----- Original Message -----
From: "Torben Schl�ntz" <[EMAIL PROTECTED]>
To: "Brian J. Beesley" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>
Sent: Monday, November 25, 2002 10:04 PM
Subject: SV: SV: SV: Mersenne: Drifting UP(!) in Top Producers ranking?

> I'd rather not like the "penalty"/ punishment. A reward equal to
> the full effort of doing the TF would be much better - and under those
> circumstances no one would try to cheat because a factor found at eg. 63
> bits would reward very well.

That would allow another cheat.  Current Factoring assignments are
prefactored to 2^57, and are intended to be factored to 2^67.  If someone
were instead just to factor to 2^58, they would have about a 1/58 chance of
getting a full credit for less than 1/500th of the effort.  If not, then the
exponent could be abandoned.  This would also have the advantage (from the
cheat's POV) of 'poisoning' potential competitors' factoring efforts.

IMO people should expect (in the mathematical sense of the word) to get the
same amount of credit irrespective of what type of work they do.  Also
credit should be given for work (honestly) done irrespective of whether the
search was successful.  The first criterion (only) could be met by crediting
only found factors, and giving a higher credit for larger ones, up to the TF
limit.  I do not think there is any way to allocate credit that meets both
criteria, which wouldn't reward cheating in some way.

Brian's suggestion is a good one, but I would add that perhaps each user
could get an allowance, proportionate to the number of TF assignments
returned, that would be deemed to be 'honest' errors, and not penalised.

P-1 (which I do almost exclusively) seems to be woefully ill-rewarded.

> tsc

Daran G.


_________________________________________________________________________
Unsubscribe & list info -- http://www.ndatech.com/mersenne/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

Reply via email to