Houdini vs Stockfish Match 2014

[RIDDICK]

The problem is you don't update the pgn headers respectively.
Please name them with their timestamps at least(or internal version number if there is one)
Currently it looks as one entity is always playing.

Full details (full PGN file) will be posted at the end of the match, as I have already said. The full PGN header of last game is:

Code: Select all

[Event "HOUDINI vs STOCKFISH MATCH 2014"]
[Site "RIDDICK"]
[Date "2014.03.07"]
[Round "78"]
[White "STOCKFISH 060314 64 SSE4.2"]
[Black "HOUDINI 4 PRO X64"]
[Result "0-1"]
[ECO "B86"]
[Annotator "RIDDICK"]
[PlyCount "108"]
[TimeControl "40/7200:20/3600:900+15"]

There is no timestamp, the games are identified by the rounds number (in this case "78"). Why do you say: «Currently it looks as one entity is always playing.»?

[Guenther]

The problem is you don't update the pgn headers respectively.
Please name them with their timestamps at least(or internal version number if there is one)
Currently it looks as one entity is always playing.

Full details (full PGN file) will be posted at the end of the match, as I have already said. The full PGN header of last game is:

Code: Select all

[Event "HOUDINI vs STOCKFISH MATCH 2014"]
[Site "RIDDICK"]
[Date "2014.03.07"]
[Round "78"]
[White "STOCKFISH 060314 64 SSE4.2"]
[Black "HOUDINI 4 PRO X64"]
[Result "0-1"]
[ECO "B86"]
[Annotator "RIDDICK"]
[PlyCount "108"]
[TimeControl "40/7200:20/3600:900+15"]

There is no timestamp, the games are identified by the rounds number (in this case "78"). Why do you say: «Currently it looks as one entity is always playing.»?

Didn't you just say in an other post Houdini was updated too?
Still there is only 'HOUDINI 4 PRO X64'. Or did you mean you _might_ update Houdini during the match?

Note that I haven't nor intend to read the whole (8 pages) thread, but only
refer to this sentence:

I'm sorry but it's not fair. Stockfish is continually updated, why Houdini should remain the same? Even Martin Thoresen in the TCEC tournament follows the same criteria... What's the problem?

Guenther

[RIDDICK]

Didn't you just say in an other post Houdini was updated too?
Still there is only 'HOUDINI 4 PRO X64'. Or did you mean you _might_ update Houdini during the match?

Houdini version has always been the same so far, but If Mr. Houdart will decide to send me a new Houdini version (public or private), I'll use it.

PS. I'm sorry for my english, if I made some mistakes.

[RIDDICK]

HOUDINI: 42.0/80 [ +010/=062/-006 ] - STOCKFISH: 38.0/80 [ +006/=062/-010 ]

[pgn][Event "HOUDINI vs STOCKFISH MATCH 2014"]
[Site "RIDDICK"]
[Date "2014.03.08"]
[Round "79"]
[White "HOUDINI 4 PRO X64"]
[Black "STOCKFISH 060314 64 SSE4.2"]
[Result "1/2-1/2"]
[ECO "B81"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e6 6. g4 h6 7. Bg2 Nc6 8. h3
Bd7 9. Nxc6 Bxc6 10. Qe2 Be7 11. Be3 Qa5 12. Bd2 Rc8 13. Nd5 Qb5 14. Qxb5 Bxb5
15. Nxe7 Kxe7 16. Bc3 e5 17. O-O-O Nd7 18. Kb1 b6 19. Bd2 Nc5 20. b3 h5 21. a4
Bc6 22. f3 Ne6 23. Kb2 hxg4 24. hxg4 Rxh1 25. Bxh1 Rh8 26. Bb4 Nc5 27. Bg2 Rh2
28. Rd2 a5 29. Bf1 Rh1 30. Bxc5 bxc5 31. Bb5 Ba8 32. Bc4 Rh3 33. Rf2 Bc6 34. c3
f6 35. Bf1 Rh4 36. Ka3 d5 37. exd5 Bxd5 38. b4 cxb4+ 39. cxb4 1/2-1/2[/pgn]
[pgn][Event "HOUDINI vs STOCKFISH MATCH 2014"]
[Site "RIDDICK"]
[Date "2014.03.08"]
[Round "80"]
[White "STOCKFISH 060314 64 SSE4.2"]
[Black "HOUDINI 4 PRO X64"]
[Result "1/2-1/2"]
[ECO "B81"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e6 6. g4 h6 7. Bg2 Nc6 8. h3
Bd7 9. f4 Qb6 10. Nb3 d5 11. exd5 Nb4 12. a3 Nbxd5 13. Nxd5 Nxd5 14. Bxd5 exd5
15. Qe2+ Qe6 16. Be3 Bd6 17. f5 Bg3+ 18. Kd2 Qe5 19. Bd4 Qxe2+ 20. Kxe2 O-O 21.
Kf3 Bh4 22. a4 Rfe8 23. Rh2 b6 24. a5 bxa5 25. Rxa5 a6 1/2-1/2[/pgn]

[neelbasant]

How many will be played with this variation ?
Always the same :- e4 c5 Nf3 d6 d4 cxd4 Nxd4 ------.

Neel

[Uri Blass]

I hope that you test only public versions(versions that everybody can buy)

If Robert Houdart sell a new version of houdini and you test it during the match it is ok.

If he does not do it then I prefer to see the match continue between latest stockfish that is always available to everyone and houdini4.

I'm sorry but it's not fair. Stockfish is continually updated, why Houdini should remain the same? Even Martin Thoresen in the TCEC tournament follows the same criteria... What's the problem?

It is your tournament and your right to do what you want.

The reason that I prefer public versions is because I use results of tournaments with big number of games for decision if to buy a new program and if you test a private version then it is clearly not relevant.

So far I did not buy Houdini4 because I did not see a convincing result to cause me to believe that houdini4 is stronger than stockfish(and your result so far is not statistically significant so practically the strongest programs that I have are stockfish developement version and houdini3).

[Uri Blass]

How many will be played with this variation ?
Always the same :- e4 c5 Nf3 d6 d4 cxd4 Nxd4 ------.

Neel

see page 5 of this thread that has list of openings
post from
Thu Feb 20, 2014 7:34 am

It seems that we had
1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e6 6. g4 h6 7. Bg2 Nc6 8. h3 Bd7 1/2-1/2
We have only another one with d6 and later go for different second moves by black

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e6 6. f4 Nc6 7. Be3 e5 8. Nf3 Ng4 1/2-1/2
1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 e5 5. Nb5 d6 6. c4 Be7 7. N1c3 a6 8. Na3 h6 1/2-1/2
1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e5 6. Ndb5 d6 7. Bg5 a6 8. Na3 b5 1/2-1/2
1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 d6 6. Bg5 e6 7. Qd2 a6 8. O-O-O Bd7 1/2-1/2
1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 d6 6. Bg5 e6 7. Qd2 a6 8. O-O-O h6 1/2-1/2
1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 d6 6. Bc4 e6 7. Be3 Be7 8. Qe2 a6 1/2-1/2
1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 d6 6. Bc4 Qb6 7. Nb3 e6 8. O-O Be7 1/2-1/2
1. e4 c5 2. Nf3 Nc6 3. Bb5 e6 4. O-O Nge7 5. c3 d5 6. exd5 Qxd5 7. Re1 a6 8. Bxc6 Nxc6 1/2-1/2
1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 d6 6. g4 h6 7. h4 Nc6 8. Rg1 d5 1/2-1/2
1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 d6 6. g4 h6 7. h4 Nc6 8. Rg1 h5 1/2-1/2
1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 d6 6. Be2 Be7 7. O-O O-O 8. Be3 Nc6 1/2-1/2
1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 d6 6. Be3 a6 7. f3 b5 8. Qd2 Nbd7 1/2-1/2
1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 a6 5. Nc3 Qc7 6. g3 Bb4 1/2-1/2
1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nc6 5. Nc3 Qc7 6. Be3 a6 7. Qd2 Nf6 8. O-O-O Bb4 1/2-1/2

later we go to different opening move by white in second move

1. e4 c5 2. Nc3 Nc6 3. g3 g6 4. Bg2 Bg7 5. d3 d6 6. Be3 e6 7. Qd2 Rb8 8. Nge2 Nd4 1/2-1/2
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 1/2-1/2
1. e4 c5 2. Nc3 Nc6 3. f4 g6 4. Nf3 Bg7 5. Bc4 e6 6. f5 Nge7 7. fxe6 dxe6 1/2-1/2
1. e4 c5 2. Nc3 Nc6 3. f4 g6 4. Nf3 Bg7 5. Bc4 e6 6. O-O Nge7 7. d3 O-O 8. Qe1 d5 1/2-1/2
1. e4 c5 2. Nc3 Nc6 3. f4 g6 4. Nf3 Bg7 5. Bb5 Nd4 6. O-O a6 7. Bd3 d6 8. Nxd4 cxd4 1/2-1/2
1. e4 c5 2. Nc3 Nc6 3. Nf3 g6 4. d4 cxd4 5. Nxd4 Bg7 6. Be3 Nf6 7. Bc4 O-O 8. Bb3 d6 1/2-1/2
1. e4 c5 2. c3 d5 3. exd5 Qxd5 4. d4 Nf6 5. Nf3 Bg4 6. Be2 e6 7. h3 Bh5 8. O-O Nc6 1/2-1/2
1. e4 c5 2. c3 d5 3. exd5 Qxd5 4. d4 Nf6 5. Nf3 Bg4 6. Be2 e6 7. O-O Nc6 8. Be3 cxd4 1/2-1/2
1. e4 c5 2. d4 cxd4 3. c3 dxc3 4. Nxc3 Nc6 5. Nf3 e6 6. Bc4 a6 7. O-O Nge7 8. Bg5 f6 1/2-1/2
1. e4 c5 2. f4 e6 3. Nf3 Nc6 4. Nc3 a6 5. g3 d5 6. Bg2 d4 1/2-1/2
1. e4 c5 2. f4 e6 3. Nf3 Nc6 4. Nc3 a6 5. g3 d5 6. d3 d4 7. Ne2 Nf6 1/2-1/2
1. e4 c5 2. f4 e6 3. Nf3 Nc6 4. Nc3 a6 5. g3 d5 6. e5 Nge7 7. Bg2 Nf5 8. O-O h5 1/2-1/2
1. e4 c5 2. f4 e6 3. Nf3 d5 4. Bb5 Bd7 5. Bxd7 Nxd7 6. d3 Bd6 7. O-O Ne7 8. c4 O-O 1/2-1/2
1. e4 c5 2. f4 Nc6 3. Nf3 g6 4. Bb5 Bg7 5. Bxc6 bxc6 6. d3 d6 7. O-O Nf6 8. Nc3 O-O 1/2-1/2

later we have many games with 1.e4 e5 or different openings.

[RIDDICK]

Stockfish is able to better exploit a favorable white-side opening line and he retrieves 1 point! Curious evaluation profile charts of both engines in the two games:

81st Round:


82nd Round:


HOUDINI: 42.5/82 [ +010/=063/-007 ] - STOCKFISH: 39.5/82 [ +007/=063/-010 ]

[pgn][Event "HOUDINI vs STOCKFISH MATCH 2014"]
[Site "RIDDICK"]
[Date "2014.03.08"]
[Round "81"]
[White "HOUDINI 4 PRO X64"]
[Black "STOCKFISH 060314 64 SSE4.2"]
[Result "1/2-1/2"]
[ECO "B82"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e6 6. f4 Nc6 7. Be3 e5 8. Nf3
Ng4 9. Qd2 Nxe3 10. Qxe3 exf4 11. Qxf4 Be6 12. O-O-O Be7 13. Nd5 O-O 14. Kb1
Rc8 15. c3 Kh8 16. h4 Ne5 17. Be2 Nxf3 18. gxf3 Rc6 19. Qg3 Bxd5 20. Rxd5 Qb6
21. f4 Rc5 22. Rxc5 dxc5 23. Bc4 Qc6 24. Bd5 Qg6 25. Qf3 b5 26. h5 Qh6 27. Qg4
f5 28. exf5 Qd6 29. Rg1 Qf6 30. Be6 Rd8 31. Rd1 Qf8 32. Rxd8 Qxd8 33. Kc2 b4
34. Qg2 a5 35. Qe4 Bf8 36. Bc4 Be7 37. a4 Bf6 38. Qe3 bxc3 39. bxc3 Qf8 40. Bd5
Be7 41. Qe4 Qf6 42. Qe6 Qxe6 43. fxe6 Kg8 44. Kb3 Kf8 45. Kc4 Bf6 46. Kxc5 Bxc3
47. Kd6 Bf6 48. Kd7 Be7 49. Be4 h6 50. Bg2 Bb4 51. Ba8 Ba3 52. Bh1 Bb4 53. Bd5
Ba3 54. Ba8 Bc5 55. Bf3 Ba3 56. Bc6 Bb4 57. Bg2 Bc5 58. Bf1 Bb4 59. Bh3 Be7 60.
Bf5 Bb4 61. Bd3 Ba3 62. Be2 Bb4 63. f5 Be7 64. Bc4 Bg5 65. Bf1 Bh4 66. Bh3 Bg5
67. Bg4 Bf6 68. Bd1 Be7 69. Bf3 Bh4 70. Be4 Be7 71. Bb7 Ba3 72. Bc6 Bb4 73. Ba8
Be7 74. Be4 Bc5 75. Bf3 Bb4 76. Bg2 Be7 77. Bh1 Bg5 78. Bb7 Be7 79. Ba6 Bg5 80.
Bc4 Bh4 81. Bd3 Be7 82. Be2 Bg5 83. f6 Bxf6 84. Bd3 Bh4 85. Be4 Be7 1/2-1/2[/pgn]
[pgn][Event "HOUDINI vs STOCKFISH MATCH 2014"]
[Site "RIDDICK"]
[Date "2014.03.08"]
[Round "82"]
[White "STOCKFISH 060314 64 SSE4.2"]
[Black "HOUDINI 4 PRO X64"]
[Result "1-0"]
[ECO "B82"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e6 6. f4 Nc6 7. Be3 e5 8. Nf3
Ng4 9. Qd2 Nxe3 10. Qxe3 exf4 11. Qxf4 Be6 12. O-O-O Be7 13. Nd5 O-O 14. Kb1
Rc8 15. c3 Re8 16. Bd3 h6 17. Bc2 b5 18. h4 Ne5 19. Bb3 Kh8 20. Nxe5 dxe5 21.
Qxe5 Qd6 22. Qh5 Qg3 23. Nxe7 Rxe7 24. Bxe6 Rxe6 25. Qxf7 Rxe4 26. Qd7 Rb8 27.
Qxa7 Re2 28. Qd4 Rf8 29. Ka1 Kh7 30. Rhg1 Rff2 31. Rb1 Rf5 32. a3 Rh5 33. Qd7
Ree5 34. Rh1 Re2 35. Rhd1 Rxg2 36. Rd4 Qf3 37. Qe6 Re2 38. Qb3 Qf6 39. Rbd1
Ree5 40. c4 bxc4 41. Qxc4 Re3 42. Ka2 Rf3 43. R1d2 Re5 44. a4 Re8 45. a5 Qe5
46. Rd5 Qe6 47. b3 Qf6 48. h5 Rfe3 49. Rc5 Re1 50. Qd4 R8e5 51. Rxe5 Rxe5 52.
b4 Qe6+ 53. Ka3 Kh8 54. Qd8+ Kh7 55. Qd6 Re3+ 56. Ka4 Qe8+ 57. Qd7 1-0[/pgn]

[shrapnel]

I'm sorry but it's not fair. Stockfish is continually updated, why Houdini should remain the same? Even Martin Thoresen in the TCEC tournament follows the same criteria... What's the problem?

The 'problem' is that Uri is worried that while Houdini 4 is already beating Stockfish in LTC and on good Hardware like yours, a newer version of Houdini will COMPLETELY outclass Stockfish and CRUSH it.
So, according to him, its OK if Stockfish keeps getting updated, but he will cry "FOUL" if Houdini is updated !
There are words and phrases which everyone knows, which can be used to describe this sort of behaviour/attitude of the Stockfish Dev.Team, but I won't use them here because the Moderators here seem to be annoyed with me for speaking my mind frankly !
( So, just imagine what words I would use, and you'd be right ! )

[ernest]

Curious evaluation profile charts of both engines in the two games:

What do you feel "curious" here?

Also, I am surprised nobody tries to find an explanation as to why Houdini "seems" to perform better vs Stockfish in these games, compared to other LTC matches.

Of course it first has to be remembered that less than 100 games have been played so far.

Is it due to the narrow choice of openings? (is Stockfish not so good on Sicilians?)

Also note that a few games were played with Stockfish Syzygy compiles by Ronald de Man, dated Feb 18 to Feb 21 (1st compile), known to be buggy.
(see http://www.talkchess.com/forum/viewtopic.php?t=51378 )