Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: about rating in ICC.

Author: Bruce Moreland

Date: 09:22:52 06/30/99

Go up one level in this thread



On June 30, 1999 at 10:21:55, Robert Hyatt wrote:

>On June 30, 1999 at 09:46:06, blass uri wrote:
>
>>
>>On June 30, 1999 at 08:52:59, Peter McKenzie wrote:
>>
>>>On June 30, 1999 at 08:27:30, blass uri wrote:
>>>
>>>>I think that there should be competitions when everyone must agree to play
>>>>everyone.
>>>>
>>>>Rating of players should be based *only* on the results of these competitions.
>>>>
>>>>players who refuse to compete for a long time should have no rating.
>>>>players who win a competition(except the winner of the strongest competition)
>>>>can play next time in a competition of better rated players when players who
>>>>lose will play next time in a competition of worse players.
>>>
>>>I don't like this idea, and fortunately it has zero chance of happening:
>>>
>>>People pay good money to play on ICC, having your rating calculated is a valued
>>>part of the service provided.  The rules you suggest would obviously make it
>>>much less convenient for the majority of people to play rated games, therefore
>>>they would be less likely to shell out their subscription money.  Hmm, I wonder
>>>if the nice folks at ICC want to drastically decrease their income?
>>
>>It is possible to have 2 rating systems.
>>
>>The reason for my idea is that I find that the rating is misleading if people
>>can avoid playing specific opponents.
>>
>>Uri
>
>
>I don't see how this can be helped.  It has happened in human rating systems
>since Elo hatched the idea.  we have 'sandbaggers' (those who keep their rating
>low so they can win class prizes in a class lower than they should be playing
>in) and those with inflated ratings because they 'pick' their events to play
>in.  Always have had 'em... probably always will.
>
>ICC is _huge_.  There is no way to 'enforce' what you suggest, because there is
>no way to have an event with 50,000 players at one time.  So the problem will
>_always_ exist there...

What he suggests is impractical, but it would have a nice effect upon the rating
pool, I bet it would contract a lot.

bruce



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.