pie chart: LL attempts

[sixblueboxes]

This is a pie chart of the table here
http://www.mersenne.org/report_top_500_ll/
of LL attempts over the past year. Ordered by rank of GhzDays. Only the first 20 of ~2800 are in the legend.

So, assuming each exponent has a P=0.5 of being prime, it looks like there's still a 1/3 chance that the next discovery will go to one of the top three. Is my math right? I guess I'm also assuming that the output of each contributor is constant.

[LaurV]

Originally Posted by sixblueboxes

So, assuming each exponent has a P=0.5 of being prime

I thought they are all prime!
Regarding the chance of each exponent to generate a mersenne prime, well, that is much lower....

[axn]

Originally Posted by sixblueboxes

So, assuming each exponent has a P=0.5 of being prime

Assuming that every exponent has the same chance of being prime (which isn't actually the case, but close enough), yes, there is a better than 1/3 chance that next prime will go to one of the top 3.

EDIT:- Of course, ANONYMOUS is actually an amalgamation of all users who haven't registered an id.
EDIT2:- Hmmm... There are actually multiple ANONYMOUS there. Does that mean #3 is a single user?

[Mini-Geek]

Originally Posted by axn

Assuming that every exponent has the same chance of being prime (which isn't actually the case, but close enough), yes, there is a better than 1/3 chance that next prime will go to one of the top 3.

Didn't LaurV teach you anything? Every exponent does have the same chance of being prime: 1.

[axn]

Originally Posted by Mini-Geek

Didn't LaurV teach you anything? Every exponent does have the same chance of being prime: 1.

Of course I read LaurV's comment before posting This is my attempt at balancing the pedantry

[Aramis Wyler]

Originally Posted by sixblueboxes

So, assuming each exponent has a P=0.5 of being prime, it looks like there's still a 1/3 chance that the next discovery will go to one of the top three. Is my math right?

Your math isn't bad but the premise is flawed - Curtis Cooper already reserved the next 4 or 5 mersenne primes because he knows them in advance. :)

[Gordon]

Originally Posted by Aramis Wyler

Your math isn't bad but the premise is flawed - Curtis Cooper already reserved the next 4 or 5 mersenne primes because he knows them in advance. :)

He didn't get M2976221 in advance...

[LaurV]

@Gordon: That is false, he knew all of them but he didn't want us to see that, so he intentionally left few behind...

[R.D. Silverman]

Originally Posted by sixblueboxes

This is a pie chart of the table here
http://www.mersenne.org/report_top_500_ll/
of LL attempts over the past year. Ordered by rank of GhzDays. Only the first 20 of ~2800 are in the legend.

So, assuming each exponent has a P=0.5 of being prime,

Assuming what?

Your last statement above is complete nonsense. May I suggest that you spend some time learning some number theory and statistics?

You also need to learn how to write mathematics and to think about
what you write before you write it. For GIMPS, every
exponent that is checked is prime.