Quote from: The Vark on August 08, 2011, 11:55:28 PM

Quote from: organofcorti on August 08, 2011, 11:10:32 PM

All I'm saying is if you have more than three pools, the only hopping you have to do is when another pool has fewer shares than the one you're on. You shouldn't have to jump to backup at 43% - this number its only useful if you have one proportional pool and one pps pool. So if any pool ops are checking for you hopping off at 43%, not hopping off then won't have an effect on your long term earnings.
...
sort of. we're talking long term here, so even going over 1*diff doesn't matter. If you set the threshold to 3*diff, then for a particular round it might make a difference, but only if no other pools were available. But that's pretty rare and in the long run it wont have a significant effect on overall earnings.
...
I know this is all very counter intuitive - I wrote the sim and it took me a week to believe the results. Plus I'm probably not explaining it well, sorry.

First, I wanted to thank you for doing these sims and sharing the results, this is absolutely fascinating. It makes sense right up until going over 1*diff part. I think that is what I'm having trouble getting my head around, so maybe you could try to explain that aspect of it better?

It seems that any time that you can submit a share with a higher EV that is your best course of action. Is this specifically vs. submitting to a PPS pool? Would I be right in assuming that if you had the option to Solo mine vs submit a share at more than 1*diff that Solo mining would be the better option in the long run?

Sorry for all the questions, just a little bit mind bending....

(Oh, and if anyone has a hden invite, I'd love to not be cluttering up the thread with non-BH stuff. Thanks)

I'm glad I could confuse you as much as relearning all my uni stats confused me. In stats it seems that 'common sense' is often not a very reliable guide.

The reason that going over 1*diff is ok: Each data point is the result of at least one hundred thousand rounds using that particular hop off point (x axis). It's not the result of just one round. So while it may impact a particular round, if you have enough pools it wont happen very often. Just because you set your hop threshold to say 3*diff, it doesn't mean you'll mine the often, and in the long run won't make a significant difference to your earnings. A pleasant side effect is although you'll still be hopping, it wont be until just 43% any more so it will be harder to point you out as a hopper.

As far as your second question goes, yes that's true. But it won't happen often enough to give you many pps shares if you have lots of other pools. See the difference between the '0 other pools' graph and the '5 other pools' graph? Hopping to another pool that just found a block will happen more and more frequently when you have more pools, and sooner.

I won't clutter this thread up further, but I'll get some intro to hopping blog posts up on hoppersden using byteHopper to illustrate. When I can stop fiddling with the sims, that is. The best hop point vs efficiency for slush should be done soon.