Author: Amir Ban
Date: 16:43:58 09/30/98
Go up one level in this thread
On September 30, 1998 at 17:49:34, Enrique Irazoqui wrote: >On September 30, 1998 at 17:14:49, Bruce Moreland wrote: > >>You can find patterns in any random thing though. > >I don't think patterns found in 506 games are random. And you still have to >explain why Fritz 5 scores better in autoplayed than in manual games: Paris, >Torstein, Shep, Thorsten, KK cup... I think this pattern showing the strong >influence of Fritz learner in long sequences of auto232 games can explain it. > >Put it another way. Had the autoplayed matches been 10 games long instead of 20 >or 40, Fritz 5 would have scored much closer than in the manual tournaments it >played. > >If this is true, we are also talking about the general validity of ratings >obtained after autoplayed games. It seems quite important to me. > >Enrique > You mean obtained with learning. Autoplaying has nothing to do with it. It would be the same if the games were played manually. Just a summary of my own (Junior) experience: I play often against Fritz5, always with learning disabled, and usually with the Powerbook, and I find it to be extermely strong, probably stronger than any other program I know. Perhaps it gets even stronger with learning. Amir
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.