Jeremy Lin and Moneyball: The Problem of Identifying Talent

After watching Jeremy Lin (Knicks) score 38 points against the Lakers tonight, I’m now on the Lin bandwagon.  I don’t really even follow basketball that closely, but this seems like an intriguing story.

How on earth did someone like this go unnoticed?   Seriously.  He happened to get an opportunity to show his stuff as Carmelo Anthony and Amare Stoudemire are injured – and boy has he delivered.

Here’s a kid who didn’t get recruited for college ball, despite a tremendous record in high school.  He was a superstar at Harvard but went undrafted by the NBA after graduating from Harvard (in economics) in 2010.  He played a few games for Golden State and Houston, but was cut by both.  He has played D-league basketball this year, until a few weeks ago.  As of last week, he did not have a contract.

But come on: is basketball truly this inefficient at identifying and sorting talent?  The comparisons and transfer of ability across “levels” (high school-college-professional) of course is tricky, though you would think that with time there would be increased sophistication.

Now, four games of course doesn’t make anyone a star.  But even if Lin proves to “just” be a solid bencher, it seems that talent scouts clearly undervalued Lin (who lived in his brother’s apartment until recently).  How much latent talent is out there?  (I think that at the quarterback position in professional football – there are significant problems in identifying talent, but that’s another story.)

There are of course also some very interesting player-context/team-fit, interaction-type issues here, and I’m not sure that this really gets carefully factored beyond just individual contribution (thus not recognizing emergent positive, or negative, player*player effects).  It’ll be interesting to see what happens, for example, when Carmelo Anthony is added back into the mix.

Well, it’ll be interesting to see how all this plays out.  There is in fact a sabermetrics-type, stats-heavy, Moneyball-like thing in basketball as well – called ABPRmetrics.  I would be curious to know whether there are ways to statistically identify Lin-type undervaluation and potential, and whether phenoms like this lead to better metrics for identifying talent.

UPDATE: Here’s ONE analyst/statistician who saw Lin’s potential in 2010.

---