Author: Stuart Cracraft
Date: 15:56:28 01/05/06
Go up one level in this thread
On January 04, 2006 at 21:07:56, Vincent Diepeveen wrote: >On January 04, 2006 at 14:52:06, Stuart Cracraft wrote: > > >>http://www.robinupton.com/research/phd/cp_intro.html >> >>So, has anyone actually used these CNS and CNS-derived ideas in >>a chess program? I know about B* and Berliner. I want to know >>about things more recent that at least *someone* can speak to >>in the commercial or semi-commercial world. > >Yes i put in months of my time to get CNS to work in diep and >experimented extended with it. > >It doesn't work. > >>If you don't have actual data, how about theories on how these >>would be used? > >>It seems that determining the least number of nodes that would >>have to change to affect the root score or growing the tree >>in such a way that the conspiracy is as large or as small as >>possible, could yield some interesting deeper searches. >> >>What is actual in regards to the above for computer chess? > >You can do nullmove in CNS too, that's what i added to it. > >It speeds it up considerable. > >Certain combinations which are just like 11 ply or so, CNS never finds. > >CNS has a worst case and that is that it never searches search space which your >evaluation function doesn't understand. > >So conspiracy number search never corrects your evaluation function. >That's real ugly. > >It spends it majority of time to complete nonsense. Like 1.e4,h5 2.Qxh5 Rxh5 >Rxh5 is forced more or less and CNS will deeply extend that line. > >In depth limited PVS you nullmove, get a single cutoff and you can bury this >nonsense line. 100x more efficient. > >Vincent > >>Thanks, >> >>Stuart I now assume that Rybka does not use CNS nor CNP in any sense. That is has a new paradigm that involves some other form of probabilistic evaluation perhaps layered on top of a semi-traditional style of search rather than a radical type of search. Stuart
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.