jeremy.p.spinrad@[EMAIL PROTECTED]
wrote:
>This was a pure and simple case of ratings fraud.
>
>It is easy to say that the ratings committee should catch this, but
>very hard in practice. There are too many pieces of data coming in to
>check whether every one is legitimiate. The rating committee's chief
>job is to set up a good system, not to police it.
Is that specified in the ratings committee job description?
The rating committee should make a reasonable effort to police
the system against cheating. If the only job of the rating
committee is to set up a good system, the committee should have
been disbanded once the system was set up.
>Every rating system can be manipulated by a cheater placed in
>a position of trust.
That's not true. Systems exist where no one person can cheat
the system; any CPA can set up such a system for you.
>I do not hold the rating system or the rating committee at
>fault, just as I would not hold the national basketball
>league at fault if a referee is found to be in the pay of
>gamblers.
In both cases, whether the organization is at fault hinges
on the nature of the cheating. If it was subtle cheating that
nobody could have caught, the organization is not at fault.
If it was blatent cheating that anybody should have been able
to catch, or if they were tipped off and ignored the tip, then
the organization is at fault This does not, of course, reduce
the guilt of the atcual cheater.
Someone playing the same high-rated player over and over -- a
high rated player who never plays anyone else -- is trivially
easy to spot. Any s****ts federation that doesn't commision
someone to write software that catches such blatant cheating
is simply not doing their job.
--
Guy Macon
<http://www.guymacon.com/>
|
