NBA usage rate tells you how much of a team's offence runs through one player, and that matters for player props because points follow shots. Basketball Reference defines usage percentage as an estimate of the percentage of team plays a player uses while on the floor. When a player's usage goes up, their points line should usually go up too, but by less than the raw numbers suggest, because efficiency tends to drop as a player takes on more. Usage barely touches rebounds, and it doesn't count assists at all.
This guide covers what usage measures, what it leaves out, how to turn a usage change into a projection and a fair price, and which props it really moves. All numbers are invented examples, not selections.
What usage rate measures
A "play" in the usage formula ends one of three ways: a field goal attempt, a trip to the free-throw line, or a turnover. The Basketball Reference formula is:
USG% = 100 × ((FGA + 0.44 × FTA + TOV) × (Team MP / 5)) / (MP × (Team FGA + 0.44 × Team FTA + Team TOV))
In plain terms, it takes the plays a player used, adjusts for the minutes they were on court, and divides by the plays the whole team used. The 0.44 converts free-throw attempts into possessions, since not every free throw is a separate trip. The NBA.com stats glossary gives the same idea: the percentage of team plays used by a player when they are on the floor.
Why 20% is the average
Five players share the floor, and every play is used by one of them. So the five on court add up to roughly 100%, and the average is about 20%. A ball-dominant star might sit near 30% or higher. A corner shooter or a rim-running centre might sit around 15%.
That's the key to reading usage changes. Usage is a share, so if one player's share goes up, someone else's comes down. When a high-usage player leaves the rotation, their plays don't vanish. They get spread across the other four, and rarely evenly.
What usage leaves out
Usage only counts shots, free throws and turnovers. It ignores:
- Assists. A pass-first guard can have modest usage and still run the offence. Assist percentage (AST%) measures that: the share of teammate field goals a player assisted while on the floor.
- Rebounds. Total rebound percentage (TRB%) is the share of available rebounds a player grabbed. Usage doesn't come into it.
- Off-ball work. Screens, spacing and cuts that create shots for others don't show up.
Turnovers do count. A rise in usage includes the extra turnovers that come with handling the ball more, which is one reason efficiency often slips.
Which props usage moves
| Prop | Link to usage | What to check alongside it |
|---|---|---|
| Points | Strong: more plays used means more shots and free throws | Efficiency, minutes, pace |
| Threes | Medium: more shots, but only some are threes | Share of shots from three |
| Assists | Weak: assists aren't in the formula | AST%, who else handles the ball |
| Rebounds | Very weak | TRB%, minutes, opponent's shooting |
| Turnovers | Strong: turnovers are part of usage | Ball-handling role |
| PRA combos | Mostly through the points leg | Each leg separately |
If you price combo props, our guide to PRA betting shows why the legs aren't independent. A usage rise lifts the points leg most, and can even trim assists if the player stops setting others up.
The usage-efficiency trade-off
This is what most usage articles skip. Extra plays aren't free. A player who takes on more of the offence gets harder shots, more defensive attention and less rest on the ball. Analysts have measured this for years.
In a 2011 Basketball-Reference analysis, Neil Paine estimated how much a player's offensive rating changes for each 1% change in usage:
| Starting usage | Change in offensive rating per 1% of usage |
|---|---|
| 18% or less | 1.65 points |
| 18% to 23% | 1.24 points |
| 23% or more | 0.82 points |
Offensive rating is points produced per 100 possessions. The pattern matters more than the exact figures: low-usage players lose the most efficiency when they're asked to do more. Paine made the same point in a 2024 piece, noting that a hyper-efficient role player is efficient partly because their share of possessions is small, and that the link between usage and efficiency has been getting weaker over time. Treat the table as a rough guide, not a law.
Turning usage into a points projection
You can build a simple points projection from four inputs:
- Minutes. How long the player is on court.
- Team plays while on court. Pace (possessions per 48 minutes, as the NBA.com glossary defines it) scaled to those minutes. A regulation game is 48 minutes under NBA Rule No. 5.
- Usage. The share of those plays the player uses.
- Points per play. How efficiently they score on the plays they use.
Worked example (invented)
Player P plays 32 minutes for a team that averages 100 possessions per 48 minutes, with 20% usage and 1.12 points per play.
- Team plays while P is on: 100 × 32 / 48 = 66.7
- Plays P uses: 66.7 × 20% = 13.3
- Projected points: 13.3 × 1.12 = 14.9
Now P's role changes. A teammate drops out of the rotation and you expect P's usage to rise to 24% with the same 32 minutes. The bookmaker has already moved P's points line to 16.5, priced at $1.87 for both over and under.
The naive projection keeps efficiency the same:
- Plays used: 66.7 × 24% = 16.0
- Points: 16.0 × 1.12 = 17.9
The trade-off projection trims efficiency. Using the tiers above, going from 20% to 23% costs about 3 × 1.24 = 3.7 points of offensive rating, and 23% to 24% costs about 0.8 more. That's roughly 4.5 points per 100 plays, or 0.045 per play:
- Points per play: 1.12 minus 0.045 = 1.075
- Points: 16.0 × 1.075 = 17.2
From projection to fair price
To turn a projection into a probability you need a spread of outcomes. Assume, for this example, that P's points land in a bell curve with a standard deviation of 6 points around the projection. Then:
| Naive | Trade-off | |
|---|---|---|
| Projected points | 17.9 | 17.2 |
| Chance of over 16.5 | 59.2% | 54.7% |
| Fair price for the over | $1.69 | $1.83 |
| Expected value at $1.87 | +10.7% | +2.2% |
The same usage guess produces a 10.7% edge or a 2.2% edge depending on one assumption about efficiency. A 2.2% edge is well inside the error of guessing minutes, usage and the spread of outcomes. The honest read is "close to fair, maybe slightly in your favour", not "big value". This is also why a bookmaker can move a line a long way on role news and still have priced it sensibly.
To check the price side of the sum, the margin on a $1.87 / $1.87 market is 6.95%. Our guide to what vig is shows how to strip it out, and the BetBank calculators do the conversion from fair percentage to price for you.
Pace: the input people forget
Usage is a share. Pace sets the size of the pie. If P's game is expected to run at 104 possessions instead of 100, team plays while P is on court go from 66.7 to 69.3. At 24% usage that's about 0.6 extra plays, or roughly 0.7 extra points. Small, but on a line set at 16.5, 0.7 points moves the fair price more than the gap between most bookmakers.
Threes props: usage times shot mix
For a threes line, usage only gets you halfway. Say P's extra usage adds two field goal attempts a game, and 40% of P's shots are threes. That's 0.8 extra three-point attempts. At a 36% hit rate, it's about 0.3 extra threes made. A rise in usage can barely move a 2.5 threes line, while the same rise moves a points line by two or three points. If you bet milestone lines like 3+ threes, our guide to NBA alternate lines shows how small changes in the average shift those prices.
Mistakes to avoid with usage
- Using season usage after a role change. A season average blends games before and after the change. Look at games that match the current role.
- Trusting a tiny sample. Five games at a new usage level is a small sample. Our piece on NBA player prop hit rates explains how wide the uncertainty is over a handful of games.
- Ignoring minutes. High usage in 20 minutes off the bench can produce fewer points than average usage in 36 minutes.
- Assuming efficiency holds. The trade-off table above is the reason to trim.
- Using usage for assists or rebounds. Check AST% and TRB% instead.
- Comparing raw prices across books. Compare fair percentages after the margin is out, on an odds screen, so you're comparing like with like.
Quick checklist
- What's the player's usage in games that match tonight's role?
- How many minutes will they play?
- What's the expected pace of the game?
- How much should efficiency drop on the extra plays?
- Does the prop depend on usage at all (points, threes) or on something else (assists, rebounds)?
- What's the fair price, and does the best available price beat it after the margin?
None of this turns a prop into a sure result. Players have cold nights, coaches change rotations mid-game, and blowouts cut minutes. The aim is to know what a usage change is worth before you look at the price, so you don't overrate the obvious story.
See NBA prop prices side by side
BetBank's NBA odds board puts player prop and game prices from Australian bookmakers side by side, with no-vig fair percentages, line movement and a bet tracker to check your own results over time. First-time members can get the NBA season special at betbank.ai/nba: half price for the first 2 months on a weekly or monthly plan, until 30 November 2026.
BetBank is for adults 18+ only. Odds and outcomes are never certain, so set limits and gamble responsibly. You can exclude yourself from licensed online and phone betting through BetStop, and free support is available 24/7 from Gambling Help Online on 1800 858 858.