Author: Tord Romstad
Date: 11:21:27 01/26/01
Go up one level in this thread
On January 25, 2001 at 15:32:39, Ed Schröder wrote: >On January 25, 2001 at 14:00:38, Tord Romstad wrote: > >>On January 25, 2001 at 13:15:59, Ed Schröder wrote: >> >>>The goal of null-move is to lower the (current) depth with X-plies, in >>>CC terms called R=X. Most common is R=2. "R" stands for reduce. >>> >>>Of course you don't have to do a null-move to practise a R=2. In Rebel >>>I do R=1, R=2 and complete razoring (R=INFINITE) based on eval and some >>>special code not to prune / reduce important parts of the tree. >> >>Do you use R=INFINITE for particularly bad-looking moves everywhere in the >>tree, or just near the leaves? It seems very dangerous to do this near the >>root of the tree ... > >There is an absolute minimum depth. At the early iterations Rebel might prune >on ply 1 but not on deeper iterations. Still it easily might happen being on >iteration 11 the tree is pruned at ply 3 or 4 but then it is fully justified. My old version of Rebel for the ChessMachine card shows the search depth as two numbers, separated by a slash. In the middle game, the search depths at successive iterations are typically something like 2/4, 2/5, 3/6, 3/7, 3/8 and 4/9. Does a search depth like 4/9 mean that the type of pruning you describe would not be applied during the first four plies? Tord
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.