Author: Dan Andersson
Date: 05:20:39 03/15/01
Go up one level in this thread
I am certain that the DF Tiger games where a special book made by a human, i.e. book cooking. So I fail to see the relevance of that. As for comp-comp games as an indicator, if the program in question sees an advantage for one side it will most probably make it a self fulfilling prophecy by compromising the position of the side it does not favour. For example it might decide to sacrifice a pawn for nothing for some illusory activity since the sore is so bad anyways, the chances for that is increased (I'm assuming a normal distributon of results here). I did some quick tests too by autoplaying a program that favors black and lo and behold black scored big in this position. But there are ways of getting better results from c-c testing by using different programs using positional learning. But all in all the best results are gained by playing one side in a couple of games and/or analyzing without computer and checking it out afterwards. But if the program is much too strong (maybe 400 points higher) it might simply overwhelmyou whatever you do. I only said that I believe Gandalf and Tiger are stronger than Crafty in the opening phase (They play great there)but they migth be stronger than others too. The opening phase does not end with the book it might actually continue for quite some time later. Rgards Dan Andersson
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.