In NCAA Tournament, regular season overachievers often disappoint
An interesting analysis by Nate Silver of regular-season overachievers (relative to their pre-season rankings) in the NCAA tournament. Bad news for Notre Dame: teams that rise from being unranked preseason to the top 10 by the end of the year often disappoint in the tournament.
I’m as fascinated by the accuracy of the preseason polls as anything else in the piece. In the tourney, the team with the better preseason ranking wins 72% of the time- the same predictive accuracy as using the tournament seeding (higher seed wins 72% of the time), and the tiniest bit more accurate than the RPI (71% win percentage for the team with the better RPI). That seems like some pretty crazy Wisdom of the Crowds stuff to me.
I’m guessing this also partly explains why hoops heads get beat by secretaries in office pools. Hard-core college basketball fans are aware of the overachievers- such as Notre Dame or San Diego State this year- and end up giving them overrating them in their bracket picks. Meanwhile, casual fans aren’t as sensitive to who the overachievers are (“I thought Notre Dame was a football school, not a basketball school!”), so they aren’t as susceptible to the overachiever effect. (Side note: I might be implying that overachievers tend not to be big-name programs. That’s an interesting question on its own.)