The Doomsday of Stockfish DD

[mohzus]

So far the best I've done is a draw against SF DD running at 24% strength on my tablet. I'm trying hard to draw it at 40% strength but no success so far. My strength is FICS glicko rating of 1500 to 1600 in blitz.
[pgn][Event "?"]
[Site "?"]
[Date "2013.12.12"]
[Round "?"]
[White "Player"]
[Black "Stockfish DD (24.6%)"]
[Result "1/2-1/2"]
[TimeControl "15"]

1. e3 e6 2. d4 d5 3. Nf3 Nf6 4. Bd3 Nc6 5. O-O Be7 6. b3 a5 7. a4 Bd7 8. c4 Nb4
9. c5 O-O 10. Nc3 Rc8 11. Bb2 c6 12. Ne5 b6 13. Be2 Rb8 14. f4 bxc5 15. h3 Re8
16. g4 Qc7 17. g5 cxd4 18. exd4 Ne4 19. Nxe4 dxe4 20. Bc4 e3 21. Qh5 Rf8 22. f5
e2 23. Rf2 Nc2 24. g6 hxg6 25. fxg6 fxg6 26. Qxg6 Nxa1 27. Bxe6+ Bxe6 28. Qxe6+
Kh7 29. Qg6+ Kh8 30. Nf7+ Rxf7 31. Qh5+ Kg8 32. Qxf7+ Kh8 33. Qh5+ Kg8 34. Qf7+
Kh8 35. Qh5+ Kg8 36. Qf7+ 1/2-1/2[/pgn]

[Lyudmil Tsvetkov]

So far the best I've done is a draw against SF DD running at 24% strength on my tablet. I'm trying hard to draw it at 40% strength but no success so far. My strength is FICS glicko rating of 1500 to 1600 in blitz.
[pgn][Event "?"]
[Site "?"]
[Date "2013.12.12"]
[Round "?"]
[White "Player"]
[Black "Stockfish DD (24.6%)"]
[Result "1/2-1/2"]
[TimeControl "15"]

1. e3 e6 2. d4 d5 3. Nf3 Nf6 4. Bd3 Nc6 5. O-O Be7 6. b3 a5 7. a4 Bd7 8. c4 Nb4
9. c5 O-O 10. Nc3 Rc8 11. Bb2 c6 12. Ne5 b6 13. Be2 Rb8 14. f4 bxc5 15. h3 Re8
16. g4 Qc7 17. g5 cxd4 18. exd4 Ne4 19. Nxe4 dxe4 20. Bc4 e3 21. Qh5 Rf8 22. f5
e2 23. Rf2 Nc2 24. g6 hxg6 25. fxg6 fxg6 26. Qxg6 Nxa1 27. Bxe6+ Bxe6 28. Qxe6+
Kh7 29. Qg6+ Kh8 30. Nf7+ Rxf7 31. Qh5+ Kg8 32. Qxf7+ Kh8 33. Qh5+ Kg8 34. Qf7+
Kh8 35. Qh5+ Kg8 36. Qf7+ 1/2-1/2[/pgn]

Interesting game, Robert.
You almost mated it.

Very nice avatar btw.

[mohzus]

Interesting game, Robert.
You almost mated it.

Very nice avatar btw.

hehe, thank you.
I just beat SF DD on my tablet with full strength, although I highly suspect a bug. Time control was 15 s for the whole game. Of course I could not cope up with it and strangely I am not automatically forfeited on time on Droidfish.
So the game was normal until a point where I suspect SF ran out of time or was very close to it, it just started to make random moves.
Warning, extremely boring game: [pgn][Event "?"]
[Site "?"]
[Date "2013.12.13"]
[Round "?"]
[White "Player"]
[Black "Stockfish DD"]
[Result "1-0"]
[TimeControl "15"]

1. e3 Nf6 2. d3 Nc6 3. Nc3 e5 4. Nf3 d5 5. Be2 Be7 6. O-O O-O 7. h3 d4 8. Nb1
a6 9. e4 Be6 10. Nbd2 Nd7 11. Re1 f6 12. Nf1 Nc5 13. Ng3 Qd7 14. Kh2 a5 15. a3
a4 16. Bd2 Rad8 17. Kg1 b5 18. Kh2 Bd6 19. Kg1 g6 20. Kh2 Rfe8 21. Kg1 Qc8 22.
Kh2 Be7 23. Kg1 Qa6 24. Kh2 Qa7 25. Kg1 Qb6 26. Kh2 Qa6 27. Kg1 Qa7 28. Kh2 Bd6
29. Kg1 Qb6 30. Kh2 Rb8 31. Kg1 Bf8 32. Kh2 Rec8 33. Kg1 Be7 34. Kh2 Qa7 35.
Kg1 Qa6 36. Kh2 Re8 37. Kg1 Bd7 38. Kh2 Qb7 39. Kg1 Be6 40. Kh2 Qa7 41. Kg1 Qa6
42. Kh2 Rf8 43. Kg1 Qb7 44. Kh2 Ra8 45. Kg1 Rad8 46. Kh2 f5 47. Kg1 f4 48. Nh1
Kg7 49. Kh2 Qa8 50. Kg1 Rde8 51. Kh2 Rb8 52. Kg1 Rbd8 53. Kh2 Rfe8 54. Kg1 Qa6
55. Kh2 Qb6 56. Kg1 Qb7 57. Kh2 h6 58. Kg1 Qa6 59. Kh2 Qa8 60. Kg1 Qb7 61. Kh2
g5 62. Ng1 Kh8 63. f3 Kg7 64. Nf2 h5 65. Kh1 Qa8 66. Kh2 Rf8 67. Kh1 Rde8 68.
Kh2 h4 69. Kh1 b4 70. axb4 a3 71. Rxa3 g4 72. Rxa8 g3 73. Ng4 Nxb4 74. Rxe8 c6
75. Rxf8 Nb3 76. cxb3 c5 77. Bxb4 c4 78. Bxe7 c3 79. bxc3 dxc3 80. Nxe5 c2 81.
Qxc2 Bxb3 82. Qxb3 Kh6 83. Qe6+ Kh7 84. Qg6# 1-0
[/pgn]
I think things went wrong for SF at move 69 when it played b4. All its next moves seem totally random to me. Bug?


Edit: I've just replayed another game and won again and I can confirm that SF makes moves that look random when its time goes to 0. I have set a 1 minute time control game, moved my king on and off for almost 100 moves and SF went down on time and started to make random moves until its loss.

[Lyudmil Tsvetkov]

Interesting game, Robert.
You almost mated it.

Very nice avatar btw.

hehe, thank you.
I just beat SF DD on my tablet with full strength, although I highly suspect a bug. Time control was 15 s for the whole game. Of course I could not cope up with it and strangely I am not automatically forfeited on time on Droidfish.
So the game was normal until a point where I suspect SF ran out of time or was very close to it, it just started to make random moves.
Warning, extremely boring game: [pgn][Event "?"]
[Site "?"]
[Date "2013.12.13"]
[Round "?"]
[White "Player"]
[Black "Stockfish DD"]
[Result "1-0"]
[TimeControl "15"]

1. e3 Nf6 2. d3 Nc6 3. Nc3 e5 4. Nf3 d5 5. Be2 Be7 6. O-O O-O 7. h3 d4 8. Nb1
a6 9. e4 Be6 10. Nbd2 Nd7 11. Re1 f6 12. Nf1 Nc5 13. Ng3 Qd7 14. Kh2 a5 15. a3
a4 16. Bd2 Rad8 17. Kg1 b5 18. Kh2 Bd6 19. Kg1 g6 20. Kh2 Rfe8 21. Kg1 Qc8 22.
Kh2 Be7 23. Kg1 Qa6 24. Kh2 Qa7 25. Kg1 Qb6 26. Kh2 Qa6 27. Kg1 Qa7 28. Kh2 Bd6
29. Kg1 Qb6 30. Kh2 Rb8 31. Kg1 Bf8 32. Kh2 Rec8 33. Kg1 Be7 34. Kh2 Qa7 35.
Kg1 Qa6 36. Kh2 Re8 37. Kg1 Bd7 38. Kh2 Qb7 39. Kg1 Be6 40. Kh2 Qa7 41. Kg1 Qa6
42. Kh2 Rf8 43. Kg1 Qb7 44. Kh2 Ra8 45. Kg1 Rad8 46. Kh2 f5 47. Kg1 f4 48. Nh1
Kg7 49. Kh2 Qa8 50. Kg1 Rde8 51. Kh2 Rb8 52. Kg1 Rbd8 53. Kh2 Rfe8 54. Kg1 Qa6
55. Kh2 Qb6 56. Kg1 Qb7 57. Kh2 h6 58. Kg1 Qa6 59. Kh2 Qa8 60. Kg1 Qb7 61. Kh2
g5 62. Ng1 Kh8 63. f3 Kg7 64. Nf2 h5 65. Kh1 Qa8 66. Kh2 Rf8 67. Kh1 Rde8 68.
Kh2 h4 69. Kh1 b4 70. axb4 a3 71. Rxa3 g4 72. Rxa8 g3 73. Ng4 Nxb4 74. Rxe8 c6
75. Rxf8 Nb3 76. cxb3 c5 77. Bxb4 c4 78. Bxe7 c3 79. bxc3 dxc3 80. Nxe5 c2 81.
Qxc2 Bxb3 82. Qxb3 Kh6 83. Qe6+ Kh7 84. Qg6# 1-0
[/pgn]
I think things went wrong for SF at move 69 when it played b4. All its next moves seem totally random to me. Bug?


Edit: I've just replayed another game and won again and I can confirm that SF makes moves that look random when its time goes to 0. I have set a 1 minute time control game, moved my king on and off for almost 100 moves and SF went down on time and started to make random moves until its loss.

Hi Robert.

So many people playing Stockfish, on tablets, mobiles, computers, clusters?, etc.

Did you really play that game, because the quality of play up to a certain moment was much much higher than the one in the first game you posted? Now I might suppose you are a cheater, I will only need to do some analysis...

I suppose Stockfish at full strength and 15 seconds for the entire game plays the same level as Stockfish 24% strength at 5 minutes. Why do not you use an increment?

You disoriented the engine with your king manouvers. Usually Stockfish will easily see winning line in such position with passive play, but obviously 15 second for the entire game is simply an inquisition even for engines. I would suggest that you play it at least at 1 minute + 1 second at lower strength, but of course, you wanted to win a game.

[mohzus]

Hi Robert.

So many people playing Stockfish, on tablets, mobiles, computers, clusters?, etc.

I will not lie, I have started to play SF on my tablet around 3 days ago when I read your thread here. You inspired me. I even wonder whether this "anti bot" strategy could work against strong humans.
Personally I don't know how I would counter it, but I am nowhere close to strong.

Did you really play that game, because the quality of play up to a certain moment was much much higher than the one in the first game you posted? Now I might suppose you are a cheater, I will only need to do some analysis...

I assure you I did not cheat a single move and I terribly doubt any engine would match most of my moves. But feel free to test and report me the results, after all this would be interesting to me

I suppose Stockfish at full strength and 15 seconds for the entire game plays the same level as Stockfish 24% strength at 5 minutes. Why do not you use an increment?

I do not really know how SF is weakened. If it is weakened by levels (there are 20), then I have read that it should have the same strength on a tablet compared to a dual core, etc. I chose such a time control because I wanted SF to play almost instantly (I was literally blitzing and too impatient to wait its moves. I finished the game with something like -2 to -5 minutes, I can't remember exactly). But you are right, I should try to play a fair game and give it an increement per move.

You disoriented the engine with your king manouvers. Usually Stockfish will easily see winning line in such position with passive play, but obviously 15 second for the entire game is simply an inquisition even for engines. I would suggest that you play it at least at 1 minute + 1 second at lower strength, but of course, you wanted to win a game.

Ok, I will try later.

Here's the game I mentionned in my last post, 126 moves long, SF had 0 on the clock at move 100...b4 (the first "crazy move").
[pgn][Event "?"]
[Site "?"]
[Date "2013.12.13"]
[Round "?"]
[White "Player"]
[Black "Stockfish DD"]
[Result "1-0"]
[TimeControl "60"]

1. e3 e6 2. d4 Nf6 3. Nf3 Be7 4. Be2 c5 5. c3 O-O 6. O-O d5 7. h3 Nc6 8. a3 c4
9. Nbd2 Na5 10. Qa4 b6 11. Re1 Qc7 12. Kh1 Bd6 13. Kg1 Bb7 14. Kh1 Rad8 15. Kg1
Bc6 16. Qd1 Rfe8 17. Kh1 h6 18. Kg1 b5 19. Kh1 a6 20. Kg1 Bd7 21. Kh1 Rf8 22.
Kg1 Bc6 23. Kh1 Rb8 24. Kg1 Kh8 25. Kh1 Bb7 26. Kg1 Ra8 27. Kh1 Rfd8 28. Kg1
Kg8 29. Kh1 Rac8 30. Kg1 Bc6 31. Kh1 Re8 32. Kg1 Kh8 33. Kh1 Red8 34. Kg1 Bb7
35. Kh1 Nc6 36. Kg1 Rf8 37. Kh1 Rfe8 38. Kg1 Na5 39. Kh1 Bc6 40. Kg1 Red8 41.
Kh1 Kg8 42. Kg1 Bb7 43. Kh1 Nc6 44. Kg1 Kh8 45. Kh1 Ne7 46. Kg1 Nf5 47. Kh1 Bc6
48. Kg1 Bd7 49. Kh1 a5 50. Kg1 a4 51. Kh1 Bc6 52. Kg1 Kg8 53. Kh1 Bb7 54. Kg1
Kh8 55. Kh1 Ne7 56. Kg1 Bc6 57. Kh1 Re8 58. Kg1 Ng6 59. Kh1 Kg8 60. Kg1 Rcd8
61. Kh1 Ne7 62. Kg1 Nf5 63. Kh1 Kh8 64. Kg1 Ne7 65. Kh1 Bb7 66. Kg1 Nf5 67. Kh1
Rc8 68. Kg1 Kg8 69. Kh1 Ne7 70. Kg1 Bc6 71. Kh1 Red8 72. Kg1 Ng6 73. Kh1 Ra8
74. Kg1 Ne7 75. Kh1 Rab8 76. Kg1 Rbc8 77. Kh1 Bb7 78. Kg1 Nc6 79. Kh1 Re8 80.
Kg1 Kh8 81. Kh1 Qe7 82. Kg1 Rcd8 83. Kh1 Ra8 84. Kg1 Red8 85. Kh1 Rdc8 86. Kg1
Kg8 87. Kh1 Qc7 88. Kg1 Re8 89. Kh1 Na5 90. Kg1 Red8 91. Kh1 Nc6 92. Kg1 Qe7
93. Kh1 e5 94. Kg1 e4 95. Nh2 Rac8 96. Nhf1 Re8 97. Nh2 Qd7 98. Nhf1 Rcd8 99.
Nh2 Bc8 100. Nhf1 b4 101. axb4 a3 102. Rxa3 h5 103. Bxh5 g6 104. Be2 g5 105.
Nh2 g4 106. Nxg4 Nxb4 107. Nxf6+ Kg7 108. Nxd7 f6 109. cxb4 c3 110. bxc3 f5
111. Nc5 f4 112. exf4 e3 113. fxe3 Bxh3 114. gxh3 Bxf4 115. exf4 Rd6 116. Bg4
Ra6 117. Rxe8 Rxa3 118. Qf3 Ra1 119. Re7+ Kf6 120. Nde4+ dxe4 121. Nxe4+ Kxe7
122. Qe3 Rb1 123. Nc5+ Kd6 124. Qe6+ Kc7 125. Qd7+ Kb6 126. Qb7# 1-0
[/pgn]

[overlord]

[pgn][Event "Blitz"]
[Site "?"]
[Date "2013.12.14"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Stockfish DD"]
[Result "1/2-1/2"]
[PlyCount "199"]

1. d4 Nf6 2. c4 e5 3. dxe5 Ng4 4. Nf3 Bc5 5. e3 Nc6 6. Nc3 Ngxe5 7. b3 O-O 8.
Bb2 Re8 9. Be2 Nxf3+ 10. Bxf3 Bd6 11. O-O Be5 12. Qd2 d6 13. Rfd1 Bf5 14. Ne2
Qh4 15. g3 Qh6 16. Nd4 Bxd4 17. Bxd4 Be4 18. Bg2 Bxg2 19. Kxg2 Qg6 20. Qe2 a5
21. a4 Qe4+ 22. Qf3 Qxf3+ 23. Kxf3 f6 24. h4 Nd8 25. Rd3 Ne6 26. Rad1 Kf7 27.
R1d2 b6 28. Rd1 Nc5 29. Bxc5 bxc5 30. Rc1 Rab8 31. Rcc3 h5 32. e4 g6 33. Re3
Ke6 34. Ke2 Kd7 35. f4 c6 36. Kd2 Re6 37. Ke2 Rbe8 38. Kf3 f5 39. exf5 gxf5 40.
Rxe6 Rxe6 41. Re3 Rg6 42. Ke2 Rg4 43. Kd3 Rg7 44. Kc3 Rg8 45. Kc2 Rb8 46. Kc3
Ra8 47. Kc2 Rg8 48. Kc3 d5 49. Kd3 Kd6 50. Rf3 d4 51. Kc2 Rg4 52. Kd3 Ke6 53.
Kc2 Rg8 54. Kd3 Kd7 55. Kc2 Kd6 56. Kd3 Rg7 57. Kc2 Rg6 58. Kd3 Rg4 59. Kc2 Ke7
60. Kd3 Rg7 61. Kc2 Rg8 62. Kd3 Rg4 63. Kc2 Rg7 64. Kd3 Kf6 65. Kc2 Rg6 66. Kd3
Ke6 67. Kc2 Rg4 68. Kd3 Kd7 69. Kc2 Rg8 70. Kd3 Rg7 71. Kc2 Rg4 72. Kd3 Kc7 73.
Kc2 Kb7 74. Kd3 Rg7 75. Kc2 Re7 76. Kd2 Kb6 77. Rd3 Rg7 78. Rf3 Rg4 79. Kd3 Rg6
80. Kc2 Kb7 81. Kd3 Re6 82. Kd2 Kb6 83. Rd3 Re8 84. Rf3 Re4 85. Rd3 Kc7 86. Rf3
Kd6 87. Rd3 Re6 88. Rf3 Re7 89. Rd3 Kc7 90. Rf3 Rg7 91. Rd3 Rg8 92. Rf3 Rg6 93.
Rd3 Kb7 94. Rf3 Rg4 95. Rd3 Rg6 96. Rf3 Rg8 97. Rd3 Rg4 98. Rf3 Rg6 99. Rd3 Kb6
100. Rf3 1/2-1/2[/pgn]

[Lyudmil Tsvetkov]

Hi Robert.

So many people playing Stockfish, on tablets, mobiles, computers, clusters?, etc.

I will not lie, I have started to play SF on my tablet around 3 days ago when I read your thread here. You inspired me. I even wonder whether this "anti bot" strategy could work against strong humans.
Personally I don't know how I would counter it, but I am nowhere close to strong.

Did you really play that game, because the quality of play up to a certain moment was much much higher than the one in the first game you posted? Now I might suppose you are a cheater, I will only need to do some analysis...

I assure you I did not cheat a single move and I terribly doubt any engine would match most of my moves. But feel free to test and report me the results, after all this would be interesting to me

I suppose Stockfish at full strength and 15 seconds for the entire game plays the same level as Stockfish 24% strength at 5 minutes. Why do not you use an increment?

I do not really know how SF is weakened. If it is weakened by levels (there are 20), then I have read that it should have the same strength on a tablet compared to a dual core, etc. I chose such a time control because I wanted SF to play almost instantly (I was literally blitzing and too impatient to wait its moves. I finished the game with something like -2 to -5 minutes, I can't remember exactly). But you are right, I should try to play a fair game and give it an increement per move.

You disoriented the engine with your king manouvers. Usually Stockfish will easily see winning line in such position with passive play, but obviously 15 second for the entire game is simply an inquisition even for engines. I would suggest that you play it at least at 1 minute + 1 second at lower strength, but of course, you wanted to win a game.

Ok, I will try later.

Here's the game I mentionned in my last post, 126 moves long, SF had 0 on the clock at move 100...b4 (the first "crazy move").
[pgn][Event "?"]
[Site "?"]
[Date "2013.12.13"]
[Round "?"]
[White "Player"]
[Black "Stockfish DD"]
[Result "1-0"]
[TimeControl "60"]

1. e3 e6 2. d4 Nf6 3. Nf3 Be7 4. Be2 c5 5. c3 O-O 6. O-O d5 7. h3 Nc6 8. a3 c4
9. Nbd2 Na5 10. Qa4 b6 11. Re1 Qc7 12. Kh1 Bd6 13. Kg1 Bb7 14. Kh1 Rad8 15. Kg1
Bc6 16. Qd1 Rfe8 17. Kh1 h6 18. Kg1 b5 19. Kh1 a6 20. Kg1 Bd7 21. Kh1 Rf8 22.
Kg1 Bc6 23. Kh1 Rb8 24. Kg1 Kh8 25. Kh1 Bb7 26. Kg1 Ra8 27. Kh1 Rfd8 28. Kg1
Kg8 29. Kh1 Rac8 30. Kg1 Bc6 31. Kh1 Re8 32. Kg1 Kh8 33. Kh1 Red8 34. Kg1 Bb7
35. Kh1 Nc6 36. Kg1 Rf8 37. Kh1 Rfe8 38. Kg1 Na5 39. Kh1 Bc6 40. Kg1 Red8 41.
Kh1 Kg8 42. Kg1 Bb7 43. Kh1 Nc6 44. Kg1 Kh8 45. Kh1 Ne7 46. Kg1 Nf5 47. Kh1 Bc6
48. Kg1 Bd7 49. Kh1 a5 50. Kg1 a4 51. Kh1 Bc6 52. Kg1 Kg8 53. Kh1 Bb7 54. Kg1
Kh8 55. Kh1 Ne7 56. Kg1 Bc6 57. Kh1 Re8 58. Kg1 Ng6 59. Kh1 Kg8 60. Kg1 Rcd8
61. Kh1 Ne7 62. Kg1 Nf5 63. Kh1 Kh8 64. Kg1 Ne7 65. Kh1 Bb7 66. Kg1 Nf5 67. Kh1
Rc8 68. Kg1 Kg8 69. Kh1 Ne7 70. Kg1 Bc6 71. Kh1 Red8 72. Kg1 Ng6 73. Kh1 Ra8
74. Kg1 Ne7 75. Kh1 Rab8 76. Kg1 Rbc8 77. Kh1 Bb7 78. Kg1 Nc6 79. Kh1 Re8 80.
Kg1 Kh8 81. Kh1 Qe7 82. Kg1 Rcd8 83. Kh1 Ra8 84. Kg1 Red8 85. Kh1 Rdc8 86. Kg1
Kg8 87. Kh1 Qc7 88. Kg1 Re8 89. Kh1 Na5 90. Kg1 Red8 91. Kh1 Nc6 92. Kg1 Qe7
93. Kh1 e5 94. Kg1 e4 95. Nh2 Rac8 96. Nhf1 Re8 97. Nh2 Qd7 98. Nhf1 Rcd8 99.
Nh2 Bc8 100. Nhf1 b4 101. axb4 a3 102. Rxa3 h5 103. Bxh5 g6 104. Be2 g5 105.
Nh2 g4 106. Nxg4 Nxb4 107. Nxf6+ Kg7 108. Nxd7 f6 109. cxb4 c3 110. bxc3 f5
111. Nc5 f4 112. exf4 e3 113. fxe3 Bxh3 114. gxh3 Bxf4 115. exf4 Rd6 116. Bg4
Ra6 117. Rxe8 Rxa3 118. Qf3 Ra1 119. Re7+ Kf6 120. Nde4+ dxe4 121. Nxe4+ Kxe7
122. Qe3 Rb1 123. Nc5+ Kd6 124. Qe6+ Kc7 125. Qd7+ Kb6 126. Qb7# 1-0
[/pgn]

The strategy of closing the game, and especially the center, should work against anyone if you want to achieve a draw, provided that your opponent chooses to play optimally. As said, chess is supposed to be a draw, and closing the game with either colour successfully proves so. With white you should always be able to achieve a draw by playing e3,d4,c3,f4, etc., and with black, by playing d5,e6,c6,f5. The opponent possibly can not avoid closing the entire game and a draw, if she does not decide to play suboptimal moves. If the opponent plays optimally, as top engines do, she will gain some space advantage with a pawn closing a file, rather than not gaining space and opening a file. This is simply the optimal way of playing, but by best play it only leads to a draw.

If the opponent decides for some reason that she will not play optimally, she will not close the game, you will have better chances to win, but your opponent's chances of winning the game in case she is the stronger player will also increase. If the opponent is much stronger player, she simply opens the game (they call it here playing with contempt) and wins. But if your opponent is a weaker player than you, she simply loses easier. The standard case would be when your opponent does not know how strong are you (and in that case using contempt would be wrong) and therefore playing suboptimal moves and opening the game unduly would entail considerable risks.

So I would say, yes, the strategy works perfectly for a draw, but who is really interested in drawing? Chess is played to win a game or an event.

I do not know if you are not cheating, Robert, I have my suspicions. It is only 2 games, but you made a qualitative leap from the first to the second game. Either we have a 1 in a million case, or you are cheating, but somehow the possibility that you learned well your opening or got sudden inspiration is also not to be excluded.

126 moves, it takes loooong to beat Stockfish...

[Lyudmil Tsvetkov]

[pgn][Event "Blitz"]
[Site "?"]
[Date "2013.12.14"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Stockfish DD"]
[Result "1/2-1/2"]
[PlyCount "199"]

1. d4 Nf6 2. c4 e5 3. dxe5 Ng4 4. Nf3 Bc5 5. e3 Nc6 6. Nc3 Ngxe5 7. b3 O-O 8.
Bb2 Re8 9. Be2 Nxf3+ 10. Bxf3 Bd6 11. O-O Be5 12. Qd2 d6 13. Rfd1 Bf5 14. Ne2
Qh4 15. g3 Qh6 16. Nd4 Bxd4 17. Bxd4 Be4 18. Bg2 Bxg2 19. Kxg2 Qg6 20. Qe2 a5
21. a4 Qe4+ 22. Qf3 Qxf3+ 23. Kxf3 f6 24. h4 Nd8 25. Rd3 Ne6 26. Rad1 Kf7 27.
R1d2 b6 28. Rd1 Nc5 29. Bxc5 bxc5 30. Rc1 Rab8 31. Rcc3 h5 32. e4 g6 33. Re3
Ke6 34. Ke2 Kd7 35. f4 c6 36. Kd2 Re6 37. Ke2 Rbe8 38. Kf3 f5 39. exf5 gxf5 40.
Rxe6 Rxe6 41. Re3 Rg6 42. Ke2 Rg4 43. Kd3 Rg7 44. Kc3 Rg8 45. Kc2 Rb8 46. Kc3
Ra8 47. Kc2 Rg8 48. Kc3 d5 49. Kd3 Kd6 50. Rf3 d4 51. Kc2 Rg4 52. Kd3 Ke6 53.
Kc2 Rg8 54. Kd3 Kd7 55. Kc2 Kd6 56. Kd3 Rg7 57. Kc2 Rg6 58. Kd3 Rg4 59. Kc2 Ke7
60. Kd3 Rg7 61. Kc2 Rg8 62. Kd3 Rg4 63. Kc2 Rg7 64. Kd3 Kf6 65. Kc2 Rg6 66. Kd3
Ke6 67. Kc2 Rg4 68. Kd3 Kd7 69. Kc2 Rg8 70. Kd3 Rg7 71. Kc2 Rg4 72. Kd3 Kc7 73.
Kc2 Kb7 74. Kd3 Rg7 75. Kc2 Re7 76. Kd2 Kb6 77. Rd3 Rg7 78. Rf3 Rg4 79. Kd3 Rg6
80. Kc2 Kb7 81. Kd3 Re6 82. Kd2 Kb6 83. Rd3 Re8 84. Rf3 Re4 85. Rd3 Kc7 86. Rf3
Kd6 87. Rd3 Re6 88. Rf3 Re7 89. Rd3 Kc7 90. Rf3 Rg7 91. Rd3 Rg8 92. Rf3 Rg6 93.
Rd3 Kb7 94. Rf3 Rg4 95. Rd3 Rg6 96. Rf3 Rg8 97. Rd3 Rg4 98. Rf3 Rg6 99. Rd3 Kb6
100. Rf3 1/2-1/2[/pgn]

I want to see a blitz win, Miroslav.

Give yourself double time in blitz and try to win.
Stockfish must like playing the Budapest gambit, as against me it also frequently does so. Then I go in the same line to d5 with the knight instead of e2, and some interesting games occur. But you proved it once again independently, a reasonably strong player should be able to draw occasionally top engines without much effort by exchanging a quantity of pieces.

[mohzus]

The strategy of closing the game, and especially the center, should work against anyone if you want to achieve a draw, provided that your opponent chooses to play optimally. As said, chess is supposed to be a draw, and closing the game with either colour successfully proves so. With white you should always be able to achieve a draw by playing e3,d4,c3,f4, etc., and with black, by playing d5,e6,c6,f5. The opponent possibly can not avoid closing the entire game and a draw, if she does not decide to play suboptimal moves. If the opponent plays optimally, as top engines do, she will gain some space advantage with a pawn closing a file, rather than not gaining space and opening a file. This is simply the optimal way of playing, but by best play it only leads to a draw.

If the opponent decides for some reason that she will not play optimally, she will not close the game, you will have better chances to win, but your opponent's chances of winning the game in case she is the stronger player will also increase. If the opponent is much stronger player, she simply opens the game (they call it here playing with contempt) and wins. But if your opponent is a weaker player than you, she simply loses easier. The standard case would be when your opponent does not know how strong are you (and in that case using contempt would be wrong) and therefore playing suboptimal moves and opening the game unduly would entail considerable risks.

So I would say, yes, the strategy works perfectly for a draw, but who is really interested in drawing? Chess is played to win a game or an event.

Interesting theory. I am not sure I fully agree for open positions.
I suppose that there are open positions that leads to a draw with perfect play; or openings that opens the position which also lead to a draw with perfect play. But I think you are right that when you close the position the outcome is a draw with perfect play. Of course I/we cannot demonstrate these statements.

I do not know if you are not cheating, Robert, I have my suspicions. It is only 2 games, but you made a qualitative leap from the first to the second game. Either we have a 1 in a million case, or you are cheating, but somehow the possibility that you learned well your opening or got sudden inspiration is also not to be excluded.

126 moves, it takes loooong to beat Stockfish...

I am not sure you realized, but my first game was a fully deserved draw (SF had 5 minutes to play, it never went into time troubles) against SF at 24% strength.
My 2nd and 3rd games were likely to be drawn (with some unfair time controls, since I would have forfeited on time) but I won only because of a bug, that is, SF plays totally random when its time is either 0 or negative. I deserve no credit at all for the wins I've shown here. At best I would have drawn if SF could manage negative times and in the worst case I would have lost. So the fact that I won does not mean that I played at a higher quality. However if you noticed that I played better in my 2nd game... well maybe, but it was not enough to beat SF. In the 2nd game the position was a dead draw before SF ran out of time and played 69...b4?
So I still have a long road to achieve your level and performance Lyudmil.
Lastly I want to point out that I don't think that moving my king back and forth confused SF in my 2nd or 3rd game. It went crazy only because it ran out of time.

[mohzus]

"Sorry" to get back to this, but I think now I understand why SF plays "crazy random moves" when its time ran up. It's not a bug but a feature.
in search.cpp one reads

Code: Select all

 // Stop the search if most of the available time has been used. We
            // probably don't have enough time to search the first move at the
            // next iteration anyway.
            if (Time::now() - SearchTime > (TimeMgr.available_time() * 62) / 100)
                stop = true;

So basically, the 2 games I've won against SF were not true wins and I won because it ran out of time (unfairly, since I would have lost on time much sooner than it!).
I have yet to draw against SF at full strength; win is for now out of reach for me.