Is NBA player tracking data public?

Partly. Raw positional data is generally restricted to teams and league partners, but the league publishes many derived tracking statistics, such as speed, distance, touches, drives, and defender distance on shots, on its official stats site for public use. Check the site's terms before reusing data.

How do I get NBA tracking stats in Python?

Community-maintained Python libraries can retrieve the tracking tables published on the league's stats site. Respect rate limits and terms, cache responses, and record when you retrieved data, since endpoints and definitions can change between seasons. Keep requests modest and store raw responses so you can reprocess them later.

What can tracking data tell you that box scores cannot?

It shows how players move, how often they touch the ball, how they create shots, how tightly they are defended, and how they defend others. These details explain role and impact in ways counting stats such as points and rebounds do not.

NBA Player Tracking Data: What It Is and How to Use It

NBA player tracking data comes from camera systems in every arena that record the positions of players and the ball many times per second. Raw coordinates are not public, but the league publishes derived tracking stats such as speed, distance, touches, drives, passes, and shot defence. Analysts use them to study usage, movement, defence, and fatigue.

What tracking data captures

Camera systems in NBA arenas track the location of every player and the ball throughout the game. From those positions, analysts can calculate speed, distance travelled, spacing, who is guarding whom, how close defenders are on shots, and how possessions develop. This turns games into detailed spatial data rather than just events in a box score.

Who has access to what

The full positional data, recorded many times per second, is generally available to teams and the league's partners rather than the public. The league publishes aggregated tracking statistics on its official stats site, and occasional academic or competition datasets have been released. Most public analysis relies on these derived stats.

Publicly available tracking stats

The league's stats site groups tracking-derived statistics into several families. Most analysis draws on these.

Speed and distance

Average speed and distance travelled per game, split by offence and defence.

Touches and possession

Number of touches, time of possession, and touches in specific areas such as the paint or elbows.

Drives and passing

Drives to the basket, passes made, potential assists, and assist opportunities.

Shooting context

Catch-and-shoot versus pull-up shooting, shot frequency by defender distance.

Defence and rebounding

Defended field goal percentage, rim protection, and contested and uncontested rebounds.

How to access the data

The league's stats website displays tracking tables by player and team. Analysts commonly retrieve data programmatically using community-maintained Python libraries that call the same endpoints the website uses. Respect usage terms and rate limits, cache results, and store data with dates, since stats are revised and endpoints can change.

Uses in analysis

Tracking stats help explain how a player contributes beyond points, rebounds, and assists. A high number of potential assists with few actual assists may indicate teammates missing shots. Drives and paint touches show how a player creates pressure. Defender distance on shots separates difficult shot-making from open looks. Speed and distance can inform workload and fatigue analysis.

Uses in modelling

In projection models, tracking stats can improve estimates of player roles and usage, especially when lineups change. For example, touches and time of possession help project how usage shifts when a primary ball-handler is out. Careful feature selection matters, since many tracking metrics are correlated with each other.

Limitations

Tracking stats describe what happened, not always why. Defensive metrics depend on matchups and scheme. Small samples early in a season are noisy. Changes in tracking providers or calculation methods over time can make long-term comparisons tricky, so check definitions before comparing seasons.

Beyond the NBA

Other leagues use tracking too, including the NFL's Next Gen Stats from player-worn chips and various soccer tracking systems. The analytical ideas transfer across sports: positions over time reveal movement, spacing, and decision-making that event data cannot.

Storing tracking stats

Store each stat with player ID, team, game or date range, and the definition version. Join with play-by-play, lineups, and odds to analyse how tracking-based roles relate to performance and to market prices.

Example analyses

Analysts compare a player's catch-and-shoot and pull-up shooting to understand shot creation, track how a centre's rim protection changes opponent shooting near the basket, or study how a team's pace and distance travelled change on the second night of back-to-backs. Each combines tracking stats with traditional box score and play-by-play data.

Tracking stats and betting markets

Player prop markets depend on projected minutes and usage. Tracking-derived measures, such as touches and time of possession, can help estimate how a player's role changes when teammates are injured. As with any data, test whether these features improve projections against closing prices before relying on them.

Data quality checks

Check games played and minutes alongside tracking stats, since per-game averages from few games mislead. Compare season totals with sums of game logs to catch missing data. Note mid-season trades, which split a player's stats across teams.

Getting started with tracking data

Pick one question, such as which guards drive most often and how efficient those drives are, pull the relevant tracking tables for a season, join them with box score data, and visualise the results. Expand to multi-season analysis once the pipeline is reliable.

What tracking data means in practice

Use public tracking stats to understand role, movement, and defence, combine them with box scores and lineups, and treat them as context rather than complete measures of value.

This page describes what tracking data contains and is not betting advice.

Frequently asked questions

Is NBA player tracking data public?: Partly. Raw positional data is generally restricted to teams and league partners, but the league publishes many derived tracking statistics, such as speed, distance, touches, drives, and defender distance on shots, on its official stats site for public use. Check the site's terms before reusing data.
How do I get NBA tracking stats in Python?: Community-maintained Python libraries can retrieve the tracking tables published on the league's stats site. Respect rate limits and terms, cache responses, and record when you retrieved data, since endpoints and definitions can change between seasons. Keep requests modest and store raw responses so you can reprocess them later.
What can tracking data tell you that box scores cannot?: It shows how players move, how often they touch the ball, how they create shots, how tightly they are defended, and how they defend others. These details explain role and impact in ways counting stats such as points and rebounds do not.