Author: Roy Eassa
Date: 16:27:05 04/21/02
Go up one level in this thread
On April 21, 2002 at 19:13:01, Mike S. wrote: >On April 21, 2002 at 19:02:13, Uri Blass wrote: > >>(...) > >>If chessbase cannot know everything about kramnik's preperation in order to >>change the program based on the weaknesses that kramnik finds during the >>preperation than Fritz does not get optimal playing conditions. > >I really don't see why Fritz would have to worry. :o) > >There are so many reasons why a program's moves are unpredictable. Kramnik can >only try to choose an opening variation (which ChessBase will undoubtly be well >prepared for, at least a far as all public GM games are concerned). And as soon >as the engine calculates, there could be: > >1. slightly different response time from Kramnik -> different comp move >2. slightly different time control setting, i.e. to cover handling time -> >different comp move >3. much bigger hash tables than Kramnik used in preparation -> different comp >move >4. 8 CPUs Kramnik probably won't have available in preparation -> different comp >move > >He can try to adapt to the style, try to find more "general" weaknesses. But for >that, I think it won't matter much if this is Fritz 7.0.0.6 or Deep Fritz 7.x or >whatever (probably it doesn't even matter if it's Fritz, Junior or Nimzo). > >Does somebody really think Kramnik can prepare "Anti-Fritz" especially, i.e. >thing which wouldn't work against any other program too? Forget it... > >Regards, >M.Scheidl Well said! It's playing against strong computers IN GENERAL that the GMs, including Kramnik, must master. The strengths and weaknesses of computers are so completely different from those of human opponents that it's likely to be very beneficial for a GM to substantially refocus his (or her) preparation to meet this very different type of opponent.
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.