Author: John Merlino
Date: 13:24:44 10/06/00
Go up one level in this thread
On October 06, 2000 at 15:28:45, Jonathan Lee wrote: >On October 06, 2000 at 00:33:15, John Merlino wrote: > >>On October 05, 2000 at 20:51:44, Jonathan Lee wrote: >> >>>On October 05, 2000 at 15:59:26, John Merlino wrote: >>> >>>>On October 05, 2000 at 02:12:58, Timothy J. Frohlick wrote: >>>> >>>>>On October 04, 2000 at 23:55:37, Tim OLena wrote: >>>>> >>>>>>Sorry if this is an old topic (I don't get on here very often). >>>>>> >>>>>>Chessmaster 6000 tells me it thinks it's rating is 2666. >>>>>> >>>>>>How does it arrive at this estimate? >>>>>> >>>>>>Thanks! >>>>>> >>>>>>-TO' >>>>> >>>>>Tim, >>>>> >>>>>It is just a best guess but not very accurate. All of the programs do a little >>>>>computation based only on the speed of your processor. I would not give much >>>>>credence to anything over 2550 though. John Merlino could probably tell you >>>>>more. It is really just a marketing gimmick. I own CM6000 and it is a fine >>>>>program but it is no grandmaster. Too many holes in the knowledge. >>>>> >>>>> >>>>>Tim Frohlick >>>> >>>>As Tim said, it is a very simple calculation based on SSDF's rating of >>>>Chessmaster 6000 on a Pentium 90 (which was 2473). An additional 70 points >>>>(approximately) is added for each doubling of processor speed. I do not know if >>>>this last part is particularly accurate, but that's how it arrives at its >>>>rating. >>>> >>>>jm >>>To John Merlino, >>>On the August 2000 SSDF list by icdchess, your CM6K should be 2600+ SSDF if on >>>450 MHZ K6-2 by way of doubling. >>>Oh by the way, I heard of pirating of CM6K (burning the compact disc), but I >>>bought it ethically; it was only 6 dollars at CompUSA. >>>Jonathan (77th message) >> >>I think you're confused about something (or, perhaps I am, which is certainly >>possible). But, first, I must make a clarification: CM6K was tested by the SSDF >>on a Pentium 200, rather than a Pentium 90. >> >>The SSDF does not change the processor that an engine is tested on. So, CM6K's >>rating will ALWAYS be listed as being tested on a P-200. Since the SSDF will >>never test CM6K with its next set of hardware (the K6-2 450 that you speak of), >>we have to make some estimation of what CM6K would be on today's hardware. This >>is where the above calculation (70 points for every doubling of speed) comes in. >> >>Please correct me if this is not what you were talking about, >> >>jm >Thanks for the correction. > >I'll make an assumption that CM7K (or CM8K) could attain 2700 FIDE like >Junior 6 at Dortmund (5.6 GHZ and tournament time control) and >Fritz 6 at Frankfurt (4 GHZ and game in 25 minutes). > >I don't know what the doubling of speed is like for FIDE ratings compared to >SSDF ratings, but if FIDE is also 70 points then 10 GHZ could almost equal the >rating of the 2nd best player in the world which is Kramnik on tournament >control. > >Hardware tends to double in speed every 18 months. >Also, parallel processors with 1 GHZ multiply by 8 could make 8 GHZ just like >Dortmund (700 MHZ x 8) and Frankfurt (500 MHZ x 8). >Jonathan (78th message) Well, CM7K uses the same engine as CM6K, which is why CM7K's engine has never been (and will never be) tested by the SSDF. Nor should it be. As for CM8000, only time will tell.... 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.