Author: Uri Blass
Date: 11:20:33 01/12/01
Go up one level in this thread
On January 12, 2001 at 13:46:35, Ratko V Tomic wrote: >> They _still_ have a lot of weaknesses. > >If Van der Wiel (or any GM) were to play 50 or 100 >games against the program, these would become quite >obvious and program would start losing badly. Even >much weaker players, like myself (around 2100 USCF >from early 1980s), after playing daily for a few >weeks with a program improve score substantially. The real contest is if the programmer can also change the opening book of the program and other parameters between the games. I did not try to play seriously against chess programs but I guess that if I try when the result is important for me then I may try to prepare a killer line against the program with white(for example I may play 1.a3 and 2.c3 and try to prepare all the moves before the game in some lines 1.a3 e5 2.c3,1.a3 d5 2.c3 when I hope that the program will be out of book after 2 moves) If the program is deterministic I may win every game with white by this way and even if it is not determinstic I may get a structure that I expect. It is not going to work if the programmer can change parameters between the games. He can change the evaluation function to avoid the same structure even if the program is objectively not better. 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.