Author: Jorge Pichard
Date: 09:16:40 01/26/04
Go up one level in this thread
On January 26, 2004 at 11:21:09, Tord Romstad wrote: >On January 26, 2004 at 10:49:02, Jorge Pichard wrote: > >>Since the Opening is NOT that important in FRC and HIARCS is leading, the >>positional understanding of HIARCS is much higher than the rest. > >No, this does not follow. It is perfectly possible that the evaluation function >in Hiarcs is just better at handling the unusual opening positions you see in >FRC than the evaluation function in the other top engines. > >For instance: If I make an FRC version of my engine some time soon (which is >likely), I will rewrite parts of my evaluation function. In particular, the >way I evaluate development makes no sense at all in FRC, and should be replaced >with new and very different code. This change will help greatly in FRC, but >make my engine very slightly weaker in normal chess. > >>There is clear indication that with a better opening book, HIARCS could be >>the leader in SSDF as well. > >I disagree again. A better way to do this is to run a series of Nunn matches >against the other top programs. > >Don't get me wrong; I am not trying to knock Hiarcs. Hiarcs is an >awesome chess program, and based on the games I have seen it play it is >easily my favorite among the commercial engines. I don't think I have >ever looked forward to the release of a new chess program as much as >I currently look forward to Hiarcs for PalmOS. In short, I am a great >fan of Hiarcs. But that it narrowly defeats the other top engines in >an FRC tournament only indicates that it is probably a better FRC >engine than its competitors, nothing more. You can't use this to make >conclusions about its positional understanding. > >Tord HIARCS was not tuned for playing FRC, neither Fritz, nor Shredder and they are all handling these ugly 15 positions. I'm not concluding that HIARCS has a bad opening book simply because it plays good with these 15 positions, but because I have seen in several of my matches and other people who post games of HIARCS that HIARCS Openings was NOT tuned for HIARCS. And strongly believe that HIARCS could gain another 15 or 20 points if it had a better opening book. Jorge
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.