OpenChess

Would there be any reason to do a tournament with a fixed search depth and unlimited time? My thought is that a fixed depth rating list would indicate more accurate evaluation functions. The problem is that sometimes speed compensates for accuracy and those two things work in concert. Anyone else have thoughts on this?

And how we know that the depth reported is consistent or uniformed between all engines?
Better explained are we sure depth 18 is the same for engine A and B?

We actually know that depth 18 is NOT the same between engine A and B, sometimes even among different versions of the same engine any depth is NOT the same (for instance, Rybka 4's multiprocessor version's depth is much thicker than Rybka 4's single processor, when ran at >1CPU).

Ovyron wrote:(for instance, Rybka 4's multiprocessor version's depth is much thicker than Rybka 4's single processor, when ran at >1CPU).

You mean the search tree is thicker?...
Well, Vasik said something about that (maybe that's why he is so intent on his cluster version), but I am not sure if it is really very significant.

Yes, and it is significant enough for the MP version to beat the SP version in any match of reasonable length, at fixed depth.

Ovyron wrote:...for the MP version to beat the SP version in any match of reasonable length, at fixed depth.

Do you have a link for such (Rybka) matches?

No, AFAIK those were private Vas's matches, but should be reproducible by everyone, you can try playing Rybka MP (using all your cores) - Rybka SP at fixed depth 1, it should give you thousands of games in a very short time.

Ovyron wrote:No, AFAIK those were private Vas's matches, but should be reproducible by everyone, you can try playing Rybka MP (using all your cores) - Rybka SP at fixed depth 1, it should give you thousands of games in a very short time.

Well, I did it (depth 12), no success!...
see http://rybkaforum.net/cgi-bin/rybkaforu ... #pid286083

Heh, I guess my lesson is not to take whatever Vas says at face value