From: "Stouten, Pieter [R&D/0467]" <pieter.stouten <at> PHARMACIA <dot> COM>
Date: 12 dec 2002
Subject: Re: Elo ratings and grades

--- Hugo Hollanders <H.Hollanders <at> MERIT <dot> UNIMAAS <dot> NL> wrote:

> It seems odd that recognizing grades "earned" in Japan is now being
> questioned by FESA. In the past these grades were always recognized
> in Europe (at least I think they were).
>
I don't think so. Eric applied his own judgement in that he accepted the NSR
grade of a winner/runner-up of a major tournament, but not the grades of
players that were eliminated earlier on (despite the fact they got official
diplomas, e.g., Asle at the first ISF). He also indicated that the grades of
players obtained on the internet should not be accepted, despite the fact
that these players got an official NSR diploma. I would tend to agree with
Eric that these are the right things to do, but it does mean we have NOT
automatically accepted NSR grades in the past. Eric's and Asle's policies
were and are unwritten, which led to the current confusion and discussion.
So, I would favour establishing clarity by writing up the rules, but of
course FESA needs to agree first what these rules should be. Whether odd or
not, I hope you agree this is the right thing to do.

> Furthermore, there is NO link between a player's current elo-rating
> and his playing grade.
> [...]
> If one thinks otherwise, a close look at the current elo-list proves
> that there is no link between dan/kyu-grade and elo-rating. And this
> can also not be achieved, unless de-promotions would be introduced
>
Looking at the Elo list, you will notice significant correlation between
ratings and grades. In fact, I hope that Asle can determine the statistical
correlation when he gets back from vacation.

> Dan- and kyu-grades reflect a past "best" performance,
> and as such these will always on average overestimate a player's
> strength. To put it more simply, a dan- or kyu-grade is more of
> an honorary nature, [...]
>
Indeed more simply put, but also inaccurately. The Elo-based grades are in
no way honorary. They are tough to get titles that are exclusively based on
actual sustained (!) peak performance. Elo ratings predominantly reflect
current strength; grades reflect peak performance. I think there is a place
for both, and I think we should take both seriously. Therefore, FESA and
USSF need to decide how they want to do that. And that may very well include
accepting NSR grades.

To summarize all my mails about this issue: I have no fixed opinion about
the contents of a policy that governs accepting non-Elo-based grades, but I
strongly believe FESA needs to discuss this matter and prepare such a
policy.

Ciao,

Pieter