Well, here we go. The Super Bowl is upon us, and we’re going to hear all sorts of claims about data and analytics. How some statistic tells us that Tom Brady made a good decision. Or that the Rams’ coach went with his gut, even though the data told a different story. Many (most?) of these claims are going to reveal common fallacies about data+sport, some of which seem sort of correct at first glance.
Send me a description of what you think is the most egregious use of data during Super Bowl Sunday. (Email link is at the top right of the page.) Winner gets glory. Maybe a prize. But probably just glory.
Last year, I ran this contest with some Kellogg students and friends, and the winner relayed this nugget: The Eagles kicker, Jake Elliott was setting up for a 46-yard FG. In Elliott’s career, he had made 88% of FGs < 50 yards, and 64% >=50 yards. Claim: "It's a good thing this is less than 50 yards, because Elliott is clearly a lot better at under 50 yards." (The point? He probably doesn’t have an 88% chance of making a 46 yarder – you can’t lump a 46 yarder in with all kicks under 50. If you need me to explain further, then no worries – you probably won’t win this year anyways.)