Author: Albert Silver
Date: 15:32:30 02/22/01
Go up one level in this thread
On February 22, 2001 at 15:17:42, John Merlino wrote:
>On February 22, 2001 at 05:41:25, Sune Larsson wrote:
>
>>On February 21, 2001 at 21:12:40, John Merlino wrote:
>>
>>>On February 21, 2001 at 20:42:35, Alastair Allan wrote:
>>>
>>>>Some personalities' playing levels in CM8000 have been validated against human
>>>>players with known ratings. This is a valuable feature for players in the lower
>>>>range of ability who would like to have a "dependable" estimate of where they
>>>>stand. I don't know of any other chess program that has undergone this kind of
>>>>ecological validation. I would venture to say that this unique feature is as
>>>>good a selling point as any autoplayer function that might be considered in
>>>>future releases.
>>>
>>>Thank you for saying so. However, there were only five personalities that were
>>>used in the test: Chessmaster, Josh Age 9, Willow, Sonja and Skippy -- basically
>>>personalities with ratings of approximately 2500, 2000, 1500, 1000 and 500
>>>respectively. With these five "real world" anchor points, the ratings data
>>>probably became a fair bit more accurate, but I would still venture to say that
>>>some, if not many, of the personalities are between 50 and 150 points off the
>>>mark.
>>>
>>>As for it being a selling point, I'm sure it's mentioned somewhere on the box.
>>>;-)
>>>
>>>jm
>>
>>Ok, I get it - five personalities were used in the test. Among them Chessmaster
>>itself with rating of approximately 2500 (as stated above). But where does
>>this Chessmaster rating of 2861, on my PIII 800, come from ? On the box is
>>written "The Most Powerful Chess Engine Available on a PC". Oh, maybe CM
>>increases its rating by beating me ?! ;) As stated at the back of the box:
>>**Rated 1** "The Chessmaster engine achived the No 1 ranking by the SSDF".
>>
>>Sune
>
>The reason the Chessmaster personality was chosen as the representative of the
>2500 category was that it was absolutely necessary to test the theoretically
>"best" settings for the engine. From the results of the test, and the new
>minimum spec machine (going from a P/90 to a P2/233), Chessmaster's minimum
>rating on the P2/233 turned out to be approximately 2725 (remember that these
>are USCF ratings, which are typically in the neighborhood of 100 points higher
>than FIDE).
I presume the numbers were conjured up by the marketing department. Of course it
would be easy to justify too: let's see, after 5 games, the program performed at
2450 (on the P2), which means there is an error margin of... about 200 points so
we'll say it's rated "roughly" 2625 according to tests... :-)
Albert
>
>So, with a rating of 2725 on the low-end machine, your P3/800 increases the
>strength by an additional 140 points or so.
>
>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.