Will we see RYBKA 2.3.2 4CPU in cegt?

[ArmyBridge]

Ray how many games do you think CCRL must do in Zanzibar vs Rybka 2.3.2 to get a good reference of the strength between both engines?, recently M ANSARI said " While trying to get new lines for my book I decided to let Rybka play Zap Zanzibar on my Quad running 64bit OS and 3.2Ghz. Zap is no push over ... but from the 100 games I put it to play against Rybka ... Zap only managed 3 wins vs 41 losses and rest were draws", look Zap only get 3 wins!!!! the CCRL result is a early result but.... the diference is huge ... I guess 100 games maybe is not enough to get a good parameter, what do you think?

[Spock]

50 games per pair is enough. Then multiply that by say 10 pairs = 500 games which is enough for a reliable rating

[ArmyBridge]

Humm...In CEGT :Rybka 2.3.2 x64 2CPU vs Zap!Chess Zanzibar x64 4CPU - 2995 50 + 12 = 30 - 8 54.0 % , very close to CCRL result in the same match, it is a pity that Zanzibar will not be more update for Anthony Cozzie maybe Zanzibar is the best reference for Rybka development, I guess until Rybka don´t smash Zap!Chess Zanzibar x64 4CPU we would not say that Rybka did a great improvement is just a good improvement

[Spock]

I did make another post about 2.3.2a quad performance but deleted it. Decided Rybka forum was a better place for it

[M ANSARI]

Yes I tested Rybka 2.3.2a vs Zap Zanzibar and results were very very lopsided in favour of Rybka. But please note that I used Rybka II.ctg versus Zappa's book. I think CCRL and CEGT do not use these books and have certain neutral books.

[ArmyBridge]

Hi ANSARI
I ´have very interest to see your test , would you send it to me?
Regards