Redskin Loss Predicts Kerry Election Victory
Politics via mefi
The Washington Redskins’ defeat yesterday predicts that Kerry will win the presidential election. Why? Every election year since the Redskin’s inception in 1936 (when they were known as the Boston Redskins), their victory or defeat during the final game before a presidential election has matched the outcome of the election.
MeFi says ascribing predictive powers like this is an example of post hoc ergo propter hoc (fallacious) reasoning, and I think that’s correct—although there’s not an assumption of direct causality being made, in saying ”a predicts b” there’s a causality assumption being made somewhere in the chain of events between a and b. Still, it’s hard to get around the fact that it’s a pretty amazing correlation. Elementary probability theory tells us that there’s a 0.5^17 = .00076% chance of the Redskin’s final victory correctly predicting the election for 17 elections in a row, which is pretty damn small.
But the insight here is that the chance of finding some correlating event is much, much higher. If Bush wins this election, people will simply go out and find another correlating event that does match—go “fishing”, to use the technical term.
So don’t get your hopes up.