Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Does SSDF list show diminishing returns? Some calculations

Author: Dann Corbit

Date: 22:48:46 10/08/01

Go up one level in this thread


On October 08, 2001 at 05:48:55, Daniel Clausen wrote:

>Hi
>
>On October 08, 2001 at 03:13:22, Jouni Uski wrote:
>[snip]
>
>>In average 3,70. This gives (log2/log3,7)*84 = 45 points for speed doubling.
>>Previously it has been spoken about 60-70 points for doubling, so there is
>>diminishing returns now!
>>
>>Jouni
>
>That or either the previous or the new claim is based on shaky data. ;)

That all depends.  What is the size of the error bar for those figures?

If it is +/- 100 ELO, we should expect quite a broad range of possible outcomes.
If it is +/- 2 ELO, then something may be amiss.
I expect that the error bar more closely resembles the former than the latter.

I am very sure that the very best ELO figures available are not more accurate
than +/-30 ELO or so (based on the best SSDF data) and therefore, the
extrapolated gains based on a few points with these data will be on even more
shaky ground.




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.