Author: Vincent Diepeveen
Date: 13:57:35 04/25/01
Go up one level in this thread
On April 25, 2001 at 08:46:48, Mogens Larsen wrote: >On April 25, 2001 at 08:39:19, Vincent Diepeveen wrote: > >>Diep doesn't have learning. With a small tournament book you keep >>on losing if you lose the first game. The same happens when opponent has >>better learning. All commercials that test for SSDF know *exactly* how >>the learning of others reacts on situations. >> >>Very easy to measure. > >I know, but it doesn't mean that it's unfair. The same applies to good or bad >books or EGTB support. If Diep doesn't have learning and loses because of it, >then it's your own fault and nobody elses. Competing is rarely about the lowest >common denominator. > >Mogens. Fact is that this learning is not relevant for tournaments where Shredder joins, nor is it relevant for matches versus Kramnik, as you can each night turn on a different variation by hand after some analysis. Here they decide the player of Kramnik-Kasparov partly also by using an auto232 player where learning is of major importance. A program that has to learn without being able to touch it by hand in the meantime is an unguided missile! The lucky coincidence in this respect is that both progs probably use the same learning.
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.