Author: David Rasmussen
Date: 14:29:52 11/23/00
Go up one level in this thread
On November 21, 2000 at 21:33:14, Robert Hyatt wrote: >On November 21, 2000 at 21:19:13, Michael Neish wrote: > >> >>Hi, >> >>Just a question about Quiesce(). I find that my program searches about 80% of >>its nodes in Quiesce() -- sometimes over 90% -- and only 20% in Search(). Is >>this normal? I'm doing checks and quick return if the material score is about a >>Pawn below alpha. >> >>Thanks. >> >>Mike. > > >Here is the thing you are probably counting wrong if you want to match the >numbers some give here... when depth=0, you call quiesce(), and _that_ node >is _not_ a q-search node, because there is _no_ choice involved in whether >to reach that node or not. Nodes _below_ that are 'optional' depending on >your capture rules... I don't count like that as it is more difficult, but >if you wanted to have a Search(), Leaf() and Quiesce(), then you could >count things and discover that it is possible to really get pretty low on >nodes searched by Quiesce() above... I've never counted this way before, but I see what you mean. Its basically a question of what you call quiescence search nodes. With that way of counting I get 20-40% instead of 70-80% Now Im happy :)
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.