Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Search Depth

Author: Amir Ban

Date: 22:45:06 08/02/98

Go up one level in this thread


On August 03, 1998 at 00:55:03, Steven Juchnowski wrote:

>When playing programs such as Fritz5, one frequently sees depth search values
>such as "12/32". Is it possible that any program on a micro (or any computer for
>that matter) can see 32 ply ahead, particularly in a complicated middle game?
>
>Or is the program conducting a selective search that runs 32 ply deep?
>

"32" is the deepest that the program looks into any line. It will extend some
lines according to its extension rules, but not more than 32 ply. It will
actually reach 32 ply on very few branches, and perhaps not at all.

Many programs print search depth in this form of lower-bound/upper-bound. I
don't know why they do that. The upper bound is of no practical importance.


>Also is the program conducting a nominal 12 ply search, or in this instance a
>minimum search depth of 12 ply.
>
>Regards

In the Fritz case it is 12-ply minimum with null move pruning. Null move means
that the program in most lines doesn't reach even the minimum 12, but by
convention null-move programs show their nominal depth as if it was full-width.

Amir



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.