One Stat Tracks Offensive Endings; the Other Tracks Ball Contact

NBA usage rate and touches measure different parts of offense. Usage rate estimates a player’s share of shooting, free-throw, and turnover responsibility while on the court; touches count how often the player receives or contacts the ball within the tracking system. A player can therefore touch the ball frequently without recording many usage-counted events—or generate a shot, free-throw action, or turnover after only a brief touch.

The short answer: usage measures endings, touches measure involvement

The key distinction is offensive-chance responsibility versus contact frequency. USG% focuses on field-goal attempts, free-throw activity, and turnovers attributed to a player. Touches focus on how often the ball reaches that player, whether the player shoots, passes, dribbles, or quickly moves it elsewhere.

The NBA Stats Glossary defines usage, touches, and potential assists as separate metrics. They answer different questions and use different units and denominators, so they should not be treated as interchangeable versions of “ball dominance.”

Comparison	Usage rate (USG%)	Touches
Core question	Who assumed the team’s shot, free-throw, and turnover responsibility while on the floor?	How often did the player receive or contact the ball?
Counted activity	Field-goal attempts, free-throw activity, and turnovers	Tracked instances of ball contact
Unit	Percentage	Count, commonly shown as a total or per game
Major omissions	Routine passes, assists, screens, spacing, and ordinary handling	How long the ball was held and what happened during the touch
Best use case	Comparing possession-ending workload and offensive chances taken	Comparing frequency of on-ball involvement

A high number in one column does not require a high number in the other. An initiator may receive the ball on nearly every trip, organize the set, and pass without taking the shot. A finisher may touch it once near the basket and immediately record a usage-counted event.

What usage rate counts—and what it leaves out

Standard usage descriptions center on three types of box-score activity:

• Field-goal attempts
• Free-throw activity
• Turnovers

An ordinary pass does not directly increase standard USG%. Neither does an assist simply because the player handled the ball and created a shot for someone else. If that player commits a turnover, however, the turnover contributes to usage because responsibility for the lost offensive chance is assigned to that player.

This makes usage a useful estimate of workload, but not a perfectly literal count of team possessions. Common calculations infer offensive chances from box-score events, and free-throw activity requires an adjustment because not every free throw represents a separate possession or chance. The exact implementation can vary among data providers.

Offensive rebounds create another complication. A team can miss, recover the ball, and shoot again without surrendering possession. That sequence contains multiple shooting chances within one continuing possession. USG% is therefore better described as an estimated share of offensive chances or possession-ending responsibility than as an exact share of literal possessions.

Usage also omits actions that may help produce the eventual shot:

• Screens
• Spacing
• Off-ball cuts and movement
• Defensive attention or “gravity”
• Quick swing passes
• Passes that create an advantage before the assist
• Decoy actions that open space elsewhere

USG% consequently measures workload or offensive-chance responsibility, not total offensive contribution. It does not establish whether a player was efficient, selfish, valuable, or harmful. Those judgments require additional evidence about scoring efficiency, turnovers, playmaking, role, and team performance.

What the NBA touches table adds

Touches reveal how frequently a player becomes involved with the ball. By themselves, however, they do not show how long each contact lasted or what the player did with it. A one-second catch-and-pass and a prolonged isolation can each begin with one touch.

The NBA’s player touches table adds several fields that help separate those behaviors:

• Total touches and frontcourt touches
• Time of possession
• Average seconds per touch
• Average dribbles per touch
• Points per touch
• Paint, elbow, and post touch categories

Two players can record similar touch totals while producing very different possession-time figures. One might hold and dribble before initiating an action; the other might make rapid decisions and keep the ball moving. Average seconds and dribbles per touch reveal that difference more clearly than raw touch count.

Touches are also distinct from potential assists. A touch records ball contact. If the question concerns shot creation for teammates, potential assists and other passing statistics are more relevant than touch totals alone.

The NBA table is filter-dependent and attributes its tracking data to Second Spectrum. Values change with the season, season type, date range, player filters, and totals-versus-per-game setting. The table also notes that player tracking is not available for every game.

Points per touch can be informative, but it is not a complete efficiency measure. Post scorers, transition finishers, point guards, and low-volume connectors receive different types of touches in different locations. Their points-per-touch figures reflect those roles as well as execution.

A worked possession showing why the metrics diverge

Consider this hypothetical possession:

1. The point guard brings the ball up and passes to a connector.
2. The connector immediately swings the ball to a wing.
3. The wing shoots.

Player	Touch recorded?	Usage-counted event?	What happened
Point guard	Yes	No	Advanced the ball and passed
Connector	Yes	No	Made a quick pass
Wing	Yes	Yes	Attempted the shot

All three players can receive a touch, but the wing records the shot-based usage event. The earlier passes matter to the offense without directly increasing either passer’s standard USG%.

Now consider a second hypothetical outcome. The point guard receives the ball several times, resets the action, uses a screen, and makes the final pass before a teammate shoots. The guard can accumulate multiple touches and substantial possession time without recording a shot, free-throw action, or turnover. If the final pass meets the tracking definition, it may produce a potential assist.

In a third hypothetical outcome, a cutter receives one pass and immediately shoots or draws free throws. That player may have only one short touch yet still record a usage-counted event.

These examples suggest three practical role descriptions:

• High-touch facilitator: Receives and controls the ball frequently but often passes before the shot or turnover.
• Quick-decision connector: Records touches while moving the ball with little time or dribbling per touch.
• High-usage finisher: Takes a substantial share of shots, generates free-throw activity, or commits turnovers, sometimes after brief touches.

These are analytical archetypes, not official player classifications. A player can move among them depending on lineup, matchup, and assigned role.

Choose the metric that matches the basketball question

Start with the question rather than searching for one statistic that appears to describe everything.

Basketball question	Primary metric	Useful companion
Who took the largest share of offensive chances?	USG%	Field-goal attempts, free throws, turnover rate
Who received the ball most often?	Touches	Frontcourt or location-specific touches
Who controlled the ball longest?	Time of possession	Seconds and dribbles per touch
Who made quick decisions?	Seconds or dribbles per touch	Touches and passing data
Who created possible shots for teammates?	Potential assists or passing statistics	Assists and turnovers
Was the player’s workload productive?	Scoring efficiency and turnovers	USG% and lineup offense

Use USG% to compare responsibility for field-goal attempts, free-throw activity, and turnovers while players were on the court. Use touches to compare how frequently they became involved with the ball.

For control duration, add time of possession. For handling style and decision speed, examine average seconds and dribbles per touch. For teammate shot creation, use potential assists and passing statistics rather than assuming that every touch created an advantage.

When evaluating productivity, pair usage with scoring efficiency and turnover measures. A large workload can be efficient or inefficient; usage alone cannot make that distinction. Lineup and team results add context, but descriptive relationships do not prove that one player’s style caused those results.

None of these measures fully captures screening, spacing, cuts, gravity, or other off-ball contributions.

How to compare players without mixing incompatible data

A fair NBA usage rate versus touches comparison requires matched settings. Use the same:

• Season and season type
• Date range
• Player pool
• Team or opponent filters, where applicable
• Totals or per-game display
• Games and minutes criteria

Do not compare one player’s total touches with another player’s touches per game. Account for games and minutes played, and disclose any minimum-playing-time threshold. No single cutoff is automatically correct for every analytical question.

A lead guard may receive more touches because the offense requires that player to advance the ball. A faster team may create more opportunities for all five players. State whether the comparison uses totals, per game, per minute, per team possession, or another consistent basis.

Tracking tables can be incomplete or shaped by active filters. Before drawing conclusions, verify the season type, date window, display mode, and intended player pool. Remember that tracking is not available for every game and that an on-screen table may show only part of the available dataset.

Modern tracking touches should not be mixed with older or alternative definitions. Historical manual charting sometimes counted only contacts after the ball reached an attacking position. For example, an 82games charting method excluded certain rebound and backcourt contacts. Other systems estimate touches from box-score events rather than observing them directly. Those measures may be useful for their original purpose, but they are not equivalent to current Second Spectrum tracking.

Finally, do not assume a fixed correlation between touches and USG%. Establishing one would require a matched dataset covering the same players, games, minutes, filters, and definitions.

Why a passing-inclusive usage metric can double-count a possession

It may seem appealing to combine usage, touches, and potential assists into one offensive-load number. The problem is that these metrics can describe overlapping stages of the same possession.

Consider a revised hypothetical sequence:

1. A creator receives a touch and passes to a connector.
2. The connector returns the ball, giving the creator another touch.
3. The creator makes the final pass to a shooter.
4. The shooter receives a touch and attempts the shot.

The creator may receive credit for touches and, if the final pass qualifies, a potential assist. The connector records another touch. The shooter records a touch and the shot-based usage event. All of that activity occurs within one possession.

Adding those entries without an explicit weighting system and denominator treats overlapping stages of the same possession as separate workload events. The result may measure a chosen form of activity, but it is not automatically a clean share of possessions.

A fan-created Reddit proposal called LOAD experiments with adding passing activity to possession-ending usage and acknowledges the double-counting issue. It is not an official or validated NBA statistic.

A composite can still be useful for a narrowly defined research purpose. Its creator must explain what each component represents, how overlapping events are weighted, and what the denominator means. The result cannot then be interpreted as standard USG%, tracked touches, or an exact share of team possessions.

For most comparisons, a small dashboard is clearer: use usage rate for offensive-chance responsibility, touches for ball-contact frequency, possession-time measures for control duration, and potential assists or passing data for creation. Evaluate effectiveness separately with efficiency, turnovers, and lineup context.

Frequently asked questions

Does a 20% usage rate mean a player used exactly one-fifth of all possessions?

Not exactly. With five players on the court, 20% provides the intuitive baseline of an even one-fifth share, but it is not a universal benchmark for player quality. The Hoopsalytics explanation of usage rate uses 20% to illustrate that even five-player split.

USG% is generally estimated from field-goal attempts, free-throw activity, turnovers, team events, and minutes. Offensive rebounds and free-throw adjustments mean offensive chances do not always correspond one-for-one with literal possessions.

It is best read as an estimated share of the team’s relevant offensive chances while the player was on the floor—not proof that the player controlled the ball on exactly one-fifth of possessions.

Can touches be estimated from a box score if tracking data are unavailable?

Yes, but the result is an estimate under a different definition. One published box-score touch estimate combines field-goal attempts, turnovers, adjusted free-throw activity, and an assist-based estimate of passes in attacking positions.

Such a formula depends on assumptions and is not equivalent to directly tracked Second Spectrum touches. Estimated touches should be labeled clearly, with the formula and definition disclosed, and should not be combined with modern tracking counts as though they measured the same thing.