I'm not sure which random distribution you mean ... But if you assume a normal distribution with constant mean, and, say, 100 years of records, the chances of you still getting the hottest now are very very small