Like I said, it’s an estimate that relies on historical data. You can account for choices being left blank, etc., and guess at hunter numbers based on trends. The most difficult is when they add or remove hunts. Even though the actual draw is done once, if I want to estimate the percent likelihood of drawing with a set of 1st, 2nd, 3rd choices, I have to run the simulation a large number of times, with different simulated hunter populations each time. You are right that there are many bad ways to estimate odds, but the way I have found satisfies my needs and has been reliable. I believe it to be mathematically sound, and am happy to discuss the particulars of you are interested.