Author: James T. Walker
Date: 14:35:52 12/21/01
Go up one level in this thread
On December 21, 2001 at 16:31:42, Uri Blass wrote: >On December 21, 2001 at 15:33:15, Bruce Moreland wrote: > >>On December 21, 2001 at 10:34:22, Graham Laight wrote: >> >>>All my chess computers/programs only allow you to set time control per move to >>>the nearest second. Even with "permanant brain" (where did THAT expression come >>>from?) off, this still makes them too strong for an enjoyable game of chess. >>> >>>A nice simple idea - why not allow the player to specify the time in >>>milliseconds? We could then level down the play by throttling off almost all of >>>the computer's thinking time. I'm sure that they'd be much more fun to play >>>against then. >>> >>>-g >> >>I had a program set up to play on ICC with 167 milliseconds per move, and no >>thinking on the opponent's time. It played 5 0 only, was typically rated 2300, >>and won matches with IM's. >> >>I had another one that thought for 20 milliseconds, and it was usually over >>2000. >> >>That was as well as I could do given my machine's clock resolution. > >I think that you can get more than it if you use the number of nodes and other >factors to get a good estimate for the time per move. > >Uri Hello Uri, When I had a computer account on ICC I let CSTal play some blitz on level 1. At that level it averaged about 200-300 nps and carried a blitz rating of around 2050. People just don't realize the blunders they make at blitz time controls. Many people seem also to be intimidated by computers and their ratings without realizing the kind of mistakes they (computers) make at fast time controls. Jim
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.