22

NBA Hunting Grounds: An Introduction to TeamSPACE

In animal movement theories, a Hunting Ground is an area a predator regularly visits find prey. Applied to principals of law enforcement and crime analysis, a Hunting Ground is an area a serial criminal attacks victims in a regular, patterned, and predictable manner.

Applying notions of unique decision-making processes and rational choice theory, the concept of a Hunting Ground directly applies to the way a basketball player “hunts” for scoring opportunities. The areas on the court an individual regularly visits (either by personal choice, coaching discretion, or the guidance of teammates) are typically driven by the likelihood of success. The extremes are easy to identify: Andre Drummond scores exclusively in the paint, James Harden attacks from the top of the arc, and Al Jefferson is deadly from the left block. The point is this — all players have preferred areas they will consistently (and predictably) attempt shots from; the clusters of continued successes over the course of a game, season, and career become their Hunting Grounds.

To extend the metaphor a bit further, animals, criminals, and basketball players all operate relative to their counterparts. Every lion cannot attack the same watering hole; each bank robber cannot rob the same branch in Times Square; every player cannot exclusively shoot corner 3-pointers. Each is bounded by its associates, either other predators, offenders, or teammates. Thus, their activity space is shared, and the Hunting Grounds of fellow teammates specifically address floor spacing and lineup continuity issues.

Further, each group seeks to avoid countermeasures from adversaries. These opponents include hunters, law enforcement, and the other team’s defense.

Unique to basketball, however, is the notion of a fixed spatial dimension. Lions and bank robbers can self-organize across terrain and cities; each is able to operate in naturally defined activity space. Conversely, a five-player lineup is bounded by the court. As such, the distinct space a player occupies for scoring opportunities should theoretically be as unique as possible, minimizing overlap with teammates.

In any case, we can extrapolate from point-based events (attacks/crimes/shots) using kernel densities based on observed mean nearest neighbor distances of each individual’s collection of points across the study area, weighted by success (in this case, a made shot), resulting in amorphous areas that cut across traditional boundaries to identify the true clusters of activity. (Sorry, got REALLY nerdy there.) I cautiously call these “hot spots,” although a Hunting Ground is really a more comprehensive representation of activity.

So what? The point is this — if we render the Hunting Grounds for a given lineup on the court at the same time, a successful team should exhibit little-to-no overlap; a rainbow of unique color across the court (the past two San Antonio Spurs teams come to mind). Less successful teams/lineups should experience chronic overlap and/or a general lack of activity; anecdotally this is the 2013-2014 Detroit Pistons (The Josh Smith Experience).

Having said all this, this offseason I will be examining the starting lineups of all 30 teams, past and present. Our journey begins with the 2013-2014 New York Knicks, considered by many an underachieving mess.

2013-2014 New York Knicks Starting Lineup

2013-2014 New York Knicks Starting Lineup

[Before I go any further it must be stated that this analysis is not even remotely possible without the help of Austin Clemens. At a minimum, go follow him on Twitter immediately (@AustinClemens2) as mild compensation for his hard work.]

So what does this TeamSPACE chart mean? In this case, it highlights the general LACK of successful clusters of shooting activity (aside from everyone clustering near the rim). There’s Melo (in shades of red) the midrange monster, and everyone else. Iman Shumpert has some good activity in the corners; however, Shump, Raymond Felton and Andrea Bargnani are basically just in Melo’s way. There’s no consistency from beyond the arc; merely small splotches of overlapping, crowded activity. Someone needs to dominate those wings, and they are not. Ditto for the midrange; Melo freely dominates the baseline, but free throw line and above is crowded by Shump, Felton, and Bargs. Overall, last season’s Knicks exhibit an inefficient and unbalanced distribution.

This is especially compelling when compared against last year’s champs, the Spurs:

2013-2014 San Antonio Spurs Starting Lineup

2013-2014 San Antonio Spurs Starting Lineup

This TeamSPACE could not be more different (well it could, but it is still significantly different). It brings a tear to my eye. The Spurs represent an incredibly balanced and spatially diverse starting lineup. Danny Green lights in up from some many areas around the arc, the overlap in the corners with Kawhi Leonard is negligible. Tony Parker, Kawhi, and Tim Duncan are equally active in the mid-range, each with dominant (but not exclusive) areas of influence. And at the rim is a microcosm of the rest of the court – diversity, equality, and areas of opportunity for many. Obviously everyone overlaps at the rim, but the Hunting Grounds noticeably extend beyond the rim (while staying in the paint) for Splitter, Duncan, Leonard, and Parker. Overall, it’s the result of five artists painting on the same canvas, sharing one brush.

Kirk Goldsberry raised the collective intelligence of everyone’s ability to examine spatial shot data over the past two seasons. Austin Clemens just changed the game with user-searchable shot charts since the mid-1990s. I’ve previously introduced this topic (here, for starters), but now it’s time to take things to another level. How can we better understand individual shot selection within team dynamics? My answer is TeamSPACE, based on Hunting Grounds. As free agency finishes up, look for more TeamSPACE charts and analysis throughout the offseason and into the 2014-2015 NBA regular season.

Matt D'Anna