Ytterbium, I believe there are those that hop pools to maximize their returns....or mine more BTC. I imagine it is a risk to jump from mining pool to mining pool. no risk no reward!
Loredo, please expand if this is the case.
With regard to pool hopping, I'll leave that to others who better understand the dynamics (or reference threads on the topic, which undoubtedly exist). But yes, there are advantages to pool hopping, for the same underlying reason there can be more or less than 3600 coins mined per 24 hours: one factor in
ex-post revenue is fixed while you, yourself, may have a dynamic decision you can make against that factor.
I was referring to the following, in which, again, I'm not anybody's idea of an expert. You can find, in the wiki's or elsewhere on line, the mathematics of the instantaneous probabilities of success in mining on any one hash. That's a function of difficulty, and the actual network hash rate at that time.
A business reckons gross revenue per time interval. Quite uniquely in bitcoin, gross revenue across a time interval is a tractable stochastic function of the above probability; the linkage between that probability and revenue over a time interval is the outcome (actually, I believe, a stochastic integral of outcomes) of the binomial distribution of the number of successes, conditional on the input factors.
Even that, however, is a one-stage calculation. If someone (think ASICminer) has the ability to increase their - and thus, the network's - hash rate in a meaningful way, they can affect an increase in their current revenue flow. But they also -for the next stage- affect the subsequent difficulty, which, all equal, might reduce their future revenue flow.
Thus, entities who have significant hash power, and the ability to throttle that up and back, can view this as a multi-stage optimization problem. But it's not only stochastic; it's also game-theoretic if they are not the only actor in that play.
That's what I meant when I used the term "gaming." Again, I'm not an expert on stochastic optimization, but the above is my somewhat naive understanding of the issue.
