Author: blass uri
Date: 03:24:16 11/16/98
Go up one level in this thread
On November 15, 1998 at 16:53:02, Amir Ban wrote: >On November 15, 1998 at 14:29:41, Jürgen Hartmann wrote: > >>Two or three months ago I read in a usenet discussion group that an english >>programmer (Whittington) claimed that with his new version of an autoplayer >>connection he had found ways to disable the learning of all his opponents. He >>said he did it because the introduction of learning was a consequence of the >>somewhat artificial autoplayer match testing and had not much to do with real >>world chess. >> >>I thought his solution was amusing. Those new learning and anti-learning >>features of chess software remind me of a game we used to play as students on >>the Atari ST: It was called "Core Wars". You had to write an assembler program >>in a software-emulated address space and the only purpose was to overwrite >>'enemy' programs. >> >> >>Jürgen Hartmann > >What was the method he suggested ? > >I've also written posts here about "what to do against a learner". A wide book >and a little randomization is all you need. > >Amir This is enough in games against opponents with a learner but it is not enough in games against opponent without learner and without a wide book and a little randomization. I saw fritz5 repeat the same win in the ssdf games 5 times(against Rebel8(p90)) If Junior5 does not do the same then it gives fritz5 an advantage in the ssdf rating 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.