We spend a lot of time on the site talking about statistical models. Statistical models are great for identifying trends and relationships between variables when we have a significant amount of data. Models are also useful for moving us beyond arguments based on examples and anecdotes. We think this is particularly important when discussing sports. Every guy in every bar has a theory that they can support with an example.
In our current series on college basketball programs’ abilities to convert recruits into NBA draft picks, we have decided to start with summary data for each school. We plan on concluding the series with a statistical model that predicts the likelihood of a player being drafted based on the player’s recruiting ranking, the school’s investment in the program, the rankings of the player’s teammates and other factors. We decided to start with the summary efficiency rankings simply because these rankings are more accessible to fans and tend to generate more conversation.
I wanted to use today’s rankings of the Big Ten schools as an excuse to delve into a specific comparison between two schools. I have two reasons for this. The first is that looking at the data for a couple of schools will highlight why our statistical model gives the results it does. The other reason is that I (Lewis) want to provide some recruiting material for my Illini.
The chart below lists our efficiency rankings for the Big Ten. At the top, we have two solid programs in Purdue and Illinois. These two are followed by the recent and traditional powers: Ohio State, Indiana and Michigan State. While Ohio State has the most draft picks, they also had the greatest recruiting success with players like Greg Oden, Mike Conley, BJ Mullens, and Jared Sullenger coming through Columbus in the last decade.
(For more details about the methodology, click here)
Now back to my second motive. As an aside, I thought about titling this piece “Why Jabari Parker, Cliff Alexander and Jahlil Okafor Don’t Need to Travel Far from Home.” In our rankings of the ACC, the Duke Blue Devils finished in the middle of the pack. What I’d like to do (and I know this is self-indulgent) is to compare the Illini with Duke. In the table below I give the rankings of members of Duke’s and Illinois’ recruiting classes from 2001 to 2002 (I collected these by hand so please excuse any omissions).
Over the relevant drafts, Duke had 11 players selected compared to 4 for Illinois. While this may seem to be a reason for a student athlete to choose Duke, when we look at the input, things are much less clear. From 2002 to 2010 Illinois had 1 top twenty recruit. In contrast, Duke had 13. If we look at top thirty recruits, Illinois still had 1 while Duke had 15.
I think the explanation for these results is pretty simple. When an athlete chooses a school in a power conference, but without a roster loaded with McDonald’s All-Americans, that athlete has more chances to see the floor, and even when on the floor the athlete has a better chance to be the focal point for the offense. Going all the way back to 2002, Dee Brown was a featured star at Illinois while the similarly rated Sean Dockery was a role player for Duke. Another highly rated player from Illinois Michael Thompson ended up transferring from Duke to Northwestern. And while some attrition is natural, it is interesting that Thompson was rated higher (30th) than every single Illinois recruit in the period from 2002 to 2010.
So what is the take away? In terms of the preceding comparison, it is that what the glitz and glamour of playing at a high profile school is attractive, the high profile nature of a Duke is likely meaningless when it comes to getting to the next level. In fact, the tendency of very highly rated players to choose schools like Duke means that the player’s chances of making the pros might actually be a bit less at a Duke than an Illinois.
But, as noted, the comparison of Duke and Illinois is anecdotal. What we really need to reach the preceding conclusions is more data. My comparison of Illinois and Duke is mainly intended to foreshadow the statistical analysis we will provide next week. This analysis is designed to tease out the effects of player quality, within roster competition, school investment and on-court success on player development.
PREVIOUS POST: RANKING THE ACC
NEXT POST: RANKING THE PAC-12