Author: David Rasmussen
Date: 09:09:01 12/23/00
Go up one level in this thread
On December 22, 2000 at 15:58:17, Ernst A. Heinz wrote: >Hi Brian, > >>What are the pro's and con's of doing null moves "at the root", e.g. after a ply >>1 move and then another null move by the same side. It seems to help >>dramatically in some tactical situiations, but not others. Thus far, Tinker has >>only used null moves starting at ply 2 (actually ply 3) moves, e.g, Tinker makes >>a move, then opponent replies, then another null move reply. > >Why do you deem making a null move directly after the first move >from the root dangerous or not appropriate? > >I still wonder what, in particular, "Junior" uses the null-move >score for which it seems to calculate for the root node _itself_. > >Any ideas? > >Amir? > >=Ernst= Maybe it could tell something about the nature of the position, such as it being a strategical position more than a tactical one. If doing a nullmove doesn't fail low, maybe the position is relatively more quiet, and maybe it then deserves extended time for the program to see the right "plan". I find in general that programs have the hardest time assessing positions that are roughly equal. If there are clear advantages or disadvantages, the program will adress these, whereas, if there isn't, they will just wander about, trying to improve piece placement etc.
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.