9 2} Qg4+ {
(Qh5-e2+) -22.74/14 1} 81. Kh1 {-4.53/8 2} Qh3+ {(Qg4-h4+) -25.57/14 1} 82. Kg1
{-4.53/7 0} Qe3+ {-25.85/13 1} 83. Kh1 {-4.53/7 2} Rf2 {(Rf6-f4) -91.75/15 1}
84. Qa8+ {-11.88/7 1} Kf7 {(Kf8-g7) -92.00/16 1} 85. Qd5+ {(Qa8-g2) -14.19/6 1}
Kg6 {-93.15/16 1} 86. Qc6+ {-14.89/8 1} Kh5 {-93.15/16 1} 87. Qg2 {-14.40/6 1}
Rxg2 {-93.05/10 0} 88. Kxg2 {-15.62/7 1} Qxc3 {-93.15/15 1} 89. Rxa3 {
(Ra1-h1+) -20.27/7 1} Qxa3 {-#8/12 1} 90. d5 {(Kg2-f2) -22.35/7 1} Qe3 {
(Kh5-g4) -#6/8 0} 91. d6 {-21.60/6 1} g4 {(b5-b4) -#5/6 0} 92. d7 {-#4/5 0}
Qf3+ {(g4-g3) -#4/6 0} 93. Kh2 {-#4/5 0} Qf2+ {(g4-g3+) -#3/6 0} 94. Kh1 {
-#3/3 0} g3 {-#2/6 0} 95. d8=Q {-#2/3 0} Qf1# {-#1/6 0} 0-1

[yanquis1972]

What I noticed most is that the evaluation of SF 1.7 is very close to R3 ... usually within .10 cp.

I do not see it

i dont see it either unfortunately. in fact i've barely seen a position where the eval is within .10 cps oddly enough. of course stockfish fluctuates like crazy from ply to ply.

[Trahald]

Core i5 430M (Quad) I get threads = 2 by default.

Thanks to Tord, Marco and Joona as always.

[zamar]

... we don't expect a big ELO increase from 1.6.X series (if any),..

That is the best understatement I have read for a long time!

You missed the important point:

Indeed new search is completely untested with long time controls, so it might scale much better or much worse than 1.6 series.

Honestly speaking I think that we were really lucky this time. Based on early results it seems that quickly dropping effective branching factor is a very good thing at long time controls, but we had no idea about this.

[M ANSARI]

What I noticed most is that the evaluation of SF 1.7 is very close to R3 ... usually within .10 cp.

I do not see it
Here are examples:

I see significant higher evaluations by stockfish(see moves near move 40 in the first game)

I see strong disagreement in evaluation in the second game that I post
in moves 20-30 of the second game stockfish evaluates itself as better when rybka evaluates the position as close to equal.

[Event "URI-AMD, Blitz:5'+5""]
[Site "URI-AMD"]
[Date "2010.04.09"]
[Round "2"]
[White "Stockfish 1.7 JA"]
[Black "Rybka 3 1-cpu 32-bit"]
[Result "1-0"]
[ECO "A30"]
[Annotator "0.32;0.22"]
[PlyCount "181"]
[TimeControl "300+5"]

{W=17.3 ply; 539kN/s B=11.5 ply; 23kN/s} 1. c4 c5 2. Nf3 Nf6 3. Nc3 e6 4. g3
b6 5. Bg2 Bb7 6. O-O Be7 7. d4 cxd4 8. Qxd4 d6 9. Rd1 a6 10. b3 Nbd7 11. e4 {
Both last book move} Qc8 {0.22/12 25} 12. Ba3 {0.32/16 16} Nc5 {0.10/12 8} 13.
Qe3 {(e4-e5) 0.40/17 11} O-O {-0.05/11 13} 14. Nd4 {0.28/17 16} Qc7 {
-0.02/12 32} 15. f3 {0.40/16 10} h6 {(Nf6-d7) -0.01/12 39} 16. Rd2 {
(Ba3-b2) 0.40/16 13} Rac8 {(Rf8-c8) 0.00/12 31} 17. Rad1 {0.48/17 17} Rfe8 {
0.02/11 11} 18. Nc2 {(Qe3-f2) 0.52/16 11} Rb8 {(Nf6-h5) 0.03/10 12} 19. Qe2 {
0.40/17 17} Rbc8 {(Re8-c8) 0.02/11 11} 20. Bb2 {0.48/18 14} Ncd7 {
(Bb7-c6) 0.03/11 14} 21. Ne3 {0.44/17 9} Red8 {(Qc7-b8) 0.04/12 52} 22. Ba3 {
(f3-f4) 0.48/17 11} Nc5 {0.00/11 5} 23. Qf2 {(Ba3-b2) 0.44/17 12} Bc6 {
0.00/12 26} 24. Rc1 {(Qf2-e2) 0.44/17 11} Qb8 {-0.05/11 11} 25. Nc2 {
(Ba3-b2) 0.40/16 13} Qc7 {(Bc6-d7) 0.00/12 29} 26. Nd4 {(Nc2-e3) 0.44/16 11}
Ba8 {(Bc6-b7) 0.00/11 6} 27. Bh3 {(Ba3-b2) 0.44/16 12} Kh7 {(Ba8-b7) 0.00/10 10
} 28. b4 {(Ba3-b2) 0.64/16 12} Ncd7 {0.32/11 7} 29. Bxe6 {0.72/17 13} fxe6 {
0.32/10 0} 30. Nxe6 {0.68/18 11} Qa7 {(Qc7-b8) 0.23/11 9} 31. Nxd8 {0.72/15 17}
Rxd8 {0.22/12 11} 32. Nd5 {0.60/15 10} Rf8 {0.31/12 29} 33. Qe3 {
(Rd2-d1) 0.56/16 8} Nxd5 {0.37/10 14} 34. exd5 {0.44/16 22} Bg5 {
(Be7-f6) 0.45/10 6} 35. f4 {0.60/15 7} Bf6 {0.38/11 8} 36. g4 {
(Qe3-d3+) 0.84/16 13} g6 {0.34/9 11} 37. Rf1 {(Rd2-e2) 1.37/15 10} Rf7 {
(Qa7-c7) 0.58/8 13} 38. Qe6 {(g4-g5) 1.49/16 14} Re7 {0.88/9 8} 39. Qxd6 {
1.61/12 1} Re4 {0.48/10 7} 40. f5 {1.25/16 10} gxf5 {0.48/11 12} 41. Rxf5 {
(g4xf5) 1.53/15 7} Rxc4 {0.54/11 10} 42. h3 {1.49/16 10} b5+ {0.67/11 8} 43.
Kg2 {(Rd2-f2) 1.53/17 9} Qc7 {0.49/10 7} 44. Rxf6 {1.41/18 10} Nxf6 {0.77/11 6}
45. Qxf6 {1.33/10 1} Bxd5+ {0.77/11 7} 46. Rxd5 {1.33/13 1} Rc2+ {0.77/9 0} 47.
Kf3 {1.41/17 7} Rc3+ {0.71/11 8} 48. Ke4 {1.65/18 9} Rxa3 {0.71/11 3} 49. Qf5+
{1.61/18 9} Kh8 {0.71/10 0} 50. Rd7 {1.61/17 7} Qc6+ {0.71/11 9} 51. Kd4 {
1.85/19 10} Qc3+ {(Qc6-c4+) 1.03/12 8} 52. Kd5 {1.89/16 0} Qc4+ {0.92/12 7} 53.
Ke5 {1.85/20 8} Re3+ {0.92/13 6} 54. Kf6 {1.85/21 6} Qc6+ {0.91/13 4} 55. Kf7 {
1.85/17 1} Qc4+ {0.91/12 0} 56. Rd5 {1.81/23 10} Qc7+ {0.91/12 0} 57. Qd7 {
1.61/22 7} Qb8 {0.91/12 0} 58. Qd8+ {1.57/21 8} Qxd8 {0.91/12 0} 59. Rxd8+ {
1.73/16 0} Kh7 {0.91/17 12} 60. Rd6 {1.49/22 8} Rf3+ {1.11/16 7} 61. Rf6 {
1.69/22 8} Rxh3 {1.11/17 2} 62. Rxa6 {1.77/21 9} Rg3 {(Rh3-f3+) 1.11/18 10} 63.
Rb6 {2.54/18 22} Rxg4 {(Rg3-f3+) 1.11/17 3} 64. Rxb5 {2.34/17 6} Rg7+ {
(Rg4-f4+) 1.11/17 6} 65. Ke6 {3.23/17 5} Rg6+ {1.11/17 6} 66. Kd7 {3.23/19 8}
Rg7+ {1.11/17 9} 67. Kd6 {(Kd7-c6) 3.31/17 8} Rg6+ {1.11/14 8} 68. Kc7 {
3.11/19 8} Rg7+ {1.11/13 7} 69. Kb6 {3.67/18 8} Rg6+ {2.26/12 6} 70. Ka5 {
3.59/18 5} Rg8 {(Rg6-g7) 2.26/14 6} 71. Rb6 {(Rb5-d5) 4.12/17 6} h5 {0.75/13 6}
72. b5 {4.04/16 7} Rg2 {(h5-h4) 1.23/12 6} 73. a4 {(Rb6-c6) 6.22/16 6} h4 {
(Rg2-g8) 3.32/10 5} 74. Rc6 {(Rb6-d6) 7.95/17 5} Rg5 {(h4-h3) 3.32/11 5} 75.
Rc1 {10.14/18 8} h3 {(Rg5-e5) 3.32/10 5} 76. Rh1 {11.91/19 8} Kg6 {4.91/10 5}
77. Ka6 {(Rh1xh3) 13.37/17 6} Re5 {(Rg5-c5) 5.11/8 5} 78. b6 {
(Rh1xh3) 19.71/17 10} Re6 {(Re5-e4) 5.58/8 5} 79. Rxh3 {77.09/17 6} Kg5 {
(Re6-d6) 6.75/9 5} 80. Ka7 {(a4-a5) 82.87/22 7} Kg4 {6.92/7 4} 81. Rh1 {
(b6-b7) 82.92/23 6} Kf5 {(Kg4-f3) 10.29/9 5} 82. b7 {(Rh1-f1+) 95.45/16 7} Re7
{10.07/9 5} 83. a5 {(Rh1-f1+) 101.67/15 11} Rd7 {(Re7-c7) 11.80/8 5} 84. Ka8 {
(a5-a6) 95.81/17 7} Rd3 {(Rd7xb7) 17.31/6 5} 85. b8=Q {#11/11 2} Rd2 {
(Rd3-e3) 18.12/7 5} 86. Rf1+ {(Qb8-b5+) #8/10 1} Rf2 {(Kf5-e6) #11/8 4} 87.
Rxf2+ {#6/6 0} Ke6 {(Kf5-e4) #11/7 2} 88. Re2+ {(a5-a6) #5/6 0} Kd5 {
(Ke6-f5) #6/3 0} 89. Qb4 {#3/6 0} Kc6 {#3/3 0} 90. Rd2 {(Qb4-c4+) #2/6 0} Kc7 {
#2/3 0} 91. Qc3# {#1/6 0} 1-0


[Event "URI-AMD, Blitz:1'+1""]
[Site "URI-AMD"]
[Date "2010.04.08"]
[Round "11"]
[White "Rybka 3 1-cpu 32-bit"]
[Black "Stockfish 1.7 JA"]
[Result "0-1"]
[ECO "B25"]
[Annotator "0.03;0.04"]
[PlyCount "190"]
[TimeControl "60+1"]

{W=8.8 ply; 23kN/s B=13.3 ply; 426kN/s} 1. e4 c5 2. Nc3 Nc6 3. g3 g6 4. Bg2
Bg7 5. d3 d6 6. f4 e6 7. Nf3 Nge7 8. O-O O-O 9. Be3 Nd4 {Both last book move}
10. Rb1 {0.03/9 7} Bd7 {(Qd8-a5) 0.04/13 2} 11. Qd2 {(Nc3-e2) 0.08/10 4} Rc8 {
(Qd8-c7) -0.16/13 4} 12. Ne2 {0.05/9 2} Nxf3+ {(Ne7-c6) 0.00/12 4} 13. Bxf3 {
0.06/10 3} d5 {(b7-b6) -0.20/15 3} 14. b4 {(Rf1-d1) 0.09/10 5} cxb4 {-0.12/13 4
} 15. Rxb4 {(Be3xa7) 0.08/11 7} b6 {(Qd8-c7) -0.24/14 3} 16. e5 {
(Be3-d4) -0.07/9 4} f6 {(Ne7-f5) -0.52/13 2} 17. exf6 {0.02/9 1} Bxf6 {
-0.52/14 2} 18. Bd4 {(Be3-f2) 0.00/10 2} Nf5 {-0.52/14 3} 19. Bxf6 {0.00/11 2}
Qxf6 {-0.76/14 3} 20. c3 {(c2-c4) -0.03/11 2} Qe7 {-0.64/13 2} 21. Re1 {
(g3-g4) -0.04/11 4} Qc5+ {-0.68/14 2} 22. d4 {-0.05/12 2} Qc7 {-0.84/16 4} 23.
Qd3 {(Rb4-b3) -0.04/11 3} Kg7 {(Nf5-d6) -0.72/15 2} 24. Bg2 {
(Rb4-b3) -0.04/10 3} h6 {(Nf5-d6) -0.76/16 3} 25. Rbb1 {(Rb4-b3) -0.04/10 3}
Nd6 {-0.84/14 3} 26. Rb4 {-0.08/9 3} Rf5 {(Nd6-c4) -0.88/14 7} 27. Qa6 {
(Bg2-f3) -0.05/9 2} Bc6 {(Rf5-f7) -0.80/12 2} 28. Qd3 {-0.05/9 3} Bb7 {
-0.88/13 2} 29. Rb3 {(Bg2-h3) -0.03/9 2} Rf6 {(Qc7-d7) -0.80/12 1} 30. Nc1 {
(Bg2-f3) -0.04/9 2} Ne4 {-0.80/13 1} 31. Ne2 {-0.08/9 2} Bc6 {
(Rf6-f7) -0.92/14 7} 32. Rbb1 {(Bg2xe4) -0.01/9 3} Nd6 {(Rf6-f8) -1.05/12 5}
33. Nc1 {(Rb1-b3) 0.00/10 2} Ba4 {(Bc6-b7) -1.05/13 1} 34. Ne2 {
(g3-g4) -0.16/10 4} Qc4 {(Qc7-d7) -1.13/14 2} 35. Qd2 {(Qd3xc4) -0.24/9 1} Bb5
{-1.33/13 1} 36. Qe3 {(Rb1-b4) -0.34/9 3} Qd3 {-1.37/12 1} 37. Qf2 {
(Qe3xd3) -0.36/9 1} Ba4 {(Qd3-f5) -1.73/12 2} 38. Ra1 {(Re1-c1) -0.44/9 3} Nc4
{(Rf6-f8) -1.81/12 1} 39. Rab1 {(a2-a3) -0.57/9 4} Ne3 {(Qd3-e3) -2.26/13 2}
40. Qf3 {(Re1-c1) -0.63/8 3} Qd2 {-2.98/12 2} 41. Ra1 {(Re1-c1) -1.72/7 3} Bc2
{-3.55/14 1} 42. Bh3 {-2.19/8 1} Nd1 {-3.23/15 2} 43. Rf1 {(Kg1-f1) -2.28/9 1}
Be4 {-3.83/15 1} 44. Raxd1 {-2.28/10 1} Qxd1 {-3.87/13 1} 45. Qe3 {
(Rf1xd1) -2.28/11 1} Qc2 {(Qd1-d3) -3.91/14 1} 46. a3 {-2.24/10 1} Qa2 {
-3.75/15 1} 47. Bg2 {(g3-g4) -2.25/10 1} Bxg2 {-3.67/15 4} 48. Kxg2 {-2.25/9 0}
Qxa3 {-3.39/14 3} 49. Qe5 {-2.23/10 2} Kf7 {-3.39/13 1} 50. g4 {-2.13/10 2} Qe7
{-3.35/12 1} 51. h4 {(Rf1-f3) -2.24/9 1} Qd7 {(a7-a5) -3.47/12 3} 52. Ng3 {
(Qe5-e3) -1.99/8 2} Ke7 {-3.35/11 1} 53. Qe3 {-2.09/7 0} Qc6 {-3.27/12 2} 54.
Rf3 {(Rf1-e1) -2.13/7 2} a5 {-3.63/11 2} 55. h5 {(f4-f5) -2.26/8 1} Rg8 {
(Ke7-d7) -3.87/11 2} 56. Kh3 {-2.38/7 1} g5 {-3.79/12 1} 57. f5 {-2.56/8 1} Re8
{(a5-a4) -3.95/13 1} 58. Qc1 {-2.50/7 1} a4 {-4.16/12 1} 59. Rf2 {
(Qc1-a3+) -2.50/8 1} Rc8 {-3.91/11 1} 60. Qa3+ {(Rf2-f3) -2.47/8 1} Qd6 {
-4.32/11 1} 61. Qb2 {(Qa3-c1) -2.47/8 1} Kd8 {-4.08/12 2} 62. Qa1 {
(Rf2-e2) -2.63/7 1} b5 {-4.24/11 1} 63. Kg2 {(Qa1-e1) -2.66/7 1} Kd7 {
(a4-a3) -4.44/10 1} 64. Qb2 {(Rf2-f3) -2.44/7 1} Kc6 {-4.16/11 1} 65. Rf1 {
-2.65/8 1} a3 {-4.12/14 1} 66. Qb3 {-2.58/9 1} Ra8 {-4.00/14 1} 67. Rb1 {
(Rf1-a1) -2.58/9 1} Ra5 {-4.00/12 1} 68. Ra1 {-2.58/10 1} exf5 {
(Kc6-b6) -4.88/13 1} 69. Nxf5 {-2.58/8 1} Qf4 {-4.84/11 1} 70. Ne7+ {
(Qb3-d1) -2.58/8 1} Kd7 {-3.91/12 1} 71. Qxd5+ {-2.38/9 1} Ke8 {
(Kd7xe7) -3.67/12 1} 72. Qg8+ {(Ne7-c6) -3.15/7 1} Rf8 {-4.96/11 1} 73. Qe6 {
-3.15/9 1} Qf2+ {-6.02/12 1} 74. Kh1 {-3.15/8 0} Rf6 {-5.53/13 1} 75. Qc8+ {
-3.15/9 1} Kxe7 {-9.57/12 1} 76. Qc7+ {-3.15/9 1} Kf8 {-8.00/14 1} 77. Qxa5 {
-3.15/10 1} Qh4+ {-16.12/14 1} 78. Kg1 {-3.15/8 0} Qxg4+ {-17.05/14 1} 79. Kh1
{-3.15/7 0} Qxh5+ {-19.91/14 1} 80. Kg2 {(Kh1-g1) -3.15/9 2} Qg4+ {
(Qh5-e2+) -22.74/14 1} 81. Kh1 {-4.53/8 2} Qh3+ {(Qg4-h4+) -25.57/14 1} 82. Kg1
{-4.53/7 0} Qe3+ {-25.85/13 1} 83. Kh1 {-4.53/7 2} Rf2 {(Rf6-f4) -91.75/15 1}
84. Qa8+ {-11.88/7 1} Kf7 {(Kf8-g7) -92.00/16 1} 85. Qd5+ {(Qa8-g2) -14.19/6 1}
Kg6 {-93.15/16 1} 86. Qc6+ {-14.89/8 1} Kh5 {-93.15/16 1} 87. Qg2 {-14.40/6 1}
Rxg2 {-93.05/10 0} 88. Kxg2 {-15.62/7 1} Qxc3 {-93.15/15 1} 89. Rxa3 {
(Ra1-h1+) -20.27/7 1} Qxa3 {-#8/12 1} 90. d5 {(Kg2-f2) -22.35/7 1} Qe3 {
(Kh5-g4) -#6/8 0} 91. d6 {-21.60/6 1} g4 {(b5-b4) -#5/6 0} 92. d7 {-#4/5 0}
Qf3+ {(g4-g3) -#4/6 0} 93. Kh2 {-#4/5 0} Qf2+ {(g4-g3+) -#3/6 0} 94. Kh1 {
-#3/3 0} g3 {-#2/6 0} 95. d8=Q {-#2/3 0} Qf1# {-#1/6 0} 0-1

Actually after changing the parameters as pointed out in the read me file it now seems that it is evaluating back to how older Stockfishes were evaluating. Something weird is going on in my system, as now the evaluations are also jumping again all over again, while initially they were quite stable. It could be a F12 GUI anomaly where it was showing both sides the same evaluation. At least after the read me changes it seems stable on 8 cores, but it still shows 5 CPU's loaded in the F12 GUI. I have a feeling something is not right with the F12 GUI and I might give my older GUI's a try. I managed to get 20 games against R3 in a Silversuite this time at 5_1 but it seems that R3 is still a little stronger (3 more wins). Obviously that is not enough games but the play analysis seems to indicate that as well. We will have to see how things pan out with CEGT and CCRL ... what is sure is that it is very close and most likely SF 2.0 will pull out ahead.

A good thing is that 8 core efficiency seems very good with all 8 cores loaded very close to 100%. I just have to make sure stability is good and what the reasons of the hard crashes are. I just upgraded to F12 GUI and that could be the culprit as they usually have all kinds of eye candy and 3D sound that can cause hardware conflicts that have to be sorted out.

[IWB]

You missed the important point:

Indeed new search is completely untested with long time controls, so it might scale much better or much worse than 1.6 series.

Honestly speaking I think that we were really lucky this time. Based on early results it seems that quickly dropping effective branching factor is a very good thing at long time controls, but we had no idea about this.

You missed to highlight the second part of the quote ...

Nevertheless a BIG THANK YOU to you as well! In a few hourse we will see if this version can become the new No.1!

Bye
Ingo

[Gian-Carlo Pascutto]

The possibility that cross my mind is that the JA builds gave an unexpected push to ELO. Indeed we do all our test with self compiled versions and only the final one is sent to Jim to prepare the release.

Could be that the JA compile this time is really super fast, much faster then ours.

I compiled it myself with gcc so that cannot be the difference.

[Milos]

Honestly speaking I think that we were really lucky this time. Based on early results it seems that quickly dropping effective branching factor is a very good thing at long time controls, but we had no idea about this.

If you just did what Juri did up there (which takes less than hour to test) you would know it is a good idea. Therefore, bad excuse again...

[muxecoid]

Question: What does the value of randomness UCI parameter change? Does it mean randomness in decipawns? Or percents? Rating degradation percentage?

[Tord Romstad]

Question: What does the value of randomness UCI parameter change? Does it mean randomness in decipawns? Or percents? Rating degradation percentage?

The non-technical explanation is that it adds a small amount of randomness to the play, in order to avoid repeating exact games, at the price of slightly weaker play.

The precise explanation is that it adds a small random number to all entries of the piece square tables at the beginning of each new game. The value given in the UCI parameter indicates the maximum magnitude of these random numbers. One unit is roughly 1/200 of a middle game pawn.