Originally Posted by
Dreyo
Your analysis is flawed, but in a subtle way. You have the pieces there, but you're fitting them together wrong. You note, accurately, that your chance on not having generated a chi up to and including the 5th tick is very small - .385%. But here's the thing: in order for that to happen, you had four ticks previous to the fifth where you now have not generated chi. The conditional probability for that is (not coincidentally) the number for the 4th tick, 3.85%. Which means that these allegedly high-likelihood chi generation events themselves don't happen very frequently. With the cap of 100%, we can actually model exactly what the long-term average proc rate would be. But just looking at the numbers that you gave, you can't just say that it's "ridiculous" - because you need to tie it in with the conditional probabilities.