Author: Mike S.
Date: 19:09:12 05/07/03
Go up one level in this thread
On May 07, 2003 at 21:10:51, Matthew White wrote: >On May 06, 2003 at 15:36:01, Mike S. wrote: >>It's astonishing how effective the *position learning* can be: I've just run >>some long matches - up to 60 games! - with 2 engines each, from the *same >>opening variant* each, same conditions for each game (alternate colors). You'd >>think this will give many doubles. But it turned out that every game was >>different! Most often, the alteration came relatively early, not just somewhere >>at the end of the game. >This makes the assumption that the only cause of variance in a game is >positional learning. In many positions, hashing differences, time management >algorithms, or even gnomes :) could effect the move played. This is true. I didn't assume that the position learning was the only reason (although I think it was the main reason by far). Since hash tables aren't deleted anymore, it may be that entries from the previous games are still present, thus having influence... I've seen an engine play a different move in the same position which was repeated during a match, which did *not* have position learning. Most probably this was due to the different hash tables content. I've also read once, that some engines use some randomness determing their time consumption after the opening book (maybe Shredder?), to make repetition of games more difficult. Not to forget the gnomes :o)) Regards, M.Scheidl
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.