Computer Chess Club Archives


Search

Terms

Messages

Subject: Internal Iterative Deepening

Author: David Rasmussen

Date: 03:17:30 01/24/02


IID is activated in my program when I am on the PV, and I have no hash move and
other sensible conditions are met. I have found that in many of these cases,
there is actually a hash entry, but it is an upper bound, that is, a fail low.
So there is no best move. In this case, it doesn't help to do IID, because when
I do the recursive call to Search() with depth - 2, I of course get a hash hit
immediately, returning this upper bound, so no more best move information
becomes available. This makes IID a waste of time in these cases. Is there a way
to deal with this, so that IID works, or would it be smart to check as a
condition for IID that there isn't an upper bound in hash that would make the
recursive call return immediately?

/David



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.