Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Doctor? 3.0

Author: Steve Lopez

Date: 07:21:56 03/28/99

Go up one level in this thread


On March 28, 1999 at 06:16:53, Phil Dixon wrote:

>How, and why, does it get so deep (plies), so fast?  What would you estimate the
>rating to be?

Because it's not showing plies -- it's showing 1/2 or 1/3 plies (I forget which
offhand). It consequently *looks* faster than it really is.

I can't give rating estimates -- these are extremely hardware-dependant (not to
mention other factors, like hash table size, switching off "permanent brain",
etc.). I get phone calls nearly daily from people asking questions like "How
many Elo points stronger will [Fritz, Junior, Nimzo, etc.] be if I add 32 Mb RAM
to my computer?" Good question. Beats me. It'll be stronger, but who knows how
much? I'm OK at what I do, but I'm not *that* good. ;-)

Plus there's the human (psychological) element. Different chess playing programs
play in different styles and sometimes a program will play in a style that's
just plain hard to figure out. My problems with getting my butt kicked by
Doctor? 2.0 are pretty well-documented in Electronic T-Notes on the ChessBase
USA web site. There was just something about that $%^$# program that caused me
endless grief. Even though Doctor? 3.0 is arguably stronger (from the results of
my Engine tournaments), I don't have nearly as tough a time against it as I did
against v.2.0. It's an experience many chessplayers have shared -- Kasparov had
a minus score against Boris Gulko for many years and I used to get my clock
cleaned by a kid at chess club who was rated a couple of hundred points lower
than me. Go figure...



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.