Super Fruit Example Game

[NJDenson]

[Event "Rated game, 60m + 15s"]
[Site "Engine Room"]
[Date "2008.11.28"]
[Round "?"]
[White "EL-SHADDAI, Fruit 2.3.5m Test B" 4 cores 9775]
[Black "Morula, ClusterToga 1.4 bet" 24 cores ]
[Result "1-0"]
[ECO "B80"]
[WhiteElo "2420"]
[BlackElo "2560"]
[Annotator "0.12;-0.21"]
[PlyCount "192"]
[EventDate "2008.11.29"]
[TimeControl "3600+15"]

{ClusterToga 1.4 beta5c br5: 21.0 ply; 10,567kN/s Intel(R) Xeon(R) CPU
X7460 @ 2.66GHz 2660MHz, (24 threads), Compmaster Beta 2.0.ctg,
PieceContainer->RestoreDeviceObjects} 1. e4 {B/0 0} c5 {B/0 0} 2. Nf3 {B/0 0}
d6 {B/0 0} 3. d4 {B/0 0} cxd4 {B/0 0} 4. Nxd4 {B/0 0} Nf6 {B/0 0} 5. Nc3 {B/0 0
} a6 {B/0 0} 6. Be3 {B/0 0} e6 {B/0 0} 7. f3 {B/0 0} b5 {B/0 0} 8. Qd2 {B/0 0}
Nbd7 {B/0 0} 9. g4 {B/0 0} b4 {B/0 0} 10. Na4 {B/0 0} h6 {B/0 0} 11. O-O-O {
0.12/19 184} Ne5 {(Qa5) B/0 0} 12. Qxb4 {0.30/20 263} Bd7 {B/0 0} 13. Nc3 {
0.40/20 197} Qc7 {(Rb8) B/0 0} 14. Qb3 {0.49/20 125} Rb8 {B/0 0} 15. Qa3 {
0.57/21 100} Nc4 {B/0 0} 16. Bxc4 {0.27/21 89} Qxc4 {B/0 0} 17. Qa5 {
0.23/21 208} d5 {B/0 0}

[d] 1r2kb1r/3b1pp1/p3pn1p/Q2p4/2qNP1P1/2N1BP2/PPP4P/2KR3R w k - 0 18

On move 18 Fruit makes 18. a3!!
18. a3 {0.10/20 253} Be7 {(Bd6) -0.21/21 138} 19. Nb3 {
(b3) 0.18/20 154} Bd8 {-0.84/22 135} 20. Qc5 {0.00/23 0} Bb6 {-0.77/22 161} 21.
Qxb6 {0.00/22 0} Rxb6 {-0.88/21 155} 22. Bxb6 {0.01/22 0} dxe4 {-0.89/21 150}
23. fxe4 {(Sa5) 0.09/22 0} Qc6 {-1.12/21 121} 24. Bd4 {0.13/22 50} Nxe4 {
-0.79/21 91} 25. Na5 {0.00/22 0} Qa8 {-1.05/21 93} 26. Bxg7 {(Sc4) 0.00/21 56}
Rg8 {-1.19/19 135} 27. Rhe1 {0.00/21 19} Nxc3 {-0.65/20 94} 28. Bxc3 {0.00/21 0
} Rxg4 {-0.69/20 81} 29. b3 {(h3) 0.00/20 52} Qf3 {
EL-SHADDAI,Fruit 2.3.5m Test Be offers a draw -0.58/18 105} 30. Kb2 {0.00/20 0}
f6 {(Rg5) -0.37/19 162} 31. Rd3 {(Tf1) 0.30/19 78} Qf2 {(Qf4) -0.20/21 119} 32.
Red1 {0.70/21 124} Bb5 {-0.13/21 0} 33. R3d2 {0.74/20 115} Qe3 {-0.18/20 0} 34.
a4 {0.64/20 266} Be2 {0.10/22 0} 35. Rd8+ {0.67/20 105} Kf7 {0.00/21 0} 36.
R1d7+ {0.67/20 109} Kg6 {0.00/20 0} 37. Nc6 {0.73/20 104} Bf3 {(Kh5) 0.21/20 0}
38. Ne7+ {(Td3) 1.09/18 67} Kh5 {0.00/21 84} 39. Ng8 {(Td3) 0.95/19 6} Rg1 {
0.11/20 161} 40. Nxf6+ {1.02/20 0} Kh4 {0.14/19 105} 41. Rd2 {1.00/19 5} e5 {
0.18/20 85} 42. R8d6 {(Sd5) 0.89/19 20} e4 {0.00/19 65} 43. Nd7 {0.62/18 251}
Kh3 {(h5) 0.00/23 0} 44. Nb6 {0.80/20 116} h5 {0.00/23 0} 45. Nd5 {0.64/21 87}
Qe1 {0.00/22 4} 46. Nf4+ {0.72/20 74} Kh4 {0.00/22 25} 47. Ka3 {
(Sg6+) 0.75/19 41} Qe3 {0.00/19 82} 48. Nd5 {0.69/20 0} Qc5+ {0.19/19 69} 49.
Bb4 {0.57/21 0} Qa7 {(Ra1+) 0.15/19 93} 50. Re6 {0.82/17 42} Rg8 {
(Rd1) 0.22/18 35} 51. Be7+ {(Kb2) 1.67/17 74} Kh3 {0.29/18 63} 52. Kb2 {
1.69/17 6} Rg4 {(Rg1) 0.54/19 84} 53. Ka2 {(Te5) 1.47/17 93} Rg7 {
(Qg1) 0.24/18 58} 54. Bd6 {(Sf4+) 1.66/17 36} Rf7 {(Rd7) 0.36/18 64} 55. Nc3 {
(Lf4) 1.39/18 60} Qa8 {(Bg4) -0.25/19 63} 56. Bg3 {(Kb2) 0.90/18 89} a5 {
(Rg7) -0.30/18 80} 57. Ne2 {(Kb2) 1.09/18 47} Qc8 {0.00/17 76} 58. Ng1+ {
0.83/19 0} Kg4 {0.00/14 0} 59. Red6 {0.77/19 85} h4 {(Qc3) 0.00/18 0} 60. Bf2 {
1.35/17 35} Qc3 {0.15/18 60} 61. h3+ {1.58/19 0} Kf4 {0.22/18 74} 62. Bxh4 {
1.38/20 0} Rf5 {(Qg7) 0.09/19 231} 63. Rd8 {(Sxf3) 1.56/17 50} Qc5 {
(Bh5) 0.26/17 69} 64. R8d4 {1.70/17 29} Bh5 {0.39/17 10} 65. Kb2 {
(c4) 1.32/18 57} Qb6 {(Ke5) 0.36/17 43} 66. Bf2 {(c3) 1.80/17 37} Qf6 {
(Qe6) 0.15/16 60} 67. c3 {(Ka2) 1.81/16 32} Qa6 {(Qe6) 0.46/17 46} 68. h4 {
2.13/17 33} Qf1 {0.41/19 10} 69. Rd6 {(Td7) 2.07/17 21} Bf7 {(Bg4) 0.21/15 58}
70. Ba7 {(c4) 2.20/16 38} Kg4 {0.73/18 34} 71. Be3 {2.21/18 8} Kh5 {0.65/18 49}
72. Rh6+ {2.01/19 0} Kg4 {0.31/11 0} 73. Rh7 {(c4) 1.80/18 26} Bg6 {0.98/18 33}
74. Rg7 {1.95/19 0} Rf6 {(Kh5) 0.84/19 53} 75. h5 {(Tg8) 2.28/18 73} Kxh5 {
0.96/19 52} 76. Rg8 {(Th2+) 2.16/18 60} Kg4 {0.80/18 41} 77. c4 {
(Td5) 2.16/17 0} Kg3 {0.49/18 31} 78. c5 {(Td5) 1.98/17 16} Qe1 {0.30/19 56}
79. Re2 {1.62/19 0} Qd1 {0.33/20 35} 80. Rc2 {1.64/18 3} Qd5 {0.58/19 28} 81.
Rh8 {1.95/18 13} Qe5+ {(Bh5) 0.68/19 29} 82. Ka2 {2.31/19 41} Bh5 {0.86/19 4}
83. Rd8 {(Tg8+) 2.08/17 69} Rc6 {0.94/18 79} 84. Rg8+ {2.25/19 0} Rg6 {
1.34/21 44} 85. Ne2+ {(Txg6+) 1.68/17 21} Kf3 {0.97/18 40} 86. Bd4 {1.69/20 0}
Qf5 {1.06/18 31} 87. Bc3 {1.72/18 18} Ke3 {0.88/18 7} 88. Nd4 {1.80/18 23} Qd5
{(Qxc5) 1.30/18 46} 89. Rh8 {(Txg6) 2.06/17 23} Bg4 {(Rg8) 2.05/17 40} 90. c6 {
3.02/14 30} Kd3 {(Rg8) 2.49/18 5} 91. c7 {3.24/17 26} Rg8 {2.46/19 6} 92. Rh4 {
(Lxa5) 3.69/17 18} Qg5 {(Bc8) 2.36/15 27} 93. Rhh2 {4.14/17 30} e3 {3.66/16 0}
94. Rhg2 {4.53/17 20} Rc8 {4.68/19 19} 95. Bb2 {4.57/18 7} Qh4 {5.01/19 27} 96.
Rc4 {5.12/19 0} Bd1 {(Bh5) Morula,ClusterToga 1.4 beta gibt auf (Lag: Av=0.
31s, max=0.8s) 5.05/18 39} 1-0

[Graham Banks]

Super Fruit is still private though and likely to remain that way from what I understand?

[NJDenson]

Never say never ... maybe they might bring it back. I dont think I would have found these settings if they had been still working on it. Reason I say this is ..... I have tried just about every single trick to get this thing to play good chess. Only recently its playing very good a long time controls. In fact with the new settings it has not lost a game.

[AdminX]

[d]1r2kb1r/3b1pp1/p3pn1p/Q2p4/2qNP1P1/2N1BP2/PPP4P/2KR3R w k - 0 1

Analysis by Deep Rybka 3 64-bit:

1.a3 Bd6 2.b3 Bc7 3.bxc4 Bxa5[] 4.Nde2 dxe4 5.Nxe4 Nxe4 6.fxe4 Bc6 7.Ng3 0-0 8.Rd6 Rfc8 9.Rhd1 Rb7 10.R1d4 Rcb8 11.Kd1
= (0.17) Depth: 6 00:00:00 20kN
= (-0.22) Depth: 17 00:01:12 14885kN

Why do you give 18. a3 (!!)? Is it new to theory? My book (Powerbook 08) has 17. Qa5 as the novelty here and the move (a3) does not seem hard to find.

Thanks

[Eelco de Groot]

[Event "Rated game, 60m + 15s"]
[Site "Engine Room"]
[Date "2008.11.28"]
[Round "?"]
[White "EL-SHADDAI, Fruit 2.3.5m Test B" 4 cores 9775]
[Black "Morula, ClusterToga 1.4 bet" 24 cores ]
[Result "1-0"]
[ECO "B80"]
[WhiteElo "2420"]
[BlackElo "2560"]
[Annotator "0.12;-0.21"]
[PlyCount "192"]
[EventDate "2008.11.29"]
[TimeControl "3600+15"]

{ClusterToga 1.4 beta5c br5: 21.0 ply; 10,567kN/s Intel(R) Xeon(R) CPU
X7460 @ 2.66GHz 2660MHz, (24 threads), Compmaster Beta 2.0.ctg,
PieceContainer->RestoreDeviceObjects} 1. e4 {B/0 0} c5 {B/0 0} 2. Nf3 {B/0 0}
d6 {B/0 0} 3. d4 {B/0 0} cxd4 {B/0 0} 4. Nxd4 {B/0 0} Nf6 {B/0 0} 5. Nc3 {B/0 0
} a6 {B/0 0} 6. Be3 {B/0 0} e6 {B/0 0} 7. f3 {B/0 0} b5 {B/0 0} 8. Qd2 {B/0 0}
Nbd7 {B/0 0} 9. g4 {B/0 0} b4 {B/0 0} 10. Na4 {B/0 0} h6 {B/0 0} 11. O-O-O {
0.12/19 184} Ne5 {(Qa5) B/0 0} 12. Qxb4 {0.30/20 263} Bd7 {B/0 0} 13. Nc3 {
0.40/20 197} Qc7 {(Rb8) B/0 0} 14. Qb3 {0.49/20 125} Rb8 {B/0 0} 15. Qa3 {
0.57/21 100} Nc4 {B/0 0} 16. Bxc4 {0.27/21 89} Qxc4 {B/0 0} 17. Qa5 {
0.23/21 208} d5 {B/0 0}

[d] 1r2kb1r/3b1pp1/p3pn1p/Q2p4/2qNP1P1/2N1BP2/PPP4P/2KR3R w k - 0 18

On move 18 Fruit makes 18. a3!!

Interesting the high evaluations by Fruit later in the game Nolan!

I could not get Toga to play 18. a3, the eval stays too positive for Black. I'm not really up to date with any theory, certainly not Playchess Najdorf, is 18. a3 the best accepted move? The old HS for Rybka book by Harry Schnapp gives both 18. Bf4 and 18. a3 but only a few games were used for the statistics in the book.

Analysis on very slow hardware and testing some of Bill Mowery's suggestions for sort.cpp in Toga, details on Toga Developers Discussion Board but private as yet;

Build 157 , this build limiting single reply extensions if next call to extended search_full () would start at height (+1)> iteration +6 (+1). Not sure if single reply extensions are more effective near root or near tips, near tips maybe more risk of "exploding" single reply extensions as maybe sometimes happens in Rybka (Seeing all those [] only move signs in Rybka's output, I'm not sure Vas does not just merely want us to believe he is doing single reply extensions. :lol Halve truths and Red Rybkas.. )

// [EdG: single reply extensions hack]

if (best_move != MoveNone && nextbest_value != ValueNone /*&& thirdbest_value != ValueNone*/ && node_type == NodePV && best_value < beta
&& depth_best_move < depth && ((best_value - nextbest_value) >= (300 - 10 * depth)/*||
((best_value - thirdbest_value) >= (500 - 5 * depth))*/) && depth <= 30
&& height <= SearchCurrent[ThreadId]->act_iteration + 6){// [EdG:[]'only move' hack]

[d]1r2kb1r/3b1pp1/p3pn1p/Q2p4/2qNP1P1/2N1BP2/PPP4P/2KR3R w k -

Engine: Toga 1.4 Checkov Beta 4/Toga Mara Beta 6 Build 157 (Athlon 2009 MHz, 256 MB)
by Fabien Letouzey, Thomas Gaksch, Taban Juma, Eelco de Groot

9/35 0:03 +0.40 1.Bf4 Bb4 2.Qc7 Qxc7 3.Bxc7 Rb7
4.Be5 Bxc3 5.bxc3 Ke7 6.exd5 exd5 (1.687.361) 509

10/47 0:13 +0.52 1.Bf4 Bb4 2.Qc7 Qxc7 3.Bxc7 Rb7
4.Be5 Bxc3 5.bxc3 Ke7 6.Rhe1 dxe4
7.Bxf6+ Kxf6 8.Rxe4 (6.745.990) 509

11/47 0:27 +0.37 1.Bf4 Bb4 2.Qc7 Qxc7 3.Bxc7 Rc8
4.Be5 Bxc3 5.bxc3 dxe4 6.Bxf6 gxf6
7.fxe4 e5 8.Nf5 Bxf5 9.gxf5 Rxc3 (14.358.506) 526

12/47 1:05 +0.37 1.Bf4 Bb4 2.Qc7 Qxc7 3.Bxc7 Rc8
4.Be5 Bxc3 5.bxc3 dxe4 6.Bxf6 gxf6
7.fxe4 e5 8.Nf5 Bxf5 9.gxf5 Rxc3 (34.976.344) 537

13/48 3:00 +0.25 1.Bf4 Bb4 2.Qc7 Qxc7 3.Bxc7 Rc8
4.Be5 Bxc3 5.bxc3 dxe4 6.Bxf6 gxf6
7.fxe4 h5 8.gxh5 Rxh5 9.Kb2 Rh4
10.Rhf1 Rxh2 11.Rxf6 Rb8+ 12.Nb3 (97.553.682) 538

14/55 6:55 +0.25 1.Bf4 Bb4 2.Qc7 Qxc7 3.Bxc7 Rc8
4.Be5 Bxc3 5.bxc3 dxe4 6.Bxf6 gxf6
7.fxe4 h5 8.gxh5 Rxh5 9.Kb2 Rh4
10.Rhf1 Rxh2 11.Rxf6 Rb8+ 12.Nb3 (223.901.456) 539

15/55 16:14 -0.04 1.Bf4 Bb4 2.Qc7 Qxc7 3.Bxc7 Rc8
4.Be5 Bxc3 5.bxc3 dxe4 6.Bxf6 gxf6
7.fxe4 h5 8.h3 Rxc3 9.Kb2 Rc8 10.c3 Ke7
11.Kb1 hxg4 12.hxg4 Rxh1 13.Rxh1 (528.638.787) 542

Entering 1.a3 (18. a3 in the game)

[d]1r2kb1r/3b1pp1/p3pn1p/Q2p4/2qNP1P1/P1N1BP2/1PP4P/2KR3R b k -

Engine: Toga 1.4 Checkov Beta 4/Toga Mara Beta 6 Build 157 (Athlon 2009 MHz, 256 MB)
by Fabien Letouzey, Thomas Gaksch, Taban Juma, Eelco de Groot

9/41 0:04 +0.43 1...Be7 2.exd5 Bd8 3.Nf5 exf5 4.Qc5 Rc8
5.Qxc4 Rxc4 6.h3 Ba5 7.Bd4 Bxc3
8.Bxc3 (2.142.658) 470

10/52 0:16 -0.06 1...Be7 2.Nb1 dxe4 3.Bf4 Rb7 4.b3 Qc8
5.fxe4 O-O 6.Nd2 Bc5 7.Nc4 Nxe4 (8.606.741) 519

11/52 0:41 +0.22 1...Be7 2.exd5 Bd8 3.Nf5 exf5 4.Qc5 Rc8
5.Qa7 O-O 6.Rd4 Qc5 7.Qxc5 Rxc5 8.h3 fxg4
9.hxg4 (21.890.675) 529

12/53 1:20 -0.02 1...Be7 2.b3 Qc8 3.Nde2 dxe4 4.fxe4 Nxg4
5.Ba7 Rb7 6.Rhg1 e5 7.Bb6 O-O 8.Nd5 Bg5+
9.Kb1 (43.143.695) 536

13/57 6:00 +0.51 1...Be7 2.exd5 Bd8 3.Nf5 exf5 4.Qc5 Rc8
5.Qxc4 Rxc4 6.Rhe1 Kf8 7.h3 fxg4
8.hxg4 Ba5 9.Bd2 Bc7 (193.284.941) 536

14/64 31:54 +0.14 1...Be7 2.Nde2 Bd8 3.b3 Qc6 4.Qa4 Qxa4
5.bxa4 dxe4 6.Nxe4 Nxe4 7.fxe4 Be7
8.Bf4 Bxa3+ 9.Kd2 Rb2 10.Be5 Bb4+
11.Kc1 (1.069.157.446) 558

15/64 36:50 +0.12 1...Be7 2.Nde2 Bd8 3.b3 Qc6 4.Qa4 Qxa4
5.bxa4 dxe4 6.Nxe4 Nxe4 7.fxe4 Bg5
8.Bxg5 hxg5 9.Rd6 e5 10.Rd5 Bxg4
11.Rxe5+ Kf8 (1.230.892.518) 556

16/64 56:22 +0.48 1...Be7 2.Nde2 Bd8 3.b3 Qc6 4.Qa4 Qxa4
5.bxa4 dxe4 6.Nxe4 Nxe4 7.fxe4 Be7
8.Nc3 Bc6 9.Kd2 Bf6 10.Kd3 Bxc3
11.Kxc3 Bxe4 (1.844.679.927) 545

17/64 69:26 +0.38 1...Be7 2.Nde2 Bd8 3.b3 Qc6 4.Qa4 Qxa4
5.bxa4 dxe4 6.Nxe4 Nxe4 7.fxe4 Be7
8.Nc3 Bc6 9.Bf4 Bxa3+ 10.Kd2 Rd8+
11.Ke2 Rxd1 12.Rxd1 Bc5 13.h4 (2.270.636.041) 544

After 1. a3 Be7 2.Nb3, White moves of Fruit in game 18. a3 Be7 19. Nb3

[d]1r2k2r/3bbpp1/p3pn1p/Q2p4/2q1P1P1/PNN1BP2/1PP4P/2KR3R b k -

Engine: Toga 1.4 Checkov Beta 4/Toga Mara Beta 6 Build 157 (Athlon 2009 MHz, 256 MB)
by Fabien Letouzey, Thomas Gaksch, Taban Juma, Eelco de Groot

9/29 0:01 +0.28 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.Na5 Qc8 7.Nxe4 Nxe4 8.fxe4 e5 (527.635)

10/33 0:03 +0.40 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.Na5 Qc8 7.Nxe4 Nxe4 8.fxe4 h5 9.h3 hxg4
10.hxg4 (1.555.075) 513

11/41 0:09 +0.40 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.Na5 Qc8 7.Nxe4 Nxe4 8.fxe4 h5 9.h3 hxg4
10.hxg4 (5.096.182) 512

12/41 0:29 +0.81 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.Na5 Qc8 7.Nxe4 Nxe4 8.fxe4 O-O
9.Bd4 e5 10.Bxe5 Bxg4 (14.941.041) 511

13/49 1:10 +0.34 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.fxe4 e5 7.Nc5 Rf8 8.h3 Bc6 9.Rd8+ Ke7
10.Rhd1 a5 (36.450.339) 518

14/49 3:22 +0.26 2...Bd8 3.Qc5 Bb6 4.Qxb6 Ke7 5.exd5 Rxb6
6.Bxb6 exd5 7.Kb1 Qf4 (106.130.343) 524

15/71 14:32 +0.35 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.fxe4 Qc6 7.Bd4 Nxe4 8.Na5 Qc7
9.Rd3 Nxc3 10.Bxc3 f5 11.gxf5 exf5
12.Re1+ Kf8 13.h3 Kg8 14.Bxg7 Kxg7
15.Re7+ Kf6 (468.576.883) 536

16/71 19:39 +0.42 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.fxe4 Qc6 7.Bd4 Nxe4 8.Na5 Qc7
9.Rd3 Nxc3 10.Bxc3 f5 11.gxf5 exf5
12.b4 (628.293.386) 532

17/71 29:10 +0.26 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.fxe4 Qc6 7.Bd4 Nxe4 8.Na5 Qc7
9.Nxe4 Qf4+ 10.Kb1 Qxe4 11.Nc4 O-O
12.Rhe1 Qh7 13.Bxg7 Qxg7 14.h3 h5
15.Rxd7 hxg4 (924.690.106) 528

18. a3 {0.10/20 253} Be7 {(Bd6) -0.21/21 138} 19. Nb3 {
(b3) 0.18/20 154} Bd8 {-0.84/22 135} 20. Qc5 {0.00/23 0} Bb6 {-0.77/22 161} 21.
Qxb6 {0.00/22 0} Rxb6 {-0.88/21 155} 22. Bxb6 {0.01/22 0} dxe4 {-0.89/21 150}
23. fxe4 {(Sa5) 0.09/22 0} Qc6 {-1.12/21 121} 24. Bd4 {0.13/22 50} Nxe4 {
-0.79/21 91} 25. Na5 {0.00/22 0} Qa8 {-1.05/21 93}

Toga predicted 25... Qc7 here in its last iteration when analyzing Fruit's 19. Nb3. Eval somewhere in the middle between Fruit and Cluster Toga. Now in latest iteration +0.26 for Black at 17 plies, see above. Twelve half-moves of Super Fruit and Cluster Toga on 24 core hardware correctly predicted by Toga 1.4 Checkov Beta 4 / Toga Mara Beta 6 up to this point

Edit: latest output increasing a bit for Black again

18/71 61:56 +0.66 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.fxe4 Qc6 7.Bd4 Nxe4 8.Na5 Qc7
9.Nxe4 Qf4+ 10.Kb1 Qxe4 11.Nc4 O-O
12.Rhe1 Qh7 13.Bxg7 Ba4 14.Bxf8 Bxc2+ (1.944.521.288) 523

26. Bxg7 {(Sc4) 0.00/21 56}
Rg8 {-1.19/19 135} 27. Rhe1 {0.00/21 19} Nxc3 {-0.65/20 94} 28. Bxc3 {0.00/21 0
} Rxg4 {-0.69/20 81} 29. b3 {(h3) 0.00/20 52} Qf3 {
EL-SHADDAI,Fruit 2.3.5m Test Be offers a draw -0.58/18 105} 30. Kb2 {0.00/20 0}
f6 {(Rg5) -0.37/19 162} 31. Rd3 {(Tf1) 0.30/19 78} Qf2 {(Qf4) -0.20/21 119} 32.
Red1 {0.70/21 124} Bb5 {-0.13/21 0} 33. R3d2 {0.74/20 115} Qe3 {-0.18/20 0} 34.
a4 {0.64/20 266} Be2 {0.10/22 0} 35. Rd8+ {0.67/20 105} Kf7 {0.00/21 0} 36.
R1d7+ {0.67/20 109} Kg6 {0.00/20 0} 37. Nc6 {0.73/20 104} Bf3 {(Kh5) 0.21/20 0}
38. Ne7+ {(Td3) 1.09/18 67} Kh5 {0.00/21 84} 39. Ng8 {(Td3) 0.95/19 6} Rg1 {
0.11/20 161} 40. Nxf6+ {1.02/20 0} Kh4 {0.14/19 105} 41. Rd2 {1.00/19 5} e5 {
0.18/20 85} 42. R8d6 {(Sd5) 0.89/19 20} e4 {0.00/19 65} 43. Nd7 {0.62/18 251}
Kh3 {(h5) 0.00/23 0} 44. Nb6 {0.80/20 116} h5 {0.00/23 0} 45. Nd5 {0.64/21 87}
Qe1 {0.00/22 4} 46. Nf4+ {0.72/20 74} Kh4 {0.00/22 25} 47. Ka3 {
(Sg6+) 0.75/19 41} Qe3 {0.00/19 82} 48. Nd5 {0.69/20 0} Qc5+ {0.19/19 69} 49.
Bb4 {0.57/21 0} Qa7 {(Ra1+) 0.15/19 93} 50. Re6 {0.82/17 42} Rg8 {
(Rd1) 0.22/18 35} 51. Be7+ {(Kb2) 1.67/17 74} Kh3 {0.29/18 63} 52. Kb2 {
1.69/17 6} Rg4 {(Rg1) 0.54/19 84} 53. Ka2 {(Te5) 1.47/17 93} Rg7 {
(Qg1) 0.24/18 58} 54. Bd6 {(Sf4+) 1.66/17 36} Rf7 {(Rd7) 0.36/18 64} 55. Nc3 {
(Lf4) 1.39/18 60} Qa8 {(Bg4) -0.25/19 63} 56. Bg3 {(Kb2) 0.90/18 89} a5 {
(Rg7) -0.30/18 80} 57. Ne2 {(Kb2) 1.09/18 47} Qc8 {0.00/17 76} 58. Ng1+ {
0.83/19 0} Kg4 {0.00/14 0} 59. Red6 {0.77/19 85} h4 {(Qc3) 0.00/18 0} 60. Bf2 {
1.35/17 35} Qc3 {0.15/18 60} 61. h3+ {1.58/19 0} Kf4 {0.22/18 74} 62. Bxh4 {
1.38/20 0} Rf5 {(Qg7) 0.09/19 231} 63. Rd8 {(Sxf3) 1.56/17 50} Qc5 {
(Bh5) 0.26/17 69} 64. R8d4 {1.70/17 29} Bh5 {0.39/17 10} 65. Kb2 {
(c4) 1.32/18 57} Qb6 {(Ke5) 0.36/17 43} 66. Bf2 {(c3) 1.80/17 37} Qf6 {
(Qe6) 0.15/16 60} 67. c3 {(Ka2) 1.81/16 32} Qa6 {(Qe6) 0.46/17 46} 68. h4 {
2.13/17 33} Qf1 {0.41/19 10} 69. Rd6 {(Td7) 2.07/17 21} Bf7 {(Bg4) 0.21/15 58}
70. Ba7 {(c4) 2.20/16 38} Kg4 {0.73/18 34} 71. Be3 {2.21/18 8} Kh5 {0.65/18 49}
72. Rh6+ {2.01/19 0} Kg4 {0.31/11 0} 73. Rh7 {(c4) 1.80/18 26} Bg6 {0.98/18 33}
74. Rg7 {1.95/19 0} Rf6 {(Kh5) 0.84/19 53} 75. h5 {(Tg8) 2.28/18 73} Kxh5 {
0.96/19 52} 76. Rg8 {(Th2+) 2.16/18 60} Kg4 {0.80/18 41} 77. c4 {
(Td5) 2.16/17 0} Kg3 {0.49/18 31} 78. c5 {(Td5) 1.98/17 16} Qe1 {0.30/19 56}
79. Re2 {1.62/19 0} Qd1 {0.33/20 35} 80. Rc2 {1.64/18 3} Qd5 {0.58/19 28} 81.
Rh8 {1.95/18 13} Qe5+ {(Bh5) 0.68/19 29} 82. Ka2 {2.31/19 41} Bh5 {0.86/19 4}
83. Rd8 {(Tg8+) 2.08/17 69} Rc6 {0.94/18 79} 84. Rg8+ {2.25/19 0} Rg6 {
1.34/21 44} 85. Ne2+ {(Txg6+) 1.68/17 21} Kf3 {0.97/18 40} 86. Bd4 {1.69/20 0}
Qf5 {1.06/18 31} 87. Bc3 {1.72/18 18} Ke3 {0.88/18 7} 88. Nd4 {1.80/18 23} Qd5
{(Qxc5) 1.30/18 46} 89. Rh8 {(Txg6) 2.06/17 23} Bg4 {(Rg8) 2.05/17 40} 90. c6 {
3.02/14 30} Kd3 {(Rg8) 2.49/18 5} 91. c7 {3.24/17 26} Rg8 {2.46/19 6} 92. Rh4 {
(Lxa5) 3.69/17 18} Qg5 {(Bc8) 2.36/15 27} 93. Rhh2 {4.14/17 30} e3 {3.66/16 0}
94. Rhg2 {4.53/17 20} Rc8 {4.68/19 19} 95. Bb2 {4.57/18 7} Qh4 {5.01/19 27} 96.
Rc4 {5.12/19 0} Bd1 {(Bh5) Morula,ClusterToga 1.4 beta gibt auf (Lag: Av=0.
31s, max=0.8s) 5.05/18 39} 1-0

[NJDenson]

In my books I have a3 is losing as well. I had fruit playing with a smaller book so i could see if it can find a novelty some where. From a few of the other games I played I knew it would find something theory did not think was good because of earlier games I played against Rybka 3. The settings i used for Rybka 3 were not exactly the same as the toga game settings .... I made fruit prune less .... perhaps i need to play a match between the two settings. I used these settings in my KOE game and played naum 3.1 x64. I had a bad book line that put me .83 for an eval after coming out of book. Fruit was able to find a draw out of what I believe to be a lost position. What I am noticing about this engine is that it loves to attack. When it gets into a difficult position it finds some kind of attack to salvage game.

[Eelco de Groot]

Thanks for the explanation Nolan! An engine that can come back in a worse position sounds certainly valuable! About 18.a3, if it is really worse then Cluster Toga at least did not manage to show it.

It is a pity we do not hear more about Fabien Letouzey and Ryan, I hope they are continuing with Chess 64. I suppose we can only dream about open source versions of Fruit...

More as a test of Toga Checkov I let the analysis after 19. Nb3 continue. I was happy that I got some deeper plydepths because with the previous versions the single reply extensions dominated such a big part of the searchtree that I could not get past 14 ply often. It depends on the position I suppose, but at least I got a bit deeper now with the cap on the extensions. I think that Bill Mowery's code for capture_is_good may have helped too but I would have to check with previous versions to make sure how much that helps here;

After 1. a3 Be7 2.Nb3

1r2k2r/3bbpp1/p3pn1p/Q2p4/2q1P1P1/PNN1BP2/1PP4P/2KR3R b k -

Engine: Toga 1.4 Checkov Beta 4/Toga Mara Beta 6 Build 157 (Athlon 2009 MHz, 256 MB)
by Fabien Letouzey, Thomas Gaksch, Taban Juma, Eelco de Groot

17/71 29:10 +0.26 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.fxe4 Qc6 7.Bd4 Nxe4 8.Na5 Qc7
9.Nxe4 Qf4+ 10.Kb1 Qxe4 11.Nc4 O-O
12.Rhe1 Qh7 13.Bxg7 Qxg7 14.h3 h5
15.Rxd7 hxg4 (924.690.106) 528

18/71 61:56 +0.66 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.fxe4 Qc6 7.Bd4 Nxe4 8.Na5 Qc7
9.Nxe4 Qf4+ 10.Kb1 Qxe4 11.Nc4 O-O
12.Rhe1 Qh7 13.Bxg7 Ba4 14.Bxf8 Bxc2+ (1.944.521.288) 523

19/71 85:27 +0.47 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.Na5 Qc8 7.Nxe4 Nxe4 8.fxe4 e5 9.h3 h5
10.g5 Bxh3 11.Rd5 f6 12.gxf6 gxf6
13.Rhd1 f5 14.exf5 Kf7 15.Rxe5 Bxf5 (2.672.504.337) 521

20/76 203:20 +0.72 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.Na5 Qc8 7.Nxe4 Nxe4 8.fxe4 O-O
9.Rd6 Ba4 10.c4 Qb8 11.c5 Rc8 12.Kd2 Rd8 (6.389.554.246) 523

21/76 263:48 +0.71 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.Na5 Qc8 7.Nxe4 Nxe4 8.fxe4 O-O
9.Rd6 Ba4 10.c4 Qb8 11.c5 Rc8 12.Kd2 f6
13.b3 Bb5 14.a4 Be8 15.Rf1 Rc7 (8.323.170.225) 525

22/85 572:17 +0.58 2...Bd8 3.Qc5 Bb6 4.Qxb6 Rxb6 5.Bxb6 dxe4
6.Na5 Qc8 7.Nxe4 Nxe4 8.fxe4 O-O
9.Rd6 Ba4 10.c4 Qb8 11.c5 Rc8 12.Kd2 Qa8
13.Re1 Kf8 14.h3 f5 15.exf5 exf5 (18.456.997.822) 537

Eelco