Xref: utzoo comp.sys.ibm.pc.misc:2578 rec.games.chess:5304 Path: utzoo!attcan!uunet!snorkelwacker!ai-lab!life.ai.mit.edu!cracraft From: cracraft@mole.ai.mit.edu (Stuart Cracraft) Newsgroups: comp.sys.ibm.pc.misc,rec.games.chess Subject: Re: Chess program recommendation wanted (for IBM) Message-ID: Date: 12 Oct 90 15:05:36 GMT References: <1399@pedsga.UUCP> <6302@hub.ucsb.edu> <85.26ff813b@inland.com> <1990Sep27.121239.5258@cs.rochester.edu> <4252@disk.UUCP> Sender: news@ai.mit.edu Organization: Project GNU Lines: 60 In-reply-to: jcsewell@disk.UUCP's message of 9 Oct 90 02:49:13 GMT In article <4252@disk.UUCP> jcsewell@disk.UUCP (jcsewell) writes: In article <12679@accuvax.nwu.edu> kaufman@delta.eecs.nwu.edu (Michael L. Kaufman) writes: >pals@inland.com wrote: (Chessmaster 2100 is a) >>misleading name. On a PC XT the program runs about 1900 USCF rating points, > >I may be wrong, but it seems to me that I read on the package that 2100 was its >strength on a 20 MHtz 386 computer. > The box I have brags that they took Chessmaster 2000, added 10 man years of work... that Fidelity Electronics is the "developers of the world's Only USCF Certified Master Rated (2325) Chess Program!!!" ... Lots of doubletalk and impress-me phrases, but nothing solid about the end product itself. -- J.C. Sewell DISK: Digital Information Systems of Kentucky uunet!disk!jcsewell (502) 968-5401 thru 968-5406 "Abortion is killing babies; it's killing lives." - Michael Peace Matthew 11:12 Christian Rapper International Master Larry Kaufman estimates it would take a 33mhz 386 with zero wait state to run on before Chessmaster 2100 would play at the 2100 level. On an XT, he says its something like 1750! The story of the Chessmaster 2100 box rating of 2325 is complicated. Originally Fidelity, in the form of programmers Dan and Kathe Spracklen, had a good program called the Mach 3 which achieved a 2265 rating on a 68000 and a 2325 rating on a 68020. Software Toolworks, the authors of the Chessmaster 2100 program, bought out, purchased, or got rights to the 2325 program for purposes of running it on a different (Intel) architecture. The port was done, but just about everyone I know, and Larry all feel the Chessmaster 2100 program is no where near as strong as the Mach 3 on a 680020 (called the Mach 4). I'll take a moment to plug a program called "M Chess". It is an IBM (Intel) based program. But its forebears (earlier versions of the same code) managed to defeat IM John Grefe in across the board tournament time control play, and achieve a 2500+ performance rating over a short series of games at the U.S. Amateur Team last year (all of the above while running on a 386 chip under DOS.) Recently M Chess played the programs Rex (version 2.3) by Kaufman and Zarkov (version 2.05) by Stanback. These two programs ran on 12mhz 386's while M Chess was purposely handicapped to a 4.77mhz XT. The score was 4w-2d-0l and 4w-0d-0l in M Chess's favor. Also, M Chess played and defeated Mephisto Polgar which ran on high speed bit-sliced configuration. The margin was about 3-1. M was running on a 25mhz 486 so both programs were running on fast hardware. And a forebear of M defeated Mephisto Portorose 9.5-6.5 in tests by Larry Kaufman. Against Fidelity Mach 3 at 40/2, the score is 6-0 M's favor. Against Mephisto Mondial, the score is 3-1, M's favor, etc. Anyway if you want information about M, contact The Computer Chess Gazette at 714-770-8532... Stuart