On Sun, 12 Jun 2005 21:26:29 GMT, "BrianW" <brian@nospam.net> wrote:

Nick wrote:

Hello.
I've been using SETI classic for about 5 years and have just started
to use BOINC/SETI on a couple of PCs, before seti classic goes the
way of all things.
I'm confused about how the BOINC credit system works. Is there a web
page that explains this?
I have looked. Without success, but I am a newb.

Thank you

Nick

SETI@home member since 30 Oct 1999
Total credit 39,235.68
Recent average credit 637.95
SETI@home classic workunits
as of 15 March 2005 22,865
SETI@home classic CPU time
as of 15 March 2005 189,199 hours

They send each WU to about 5 people. They then award the average credit 
claimed by the first three to return a valid result - provided that they are 
returned within the deadline. Other later results get the already awarded 
average - again provided they are on time.
HTH
Brian

Brian is right...here is the expanded version.....each unit is sent to
4 different computers, the validator then keeps a check on the units
as they return, after 3 of the 4 computers send the unit back a notice
is sent to the validator computer. If it has time it will check and
see if the 3 results are within statistical limits, no two computers
ask for the same amount of credits, and if they are within those
limits the middle of the 3 credits requested is what is granted to all
computers that crunch that same unit. Yes even the 4th computer that
returns its data after the granting of credit.
IF however the validator computer does not get to the unit before the
4th computer returns its data, then the computer again checks that the
results are within statistical limits and if all 4 are, then the top
and bottom are thrown out and the middle 2 are averaged, (b+c)/2.
IF when the validator computer checks after 3 returns are in and they
are not within the limits then it will automatically wait for the 4th
return and then if it is within the limits it will throw out the
highest and lowest of the good ones and grant credit to all. If the
4th is also outside the limits then the unit is sent out to more
computers until results are returned that are within the limits.
IF the units are not returned within 14 days then the validator
computer tell another computer that then sends the unit out again to
more computers until at least 3 results are returned and credit is
granted.
The statistical limits is a moving number that is determined by
Berkeley and is the amount of difference allowed between requested
amount of credits. If computer a asks for 23.34 credits and computer
asks for 23.78 credits and computer c asks for 26.54 credits, they are
within the limits. If however computer a asks for 56.89 credits and
computer b and c above stay the same, computer a is outside the
limits. To expand a little more, if computer d comes in with 53.90
then both b and c are setting there own set of limits and computers a
and d have their own. The unit would be sent to mroe computers and
then the results would be calculated based on what it returns.

Paul ahs a GREAT site with a ton of info about Boinc:
http://boinc-doc.net/index.php
it even has a search feature, and well over 100 pages on Boinc.