Houdini 3 gauntlets ;

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

Hugo
Posts: 782
Joined: Tue Dec 01, 2009 11:10 am

Re: Houdini 3 gauntlets ;

Post by Hugo »

Hi ALL

the first opening possition is played. 10 games ECO A22.

Now starts one of my favourite possitions in B90. also 10 games:
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6
There are many ways to continue. I hope for thrilling games!

C.K.
beram
Posts: 1187
Joined: Wed Jan 06, 2010 3:11 pm

Re: Houdini 3 gauntlets ;

Post by beram »

thats my favorite too :P
Hugo
Posts: 782
Joined: Tue Dec 01, 2009 11:10 am

Re: Houdini 3 gauntlets ;

Post by Hugo »

Two nice white wins in the B90!

[Event "Houdini 3 gauntlet 120-40 + 60-20 + 60-"]
[Site "?"]
[Date "2013.01.21"]
[Round "3.3"]
[White "Houdini 3 x4remote"]
[Black "Deep Rybka 4.1 x4"]
[Result "1-0"]
[ECO "B92"]
[Annotator "0.24;0.15"]
[PlyCount "135"]
[EventDate "2013.01.18"]
[EventType "simul"]
[Source "Doe"]

{Intel(R) Core(TM) i5-2500K CPU @ 3.30GHz 3310 MHz W=24.7 plies; 9.067kN/s
B=20.3 plies; 402kN/s} 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6.
Be2 {0.24/26 244} e5 {0.15/23 217} 7. Nb3 {0.16/27 415} Be7 {0.17/23 0} 8. O-O
{0.21/26 152} O-O {0.17/22 34} 9. Qd3 {0.22/26 352} b6 {(Dc7) 0.22/22 1084} 10.
Be3 {0.27/26 269} Nbd7 {(Lb7) 0.24/22 513} 11. Rfd1 {(a4) 0.30/26 161} Qc7 {
0.25/24 490} 12. a4 {0.29/27 47} Bb7 {(Td8) 0.25/23 91} 13. f3 {0.35/25 436} h6
{(Tfc8) 0.26/22 111} 14. Qc4 {(Dd2) 0.38/27 215} Qd8 {0.28/22 442} 15. Bf1 {
(Sc1) 0.39/27 0} Rc8 {(Te8) 0.28/23 394} 16. Qe2 {0.46/24 155} Qc7 {
(Sh5) 0.28/21 0} 17. Rd2 {(Df2) 0.51/26 216} Ra8 {0.28/20 0} 18. Qc4 {
(Df2) 0.49/24 0} Qd8 {(Dxc4) 0.33/21 247} 19. Rad1 {0.52/26 175} Rc8 {
(Lc8) 0.37/21 55} 20. Qe2 {0.59/27 796} Re8 {(Tc7) 0.40/21 0} 21. Qf2 {
(g3) 0.57/24 151} Rc6 {(Tc7) 0.42/20 301} 22. Nc1 {(Ld3) 0.60/25 199} Qc8 {
0.49/19 242} 23. N1e2 {(Kh1) 0.58/26 0} Kh7 {(h5) 0.44/19 173} 24. Ng3 {
(Kh1) 0.70/24 185} g6 {0.48/21 137} 25. Nge2 {(Kh1) 0.61/24 558} Kg7 {
0.46/22 204} 26. g3 {(Kh1) 0.60/23 181} Qb8 {(h5) 0.44/20 319} 27. Nc1 {
0.72/24 0} Rc7 {0.52/19 57} 28. h4 {(Lh3) 0.69/24 302} Rf8 {(Sc5) 0.43/18 217}
29. N1a2 {0.80/24 204} Rh8 {0.46/19 2} 30. Kh2 {(Lh3) 0.80/24 60} Rhc8 {
0.46/21 339} 31. Kh1 {(Lg2) 0.76/25 0} Re8 {(Th8) 0.46/19 164} 32. Nb4 {
(Lh3) 0.85/24 192} Nc5 {(Da8) 0.58/21 156} 33. Bh3 {(Lxc5) 0.90/24 191} Rd8 {
(Th8) 0.58/18 243} 34. Ra1 {0.92/23 137} Qa8 {(a5) 0.60/19 0} 35. Nbd5 {
(Kg1) 0.82/24 150} Nxd5 {0.63/19 153} 36. Nxd5 {0.83/26 58} Bxd5 {0.63/19 0}
37. Rxd5 {0.84/26 202} Qc6 {(h5) 0.63/19 0} 38. b3 {0.81/24 120} h5 {0.63/20 0}
39. Qd2 {0.81/25 143} Qb7 {(Tb8) 0.63/19 0} 40. Kg1 {(Kh2) 0.81/24 202} Ne6 {
(Tc6) 0.63/19 301} 41. Rd1 {(c4) 0.82/22 233} Nc5 {(Te8) 0.66/20 237} 42. Bf1 {
0.83/23 418} Rdc8 {(Sd7) 0.64/20 0} 43. Qe2 {(Ta1) 0.85/23 127} Ra8 {
(Tb8) 0.64/20 417} 44. Kh2 {(Df2) 0.82/23 244} Rac8 {0.63/19 491} 45. Qf2 {
(De1) 0.85/24 0} Rb8 {(Ta8) 0.66/20 499} 46. Qd2 {(Kg1) 0.82/24 332} Rbc8 {
0.63/20 161} 47. Kg1 {(De1) 0.82/24 0} Rd8 {(Tc6) 0.64/19 305} 48. Kh1 {
(a5) 0.81/24 158} Rc6 {0.63/19 168} 49. Kg2 {(Kh2) 0.79/24 7} Nd7 {
(Se6) 0.57/20 220} 50. Qf2 {0.73/25 169} Rdc8 {0.60/19 0} 51. c4 {
(Tc1) 0.64/0 263} Nc5 {0.54/21 272} 52. Qb2 {(Dc2) 0.58/27 124} a5 {0.51/21 137
} 53. Kh2 {0.60/27 443} Kh7 {(Kg8) 0.53/21 0} 54. Bh3 {0.70/25 400} Re8 {
(Tb8) 0.55/22 0} 55. f4 {0.71/26 207} Bf8 {0.55/20 17} 56. f5 {0.83/25 68} Qc7
{(De7) 1.01/19 29} 57. Rf1 {(De2) 0.83/24 144} Bh6 {0.66/20 176} 58. Bxh6 {
(Lg1) 1.12/26 0} Kxh6 {0.55/19 45} 59. Qe2 {1.10/27 89} Rg8 {(Th8) 0.85/21 0}
60. Rdd1 {1.37/23 177} Kg7 {(Kh7) 1.13/21 123} 61. g4 {(Tf3) 2.19/23 209} Rh8 {
0.83/19 0} 62. g5 {2.41/24 0} Rd8 {(Tf8) 1.61/17 60} 63. Rf2 {(fxg6) 3.63/23 97
} gxf5 {2.75/16 99} 64. exf5 {3.93/24 0} Rg8 {(Th8) 3.64/18 168} 65. f6+ {
12.13/22 88} Kf8 {3.64/0 0} 66. Qxh5 {(De3) 13.45/20 37} d5 {(Ke8) 5.55/16 96}
67. Qh6+ {(Txd5) 23.11/20 52} Ke8 {11.00/14 35} 68. g6 {27.19/22 87} 1-0



[Event "Houdini 3 gauntlet 120-40 + 60-20 + 60-"]
[Site "?"]
[Date "2013.01.21"]
[Round "3.4"]
[White "Bouquet _1.6 x4"]
[Black "Houdini 3 x4remote"]
[Result "1-0"]
[ECO "B92"]
[Annotator "0.19;0.00"]
[PlyCount "149"]
[EventDate "2013.01.18"]
[EventType "simul"]
[Source "Doe"]

{Intel(R) Core(TM) i5-2500K CPU @ 3.30GHz 3310 MHz W=29.0 plies; 10.657kN/s
B=28.4 plies; 9.975kN/s} 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6.
Be2 {0.19/25 300} e5 {0.00/28 359} 7. Nb3 {0.13/26 0} Be7 {(Le6) 0.05/27 240}
8. Qd3 {(0-0) 0.09/26 356} O-O {0.02/26 341} 9. O-O {0.06/25 0} Qc7 {
(Le6) 0.04/26 210} 10. Bd2 {(Le3) 0.10/25 778} Nbd7 {(b5) 0.02/26 1095} 11. a4
{0.20/26 500} b6 {0.01/26 0} 12. Rfd1 {(Df3) 0.14/26 266} Bb7 {0.06/24 190} 13.
Be3 {(Df3) 0.14/26 0} Rac8 {0.08/26 185} 14. f3 {0.13/25 0} Rfd8 {0.08/25 134}
15. Bf2 {0.15/24 0} Nh5 {(h6) 0.11/25 200} 16. g3 {0.16/26 169} Nhf6 {
0.15/25 47} 17. Rd2 {0.21/25 157} h5 {(h6) 0.14/24 136} 18. Bf1 {
(Tad1) 0.15/24 292} h4 {0.13/25 135} 19. gxh4 {0.22/24 193} Nh5 {0.19/25 0} 20.
Rad1 {(De3) 0.29/24 178} Nf4 {0.22/24 328} 21. Qe3 {0.28/26 0} Rf8 {0.26/25 319
} 22. Bg3 {0.33/25 0} f5 {0.28/24 178} 23. Bxf4 {0.25/25 0} exf4 {0.26/23 1}
24. Qe2 {0.30/25 229} fxe4 {(Lxh4) 0.26/25 76} 25. Nxe4 {0.43/24 84} Rfe8 {
(Se5) 0.30/24 77} 26. Bh3 {0.37/25 217} Ne5 {0.30/25 0} 27. Be6+ {
(Kh1) 0.36/24 328} Kh8 {0.50/26 963} 28. Nd4 {(c3) 0.34/27 0} Bxh4 {0.49/24 174
} 29. Qf1 {(Lxc8) 0.36/26 0} Rcd8 {0.49/24 60} 30. Qh3 {0.45/26 0} Qe7 {
0.49/27 232} 31. Rg2 {(Lg4) 0.41/26 0} Bxe4 {-0.16/9 1} 32. fxe4 {0.41/26 151}
Qf6 {(g5) 0.52/27 89} 33. Kh1 {0.45/28 581} f3 {0.55/28 0} 34. Rg3 {0.44/27 323
} g5 {0.48/28 60} 35. Rf1 {0.42/27 424} f2 {0.56/28 0} 36. Bd5 {0.57/28 443}
Rd7 {(Te7) 0.51/29 0} 37. Nf5 {0.51/27 241} Rh7 {0.53/29 580} 38. Rg2 {
0.51/27 0} Rf8 {0.54/27 483} 39. Qe3 {(Dc3) 0.47/28 0} g4 {0.42/25 193} 40. Qc3
{(b3) 0.55/28 987} b5 {0.61/22 56} 41. a5 {0.53/27 197} Qg5 {0.71/26 25} 42. b4
{0.58/27 108} Qh5 {0.60/28 84} 43. Nxh4 {0.58/26 0} Qxh4 {0.58/26 0} 44. Be6 {
0.44/27 146} Qf6 {(Dh3) 0.59/29 248} 45. Bf5 {0.59/27 214} Rh3 {0.64/28 59} 46.
Qd4 {0.59/27 33} Rh4 {0.65/28 247} 47. Qxf2 {0.59/27 0} Qh6 {0.61/28 218} 48.
Qe2 {0.65/28 0} Rg8 {(Df6) 0.63/29 882} 49. Qd2 {0.71/27 264} Qxd2 {0.73/28 344
} 50. Rxd2 {0.74/29 0} Nc4 {(Td8) 0.57/30 110} 51. Re2 {0.74/30 202} Ne5 {
0.62/30 80} 52. Rd1 {0.74/30 59} Rd8 {0.63/30 77} 53. Rg2 {(Ted2) 0.81/29 23}
Kg7 {0.70/31 141} 54. Rdg1 {(Lxg4) 0.81/30 0} Kf6 {(Tdh8) 0.68/32 144} 55. Bxg4
{0.80/31 305} Nxg4 {0.78/32 0} 56. Rxg4 {0.88/32 522} Rdh8 {0.88/33 0} 57. Rxh4
{0.91/31 178} Rxh4 {0.93/32 12} 58. Re1 {0.90/33 271} Ke5 {0.81/35 349} 59. Kg2
{1.04/34 0} Rh8 {(Th6) 1.04/36 436} 60. Re3 {1.28/35 943} Rg8+ {1.21/36 0} 61.
Kf2 {1.30/35 344} Rc8 {(Th8) 1.23/36 0} 62. c3 {1.39/34 152} Rh8 {1.68/33 420}
63. h3 {(Kg2) 1.55/36 0} Rh6 {1.80/33 0} 64. Ke2 {1.56/36 0} Rh5 {
(Th8) 1.77/32 60} 65. Kd3 {1.69/35 206} d5 {2.54/36 0} 66. exd5+ {1.77/35 0}
Kxd5 {2.79/34 0} 67. Rg3 {1.97/35 0} Kc6 {2.80/33 60} 68. Ke4 {(Kd2) 2.18/33 0}
Kd7 {(Th7) 2.79/31 0} 69. Rg6 {(Kf4) 2.89/31 92} Rxh3 {(Th4+) 2.92/0 50} 70.
Kd4 {3.07/33 83} Rh4+ {(Kc7) 5.28/33 206} 71. Kd3 {3.11/36 119} Rh3+ {5.21/33 0
} 72. Kc2 {3.11/37 87} Rh8 {(Th2+) 4.38/34 0} 73. Rxa6 {3.11/35 0} Kc7 {
(Th2+) 5.26/32 0} 74. Ra7+ {(Tf6) 3.11/36 115} Kc6 {5.79/29 42} 75. Rg7 {
(Tf7) 3.11/35 0} 1-0
Hugo
Posts: 782
Joined: Tue Dec 01, 2009 11:10 am

Re: Houdini 3 gauntlets ;

Post by Hugo »

Another nice game in the B90.
Stockfish continues with 6. Bg5 ! and won in an impressive way.

[pgn][Event "Houdini 3 gauntlet 120-40 + 60-20 + 60-"]
[Site "?"]
[Date "2013.01.22"]
[Round "4.1"]
[White "Stockfish 130114 x4"]
[Black "Houdini 3 x4remote"]
[Result "1-0"]
[ECO "B95"]
[Annotator "0.44;-0.02"]
[PlyCount "119"]
[EventDate "2013.01.18"]
[EventType "simul"]
[Source "Doe"]

{Intel(R) Core(TM) i5-2500K CPU @ 3.30GHz 3310 MHz W=29.3 plies; 4.053kN/s
B=24.3 plies; 8.716kN/s} 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6.
Bg5 {0.44/28 336} e6 {(e5) -0.02/26 378} 7. Be2 {(Df3) 0.28/27 424} h6 {
(Le7) -0.05/26 259} 8. Be3 {0.28/29 248} Qc7 {(Le7) -0.09/26 380} 9. O-O {
(a4) 0.16/25 235} Nbd7 {(b5) -0.02/25 256} 10. a4 {0.20/28 295} b6 {-0.08/25 0}
11. f3 {0.08/29 312} Be7 {(Lb7) 0.03/27 0} 12. Qe1 {0.08/30 203} O-O {
0.03/26 54} 13. Qg3 {(Td1) 0.16/31 395} Kh8 {0.03/24 262} 14. Rfe1 {
(Tad1) 0.16/32 62} Ne5 {(Lb7) -0.01/25 460} 15. Nb3 {0.16/32 155} Bb7 {
0.05/25 55} 16. Qf2 {(Tad1) 0.16/33 208} Nfd7 {(Sed7) -0.08/25 447} 17. Red1 {
(Tad1) 0.28/26 199} Nc5 {(Tab8) -0.09/23 280} 18. Qg3 {(f4) 0.12/27 363} Rad8 {
-0.08/24 152} 19. Bd4 {(Td2) 0.08/27 81} Ncd7 {-0.11/25 210} 20. Bf2 {
(Td2) 0.16/27 5} Rfe8 {(Sc5) -0.16/23 237} 21. f4 {(Ld4) 0.00/29 168} Ng6 {
-0.18/25 208} 22. Qe3 {(f5) 0.00/30 127} Kg8 {(Db8) -0.21/25 210} 23. g3 {
0.00/30 0} Ngf8 {(Tc8) -0.19/25 29} 24. Bf3 {(Td2) 0.00/26 165} e5 {
(Tc8) -0.20/26 338} 25. f5 {0.00/31 196} Nh7 {-0.17/26 173} 26. Qe2 {0.00/29 66
} Rc8 {-0.16/25 239} 27. Nd2 {(Td3) 0.00/30 0} Nhf6 {-0.16/24 876} 28. h4 {
0.00/31 0} Nc5 {(h5) -0.13/23 100} 29. Re1 {0.08/29 644} a5 {(Ted8) -0.12/23 0}
30. g4 {(Tad1) 0.00/30 529} Ba6 {(Dd8) -0.20/23 191} 31. Qd1 {(Sb5) 0.00/31 155
} Qd8 {-0.12/23 395} 32. Re3 {(Le2) 0.00/33 0} Nfd7 {-0.15/23 240} 33. Qe1 {
0.00/33 0} Nb8 {(Sf6) -0.09/24 228} 34. Nd5 {0.40/30 490} Bb7 {(Sc6) -0.03/26 0
} 35. Nxe7+ {(c3) 0.44/26 146} Qxe7 {-0.06/24 71} 36. Bg2 {(Sc4) 0.76/27 229}
Nc6 {(Sba6) 0.00/22 154} 37. c3 {1.13/26 162} Ba6 {(Sb8) 0.17/22 0} 38. Qb1 {
(Tf3) 1.57/26 105} b5 {0.20/23 124} 39. axb5 {1.81/26 69} Bxb5 {0.54/22 76} 40.
g5 {(Tg3) 1.65/26 65} Rb8 {0.92/22 117} 41. Rg3 {1.61/29 280} h5 {1.09/23 716}
42. g6 {1.97/30 0} Rf8 {(Le2) 1.25/24 165} 43. gxf7+ {(b4) 2.02/29 430} Qxf7 {
1.23/25 269} 44. Bxc5 {2.18/30 10} dxc5 {1.19/21 84} 45. Nf3 {2.34/30 496} Qe7
{(Dd7) 1.53/25 235} 46. Ng5 {(Da2+) 2.70/32 434} Rf6 {(Lc4) 1.33/26 162} 47.
Qd1 {2.74/32 308} Bc4 {2.08/24 144} 48. Bf1 {2.58/31 104} Bb3 {
(Lxf1) 2.08/25 450} 49. Qxh5 {2.74/29 197} Rh6 {2.29/24 0} 50. Qg4 {2.90/29 206
} Kh8 {1.91/22 123} 51. Rg2 {2.78/30 249} c4 {(a4) 2.01/24 115} 52. h5 {
(Le2) 3.59/27 150} a4 {2.53/23 159} 53. Be2 {3.63/28 24} Rf8 {(Sd8) 2.68/22 130
} 54. Kh1 {(Se6) 4.24/29 189} Qd6 {(Sd8) 3.72/23 255} 55. Rag1 {5.41/29 116}
Nd8 {4.90/24 60} 56. Nf3 {6.02/30 3} Rh7 {5.04/24 139} 57. Qh4 {6.30/31 13} Rf6
{(Df6) 3.89/21 94} 58. Rg6 {(Sg5) 7.55/31 175} Qe7 {(Sf7) 6.22/23 114} 59. Qg4
{(Txf6) 9.29/31 234} Bc2 {6.86/0 75} 60. Bxc4 {(Sh4) Zeit 9.29/32 23} 1-0
[/pgn]

Houdini shows some impressive wins in this tour, but to my surprise it lost also some games now.

C.K.
Hugo
Posts: 782
Joined: Tue Dec 01, 2009 11:10 am

Re: Houdini 3 gauntlets ;

Post by Hugo »

Hi All

a new opening has begun:

1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O b5 6. Bb3 Bb7 7. d3 Bc5 *
ECO C78

again 10 games to go.

Regards, C.K.
Lavir
Posts: 263
Joined: Sun Oct 28, 2012 11:45 am

Re: Houdini 3 gauntlets ;

Post by Lavir »

Hugo wrote: Houdini shows some impressive wins in this tour, but to my surprise it lost also some games now.
From what I've seen while following a lot of Houdini games and analyzing them after, is that when Houdini loses (apart when the position is already compromised) it is usually always for a refutation it didn't "see" and with more time it is found. To make an example of this behavior just look at the game against Equinox in the previous tournament.

It seems to me that the pruning of Houdini, while giving it that tremendous strength, at the same time is the major cause of its losses. If Robert could find a way to tailor the pruning in a way that these sort of things don't happen then Houdini would become almost unbeatable. Naturally it's easier said than done, just because the way Houdini prunes (as I've said) it is probably one of the major factors of its strength, but I think that if there's one that could do it, then that's Robert.

If then I was Robert I would actually try to find a way to have the Tactical Mode work automatically under certain patterns, because I think that if this could done correctly it could be a good advancement in strength in a lot of situations. In fact the positions where TM could be effective are not just coming from testsuites, they arise very often even in engine vs engine play. Moreover TM in itself doesn't decrease the strength of the engine dramatically (and this was somewhat of a surprise for everybody I think).

I think it would be very beneficial if Robert could find a way to automatically have TM work under certain patterns (as for example on king attacks with pawn storms or not defenders, when if you play with that mode even very strong engines get beaten in under 40 moves) at least on longer time controls. Some examples of these patterns could be when there is a piece double attacked or general contact with pieces or the aforementioned king attacks.

I in fact think that Robert has a very good winner on that Tactical Mode and if he can find a way to make it work in situations were if activated it makes a lot of difference (that as I said it is not just in seldom cases, but it happens in a lot of situations) he could achieve another impressive elo gain.
beram
Posts: 1187
Joined: Wed Jan 06, 2010 3:11 pm

Re: Houdini 3 gauntlets ;

Post by beram »

Lavir wrote:
Hugo wrote: Houdini shows some impressive wins in this tour, but to my surprise it lost also some games now.
From what I've seen while following a lot of Houdini games and analyzing them after, is that when Houdini loses (apart when the position is already compromised) it is usually always for a refutation it didn't "see" and with more time it is found. To make an example of this behavior just look at the game against Equinox in the previous tournament.

It seems to me that the pruning of Houdini, while giving it that tremendous strength, at the same time is the major cause of its losses. If Robert could find a way to tailor the pruning in a way that these sort of things don't happen then Houdini would become almost unbeatable. ...
hi Fabio, the game you refer to was the only loss of Houdini in its 42(!) games in the engine festival tournament. Isn't that almost unbeatable enough ?
http://www.talkchess.com/forum/viewtopi ... 80&t=46401

At the other hand in this tournament at much longer time control, Houdini already has 4 losses in 20 games. Coincidence ? we just must wait for more games to come to draw any conclusion.

grts bram
carldaman
Posts: 2287
Joined: Sat Jun 02, 2012 2:13 am

Re: Houdini 3 gauntlets ;

Post by carldaman »

+1 : let's not also forget that the tactical-mode version of Houdini is only slightly weaker overall (and stronger in many positions), so not much is lost by deploying it.

Maybe we need to see more tournaments where Houdini TM is playing.

Regards,
CL
Lavir
Posts: 263
Joined: Sun Oct 28, 2012 11:45 am

Re: Houdini 3 gauntlets ;

Post by Lavir »

beram wrote: hi Fabio, the game you refer to was the only loss of Houdini in its 42(!) games in the engine festival tournament. Isn't that almost unbeatable enough ?
There's never an "enough" ;-)

Then, as I said, if it was not for the pruning, it would not have lost that game (in fact with more time the right move is found).
beram wrote: At the other hand in this tournament at much longer time control, Houdini already has 4 losses in 20 games. Coincidence ?
No coincidence. It's just that the other time around it had IMO luck statistically speaking (42 games and only a loss is a "lucky" result even for Houdini 3 at that level of play) and now things are rounding up.

I'm sure that if you analyze the losses many are because of the motive I've said.
Hugo
Posts: 782
Joined: Tue Dec 01, 2009 11:10 am

Re: Houdini 3 gauntlets ;

Post by Hugo »

Here is a game that makes me wonder, because I always thought it is not possible in such long time controls that a game can be lost in a few moves:

[pgn][Event "Houdini 3 gauntlet 120-40 + 60-20 + 60-"]
[Site "?"]
[Date "2013.01.24"]
[Round "5.3"]
[White "Houdini 3 x4remote"]
[Black "Deep Rybka 4.1 x4"]
[Result "1-0"]
[ECO "C78"]
[Annotator "0.33;0.09"]
[PlyCount "89"]
[EventDate "2013.01.18"]
[Source "Doe"]

{Intel(R) Core(TM) i5-2500K CPU @ 3.30GHz 3310 MHz W=26.4 plies; 9.291kN/s
B=22.1 plies; 433kN/s} 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O b5 6.
Bb3 Bb7 7. d3 Bc5 8. Nc3 {0.33/27 595} O-O {0.09/21 227} 9. Bg5 {0.22/27 186
(Ld2)} h6 {0.07/22 360} 10. Be3 {0.22/27 0 (Lxf6)} Bxe3 {0.07/22 648 (Ld4)} 11.
fxe3 {0.24/25 126} d6 {0.07/22 10} 12. a3 {0.30/26 175} Rb8 {0.08/20 0} 13. Nd2
{0.22/25 310 (La2)} Kh7 {0.00/20 464 (Lc8)} 14. Nd5 {0.18/26 208} Nxd5 {0.00/
21 0} 15. Bxd5 {0.21/26 719} f6 {0.00/24 0 (Dd7)} 16. Be6 {0.21/24 207 (a4)}
Ne7 {0.00/22 633} 17. Qg4 {0.14/25 0} d5 {0.00/21 129 (Lc8)} 18. Rf3 {0.31/25
208} Qd6 {0.00/22 0} 19. Raf1 {0.24/25 158 (Tg3)} d4 {0.00/23 401 (Lc8)} 20.
Rg3 {0.28/23 165} g6 {0.00/22 0} 21. Rf2 {0.31/24 229 (Lf5)} c5 {0.00/21 284
(h5)} 22. exd4 {0.31/26 200} cxd4 {0.03/22 0} 23. b4 {0.27/28 179} Bc8 {0.03/
22 4} 24. Bf5 {0.29/30 197} Rg8 {0.07/23 0} 25. Bxc8 {0.29/28 9} Rbxc8 {0.07/
23 379} 26. Rh3 {0.25/30 0} h5 {0.07/24 439 (Th8)} 27. Nb3 {0.27/28 214} Rh8 {
0.06/24 0} 28. Qe2 {0.23/29 763 (Sc5)} Kg7 {0.00/24 582} 29. Nc5 {0.14/28 238}
Rcf8 {0.00/23 0 (Ta8)} 30. Qe1 {0.09/26 733 (Thf3)} f5 {0.00/24 375 (Tf7)} 31.
c3 {0.10/25 123 (exf5)} dxc3 {0.00/20 281} 32. Qxc3 {0.10/27 0} Rf6 {0.00/22
174 (fxe4)} 33. Qe1 {0.15/25 118 (exf5)} Rc8 {0.00/22 322} 34. Rhf3 {0.07/25
42 (d4)} f4 {0.00/21 213 (Tff8)} 35. d4 {0.25/27 167} Qxd4 {0.00/21 205 (Sc6)}
36. Rd3 {0.47/29 348} Qc4 {0.00/21 0} 37. Rd7 {0.49/27 204} Kf8 {0.25/20 360
(Tf7)} 38. h3 {0.55/26 148 (g3)} Rxc5 {0.26/19 252} 39. bxc5 {0.57/27 0} Qxc5 {
0.60/20 170} 40. Kh2 {0.92/29 60 (Db4)} Nc6 {0.56/18 164 (Td6)} 41. Rf3 {1.84/
26 150} Rd6 {0.87/21 42 (Ke8)} 42. Rc3 {6.06/23 10} Qd4 {0.99/20 1} 43. Rxc6 {
6.09/22 1} Rxd7 {5.20/20 41} 44. Qh4 {6.89/28 234} Qd6 {5.93/22 0} 45. Rxd6 {
6.89/26 6} 1-0
[/pgn]

it started with 38. .... R x c5? 7 moves later: game over.

Clemens Keck