Wednesday, November 14, 2012

Welcome to the Index!


What is the Draftability Index?

The Draftability Index is a metric designed to analyze college draft prospects and their likelihood of producing wins at the NBA-level. Prospects are graded and slotted at one of four levels based upon their likelihood of producing positive win scores. DI is based upon the Offensive and Defensive Win Share metric of Dean Oliver.

How does it work?

After conducting a fairly exhaustive study of the various 'advanced' approaches to player evaluation, I came to the conclusion that Win Shares is the strongest model for predicting and explaining player production. If you care to research it, there are passionate defenses and take-downs of WS from many credible sources. As a mathematical neophyte, my faith in the method derives from the simple fact that the things that WS says will happen actually happen pretty damn often.

With this in mind, I undertook a study to see to what extent outcomes tied to winning translated from college to professional basketball. After running simple correlations on the data, a clear trend emerged: defensive efficiency (DWS) correlates consistently to a mathematically significant degree while offensive efficiency (OWS) is conspicuously, maddeningly non-correlative. Sometimes offensive efficiency translates perfectly, more often there is no relationship whatsoever from college to professional basketball. It is a true red herring and a contributing cause to many poor draft day decisions year after year after year.

To apply these findings, I calculated a mean Defensive Win Score (DWS) that would constitute being able to effectively guard one’s position at the NBA level (by way of regression analysis) and used it as a weight to the college DWS scores of potential prospects. I added to this the most reliable (again, borrowed from Berri) efficiency-based offensive metric (eFG%) to create one all-encompassing efficiency rating score. It is calculated as follows:

                             (Positional DWS) / (NCAA mean DWS)^2 X (NCAA eFG%) =DI

The Index scores, once obtained, are divided into four categories, which indicate primarily, the safety of the prospect.

(0.00-.24) - Grays – Prospects who are highly suspect and show no discernible track record of efficient production.

(.25-.49) – Reds – Prospects who are suspect and have not shown consistency of efficient production.

(.50-.99) – Golds –  Prospects who have produced efficiently somewhat consistently, correlative to professional NBA production at a mathematically significant degree.

(1.00 + Up) Greens – Prospects have produced efficiently and consistently and do so and the majority continue to do so at the professional level.

Why Use Draftability Index?

It works, mostly. Originally, I calculated the Draftability Index score for every drafted player from 2002-2006 and of the 18 players that rated as 'Green' prospects, 16 went on to be productive NBA players. How many front offices have that high of a batting average?

Granted, some of these calls were not exactly revelations (Dwyane Wade, LaMarcus Aldridge) and others while good players, are not franchise changers (Tony Allen, Ronnie Brewer), however a goodly number of my other 'Greens' (Rajon Rondo, Paul Millsap, David West) have proven to be difference-making NBA players and weren't exactly chillin' in the green room on draft night.

On the flip side, of the 17 players selected in the first round over those five seasons who graded as 'grays', only three (Kevin Martin, Kris Humphries, and Chris Kaman) have produced anything in the Association, and really, Humphries is the only one I would actually want on my team. The rest read like a Bleacher Report slide show of biggest busts: Adam Morrison, Marcus Haislip, Ike Diogu, Marcus Banks...

Predictably, the 'reds' and the 'golds' are a lot less clear cut and, as you may have surmised, is where most players fall. There is some actionable data there, but its probably a good thing that nobody has to make every pick for every team. There is still a great deal of mystery in the process...

P.S. You said 16 of 18 'greens' were hits. So what was the big miss, huh?



Yup, that would be Tyrus Thomas. I don't feel this requires any additional explanation on our part.

-Andrew

No comments:

Post a Comment