For the first decade of their existence, modern legal lottery games in the US were a commercial failure. That all changed when New Jersey introduced Pick-It, the first successful legal game. The popular story is that Pick-It was successful because it imitated the popular illegal numbers game, but there is more to the story than that. When we examined the payouts from the game, we discovered something mysterious. The payouts started surprisingly high and then declined over the first ten years of the game, while the rules of the game remained unchanged. We'll use a combination of probability, data science, and historical spelunking to explain this strange phenomenon and highlight creative design features of the game that helped its commercial success.
The math in this talk will use some probability of the sort covered on the SOA P exam (Math 477).
This talk represents joint work with Anthony Gamst.