Author: Dan Homan
Date: 16:24:16 09/15/98
Go up one level in this thread
On September 15, 1998 at 18:51:00, John Coffey wrote:
>Let say hypothetically that I have a program that is doing a simple
>3 ply search, but will extend its look ahead for every capture. The question
>becomes, how far should the look ahead be extended? The last time I wrote
>a chess program (1987) I found that the program would examine frivilous
>captures out to infinity if I didn't put a limit on it. (I literally had
>20 captures in a row.)
>
>So say I limit the extensions to 2*N, where if N is 3 then 2*N equals 6.
>Seems to me that this might not yield an accurate result and could still give
>lots of frivilous captures.
>
>John Coffey
Extending every capture is way too much. I used to do this in my program
and it would really make the tree explode. You should probably only extend
when you have a good reason. Here are some extensions used in 'typical'
programs
re-capture: only extend captures that re-capture material - these are often
forced and a good bet for an extension
check: extend when you place the other guy in check
pawn-push: extend in the end-game when you push a pawn
And there are others.... One idea to limit extensions is to use fractional
extensions and only add a ply when the fractional extensions have added
up beyond some threshold.
- Dan
P.S. I am currently working on an extension in my program which causes
the program to extend on 'aggressive' moves toward the king if the king
appears suitably 'vulnerable'. The idea was to detect some king-side attacks
that my program was falling for... It seems to work, but my definitions for
'aggressive' and 'vulnerable' are too broad and I find that the program is
extending too much in situations where the king is not really in any danger.
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.