glaurung is kasparov, rybka is karpov on steriods: a game

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

ozziejoe
Posts: 811
Joined: Wed Mar 08, 2006 10:07 pm

glaurung is kasparov, rybka is karpov on steriods: a game

Post by ozziejoe »

ok, I decided to play a moderately long game (40 moves 40 min repeating), using the caro kann karpov variation as an opening and and giving the attacking side to glaurung 2.1 (kasparov like i think) and the defending side to rybka 2.3.2amp (karpov on steriods in my view)

Here is the game, with comments. ( copy the game intact at the bottom if you want to play through it)

[Event "HBS-HZMDN1S, 40'/40+40'/40+40'"]
[Site "University of Wollongong"]
[Date "2008.05.17"]
[Round "1"]
[White "Glaurung 2.1"]
[Black "Rybka 2.3.2a mp 32-bit"]
[Result "0-1"]
[ECO "B17"]
[Annotator "0.98;0.15"]
[PlyCount "156"]
[TimeControl "40/2400:40/2400:2400"]

{Intel(R) Core(TM)2 CPU T7400 @ 2.16GHz 2161 MHz W=18.5 ply; 1,210kN/
s B=16.9 ply; 107kN/s} 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nd7 5. Ng5 Ngf6
6. Bd3 e6 7. N1f3 Bd6 8. Qe2 h6 9. Ne4 Nxe4 10. Qxe4 Qc7 11. Qg4 Kf8 12. O-O c5
{Both last book move} 13. Qh4 {0.98/18 211}


[d]r1b2k1r/ppqn1pp1/3bp2p/2p5/3P3Q/3B1N2/PPP2PPP/R1B2RK1 b - - 0 13 [d]

Glaurung thinks it is winning, rybka thinks the position is equal. This evaluation difference will exist for a number of moves.


Kg8 {(b6) 0.15/17 106} 14. c3 {
(Re1) 0.76/18 136} c4 {(b6) 0.15/17 33}


[d]r1b3kr/ppqn1pp1/3bp2p/8/2pP3Q/2PB1N2/PP3PPP/R1B2RK1 w - - 0 15[d]

So much for gallaghar suggestion that black exchange the c pawn and thereby obtain a central pawn majority. rybka seems to prefer to take space and cramp white if possible

15. Bc2 {1.00/19 72} b6 {0.12/16 22}
16. Qh3 {(Nd2) 0.98/19 110} Bb7 {0.09/15 33} 17. Re1 {(Nd2) 0.96/19 107} Rc8 {
(Rd8) 0.10/15 66} 18. Kh1 {(Nd2) 0.90/17 81} Bd5 {(Rd8) 0.00/14 32} 19. Re2 {
(Kg1) 0.86/17 94} g6 {(b5) 0.00/15 120} 20. Bd2 {0.88/18 79} Kg7 {0.05/16 98}
21. Rae1 {0.68/18 122} Rcd8 {0.00/16 220} 22. Bc1 {(b4) 0.66/18 136} Bb7 {
(Bf4) 0.00/16 29} 23. Bb1 {(Qh4) 0.66/17 66} Bd5 {0.00/17 240}

[d]3r3r/p1qn1pk1/1p1bp1pp/3b4/2pP4/2P2N1Q/PP2RPPP/1BB1R2K w - - 0 [d]

Rybka sees equality, but glaurung still thinks it has an advantage.

24. Be4 {
(Bc2) 0.62/18 97} Bxe4 {-0.13/15 69} 25. Rxe4 {0.54/16 14} Nf6 {
(g5) -0.13/17 290} 26. R4e2 {0.74/18 66} b5 {(Qc8) -0.07/16 63} 27. Kg1 {
(Ne5) 0.56/18 156} b4 {-0.21/16 40} 28. Ne5 {0.45/17 51} Nd5 {-0.21/16 70} 29.
f4 {(cxb4) 0.31/17 62} bxc3 {(Rb8) -0.32/15 95} 30. bxc3 {0.29/17 51} Rb8 {
-0.35/15 24} 31. Rf1 {(Qg3) 0.31/17 69} h5 {(Rb1) -0.37/15 65} 32. Qg3 {
0.23/16 69} Rb6 {(Rhd8) -0.38/15 92} 33. f5 {(Rfe1) 0.27/16 55}

[d]7r/p1q2pk1/1r1bp1p1/3nNP1p/2pP4/2P3Q1/P3R1PP/2B2RK1 b - - 0 33
[d]

Glaurung, the mad-dog attacker, has been patient up to this point, but finally decides to sac a pawn and open things up on the f file. Does he get enough?

h4 {-0.34/17 57
} 34. f6+ {0.03/16 46} Nxf6 {(Kg8) -0.30/18 37} 35. Qf3 {(Qg5) 0.00/18 85} Bxe5
{-0.65/15 23} 36. Rxe5 {0.00/18 73} Nd5 {-0.68/15 23}

[d]7r/p1q2pk1/1r2p1p1/3nR3/2pP3p/2P2Q2/P5PP/2B2RK1 w - - 0 37
[d]

now glaurung thinks it has full compensation for the pawn and the position is equal, wheras rybka thinks it has a -.68 advantage (which is a big advantage in rybkas conservative evalution numbers)

37. Bg5 {-0.01/18 73} f5
{-0.68/14 24} 38. Rfe1 {(h3) 0.00/18 73} Qd7 {-0.69/15 20} 39. Bc1 {
(h3) -0.17/17 49} Ra6 {(h3) -0.74/16 75} 40. R5e2 {(Bg5) -0.15/16 49} Kf7 {
(Rh5) -0.71/17 75} 41. Bf4 {(Bg5) -0.15/17 202} h3 {(Rh5) -0.89/15 21} 42. Be5
{-0.25/17 77} Rh5 {-0.85/15 26} 43. g3 {(gxh3) -0.43/17 80} Qd8 {
(Qc6) -0.99/15 113} 44. Rc2 {(Rf1) -0.13/17 100} Rb6 {-0.97/17 60} 45. Qe2 {
(Bf4) -0.33/18 105} Qc8 {-1.01/15 15} 46. Rf1 {-0.25/19 85} Kg8 {
(Qb7) -1.01/18 110} 47. Rcc1 {(Qe1) -0.19/19 68} Kh7 {(Qc6) -1.02/16 56} 48.
Rf3 {(Rf2) 0.00/20 87} Qb7 {(Qc6) -0.91/18 86} 49. Rf2 {0.00/19 82} Qa6 {
(Qc8) -0.91/17 28} 50. Rf3 {0.00/19 78} Qa3 {-0.75/17 87} 51. Rc2 {0.00/20 78}
Qa4 {(Qe7) -0.55/17 72} 52. Rc1 {0.00/19 60} Kg8 {(Qa3) -0.56/17 82} 53. g4 {
-0.17/18 62}

[d]6k1/p7/1r2p1p1/3nBp1r/q1pP2P1/2P2R1p/P3Q2P/2R3K1 b - - 0 53
[d]

After alot of manuvering, glaurung invests another pawn in the attack (not a true pawn sac since no way that black can hold the pawn if he takes it)

Rh4 {(Qa3) -0.61/16 53} 54. Rg3 {0.33/17 50} Kf7 {
(Qc6) -0.61/15 15} 55. Qd2 {(Rf1) 0.00/17 92} Qd7 {(Rxg4) -0.61/17 66} 56. Rf1
{0.00/17 57} Ke8 {-0.61/16 28} 57. Qg5 {(Qe2) 0.00/17 50} Qh7 {-0.70/13 16} 58.
Qd2 {(gxf5) 0.00/18 51} Kd8 {(Qd7) -0.54/16 207} 59. gxf5 {(Qf2) 0.00/18 49}
gxf5 {(exf5) -0.37/18 114} 60. Qg5+ {(Qe2) -0.49/16 105}

[d]3k4/p6q/1r2p3/3nBp2/2pP3r/2P3Rp/P2Q3P/5RK1 w - - 0 60
[d]

This is the first time glaurung thinks it might have a bit of a disadvantage, at move 60. Rybka started having negative scores for glaurung back at move 24.

Ke8 {(Kc8) -1.09/17 57
} 61. Qc1 {(Qg8+) -0.35/17 33} Rh6 {-1.06/17 51} 62. Qc2 {-0.29/17 28} Rg6 {
(Kd8) -1.29/17 54} 63. Rff3 {-0.49/17 30} Rxg3+ {-1.67/16 25} 64. Rxg3 {
-1.68/18 37} Kf8 {-1.56/16 0}

[d]4k3/p6q/1r2p3/3nBp2/2pP4/2P3Rp/P1Q4P/6K1 b - - 0 64
[d]

This is the point that both engines think black has a winning position. I am having trouble finding the exact moment when glaurung's position went so downhill.

65. Bg7+ {(Kf2) -1.94/17 30} Kf7 {-1.96/15 0} 66.
Be5 {-3.13/18 59} Qh4 {-3.15/16 32} 67. Qc1 {-3.64/17 33} Qe4 {-3.64/17 14} 68.

[d]8/p4k2/1r2p3/3nBp2/2pPq3/2P3Rp/P6P/2Q3K1 w - - 0 68
[d]

the black queen gets behind white pawns, the rook is coming in on the b file, it is all over...

Kf2 {-3.50/18 46} Rb1 {-3.88/17 9} 69. Qd2 {-4.33/18 35} f4 {-4.32/17 14} 70.
Rg7+ {-5.27/19 38} Kf8 {-3.91/16 0} 71. Rg4 {-5.84/19 45} Rf1+ {-4.77/16 28}

[d]5k2/p7/4p3/3nB3/2pPqpR1/2P4p/P2Q1K1P/5r2 w - - 0 72
[d]

Here is a nice tactic that finishes glaurung off. Black's queen will capture the key c3 pawn

72. Kxf1 {-3.00/6 0} Ne3+ {-5.00/17 18} 73. Qxe3 {-6.25/20 52} Qxe3 {
-5.00/19 182} 74. Rxf4+ {-6.33/20 70} Ke7 {-5.24/18 167} 75. Bf6+ {
(Rh4) -6.39/19 49} Ke8 {-4.53/16 0} 76. Be5 {-6.58/20 37} Qxc3 {-5.48/15 33}
77. Kf2 {-6.58/19 43} Qb2+ {-5.65/16 62} 78. Kg3 {-7.19/19 35} c3 {-5.76/17 145
} 0-1

[d]4k3/p7/4p3/4B3/3P1R2/2p3Kp/Pq5P/8 w - - 0 79
[d]

It would not be rybka if it did not win with a passed pawn. Here is that C pawn ghallagher recomends you trade for the d pawn in order to have a pawn majority in the centre:)

full game below. Any comments at where glaurung went wrong? Was it a series of subtle mistakes? was the pawn sac a bad idea?

[Event "H40'/40+40'/40+40'"]
[Site ""]
[Date "2008.05.17"]
[Round "1"]
[White "Glaurung 2.1"]
[Black "Rybka 2.3.2a mp 32-bit"]
[Result "0-1"]
[ECO "B17"]
[Annotator "0.98;0.15"]
[PlyCount "156"]
[TimeControl "40/2400:40/2400:2400"]

{Intel(R) Core(TM)2 CPU T7400 @ 2.16GHz 2161 MHz W=18.5 ply; 1,210kN/
s B=16.9 ply; 107kN/s} 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nd7 5. Ng5 Ngf6
6. Bd3 e6 7. N1f3 Bd6 8. Qe2 h6 9. Ne4 Nxe4 10. Qxe4 Qc7 11. Qg4 Kf8 12. O-O c5
{Both last book move} 13. Qh4 {0.98/18 211} Kg8 {(b6) 0.15/17 106} 14. c3 {
(Re1) 0.76/18 136} c4 {(b6) 0.15/17 33} 15. Bc2 {1.00/19 72} b6 {0.12/16 22}
16. Qh3 {(Nd2) 0.98/19 110} Bb7 {0.09/15 33} 17. Re1 {(Nd2) 0.96/19 107} Rc8 {
(Rd8) 0.10/15 66} 18. Kh1 {(Nd2) 0.90/17 81} Bd5 {(Rd8) 0.00/14 32} 19. Re2 {
(Kg1) 0.86/17 94} g6 {(b5) 0.00/15 120} 20. Bd2 {0.88/18 79} Kg7 {0.05/16 98}
21. Rae1 {0.68/18 122} Rcd8 {0.00/16 220} 22. Bc1 {(b4) 0.66/18 136} Bb7 {
(Bf4) 0.00/16 29} 23. Bb1 {(Qh4) 0.66/17 66} Bd5 {0.00/17 240} 24. Be4 {
(Bc2) 0.62/18 97} Bxe4 {-0.13/15 69} 25. Rxe4 {0.54/16 14} Nf6 {
(g5) -0.13/17 290} 26. R4e2 {0.74/18 66} b5 {(Qc8) -0.07/16 63} 27. Kg1 {
(Ne5) 0.56/18 156} b4 {-0.21/16 40} 28. Ne5 {0.45/17 51} Nd5 {-0.21/16 70} 29.
f4 {(cxb4) 0.31/17 62} bxc3 {(Rb8) -0.32/15 95} 30. bxc3 {0.29/17 51} Rb8 {
-0.35/15 24} 31. Rf1 {(Qg3) 0.31/17 69} h5 {(Rb1) -0.37/15 65} 32. Qg3 {
0.23/16 69} Rb6 {(Rhd8) -0.38/15 92} 33. f5 {(Rfe1) 0.27/16 55} h4 {-0.34/17 57
} 34. f6+ {0.03/16 46} Nxf6 {(Kg8) -0.30/18 37} 35. Qf3 {(Qg5) 0.00/18 85} Bxe5
{-0.65/15 23} 36. Rxe5 {0.00/18 73} Nd5 {-0.68/15 23} 37. Bg5 {-0.01/18 73} f5
{-0.68/14 24} 38. Rfe1 {(h3) 0.00/18 73} Qd7 {-0.69/15 20} 39. Bc1 {
(h3) -0.17/17 49} Ra6 {(h3) -0.74/16 75} 40. R5e2 {(Bg5) -0.15/16 49} Kf7 {
(Rh5) -0.71/17 75} 41. Bf4 {(Bg5) -0.15/17 202} h3 {(Rh5) -0.89/15 21} 42. Be5
{-0.25/17 77} Rh5 {-0.85/15 26} 43. g3 {(gxh3) -0.43/17 80} Qd8 {
(Qc6) -0.99/15 113} 44. Rc2 {(Rf1) -0.13/17 100} Rb6 {-0.97/17 60} 45. Qe2 {
(Bf4) -0.33/18 105} Qc8 {-1.01/15 15} 46. Rf1 {-0.25/19 85} Kg8 {
(Qb7) -1.01/18 110} 47. Rcc1 {(Qe1) -0.19/19 68} Kh7 {(Qc6) -1.02/16 56} 48.
Rf3 {(Rf2) 0.00/20 87} Qb7 {(Qc6) -0.91/18 86} 49. Rf2 {0.00/19 82} Qa6 {
(Qc8) -0.91/17 28} 50. Rf3 {0.00/19 78} Qa3 {-0.75/17 87} 51. Rc2 {0.00/20 78}
Qa4 {(Qe7) -0.55/17 72} 52. Rc1 {0.00/19 60} Kg8 {(Qa3) -0.56/17 82} 53. g4 {
-0.17/18 62} Rh4 {(Qa3) -0.61/16 53} 54. Rg3 {0.33/17 50} Kf7 {
(Qc6) -0.61/15 15} 55. Qd2 {(Rf1) 0.00/17 92} Qd7 {(Rxg4) -0.61/17 66} 56. Rf1
{0.00/17 57} Ke8 {-0.61/16 28} 57. Qg5 {(Qe2) 0.00/17 50} Qh7 {-0.70/13 16} 58.
Qd2 {(gxf5) 0.00/18 51} Kd8 {(Qd7) -0.54/16 207} 59. gxf5 {(Qf2) 0.00/18 49}
gxf5 {(exf5) -0.37/18 114} 60. Qg5+ {(Qe2) -0.49/16 105} Ke8 {(Kc8) -1.09/17 57
} 61. Qc1 {(Qg8+) -0.35/17 33} Rh6 {-1.06/17 51} 62. Qc2 {-0.29/17 28} Rg6 {
(Kd8) -1.29/17 54} 63. Rff3 {-0.49/17 30} Rxg3+ {-1.67/16 25} 64. Rxg3 {
-1.68/18 37} Kf8 {-1.56/16 0} 65. Bg7+ {(Kf2) -1.94/17 30} Kf7 {-1.96/15 0} 66.
Be5 {-3.13/18 59} Qh4 {-3.15/16 32} 67. Qc1 {-3.64/17 33} Qe4 {-3.64/17 14} 68.
Kf2 {-3.50/18 46} Rb1 {-3.88/17 9} 69. Qd2 {-4.33/18 35} f4 {-4.32/17 14} 70.
Rg7+ {-5.27/19 38} Kf8 {-3.91/16 0} 71. Rg4 {-5.84/19 45} Rf1+ {-4.77/16 28}
72. Kxf1 {-3.00/6 0} Ne3+ {-5.00/17 18} 73. Qxe3 {-6.25/20 52} Qxe3 {
-5.00/19 182} 74. Rxf4+ {-6.33/20 70} Ke7 {-5.24/18 167} 75. Bf6+ {
(Rh4) -6.39/19 49} Ke8 {-4.53/16 0} 76. Be5 {-6.58/20 37} Qxc3 {-5.48/15 33}
77. Kf2 {-6.58/19 43} Qb2+ {-5.65/16 62} 78. Kg3 {-7.19/19 35} c3 {-5.76/17 145
} 0-1