How is this? The pool gives a mining ratio that is supposed to represent how many more times profitable it is to mine vanity addresses than bitcoins, and it's currently 2.78. I can see how the variability of this would be higher than a PPS or geometric method pool, but over long periods it should still pay better, right?
That ratio is a bit idealised and assumes you are mining all the addresses at the same time (additive rather than multiplicative address generation). From what I understand, samr7's miner just uses a few of the most profitable addresses and that's it.
From how I understand it, you can easily mine several addresses at once, think merged mining, when the "task" was created from the same private key.
If I ask for 10 vanity addresses, and create those 10 tasks so that they all are made from the same seed, someone can work at all 10 at once.
If 10 people each want one address they surely won't create the tasks from the same seed/key. So a vanity miner will have to mine an hour for the first guy, then an hour for the next, and can only find an address/solution for the current task.
That 2.78 assumes that you mine at *all* tasks available at once, which is unrealistic at least, and suggestive/dishonest at worst.
So, currently, I think, there are several miner out there which check periodically on the available work. Once a task or combination of tasks appear which is more profitable than regular mining, they point their miners there. Which means a lot of the current tasks may stay there waiting for a long time..
Lets see if a marketrate with more realistic prices establishes!
Ente
