If you check the latest standing of CEGT 120/40 section, both Rybka 2.3 & Rybka 2.1c 64 have the same 750 games, but their + or - margin for their elo rating is different. Zap!Chess also have 750 games but it + or - margin is 19.
If they have the same number of games, it should be equal, the more games the lower the margin of error.
Am I missing something?
Also, sometimes the hardware used is the faster Intel xeon, sometimes AMD 64, it would then be hard to compare the relative strength of the sofware. or am I missing something again?
no Program Elo + - Games Score Av.Op. Draws
1 Rybka 2.3 64 2CPU 2998 20 20 750 72.1 % 2833 38.1 %
2 Rybka 2.1c 64 2CPU 2997 21 21 750 71.9 % 2833 35.9 %
9 Zap!Chess Paderborn 64 2CPU 2839 19 19 750 50.2 % 2838 39.9 %
How come CEGT is not consistent in displaying data?
Moderators: hgm, Rebel, chrisw
-
Rémi Coulom
- Posts: 438
- Joined: Mon Apr 24, 2006 8:06 pm
Re: How come CEGT is not consistent in displaying data?
Hi,
Uncertainty margins depend on individual games results. In particular, draws are more efficient than win and losses in reducing uncertainty. Two draws against the same opponent reduce uncertainty much more than one win and one loss. At least, this is how it works in bayeselo. You'll notice that in your results: the less draws, the more uncertainty.
Rémi
Uncertainty margins depend on individual games results. In particular, draws are more efficient than win and losses in reducing uncertainty. Two draws against the same opponent reduce uncertainty much more than one win and one loss. At least, this is how it works in bayeselo. You'll notice that in your results: the less draws, the more uncertainty.
Rémi