Evaluating NBA Teams' Drafting

Behind the Project

8/22/2017

My name is William Graif, and I am entering my senior year of high school in Westchester, NY. I fell in love with both sports, especially basketball, and numbers at a very young age, and people who know me describe me as having an "analytical mind." I do not know whether the egg or the chicken came first, but I currently hold the second overall rank in chess for 11th graders in the United States and am a two-time national champion for my age group. An official US National Master by the age of 14 (accomplished by only a few-dozen other Americans ever), I've competed in Vietnam, Turkey, Dubai, and Greece in the World Youth Chess Championships, earning 56th for Under 14-year-olds internationally in Dubai.

I could not have done this project without my partner, Patrick Walther, who is my age and lives near me; the two of us met this very summer at a fantastic summer program called the "Wharton Moneyball Academy" where we in fact learned about sports analytics from Professor Wyner, an excellent statistics professor at Wharton. Pat is as well-versed in basketball analytics as I, and even more so in baseball. In addition to invaluable advice, guidance, and consultation, he was the chief coder for this project.

- Outline
- Coefficients
- Grooming the Data Set
- Final Rankings

0 Comments

Inspiration

8/18/2017

0 Comments

I can't believe it took me this long to read Moneyball by Michael Lewis.

For someone whose life (mostly, at least) revolves around sports analytics, to not read the Holy Bible of it -- and instead only watching the cinematized film version starring Brad Pitt (still awesome) -- was simply a travesty, I now realize. Even for a basketball fan that probably couldn't even recognize half of the MLB's All-Stars this year, Moneyball was enlightening in several ways, and far and away now my favorite book of all time.

But one thing (actually, several) about it really struck me. Much of the emphasis on Billy Beane's new "Moneyball Method" was in part the transition from the subjective to the objective in player evaluation, especially when it comes to prospect scouting. The more conventional baseball scouts would say things like "he's got wheels (fast runner), a nice swing, [etc.]" whereas Paul DePodesta, Billy's right-hand man and computer geek, would say things like "he's got an on-base percentage that cracks 40%," which is apparently -- Lewis argues -- a significantly more important observation. It has since been agreed upon that Rockets GM Daryl Morey has brought Beane's analytical ways into the basketball sphere with what some call "Moreyball" and that now the transition is more or less fully complete.

But is it? Take Lonzo Ball's scouting report from this year, published by what I actually consider an excellent basketball prospect evaluator, nbadraft.net:

You don't have to read the whole thing unless you want to, but you get the basic idea if you just glance at a few of the notes. While an incredibly detailed report written by someone who has clearly watched Lonzo play and definitely knows what he's talking about, these are the types of "arbitrary evaluations" that teams use that make Billy Beane throw chairs across rooms (read the book!). But what else can scouting departments do? It seems farfetched that college basketball statistics might actually be an accurate predictor of pro-level stats, for how can you possibly compare someone who plays in the ACC versus something like the Ivy League?

So that's ultimately the question that sparked this research project: are NBA teams, which are reliant on these "subjective" evaluations of prospects, good at selecting players or not? Lucky for those of you whose curiosity I piqued, we have an answer.

(click the link below to continue!)

Next: Outline

0 Comments