Alex_Clover I'm working on the assumption that which spot we pick is irrelevant; why would that matter if the odds were all even? [edit: having thought about it some more, "I'll get better odds if I keep picking the same answer" is just another form of the gambler's fallacy. which tile you pick doesn't matter if the odds are even. if they wrote the code so that it does matter which you pick, the game is even more rigged than I'd expect 😉 ]
https://timothyweber.org/node/255 suggests that doing at least 5 * faces tests is necessary, so 45 in our case, but 100 * f is better. anyone remember how long this thing has been running? I know I've done it every day and never got 50k.
we've also got some sampling bias because people who've never got the 50k are more likely to read this post, but I don't know how to compensate for that. still, >45 non-50k rolls on a single account should be enough to make the matter worth considering.
also, I don't believe that games of chance aren't commonly influenced by the house (anything involving real money will have the odds set so the house reliably makes a profit). it's not like PF has made any claims about what the odds actually are; it's not far-fetched to imagine that a programmer implemented it as a table of weighted odds for what you actually win, and then the graphic designers decided to put that number on a card and scatter the other possible results on the other tiles.
edit: now that I think about it, I bet they re-used some of the code from the other daily reward where you don't pick tiles. 🙂 that one I did get the 40k reward once or twice, but it also feels like it's weighted towards the middle numbers.
edit: additionally, the randomness in question here is that of computers and mobile phones, which is often actually pseudorandomness, with no quantum effects involved. Getting real randomness in a computer is actually a nontrivial problem 🙂