Author: Robert Hyatt
Date: 16:57:54 11/28/02
Go up one level in this thread
On November 27, 2002 at 18:46:38, Joshua Haglund wrote: >What is Robert's initial purpose for a "standard" rating increase? >Maybe the purpose is to focus on long matches with humans? > >How does the top 25 engines do at 3 minutes / move? >Would Crafty be at the top? :) > >Joshua >toneewa@yahoo.com I think the human GM "problem" is an interesting one. Blitz is "almost solved" in the context that I don't think a GM can go toe-to-toe with a comp in blitz and win a match of any duration. A game here and there, yes. But not a match. But long time controls are something different. If you watch comp vs GM blitz games, the GM often builds up a completely winning position but makes a positional blunder or tactical blunder that loses the game. In longer games, these are much less frequent, which means such games are going to depend on the programm winning them, not on the human losing them. That is where it is "interesting" to work...
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.