Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Who is the real NPS champion?

Author: Bruce Moreland

Date: 09:35:28 11/08/99

Go up one level in this thread


On November 08, 1999 at 11:11:35, Bas Hamstra wrote:

>And what about The King? An extremely tactical genius, while NPS is only a
>fraction of Fritz. I would definitely categorize The King as slow & smart.

My program is "fast".  I don't know how the "slow" programs work.  If they are
doing a lot of tactical work in the eval function rather than in search, I don't
know why this makes them "smart".

Example:

Let's say I take my quiescent search, call it "eval", and count every entry into
that function as one node.

The node rate of my program would probably drop by something like 50%.  This
doesn't make it any "smarter", but perhaps some would say it was?

I could make a special full-width search that does the last several plies, call
that "eval", and suddenly I'm slower than Hiarcs.

What are the "slow" programs doing?  Are they statically evaluating features, or
are they trying to resolve tactics?  If they are doing the latter, this
distinction between slow and fast may be a moot distinction, much like comparing
a two year old kid with a 48 month-old kid.  The second one is not any older.

bruce



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.