Are you saying simply that over 3 btc a day is being claimed from the faucet? Or are you actually paying out (people withdrawing) over 3 btc a day that can be traced back to an initial faucet claim?
Two quite different things. The former doesn't say much as most people claim faucet, gamble it, lose it and repeat - you are not out of pocket at all.
In theory there shouldn't be a huge difference between the two numbers. The house edge is pretty small.
If you give out 1000 satoshis, but don't let people withdraw until they have 100k satoshis, you're going to see about 99 out of ever 100 gifts be lost, but 1 in 100 will turn into 100k and get withdrawn.
As we've seen on the various sites that make their actual profit vs. turnover public, the actual profit made by dice sites is often very different than the expected profit, but as a first approximation you can expect to lose roughly the same amount that you give out in the long run.
Agreed. But theory is just that...theory.
Reality seems to find a way to never quite match what's expected lol.
The differing psychology in play and utility value we have with the faucet amount is what throws normal theory out the door here imo. As evidenced by normal house edge of ~1% v faucet edge of ~4%.
I'm no statistician, but isnt that a massive difference?
The amount is so low, endless and free that no one hesitates to lose it almost immediately. I truly believe that the actual payout from the faucet is alot less than the expected payout. And all faucet claims are returned to the house rather quickly.
I also think the faucet played a part in PDs success, and should be considered an asset to the company rather than a simple liability. Just my opinions though. Im not privy to any hard data.