A Data Science Central Community
In particular, do casinos use statisticians to detect well disguised schemes? For instance, what about clients that spend a lot of time and money in online casinos, and on average win 53% of the time (odds should be below 50% if you participate in a large number of games)? Or those who lose most of the time but usually win big when they win?
Do casinos use Monte-Carlo simulations, confidence intervals and experts in statistics and extreme value theory to catch these fraudsters?
Other big question: What about Wall Street? How do low frequency (illegal) schemes get detected, and how can you prove that the intent was illegal? With one million traders, assuming the chance to win or lose on any single transaction is 50% if you trade randomly, it is expected that one trader will win 20 times in a row - without using insider information or fraudulent schemes. You can expect that that there will be a few smart ones winning 75% of the time over 100 trades, without using fraud.