Legend has it that storks deliver babies, and there are numbers to prove it. In 2000, British mathematics professor Robert Matthews found a correlation between birth rates in 17 European countries and the number of storks nesting in them. Coincidence? Why, yes, of course. Matthews had set out to demonstrate to his students the perils of equating correlation and causation. Yet his findings could be misconstrued to argue, albeit somewhat unconvincingly, that there is only a one in 125 chance that storks do not, in fact, bring human babies into the world. The correlation between breeding storks and birth rates in Europe, he concluded, was “statistically significant, not immediately explicable and causally nonsensical.”

It is easy to believe that when two variables follow the same tre...