Until now I have thought, that it is fair and square 1 out of 10 000 probability for free rolls to win USD200. But then with more than 4,3 million users (lot less active users of course, still should be at least tens of thousands daily active) there should be several USD200 hits + lower wins DAILY.

Let's say cautious 20 000 users × 2 rolls daily = 40 000 free rolls daily, that makes 40 000/10 000 = 4 × USD200 wins cautious daily average. It could be even much more as you can imagine. Plus lower USD twenty, two and so on wins.

Hard to believe, that operator could survive that. Comments welcome.

I was very sceptical about the coin roll game and I strongly suspected there was some mathematical trick involved that made it near impossible to win the main prize. So I tried to unveil the misrepresentation and I did the maths. It seems I was wrong. Like is has been said earlier on this forum, the chance of winning the main prize is 1 in 20,000.

In every 10,000 rolls, mathematically there will be 0.5 winner of 200$, 2 winners of 20$, 4 winners of 2 $, 8 winners of 20 cents, 100 winners of 2 cents and 9885.5 winners of 0.2 cents (That is if all numbers in hash calculation have the same probability of showing up. I know too little of hashing to judge that).

For every 10,000 rolls, 171.371 $ are paid out which is about 1.71 cents per roll. That's more than most faucets pay.