Author: GuyHaworth
Date: 18:54:11 03/05/04
Go up one level in this thread
A common confusion is to try to compare FIDE ELO ratings (about a community of
human players) with SSDF ELO ratings (about a community of computers).
This is understandable, as the metric used is referred to as 'ELO' in both
cases. However, 'ELO' is the measurement system rather than the measurement,
the 'scale' rather than the 'weight'.
Following on from the above, I therefore propose a new nomenclature to
distinguish the measurement system from the specific measurements, as follows:
ELO == the measurement system, as devised by Elo
FIDE_ELO == the ELO ratings derived by FIDE for the top human players
SSDF_ELO == the ELO ratings arrived at for the 'SSDF community' by SSDF
Aficionados of the ELO system will know that it is only the _difference_ in ELO
between two players that aims to represent the difference in their strengths.
Therefore, an arbitrary number can be added to all c_ELO ratings for community
'c'.
A separate exercise might be considered to see what is the best constant to add
to the c2_ELO ratings to bring them 'into line' with the c1_ELO ratings of
community c1, based on c1-c2 games cross-community.
Unfortunately, there is evidence that:
- the constant right for the top end of the c2 community would be wrong for the
bottom end: this is perhaps to do with the 'inflation effect' of the ELO
scheme.
- the probability distribution for the 'correcting constant' is not that peaky,
i.e., c2 is not fixed with great certainty by only a few games.
However, this does not detract from the proposal to quality 'ELO' as SSDF_ELO,
FIDE_ELO etc.
I feel that less confusion would certainly be the result.
g
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.