Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: piece mobility?

Author: Tord Romstad

Date: 09:10:38 01/09/04

Go up one level in this thread


On January 08, 2004 at 19:41:57, Steve Maughan wrote:

>Tord,
>
>>This sounds really strange.  If I understand Ed's pages correctly, he doesn't
>>do what the rest of us calls lazy eval at all.  What he describes as lazy
>>eval is what is more commonly known as futility pruning in the qsearch (i.e.
>>not searching captures.
>
>This link discusses Rebel's Lazy eval:
>
>http://members.home.nl/matador/chess840.htm#LAZY%20EVAL
>
>If I read it correctly this is definitely a lazy eval, although not the normal
>method since Revel uses the true eval from the previous ply to calculate the
>margin.  I don't think this is futility pruning which Ed discusses at a
>different section.

The section about futility pruning on Ed's pages discuss the idea of using
futility pruning in the main search, which is rather popular, but more
controversial than using it in the qsearch.  If you read the section
about futility pruning more carefully you will even see that Ed explicitly
says that it is almost the same as what he calls "lazy eval":

    The routine "Futility_One" does exactly the same as the rountine
    "Lazy_Eval" (see above). MARGIN however instead of 0.50 is set to
    a higher value, basically 3.00 (3 pawn units) but is increased to
    5.00 in the end-game. Another difference in comparison with LAZY
    EVAL is that SCORE isn't updated with MARGIN.

If I understand the pseudo code on Ed's page correctly, this is how
the main loop in his qsearch looks:

score = FullEvaluate();
margin = 0.5 + score - incremental_eval;

for each move {
   if(score + change_in_incremental_eval(move) + margin < alpha)
      continue;
   make_move(move);
   qsearch();
   unmake_move(m):
}

What is, in your opinion, wrong in the above pseudo code?

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.