Suppose for the sake of conjecture that anon is a time traveler. He travels back in time to the 60s and starts making bets-- horse races, boxing matches, etc., with perfect knowledge of the outcome every time. To avoid suspicion, anon makes a wrong bet every now and then, whenever he feels like it.

Looking at the distribution of his total history of betting, how can /sci/ calculate the likelihood of anon being a time traveler?

If the question is ill-posed, how can a sketch of this situation be phrased into a mathematically interesting way?