The Champions 2012 4CPU

[Vinvin]

THE CHAMPIONS 2012 4CPU
...
Round 24

Gull II b2 64-bit 4CPU v Komodo 5 64-bit (draw)
IvanHoe 9.46h 64-bit 4CPU v Strelka 5.5 64-bit (draw)
Vitruvius 1.11C 64-bit 4CPU v Bouquet 1.5 64-bit 4CPU (draw)
Sting SF 2 64-bit 4CPU v DeepSaros 3.1a 64-bit 4CPU (draw)
Equinox 1.60 64-bit 4CPU v Chiron 1.5 64-bit 4CPU (1-0)
Stockfish 2.3.1 64-bit 4CPU v Naum 4.2 64-bit 4CPU (draw)
Rybka 4.1 64-bit 4CPU v Hiarcs 14 4CPU (draw)
Critter 1.6a 64-bit 4CPU v Houdini 3 64-bit 4CPU (0-1)


Standings after Round 24

16.5 - Critter 1.6a 64-bit 4CPU
16.0 - Rybka 4.1 64-bit 4CPU
15.0 - Houdini 3 64-bit 4CPU
14.5 - Vitruvius 1.11C 64-bit 4CPU
14.5 - Sting SF 2 64-bit 4CPU
14.0 - Strelka 5.5 64-bit
12.0 - Bouquet 1.5 64-bit 4CPU
12.0 - Equinox 1.60 64-bit 4CPU
11.5 - Hiarcs 14 4CPU
11.5 - Stockfish 2.3.1 64-bit 4CPU
10.5 - IvanHoe 9.46h 64-bit 4CPU
10.5 - Komodo 5 64-bit
10.0 - DeepSaros 3.1a 64-bit 4CPU
8.5 - Naum 4.2 64-bit 4CPU
8.5 - Chiron 1.5 64-bit 4CPU
6.5 - Gull II b2 64-bit 4CPU


Round 24 PGN - http://kirill-kryukov.com/chess/discuss ... p?id=28053

Houdini beat the 1st (Critter) ! Always a good strategy to win a tournament !
Houdini3 still 1.5 late and it still 6 rounds to play ... easy path to victory now

[S.Taylor]

Houdini3 still 1.5 late and it still 6 rounds to play ... easy path to victory now

We don't know what strategy is in machines minds. Maaybe Houdini plans to lose 2 and draw 4.
OK, that might also win the tournament if the others at the top lose all from now. But they don't seem to be in that mood.

[Leto]

After 195 games played:
Critter's (16.5/24) remaining opponents: Rybka 4.1 (16/24), Vitruvius (14.5/24), BOUQUET (12.5/24), Equinox (12/24), Komodo (11/25), Chiron (8.5/24) total opponents score: 74.5 (avg 12.41)

Rybka's (16/24) remaining opponents: Critter (16.5/24), Strelka (15/25), Sting (14.5/24), Stockfish (11.5/24), Ivanhoe (11.5/24), Deep Saros (10/25) total opponents score: 79 (avg 13.16)

Houdini's (15/24) remaining opponents: Strelka (15/25), Sting (14.5/24), Stockfish (11.5/24), Ivanhoe (11.5/25), Deep Saros (10/25), Naum (8.5/24) total opponents score: 71 (avg 11.83)

Houdini's remaining opponent field will be easier than Critter's and Rybka's.

[Ajedrecista]

Hello!

After 195 games played:
Critter's (16.5/24) remaining opponents: Rybka 4.1 (16/24), Vitruvius (14.5/24), BOUQUET (12.5/24), Equinox (12/24), Komodo (11/25), Chiron (8.5/24) total opponents score: 74.5 (avg 12.41)

Rybka's (16/24) remaining opponents: Critter (16.5/24), Strelka (15/25), Sting (14.5/24), Stockfish (11.5/24), Ivanhoe (11.5/24), Deep Saros (10/25) total opponents score: 79 (avg 13.16)

Houdini's (15/24) remaining opponents: Strelka (15/25), Sting (14.5/24), Stockfish (11.5/24), Ivanhoe (11.5/25), Deep Saros (10/25), Naum (8.5/24) total opponents score: 71 (avg 11.83)

Houdini's remaining opponent field will be easier than Critter's and Rybka's.

I did not use PGN files but I computed a rating list (with an own clumsy Fortran programme) based on the performance of each engine, with 24 games played by each engine. My ratings should be similar to EloSTAT ratings:

Code: Select all

Round Robin with 16 engines and 24 games per engine.
Total number of games: 192 games.
 
 3128.07 (Critter 1.6a 64-bit 4CPU).
 3112.62 (Rybka 4.1 64-bit 4CPU).
 3083.06 (Houdini 3 64-bit 4CPU).
 3068.80 (Vitruvius 1.11C 64-bit 4CPU).
 3068.80 (Sting SF 2 64-bit 4CPU).
 3054.79 (Strelka 5.5 64-bit).
 3000.24 (Bouquet 1.5 64-bit 4CPU).
 3000.24 (Equinox 1.60 64-bit 4CPU).
 2986.72 (Hiarcs 14 4CPU).
 2986.72 (Stockfish 2.3.1 64-bit 4CPU).
 2959.49 (IvanHoe 9.46h 64-bit 4CPU).
 2959.49 (Komodo 5 64-bit).
 2945.68 (DeepSaros 3.1a 64-bit 4CPU).
 2902.83 (Naum 4.2 64-bit 4CPU).
 2902.83 (Chiron 1.5 64-bit 4CPU).
 2839.66 (Gull II b2 64-bit 4CPU).
 
Mean of ratings:  3000.00 Elo.

Considering all these performances as true ones (which is a lot), the mean of ratings of the future opponents to the current top-3 are more less:

Code: Select all

1.- Critter 1.6a 64-bit 4CPU (16.5/24): six opponents with an average performance of 3007.37 Elo.
2.- Rybka 4.1 64-bit 4CPU (16/24):      six opponents with an average performance of 3023.93 Elo.
3.- Houdini 3 64-bit 4CPU (15/24):      six opponents with an average performance of 2986.39 Elo.

I hope no typos in my calculations; please do not consider these performance ratings very seriously but a rough idea.

------------------------

@Graham: I expect to see the new Bouquet 1.6 in your next 8 CPU tournament... please do not forget to upgrade if 1.6 version is not worse than 1.5 (which is a matter of testing)!

I am also curious about Equinox: AFAIK it is still private but I read somewhere that the author wants to make it public in a future... is it true? Thanks in advance. Please keep up your good work!

Regards from Spain.

Ajedrecista.

[beram]

THE CHAMPIONS 2012 4CPU
...
Round 24

Gull II b2 64-bit 4CPU v Komodo 5 64-bit (draw)
IvanHoe 9.46h 64-bit 4CPU v Strelka 5.5 64-bit (draw)
Vitruvius 1.11C 64-bit 4CPU v Bouquet 1.5 64-bit 4CPU (draw)
Sting SF 2 64-bit 4CPU v DeepSaros 3.1a 64-bit 4CPU (draw)
Equinox 1.60 64-bit 4CPU v Chiron 1.5 64-bit 4CPU (1-0)
Stockfish 2.3.1 64-bit 4CPU v Naum 4.2 64-bit 4CPU (draw)
Rybka 4.1 64-bit 4CPU v Hiarcs 14 4CPU (draw)
Critter 1.6a 64-bit 4CPU v Houdini 3 64-bit 4CPU (0-1)


Standings after Round 24

16.5 - Critter 1.6a 64-bit 4CPU
16.0 - Rybka 4.1 64-bit 4CPU
15.0 - Houdini 3 64-bit 4CPU
14.5 - Vitruvius 1.11C 64-bit 4CPU
14.5 - Sting SF 2 64-bit 4CPU
14.0 - Strelka 5.5 64-bit
12.0 - Bouquet 1.5 64-bit 4CPU
12.0 - Equinox 1.60 64-bit 4CPU
11.5 - Hiarcs 14 4CPU
11.5 - Stockfish 2.3.1 64-bit 4CPU
10.5 - IvanHoe 9.46h 64-bit 4CPU
10.5 - Komodo 5 64-bit
10.0 - DeepSaros 3.1a 64-bit 4CPU
8.5 - Naum 4.2 64-bit 4CPU
8.5 - Chiron 1.5 64-bit 4CPU
6.5 - Gull II b2 64-bit 4CPU


Round 24 PGN - http://kirill-kryukov.com/chess/discuss ... p?id=28053

Houdini beat the 1st (Critter) ! Always a good strategy to win a tournament !
Houdini3 still 1.5 late and it still 6 rounds to play ... easy path to victory now

The position right out of the opening was already very favorable for black
Btw I couldn't find just one game in my 3,5 million database with it.
[d] rnbqkb1r/p4p1p/2p1pp2/8/PppP4/2N2NP1/1P2PP1P/R2QKB1R w KQkq - 0 9

Stockfisch 2.2.2 gives already +1,1 for black in this position; Houdini 3 0,58; Komodo 3 0,46;
So lucky Houdini here
Not to criticize, but the openingsbook WorldClass2012-2 gave black here an (almost) easy win, as Houdini wins 80% of his games where it has a +0,5

but the tension is back in tournament

grts Bram

[Graham Banks]

@Graham: I expect to see the new Bouquet 1.6 in your next 8 CPU tournament... please do not forget to upgrade if 1.6 version is not worse than 1.5 (which is a matter of testing)!

I am also curious about Equinox: AFAIK it is still private but I read somewhere that the author wants to make it public in a future... is it true? Thanks in advance. Please keep up your good work!

Regards from Spain.

Ajedrecista.

I'll include Bouquet 1.6 in my upcoming 8 core tournament, but won't be able to test it extensively beforehand.

From what I can gather, the Equinox author intends to publicly release his engine not too far off in the future.

Cheers,
Graham.

[Graham Banks]

THE CHAMPIONS 2012 4CPU

Xeon X5430x2 Octal
ChessGUI
1024mb hash
3-4-5 piece tablebases
Ponder off
WorldClass2012-2.cgb book (limited to 8 move depth)
40 moves in 29 minutes repeating (adapted for the CCRL)
All engines 64-bit 4CPU where available
2 cycles 30 rounds


Round 25

Gull II b2 64-bit 4CPU v Strelka 5.5 64-bit (0-1)
Komodo 5 64-bit v Bouquet 1.5 64-bit 4CPU (draw)
IvanHoe 9.46h 64-bit 4CPU v DeepSaros 3.1a 64-bit 4CPU (1-0)
Vitruvius 1.11C 64-bit 4CPU v Chiron 1.5 64-bit 4CPU (0-1)
Sting SF 2 64-bit 4CPU v Naum 4.2 64-bit 4CPU (draw)
Equinox 1.60 64-bit 4CPU v Hiarcs 14 4CPU (1-0)
Stockfish 2.3.1 64-bit 4CPU v Houdini 3 64-bit 4CPU (0-1)
Rybka 4.1 64-bit 4CPU v Critter 1.6a 64-bit 4CPU (1-0)


Standings after Round 25

17.0 - Rybka 4.1 64-bit 4CPU
16.5 - Critter 1.6a 64-bit 4CPU
16.0 - Houdini 3 64-bit 4CPU
15.0 - Sting SF 2 64-bit 4CPU
15.0 - Strelka 5.5 64-bit
14.5 - Vitruvius 1.11C 64-bit 4CPU
13.0 - Equinox 1.60 64-bit 4CPU
12.5 - Bouquet 1.5 64-bit 4CPU
11.5 - Hiarcs 14 4CPU
11.5 - IvanHoe 9.46h 64-bit 4CPU
11.5 - Stockfish 2.3.1 64-bit 4CPU
11.0 - Komodo 5 64-bit
10.0 - DeepSaros 3.1a 64-bit 4CPU
9.5 - Chiron 1.5 64-bit 4CPU
9.0 - Naum 4.2 64-bit 4CPU
6.5 - Gull II b2 64-bit 4CPU


Round 25 PGN - http://kirill-kryukov.com/chess/discuss ... p?id=28056

If you install TLCV (Tom's Live Chess Viewer) on your computer, you can watch the games live move by move. You'll also be able to chat to others following the tournament in the chatroom there.
http://home.pacific.net.au/~tommyinoz/client.zip
Host - GrahamCCRL.dyndns.org Port - 16083

There is also a live broadcast in Playchess.

[Lavir]

Not to criticize, but the openingsbook WorldClass2012-2 gave black here an (almost) easy win, as Houdini wins 80% of his games where it has a +0,5

IMO the WorldClass2012 book is a very good book for testing engines in lists etc. where you play many games in reversed positions, because it has a very wide list of openings and so you can test engines in various scenarios. This gives the opportunity to ascertain real strength of an engine (the more the scenarios, the more you are sure the engine is not setted to work well under certain conditions only).

However in a tournament with little games played it's not a convenient book (as many other similar ones) just because of its wideness. Many openings are simply subpar, being drawish, having little dynamicity and so on.

For Swiss/Robin tournaments I think the best books are those with very strong limited openings so that engines don't add a "luck" factor coming from the lines chosen by the book. The more wide the book is, the more the probability a line will be not so good and the engine (no matter how good it is) will have a bad game, while maybe another engine instead get a good line and can shine. Openings are a substantial factor in all chess games, and they can indeed create a result (so to speak) before a game is even started.

The same can be said about testsuites that are made to test engines in a wide array of openings, but just for this many of them are not very good when you play a swiss/robin tournament. Books like the Rybka, Hiarcs, even Fritz ones (or even playchess books) used to come only to a certain depth (pherhaps 8/12 ply deep) are some of the better for these types of tourneys, at last IMO.

[velmarin]

Code: Select all

beram wrote:
Not to criticize, but the openingsbook WorldClass2012-2 gave black here an (almost) easy win, as Houdini wins 80% of his games where it has a +0,5 

In round nine Critter beat Houdini with the same opening.

The openings are repeated with changed colors.
http://kirill-kryukov.com/chess/discuss ... p?id=27988

[Leto]

Code: Select all

beram wrote:
Not to criticize, but the openingsbook WorldClass2012-2 gave black here an (almost) easy win, as Houdini wins 80% of his games where it has a +0,5 

In round nine Critter beat Houdini with the same opening.

The openings are repeated with changed colors.
http://kirill-kryukov.com/chess/discuss ... p?id=27988

The opening does not appear to be a guaranteed win for black, here's Fritz 6 losing with black:

[Event "Houdini 3 testing 13"]
[Site "?"]
[Date "2012.11.30"]
[Round "1.1"]
[White "Houdini 3 Pro x64 2CPU"]
[Black "Fritz 6 Light"]
[Result "1-0"]
[ECO "D44"]
[Annotator "-0.31;-0.31"]
[PlyCount "93"]
[EventDate "2012.11.30"]
[EventType "tourn"]
[Source "Leto"]
[TimeControl "120"]

{Intel(R) Core(TM)2 Duo CPU T7300 @ 2.00GHz 1994 MHz W=14.2 plies; 2,
163kN/s; testing.ctg B=9.9 plies; 2,122kN/s; testing.ctg} 1. d4 {B 0} d5 {B 0}
2. c4 {B 0} c6 {B 0} 3. Nf3 {B 0} Nf6 {B 0} 4. Nc3 {B 0} e6 {B 0} 5. Bg5 {B 0}
dxc4 {B 0} 6. a4 {B 0} b5 {B 0} 7. Bxf6 {B 0} gxf6 {B 0} 8. g3 {B 0} b4 {B 0}
9. Ne4 {-0.31/17 4} f5 {-0.31/11 6} 10. Ned2 {-0.35/18 2} c3 {-0.44/12 3} 11.
bxc3 {-0.35/17 3} bxc3 {-0.44/12 2} 12. Nb3 {-0.40/18 3 (Nb1)} Na6 {-0.69/11 5
(Nd7)} 13. Bg2 {-0.15/17 2 (Rc1)} Nb4 {-0.81/11 5} 14. Ne5 {-0.17/17 2} Bb7 {
-0.66/10 3 (Qc7)} 15. O-O {0.25/14 2 (Rc1)} Rb8 {-0.88/9 2 (Qc7)} 16. Rc1 {0.
31/16 2} c2 {-0.63/10 2} 17. Rxc2 {0.31/15 0} Nxc2 {-0.53/11 5} 18. Qxc2 {0.19/
17 3} Rc8 {-0.59/10 3} 19. Rd1 {0.11/17 13 (Rb1)} Ba8 {-0.75/10 3 (Qc7)} 20. e4
{0.08/15 2 (Nxc6)} fxe4 {-1.13/10 2 (f4)} 21. Bxe4 {0.12/14 4 (Nc5)} Bg7 {-1.
09/10 7 (Qc7)} 22. Nc5 {0.08/15 2} Qc7 {-1.00/10 6 (Qe7)} 23. f4 {0.48/15 1}
Qd6 {-0.69/9 2} 24. Bxh7 {0.65/15 2} Rc7 {-0.38/9 2 (Ke7)} 25. Be4 {0.63/15 3}
Rh6 {-0.41/9 2 (0-0)} 26. a5 {0.85/15 1 (Bg2)} Rc8 {-0.53/9 2 (Kf8)} 27. h4 {
1.01/14 3 (a6)} Rb8 {-0.56/9 2 (f5)} 28. a6 {0.99/14 1} Ke7 {-0.56/9 2 (Qc7)}
29. Bf3 {0.86/14 1} Qc7 {-0.59/8 2 (Rhh8)} 30. h5 {1.12/12 1 (Ne4)} Rg8 {-0.63/
8 1 (Rhh8)} 31. Kg2 {1.57/13 1 (g4)} Rd8 {-0.41/8 2 (Bxe5)} 32. Nb7 {2.65/15 1
(g4)} Bxe5 {0.25/10 2} 33. fxe5 {2.63/14 0 (Nxd8)} Bxb7 {-0.31/10 1} 34. Qc5+ {
5.77/15 1 (axb7)} Kd7 {0.28/12 2} 35. axb7 {5.77/14 0} Qxb7 {0.53/11 3} 36. Rb1
{5.70/13 0 (Rc1)} Qa8 {1.09/10 1 (Qc7)} 37. Rc1 {5.87/12 0 (g4)} Rc8 {3.06/12 1
} 38. Qd6+ {5.87/11 0} Ke8 {0.75/7 0} 39. Rxc6 {5.87/10 0 (g4)} Rb8 {3.78/10 1
(Rd8)} 40. Rc8+ {6.89/9 0 (Rc7)} Rxc8 {3.59/7 0} 41. Bxa8 {6.60/9 0} Rxh5 {5.
28/10 1} 42. Bc6+ {6.08/9 0 (d5)} Rxc6 {5.03/7 0} 43. Qxc6+ {6.03/9 0} Kf8 {6.
37/11 2} 44. d5 {10.98/17 1} exd5 {7.12/11 1} 45. Qc8+ {10.99/15 0 (Qf6)} Kg7 {
7.41/12 3 (Ke7)} 46. Qg4+ {11.31/15 0} Kh6 {8.44/12 1} 47. Qf4+ {11.31/14 0
(Qg8)} 1-0

Question is how much stronger must white be in order to win with this opening?