Author: Dieter Buerssner
Date: 04:09:12 11/23/01
Go up one level in this thread
On November 23, 2001 at 01:11:20, Uri Blass wrote: >>Doesn't matter, when you think about the math. Either you keep all the >>positions from the first searches, which means newer searches (going backward) >>will have no table space to write into, or vice-versa... > >It is also possible to keep always the scores of the position from the game and >the depth so the program does not need to analyze game positions when the >remaining depth is smaller when it goes backward. I agree, and this is about what I am doing in Yace in Analyze mode. >In this case if there is a long forced line in the game that is too deep for the >program then the program may analyze better when you go backward. Yes. At least it should never hurt, it can only yield better results (or when the result should be misleading, it is because the engine had no idea anway). The positional learning of engines works similar. It even hopes, that only one single analyzed position is enough, to change a move in an earlier game phase. Even without any special learning, analyzing games from back to front should allways be better. All commercial GUIs I tried do this. Regards, Dieter
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.