Computer Chess Club Archives


Search

Terms

Messages

Subject: One more post on Ruffian 2.0

Author: Djordje Vidanovic

Date: 03:21:26 12/19/03


As I had been away from my computer for a while, I only read the long threads
regarding Ruffian 2.0 last night.  I feel obliged to give some new information
regarding the thread and the advertisement in which Ruffian 2.0 is labelled as
the strongest Winboard engine available.

First of all, this very laudable description of Ruffian 2.0 was not produced by
Per-Ola.  Per-Ola actually wrote in his text that Ruffian is "a strong Winboard
and UCI engine", nothing more.  Thus, I agree with many of you who thought that
the wording was too pompous (to say the least).

However, there are a couple of mitigating circumstances.  At the time the
wording was being assembled Rebel 12 had not yet appeared as a WB engine, and
The King (a.k.a. Chessmaster) was used as a WB engine in highly questionable
ways (without its proprietary endgame bases, without its proprietary book;  but,
rather unbecomingly, like some hacked-up version of an otherwise magnificent
program (this is not laying it on thick, I really mean it).

As for the third program that was also mentioned as a strong WB engine, Deep
Sjeng 1.5, I have my own testing evidence, which I will be bold enough to
present to you (I will not supply the games as I have pitted Ruffian and DS
purely for my investigative purposes;  to find out if there are any
vulnerabilities in the Noomen book).  My evidence comprises more than 1,000
games played at G/5, G/10, G/15, G/30, G/45, played on a dual AMD MP system (MP
2000+, 1 GB RAM, Windows XP Pro, hash 128 MB each, 3-4-5 tablebases included).

So far the score of this very long match has clearly favoured Ruffian 2.0
(firmly standing between 61-62%, indicating a rating margin of about 80+ ELO
between the two).  The same dual system hosted DS against a Barton 2800+ (single
cpu system) and Ruffian was winning again (this time only 300 games played at
G/10). The dual version of Deep Sjeng 1.5 was getting 46% against the single cpu
Ruffian 2.0.  Thus, we might say that Ruffian was not only keeping DS at bay, it
was slowly and surely breaking down DS as the match went on.

These are facts and I stand behind them.  I do have all the pgns and I was the
one to have supervised the matches, or most of them, at the Faculty of
Electronics, University of Nis, Serbia.  My friend and associate, Vladan
Vuckovic (Axon programmer) was involved in the testing too.

Again, I am sorry that the ad came out the way it did, and I am also rather sad
at the fact that I was not around to try and reply to some well-founded
criticism directed at Frank Quisinsky.  I think that Frank was well-intentioned
but also that he has to learn a lot about the ways of the world, especially the
world of business.

One more thing: it would be very easy to fend off any suer who tried to complain
about Ruffian and its prospective results.  Just on the basis of the phrase used
"... strongest Winboard engine AVAILABLE", which was true at the time Ruffian's
ad came out:  no Rebel yet, DS clearly the less potent of the two, The King not
a properly distributed WB engine.  Still, let me say it again:  I believe that
many complaints made here make a lot of sense and I also believe that the text
in the ad must have been less conspicuous and less assertive.

Best regards to all,

Djordje



This page took 0.01 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.