Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Why this excitement over a new ssdf list

Author: Mark Young

Date: 16:00:54 09/27/99

Go up one level in this thread


On September 27, 1999 at 18:24:01, Rajen Gupta wrote:

>The new SSDF list is out and predictably we see the same old culprits sharing
>the top 5 places, a few pts from each other.Playing on a faster set of hardware
>config has resulted in a massive increase of 90 SSDf pts (i refuse to call them
>ELO pts)over the previous hardware-however a careful analysis of the results
>make a few things quite clear to the discerning-
>
>after the pathetic showing of REBEL against grandmasters-all of whom are toying
>with it-I wouldnt give any of the top progs more than 2475 ELO even if they were
>riding on the currently available ATHLON 800 supersystem.
>
>faster hardware merely accentuates the chess programmes strenghths ie tactical
>ability (which is anyway miles ahead of most humans)against its fellow chess
>programmes and thereby gives a false impression of rating increase.
>
>THE ACHILLES HEEL of every prog is total lack of strategy and positional
>understanding- which can be cruelly exploited by most computer savvy GMs, no
>matter how strong the hardware available-while at the same time being careful
>not to tangle with comps in tactical slugfests.
>
>Most IMs unfortunately do not have this capability- either to fully exploit the
>computers weaknesses nor to carefully avoid falling into a tactical hole-which
>results in impressive performances by comps against IM but abject failure
>against GMs
>
>MOST IMPORTANT a careful analysis will reveal absolutely no improvement in pure
>software engine terms over the last 3 years-I'm sure Genius 3, Rebel 8,CM 5555,
>hiarcs 6 will play as well as Frits 5.32, hiarcs 7, junior5 and nimzo99 if you
>start em off from a variety of fixed positions with both sides to play white and
>black and no opening books-perhaps someone should try this experiment-

This has been done, and you can argue how much improvement in pure software
strength there has been over the last three year, but if you argument is there
has been none, you will lose.

You post the opening positions that is fair to both white and black, and I will
run the games. Lets find out.

>
>rajen gupta



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.