Computer Chess Club Archives


Search

Terms

Messages

Subject: Ratio of extension

Author: Jan Pernicka

Date: 04:56:19 02/01/00


   Hi,
 I would like to know what's the ratio (approximately) n/m, where
 n = #nodes visited by exhaustive search (i.e. they were not chosen
        by selective algorhitms)
 m = #nodes visited by selective search(i.e. others)
  (ok - 1 node could be in both groups - the search depth to the horizon
  could decrease by "real number")

 Nodes visited by quiescence search I include into the 2nd group (of course).
 How is it with your (or your favourite) chess program. I would expect
 it to be 3:1 but it's only a guess ( on the contrary - for humans:
   lim(x->GM) n(x)/m(x) = 0 (+ epsilon) (:-)  )

 And the other thing - what kinds of selective search are in common?
 (quiescence,check...). It's connected with question - is it better
  to explore "this interesting move" deeply or to check them all?
  - to find deep way to win or to find amazing move - on the first sight
  unrealistic?
           Thank you,  Jan Pernicka



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.