Author: Andrew Williams
Date: 15:52:36 06/19/00
Go up one level in this thread
On June 19, 2000 at 11:35:42, Gian-Carlo Pascutto wrote: >On June 19, 2000 at 08:14:28, Andrew Williams wrote: > >>Yes. Mine is MTD too, but it doesn't print out the detailed >>information on every increment. It's interesting to see that >>SOS seems to increment by 1 most of the time. Mine "accelerates" >>the increment as it becomes clear that the score is going up a >>long way. > >It is probably using the plain MTD(n,f) algorithm as described by >Plaat et al. > >I.e. using the result of the search as the new guess. > >I remember from Aske's papers that he thought it might be possible >to improve it by doing something like you describe. > >So, I suppose it does help? > Yes. It's a very long time since I fiddled with it, but it is consistently better than going up or down by one unit at a time. The problem is that it adds a bit of complexity which isn't there for Aske Plaat's formulation; I think he showed it like that so as to make it clear to people. Andrew
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.