Home > Archive > Chess politics > April 2005 > Re: Ratings experts: Please explain





You are viewing an archived Text-only version of the thread. To view this thread in it's original format and/or if you want to reply to this thread please [click here]

Author Re: Ratings experts: Please explain
David Kane

2005-04-11, 5:59 pm


"George John" <george@neosoft.com> wrote in message
news:1112635111.039577.255590@o13g2000cwo.googlegroups.com...
quote:

> Overall, I agree with this. One minor nit pick is the score (BTW, I
> prefer score to winning) expectancy curve is never linear.


All curves are linear if you expand the scale enough. I think
in usual class tournaments (+/- 100) there wouldn't be a significant
error using a linear approximation.
quote:

> WRT performance ratings, I should also mention the online calculator
> gets into minor trouble when there is a perfect score since the
> expected rating is always less than one. For example, if a player

wins
quote:

> 4 games against all 1000 rated players, what is the performance

rating?
quote:

> It really could be anything, say 1400 or higher. The online
> calculator comes up with 1400, since 400 is commonly accepted, but
> there is no truly satisfactory way to do this, SFAIK.

Interestingly,
quote:

> sometimes the performance rating at 1/2 point less than perfect is
> higher than using the 400 number! For example, try 8 opponents with
> ratings of 2299. With a score of 8 the PR is 2699; 7.5, 2769.
>


I have a half-baked speculation that a more meaningful
performance rating in the case of a perfect score would
be to replace the actual result with something very close
to it (e.g. assume that a score of 8 in your above example
is really a 7.75 rounded up)


Copyright 2003 - 2009 gamesreviews.net Software forum  PC Hardware reviews