krinid, you are right, I am guilty of write-only

Pretty sure you just join raffles and each person that joins lowers the percentage of a win for the other. So join the raffle so the other person doesn't win;)

I will try to join as many raffles as possible to lower your chances of winning ;P

@VeganTitsy: This raffle analysis post looked at entry time rather than entry order, since I was responding to some base rate neglect in functor's original post (https://www.warzone.com/Forum/564739-secrets-rafflebot).

Just to sate my curiosity, here's the data for entry order (which has been added to the spreadsheet):


The expected win rates in that second chart account for how much competition they face- they're the sum of expected values for entries at that position in the order. On average, the first entrant to a raffle (with an unknown number of entrants) faces 6.17 opponents, while the 10th entrant faces 10.57. Of course, this is exactly what you pointed out- the first entrant might be entering a solo raffle, while the fifth entrant faces a minimum of 4 opponents.

Out of 21 different entry order positions, 2 have statistically significant divergences between expected an actual win rate: 2nd entrants won 1.61% more often than expected (p = 0.01) and 5th entrants won 1.46% less often than expected (p = 0.02). These are both, of course, spurious: out of 21 p-values, random chance easily accounts for 2 having seeming statistical significance (28% likelihood).

2nd entrants do win more often than first entrants, though- they account for a larger share of wins and had a slightly higher win rate. Beyond the aforementioned luck, this is also because single-entrant raffles are exceedingly rare- occurring about 0.7% of the time in our dataset.

Edited 11/10/2021 05:52:04