NBA Adds Hustle Stats to Playoff Coverage

At Las Vegas Summer League, the NBA experimented with tracking “hustle.” In an attempt to better capture some of the “intangibles” that correlate with winning, cataloging these events seeks to make them tangible. While the LVSL stats were never made available to the public, the test run was evidently successful enough that for 2016 Playoffs, the league will collect this information for every game, and make the totals available on NBA.com.

According to the NBA, the following will be counted through a combination of SportVU tracking and video analysis performed at the Secaucus replay center:

• Deflections – Defensive player touches the ball on a non-shot attempt
• Loose Balls Recovered – Player gains sole possession of a live ball not in the control of either team
• Contested 2’s – Defensive player closes out and raises a hand to contest a 2 point shot prior to its release
• Contested 3’s – Defensive player closes out and raises a hand to contest a 3 point shot prior to its release
• Charges Drawn – Defensive player draws a charge
• Screen Assists – Offensive player sets a screen for a teammate that directly leads to a made field goal by that teammate

The first five were the items tracked at Summer League, while “Screen Assists” are a new category.[1. No word yet on whether the “Hustle Rating” assigning arbitrary weights to individual will return, though it would be no great loss, as the weight and value of these items will need some examination to arrive at more empirically-based weights.]

This is obviously good news – more information is almost always better than less. I am particularly interested in the “Loose Balls Recovered” stat, as tracked over a full season I wonder if it could start to explain some of the value ascribed to steals in box score-based models. If having a “nose for the ball” is really a discernable skill, loose balls and perhaps even deflections might start to help identify those players who actually have it.

The Contested 2 and 3’s will be intriguing to look at, but should be interpreted with caution on an aggregate level, especially with respect to evaluating individual players. For one thing, the player who ends up contesting a shot is quite often much more the result of a rotation scheme than the play (good or bad) of the player making the contest. Further, the effectiveness of a “hands up contest” in terms of suppressing shooting percentages is unknown. As a blatant plea for more and better data, I would say that combining a “contest yes/no” variable with the late, lamented SportVU shot logs would go a ways towards answering this question, though it would probably require a much larger sample than the 80 or so playoff games we’ll get this spring. That said, it is possible on the team level, contesting shots will be useful proxy for some sort of defensive skill. That said, especially on three pointers contesting a three is most likely a distant second-best outcome to not allowing the shot to be taken in the first place.

Whether or not this is a “playoffs only” feature or will be tracked leaguewide going forward is unknown – the fact that it requires some degree of manual review might make it too time intensive to be practicable under regular season conditions when there might be 10 or more games on a particular night. If this tracking does continue into the regular 82 games schedule, it could prove some extremely interesting data points for a system like Andrew’s Player Tracking Plus-Minus model.[2. Which won the APBR win predictions contest for the second year running, so congrats on that Andrew!] Either way, it’s a good look for the league to continue to push the boundaries in terms of the information, statistical and otherwise, they continue to make available to the public.