CEGT 40/4 update online, incl. Houdini 3.0 x64 1CPU

[ThatsIt]

Hi to all !

Our new 40/4 list is online now:
http://www.husvankempen.de/nunn/blitz.htm

Code: Select all

Program                 ELO    +   -  Games  Score  Av.Op. Draws
Houdini 3.0 x64 1CPU   3088   14  14   1900  75.1%   2896  39.3%
...
Houdini 2.0c x64 1CPU  3032   11  11   2600  72.5%   2864  35.0%

Best wishes,
G.S.

[Ajedrecista]

Hello Gerhard:

Hi to all !

Our new 40/4 list is online now:
http://www.husvankempen.de/nunn/blitz.htm

Code: Select all

Program                 ELO    +   -  Games  Score  Av.Op. Draws
Houdini 3.0 x64 1CPU   3088   14  14   1900  75.1%   2896  39.3%
...
Houdini 2.0c x64 1CPU  3032   11  11   2600  72.5%   2864  35.0%

Best wishes,
G.S.

The draw ratio of Houdini 3 is wrong in your post: you wrote 39.3% (which is from Stockfish 2.2.2 x64 4CPU) instead of the correct 32.2%. The correct code box is the next one:

Code: Select all

Program                 ELO    +   -  Games  Score  Av.Op. Draws
Houdini 3.0 x64 1CPU   3088   14  14   1900  75.1%   2896  32.2%
...
Houdini 2.0c x64 1CPU  3032   11  11   2600  72.5%   2864  35.0%

Thank you very much for the tests! The current difference should be around 3088 - 3032 ± sqrt[(14)² + (11)²] = 56 ± sqrt(317) ~ 56 ± 17.8 Elo with the given confidence (I suppose 95%) and the current data (2600 games for Houdini 2.0c and 1900 games for Houdini 3).

Regards from Spain.

Ajedrecista.

[ThatsIt]

[...snip...]
The draw ratio of Houdini 3 is wrong in your post: you wrote 39.3% (which is from Stockfish 2.2.2 x64 4CPU) instead of the correct 32.2%.
[...snip... ]

Hi !

You're bang on!
Sorry for mistyping.

Best wishes,
G.S.