Stats suggestion: overall best user/team

[Troodon]

Since there are three "sub-projects" behind SoB (PRP, sieving, P-1 factoring), there could be some "overall" user/team stats to know who is the "best of the best"

[Moo_the_cow]

I don't think that it can be done anytime soon since it's very hard to compare the amount of computing time needed for sieving and factoring with the time for PRPing.

[mklasson]

IMHO a bias towards sieving wouldn't hurt. That way some stats junkies (and I'm not using that term in a derogatory fashion ;) ) might put their machines where they would do more good.

I know there are all sorts of opinions on whether more sieving is good or bad though. But really, I think that when you're sieving you're getting a factor within 500K of the prp bound about as often as you would complete a prp test. The bonus to sieving being that you've eliminated some 40 other tests as well in the same time.

[Troodon]

Some ideas:
- When you find a factor by sieving, you save the need for a PRP test AND for a P-1 test.
- With sieving, as we're getting deeper, the density of factors is decreasing. That means more CPU time (or power) to find a factor.
- As n increases, the CPU time (or power) needed for a PRP test increases.
- In contrast, the CPU time needed to P-1 factor, for example, n=4000000 and n=5000000, is almost the same.