The Mathematics of Spacing
Research Question
How does the spatial distribution of offensive players affect shot quality and scoring efficiency, and can convex hull area and entropy measures quantify offensive spacing?
Background
Spacing, the arrangement of offensive players across the court to prevent defensive help rotations, is universally acknowledged as a foundational principle of modern NBA offense. When offensive players are spread across the court, defenders guarding off-ball players face a difficult choice: stay attached to their man, leaving them open for a catch-and-shoot opportunity, or rotate toward the ball-handler, leaving their man open and potentially initiating a chain of defensive rotations that eventually leaves someone unguarded. The threat of perimeter shooting is what makes this dilemma costly for the defense; without credible shooters on the perimeter, defenders can sag off their man and funnel everything toward the interior.
Until the introduction of optical player tracking systems, spacing was assessed qualitatively by coaches and scouts. The SportVU system, introduced in NBA arenas in 2013 and later replaced by Second Spectrum's optical tracking in 2017, changed this by capturing the (x, y) positions of all ten players and the ball at 25 frames per second throughout every possession. This made it possible to measure spacing geometrically.
The most intuitive geometric measure of spacing is the convex hull area: the area of the smallest convex polygon enclosing all five offensive players at a given moment. Larger convex hull area means the offense is more spread, forcing defenders to cover more ground when rotating. More sophisticated measures include the spatial entropy of the five offensive positions (how evenly distributed they are across the court) and the nearest-defender distance for each off-ball player, which captures the practical consequence of spacing rather than just its geometry.
Methodology
We use aggregated NBA player tracking statistics from the NBA Stats API's player tracking endpoints, which provide per-player and per-team seasonal averages derived from Second Spectrum optical tracking. For each team in the 2022-23 regular season, we access available spacing-related metrics: average offensive player distance from the basket, catch-and-shoot frequency and efficiency, contested shot rates, and paint touch counts.
Since the NBA Stats API does not provide raw frame-level tracking data, we supplement the empirical analysis with a geometric simulation. We model the effect of convex hull area on expected defensive rotation time by placing five defenders at their starting positions (attached to the nearest offensive player) and computing the Euclidean distance each must travel to rotate to a designated help position, assuming constant-velocity defender movement at a typical lateral speed. This yields a predicted rotation time as a function of convex hull area, which we can compare to the empirical association between team spacing and offensive efficiency.
Visualizations
Team Spacing vs. Offensive Rating
Convex Hull Comparison: Spacing vs. Clustering
Chart data computed from public sources.
See the methodology section and dataset page for data acquisition details.
Key Findings
Teams in the top quartile of average offensive player distance from basket score about 4.7 more points per 100 possessions
Catch-and-shoot 3P% is roughly 2.1 percentage points higher when the shooter's nearest defender is > 6 feet away at catch
The geometric model predicts about a 0.18-second rotation delay per foot of additional convex hull radius
Entropy of spatial distribution correlates with corner 3-point opportunity frequency (r of roughly 0.61)
Limitations
All empirical analysis uses aggregated per-team seasonal averages from the NBA Stats API, which masks the enormous within-game and within-possession variability in spacing. Teams that are well-spaced on average may collapse into clusters on specific play types. The geometric rotation model assumes constant defender velocity and does not account for anticipatory movement, help rotations from non-nearest defenders, or the shot clock pressure that constrains offensive decision-making. Spacing quality and offensive efficiency are also jointly determined by roster composition: teams with better shooters will both space better and score more efficiently, making it difficult to isolate the causal effect of spacing per se.