Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Comments of latest SSDF list - Nine basic questions

Author: Robert Hyatt

Date: 21:38:02 06/04/02

Go up one level in this thread


On June 04, 2002 at 17:10:09, Dann Corbit wrote:

>On June 04, 2002 at 14:02:42, Rolf Tueschen wrote:
>>On June 04, 2002 at 10:59:50, Peter Fendrich wrote:
>[snip]
>>>Look at the German ratings compared to FIDE. I haven't seen them but can bet on
>>>that they are different.
>>>
>>>Peter
>>
>>Yes, much lower, about 100 points and more.
>
>This (along with USCF or BSCF and some others) is an interesting case, because
>it might be possible to create a formula which will perform a fairly accurate
>translation.
>
>That is because there are many people who are members of the national chess
>organizations and also members of FIDE.
>
>I know that there have been attempts at this.  I don't think there is a lot of
>value to the numbers, however, because the conversion functions do not give
>error bars.  Therefore (supposing for instance 100 ELO difference in the German
>case as an approximation) is it 100 ELO +/- 10 or 100 ELO +/- 150!


This is fitting a straight line thru a set of data points represented by an
nth degree polynomial function. There will be plenty of error for some, almost
no error for others.

Which is yet another way to show how the ratings between two pools are simple
impossible to compare in any accurate way...



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.