Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Branching factor, etc

Author: Tony Werten

Date: 04:47:59 09/01/02

Go up one level in this thread


On September 01, 2002 at 06:29:16, Bas Hamstra wrote:

>>Now suppose FP is right but we don't prune. What happens ? We go to quiescence;
>>evaluate; get a score above beta; and cutoff. So no nodes saved.
>>
>>FP can only save nodes when we would otherwise spend time in quiescence; ie
>>score is not above beta ie fp was wrong.
>>
>>People who report big wins with fp probably have an "inc(nodecounter)" in the
>>top of quiescence, wich doesn't get called when pruning.
>
>Couldn't agree more. Those nodecounts are very misleading, what counts is depth.
>And once you measure that, it suddenly isn't so great anymore. Heinz bases his
>research on nodecounts. But the nodes saved are "cheap" nodes, that are
>_accurately_ cut by lazy eval with little cost anyway. FP is good for picking op
>wood in blitz, not much more.
>
>But regarding SEE, not doing it MUST mean a fairly higher qrate, no? It slows
>you down and gets somewhat better in tactics. Only I hardly see that tactical
>advantage, really. Of course it might depend on your qsearch. Mine is typically
>2-3 x bigger than a Crafty like miniqsearch, not doing SEE would make it even
>worse.

It's a choice. From games XiniX played I always had the feeling it should
improve in tactics. The positional evaluation was quite exact but it happened
too often: I have a good position, I have a good position, I have a good
position, oops that costs a knight.

Now my first few ply in Qsearch are very simular to a normal search.

Tony

>
>Best regards,
>Bas.



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.