Quote from: Aexoden on August 27, 2011, 11:51:00 PM

A short period where the buffer is negative will likely not be detrimental. However, if the buffer at some point meanders quite negative, rational miners will no doubt begin to trickle away, as at that point, the chance of receiving full pay for your shares anytime in the near future is rather small. I expect this wouldn't become significant until the buffer went below -1000. But that's just a guessing game.

I was curious about the probability that the buffer would go below a certain value. So I wrote a little matlab script to calculate this. I assume constant difficulty, 500 GH/s, mining for 30 days, starting with a buffer of zero. Here is a chart that shows the probability that the buffer will be below a certain size after 30 days:

The code used to generate this is here. If my math is right, there's roughly a 25% chance that the pool will be in a -500 BTC hole after 30 days. Now, assuming that no miners leave (and difficulty is still constant), it takes another ~20% event for the pool to get back to a buffer of 0 BTC. However, if miners leave, or difficulty increases, it gets harder and harder to get out of the negative buffer already built up.

Consider this my vote for PPLNS.