Author: Uri Blass
Date: 09:04:00 04/01/02
Go up one level in this thread
On April 01, 2002 at 11:03:40, Ratko V Tomic wrote: >> At that time we were dreaming of having a chessprogram with >> GM strength at home. That seems to be a clear option now...even >> if the programs diverse heavily in regard to strategical vs >> tactical abilities... > > >It is doubtful a GM will have a GM strength program at home >beyond the initial few dozen serious games. Even a player of much >lower rating notices how programs appear much weaker after some >weeks of use. I recall taking Hiarcs 7.32 to a vacation in the >woods, where I ended up playing almost around the clock for >a week (mostly at a tempo of 30 sec/move). While at first I >was losing probably 19 out of 20, by the end I was losing >about 8-2 to 7-3, even though my rating was only 2100 (and >that is from 17+ years ago, and I haven't played on human >competitions since then). I had sense that another few weeks >of intense play and we would have played at even score. I believe that you learned something from your games against Hiarcs and I guess that you can get a performance of 2200-2300 against humans if you continue to play against humans in the same way that helped you to lose 8-2 or 7-3. I also think that computers may be more unstable because there is some chance to get them into positions that they do not understand. suppose you have a player that in 80% of the games plays like kasparov and in 20% of the games play like 1600 player. I can get 20% against that player but it does not mean that the player cannot get the GM norms and become a GM. Uri
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.