There’s a new paper out this month by Casey Ichniowski and Anne Preston concerning the NCAA tournament and the NBA draft (thanks to my PhD buddies Chris and Felipe for passing it along). Their argument is that unexpectedly strong tournament performance (especially team performance) causes players to be selected earlier in the NBA draft. This isn’t a bad thing though — in fact, they suggest that these strong tournament players tend to outperform their draft position in the NBA.
I believe their results saying that tournament performance affects draft position (this has also been shown by Chaz Thomas in an undergraduate thesis, and by David Berri, Stacey Brook, and Aju Fenn), and I mostly believe their results that strong tourney performers should be drafted even earlier, though their set up is a little odd for this second issue.
The clearest way to show that teams make mistakes in the draft is to compare NBA performance for different types of players taken at a similar position in the draft. “Types” could mean white or black, volume shooters vs. efficient scorers, 7’0″ centers vs. 6’8″ centers etc. For example, if teams that draft volume shooters tend to perform worse than teams that draft more efficient scorers — at a similar draft pick — then the first teams are probably making a mistake and drafting the volume shooters too soon.
So, I would have liked to see Ichniowski and Preston run regressions using a player’s draft pick and college statistics (regular season and tournament) to predict his individual and team performance in the NBA. Controlling for draft pick in the regression means that effects of other variables (such as tournament performance) are measured for players taken at about the same pick. For example, if they found that unexpectedly strong tournament performance leads to better NBA performance in such a regression, this would mean strong tournament players outperform weak tournament players drafted at a similar spot.
They almost do this, but instead of just using the college statistics in the regression, they first use another regression to combine all the college statistics into a single measure that they call the “March Madness bump” and use that in the regression. Roughly, this is the effect that each player’s NCAA tournament performance had on his draft position. So, instead of answering the simple question of whether strong tourney players are drafted too high or low, their regressions determine whether players who they predicted to be drafted high based on college performance were still drafted too high or low. I think that’s a confusing approach, though I bet their conclusions would be similar if they did it my way, since tournament performance is the main driver for their “bump” variable anyway.
What’s the upshot? My guess is that tournament performance is a valuable predictor of NBA performance because the games are played against strong competition. It would be illuminating to restrict regular season statistics to games against top 75 RPI opponents and see if these statistics predict just as well as tournament stats. Of course, many teams have few top 75 RPI opponents, so I still think that tournament performance is valuable. I was quite surprised that teams potentially underweight tourney play, since the NCAA tournament gets ten times the media coverage that the regular season gets. Even this increased attention must not be enough to balance the stark change in opponent quality for many players.