Author: John Merlino
Date: 10:50:40 03/29/99
Go up one level in this thread
On March 29, 1999 at 04:10:18, Timothy J. Frohlick wrote: >On March 29, 1999 at 02:22:48, Harald Faber wrote: > >>On March 28, 1999 at 14:29:45, Bruce Moreland wrote: >> >>>> THE SSDF RATING LIST 1999-03-28 65604 games played by 190 computers >>>> Rating + - Games Won Oppo >>>> ------ --- --- ----- --- ---- >>>> 1 Chessmaster 6000 64MB P200 MMX 2576 88 -71 100 78% 2363 >>>> 1 Fritz 5.32 64MB P200 MMX 2576 29 -28 627 66% 2457 >>>> 3 Hiarcs 7.0 64MB P200 MMX 2567 28 -27 676 63% 2471 >>>> 4 Nimzo 99 64MB P200 MMX 2565 37 -35 390 63% 2472 >>>> 5 Fritz 5.0 PB29% 67MB P200 MMX 2564 24 -23 946 67% 2436 >>>> 6 Junior 5.0 64MB P200 MMX 2543 30 -29 576 63% 2447 >>>> 7 Hiarcs 6.0 48MB P200 MMX 2524 25 -25 802 57% 2473 >>>> 8 Nimzo 98 57MB P200 MMX 2523 24 -23 899 59% 2459 >>>> 9 Rebel 9.0 46MB P200 MMX 2522 25 -24 843 62% 2437 >>>> 10 Rebel 8.0 46MB P200 MMX 2507 31 -30 533 58% 2448 >>>> 11 MChess Pro 6.0 41MB P200 MMX 2506 28 -28 637 57% 2453 >>>> 12 MChess Pro 7.1 42MB P200 MMX 2505 24 -24 834 56% 2460 >>>> 13 Shredder 2.0 57MB P200 MMX 2502 25 -25 765 55% 2464 >>>> 14 Genius 5.0 DOS 44MB P200 MMX 2495 24 -23 883 56% 2449 >>>> 15 MChess Pro 8.0 64MB P200 MMX 2481 30 -29 560 53% 2456 >>> >>>Why is the number in the last column < 2400 for CM6K and > 2400 for everything >>>else? >>>bruce >> >>Because it played the following opponents where some have a low rating: >> >> 1 Chessmaster 6000 64MB P200 MMX, 2576 >> >>Fritz532 P200 2-1 Hiarcs7 P200X 0.5-1.5 Junior5 P200X 2-0 >>MCP 8 P200MMX 1-1 Rebel 8.0 P90 9.5-7.5 Rebel 9.0 P90 1-0 >>Hiarcs 6 P90 16.5-5.5 Geniu4 486/66 8.5-2.5 Fritz 3.0 P90 1.5-0.5 >>Fritz3 486/66 19-1 SPARC 20 MHz 16-2 > > > >Dear Mr. Faber, > >Just as I suspected!!!! Sounds like a con-job to me. CM6000 IS NOT worthy of >the first place position.. Play the #$#%%# games with equal processors or don't >even waste our time. There is nothing useful here. Ninety two of the games out >of one hundred were played on vastly inferior hardware. And Mike Tyson can beat >up my little sister too. > >Thanks for the revelation > >Tim You are incorrect here, and this subject has come up before. The opponents that Chessmaster has played have well-established ratings with a specific set of hardware. The formula to determine chess rating is based on winning percentage AND the difference in chess rating between the two opponents. In other words, you are SUPPOSED to beat weaker opponents (obviously) and so you get less credit for doing so. The reason that some of these opponents are using inferior hardware is that they were tested some time ago with SSDF's standard computer at the time of the testing. Since hardware has increased in strength, new programs get to use the new hardware. But older opponents must still use the old hardware, otherwise their rating is meaningless. Chessmaster's 78% winning percentage against 2363 opponents is what gives it the rating equal to Fritz's 66% against 2457 opponents. As far as I know, the SSDF uses the same rating formula that the USCF uses. As the SSDF continues to test Chessmaster 6000, we will see the average opposition score go up, and Chessmaster's winning percentage go down. But whether or not Chessmaster's RATING goes up or down still remains to be seen and can only be guessed at. jm
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.