The replies here are a good reminder how easy it is to find explanations for trends that are actually explained by randomness. The best approach is to start by assuming unusual outcomes are the product of random chance and then see if there's evidence to disprove that instead of the opposite.