Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Statistical methods and their consequences - for Albert (Rolf's Way)

Author: Albert Silver

Date: 06:04:27 02/17/03

Go up one level in this thread


>Have you ever known Rolf NOT to have a problem with the way something is? :)
>
>Seriously, his stated problem is the (over) emphasis on marketing, ("we're
>number 1 on SSDF!), which he feels is used to persuade people to buy new chess
>programs, unnecessarily.

Understandably people want some kind of reference and ratings of any kind are
usually what is given. 10-15 years ago, it was done via games played against
players, but then again, there were far fewer machines, and they were quite
beatable. Nowadays things are different so the comp-comp rating lists of the
SSDF have become the norm. It is not at all surprising that they have become the
benchmark nor that the results are used for marketing purposes, but that has
nothing to do with the SSDF in itself. There may have been some controversy in
the past such as the memory rich machines to accomodate Fritz 5's inclusion,
however the machines used aren't unusual by any common standards and do give a
good feel for what to expect on said hardware. I see no reason to question their
results or methodology. Things such as the error margin can be read on every
single line, and it isn't hidden like fine print in some obscure paragraph at
the end. If one feels Chessbase's use of the SSDF to promote a program distorts
the truth about it, don't blame the SSDF, blame Chessbase. As to it's
non-inclusion of Fritz 8 so far, yes, it is very likely a commercial move, but
if someone were to buy 6-8 copies of Frtiz 8 and send it to them, I believe they
would start testing it. If they refused, THEN one would be quite correct to
question the SSDF's independence.

                                           Albert

>
>I see what he's talking about - it does somewhat distort the perception of
>strength in a new chess program, when in fact, said program may NOT be the
>strongest program after further testing. I think the average consumer is not
>going to take the margin of error into account, or understand how the ratings
>may change for a new program on SSDF.
>
>I'm sure it's no problem for folks on CCC, especially.
>
>Dave





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.