After reading the past comments I have an impression that the problem of faucet abusing by bots has not been completely solved. If it is so and if the site is still losing significant amounts because of that, maybe it would be a good idea to announce a competition for the best protection against bots. Since abusers can adjust quickly, the idea of protection have to be so good that even revealing its algorithm would not help a bit in breaking it. We all know that it’s possible, right? Satoshi Nakamoto has provided us with an example.

The first thing that comes to mind is taking advantage of the fact that all humans are different in all aspects while the bots are the same at some. Thus, after a certain amount of claims you can know almost for sure whether an account is being run by a bot or not. … Well, of course I know that most likely it’s not that simple. But I’m sure there are people much knowledgeable than me and that they can propose some really great ideas.