🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: gnu.chess
Article #5665

The BT 2450 Test

#5665
From: Andrew Levine
Date: Thu, 18 Feb 1993 08:26
37 lines
2713 bytes
Hello there again,
Here is some more information on the test.
First of all, it is called Bednorz-ToeNissen Test (my mistake). It was
published in the german magazine 'Computer Schach und Spiele', issue 5/91,
pp. 38-41. It approximates a rating in german ELO-points (I think about
100 more than the US).
For UNIX the positions and 'correct moves' are contained in 'gnutest.tar.Z'
at 'valkyries.andrew.cmu.edu'. For DOS the positions, the 'correct moves',
two batchfiles for autotesting and 3 sample files are contained in
'gnutest.exe' at the valkyries. The files are to be found in either
/pub/chess/Unix or /pub/chess/DOS on that archive.
Here is a table comparing some BT-ratings with the swedish ELO-list:

Hardware                   BT     SWED +-    (swed. speed) - CPU     (hash)
---------------------------------------------------------------------------
Fritz2                     2259   2160 +-115 (486-33)         386-40 (2 MB)
Mephisto Vancouver 32      2247   2249 +-48  (36 Mhz)         030
MChess Pro 3.13            2248   2283 +-82                   486-33
Saitek Kasparov RISC 2500  2242   2246 +-56                  ARM2-14 (128K)
Chess Genius 1.0           2209   2296 +-133 (486-33)         386-40
Chess Machine The King 512 2206   2210 +-48                  ARM2-16
Mephisto Lyon 32           2193   2259 +-59  (36 Mhz)         030
Chess Machine 512 Vanc.    2157   2222 +-36                  ARM2-16
Fidelity Elite V.9         2132   2127 +-44  (32 Mhz)         030
Mephisto Vancouver 32      2109   2157 +-32                   020-12
MChess 1.41                2100   2299 +-43  [Ver. 1.1-1.71]  486-33
...
GNU Chess 4.0 pl.60 (DJPG) 1745                               486-50

It would be interesting to see the ratings of GNU Chess 4.0 on some more
powerful platforms than the ones I tested (386-16, 486-33/50). So - test
and report. If you email me the result-files containing GNU Chess'
analysis (gnuchesr's thinking) on the positions it solved correctly I
will summarise to the net.

Andrew Levine
(levine@ldv01.Uni-Trier.de)
Technical Details
Message-ID: <levine.730023992@utrurt>
Path: arc1-ca.pugleaf.net!sparky!uunet!math.fu-berlin.de!mailgzrz.TU-Berlin.DE!news.netmbx.de!Germany.EU.net!rzultr.uni-trier.de!utrurt!levine

Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 10d71d,15a01cc6f9c6103,start
X-Google-Attributes: gid10d71d,public
X-Google-ArrivalTime: 1993-02-18 00:26:32 PST
Newsgroups: gnu.chess
Path: arc1-ca.pugleaf.net!sparky!uunet!math.fu-berlin.de!mailgzrz.TU-Berlin.DE!news.netmbx.de!Germany.EU.net!rzultr.uni-trier.de!utrurt!levine
From: levine@utrurt.uni-trier.de (Andrew Levine)
Subject: The BT 2450 Test
Message-ID: <levine.730023992@utrurt>
Summary: Some info
Keywords: chess, gnuchess
Lines: 37
Sender: news@rzultr.uni-trier.de (USENET News System)
Organization: University of Trier
Date: Thu, 18 Feb 1993 08:26:32 GMT