An NBA player prop hit rate is the share of recent games in which a player went over (or under) a line: 7 overs from the last 10 games is a 70% hit rate. It tells you how often something happened, not whether today's price is worth taking. To use one, compare it with the break-even chance built into the price (at 1.87 that's 53.5%), then ask how much a small sample can really tell you. Seven from ten sounds strong, but it's consistent with a true rate anywhere from about 40% to 89%. This guide shows the maths with invented numbers and covers the four things a hit rate leaves out.
The 2026-27 NBA regular season starts on Tuesday 20 October US time, with each team playing 82 games through to 11 April 2027 (Fox Sports). Expect hit rates everywhere on prop screens from the first week, and expect them to be built on very few games.
What a player prop hit rate is
Take an invented player, Player A, with a points line of 22.5. A prop screen might show:
| Window | Overs | Hit rate |
|---|---|---|
| Last 5 | 4 of 5 | 80% |
| Last 10 | 7 of 10 | 70% |
| Last 20 | 12 of 20 | 60% |
| Last season | 44 of 78 | 56% |
The line ends in .5, so there are no pushes: every game is either over or under. Most screens also split by home and away, by opponent, and by whether the player started.
Notice the pattern above: the shorter the window, the more extreme the number. That's normal, and it's the first clue that the short windows are mostly noise.
Start with the price: your break-even number
A hit rate means nothing without the odds next to it. The price tells you the chance a bet has to win just to break even, called the implied probability. It's 1 divided by the decimal odds, which our implied probability guide covers in full.
| Decimal odds | Break-even chance |
|---|---|
| 1.70 | 58.8% |
| 1.80 | 55.6% |
| 1.87 | 53.5% |
| 1.90 | 52.6% |
| 2.00 | 50.0% |
| 2.10 | 47.6% |
Props usually carry a bigger margin than head-to-head markets. If both the over and the under on Player A are 1.87, the two implied chances add to 106.95%, so the bookmaker's margin is just under 7%. Our guide to vig explains where that comes from.
Expected value per dollar is (true chance × decimal odds) minus 1. At 1.87:
- If the true chance is 50%, EV is -6.5%
- If it's 55%, EV is +2.85%
- If it's 58%, EV is +8.46%
So the real question isn't "is 70% high?" It's "is the true chance above 53.5%, and by how much?" A hit rate is one rough piece of evidence about that, and a noisy one.
Why 7 from 10 tells you less than it looks
Ten games is a tiny sample. A standard way to put a range on a success rate from a small sample is the Wilson score interval, which statisticians use because the simpler normal approximation breaks down on small samples (Wikipedia). Here are 95% ranges for some common hit rates:
| Record | Hit rate | Plausible true rate (95%) |
|---|---|---|
| 6 from 10 | 60% | 31% to 83% |
| 7 from 10 | 70% | 40% to 89% |
| 8 from 10 | 80% | 49% to 94% |
| 14 from 20 | 70% | 48% to 86% |
| 35 from 50 | 70% | 56% to 81% |
Every 10-game range in that table includes 53.5%, the break-even chance at 1.87. Even 8 from 10 can't rule out a coin flip. It takes something like 35 from 50 before the whole range sits above break-even, and by then the player's role may have changed.
Short streaks also fade for a plain statistical reason. Regression toward the mean says that if one sample is extreme, the next is likely to be closer to the average (Wikipedia). A player who has cleared their line 8 times in 10 has probably had some luck on top of whatever real form they're in.
The scan problem: someone always looks hot
Prop screens let you sort hundreds of lines by hit rate. That creates a second trap, known in statistics as the multiple comparisons problem: run enough tests on the same data and some will look significant by chance (Wikipedia).
Here's how big the effect is. Imagine 300 player props on a busy NBA night, and suppose every one of them is a true 50/50.
- The chance any single prop shows 8 or more overs from 10 is 5.5%. Across 300 props, that's about 16 props showing 80% or better from luck alone.
- The chance of 7 or more from 10 is 17.2%. That's about 52 props showing 70% or better.
Sort the screen by last-10 hit rate and those lucky streaks rise straight to the top. The list looks full of strong trends, even in a market where none exist.
Four things a hit rate leaves out
1. The line moved
A hit rate is usually worked out against today's line. But the player may have had a different line in most of those games. If Player A's points line was 19.5 for six games and was raised to 22.5 after a hot run, the bookmakers have already seen the same form you're looking at and priced it in. A 70% hit rate at today's line partly measures games where nobody expected them to reach it.
2. Minutes and role
Points, rebounds and assists follow minutes. A sample that mixes 36-minute games as a starter with 24-minute games off the bench is really two samples blended together. Before trusting a hit rate, check whether the minutes in the sample match the minutes the player is likely to get tonight. A teammate's injury can lift a player's role overnight, and a return from injury can shrink it.
3. Injury news and rest
NBA teams must list a participation status for any player who may be affected by injury, illness or rest by 5 p.m. local time the day before a game. For the second game of a back-to-back, the deadline is 1 p.m. local time on game day, and teams also file a game-day report between 11 a.m. and 1 p.m. local time (NBA Official). With most games played on US evenings, that news usually lands overnight or early in the morning in Australia, and prop lines can move sharply when it does. A hit rate built last week knows nothing about it.
4. Opponent and head-to-head splits
Opponent splits look specific, but teams only meet a few times a season. "3 from 3 against this opponent" is three games. Treat head-to-head hit rates as trivia, not evidence.
A worked example (invented numbers)
Player B has a rebounds line of 9.5. The screen shows 8 overs from the last 10 (80%), 13 from the last 20 (65%) and 46 from 80 last season (57.5%).
One bookmaker prices the over at 1.70 and the under at 2.10. Take the margin out:
- Implied chances: 1 ÷ 1.70 = 58.8%, and 1 ÷ 2.10 = 47.6%. Together that's 106.4%.
- Scale back to 100%: the over's fair chance is 58.8% ÷ 1.064 = 55.26%.
That fair chance sits much closer to the season hit rate than the last-10 one. Now test two prices for the over:
- At 1.75 from another bookmaker: 0.5526 × 1.75 minus 1 = -3.3%
- At 1.85 from a third: 0.5526 × 1.85 minus 1 = +2.2%
Same player, same 80% hit rate, two different answers. The hit rate is identical at every bookmaker; the price isn't. The 55.26% is still only an estimate drawn from one bookmaker's prices, so even the +2.2% is a judgement, not a certainty, and any single bet can lose.
How to read a hit rate, step by step
- Start with the price. Turn the odds into a break-even chance before you look at any trend.
- Check the sample. Prefer 20 games or more, and remember how wide the range is on 10.
- Check the line history. Was the line the same in those games, or has it already moved up?
- Check minutes and role. Do the games in the sample look like tonight's game?
- Read the injury report. Know when the next update is due and whether the line has already reacted.
- Shop the price. An odds screen shows every bookmaker's price on the same prop side by side.
- Track the result and the closing price. Over a season, your closing line value tells you more about your process than your hit rate does. A bet tracker makes that easy to check.
For more on how props are priced and which markets you'll see, see our prop betting guide.
NBA props on BetBank
BetBank's NBA board shows points, rebounds, assists, threes and combo props across Australian bookmakers with hit rates alongside, plus line movement, expected value against fair odds on the Positive EV tool, and a bet tracker that records closing line value. The hit rate sits next to the price, so you can see both at once.
For the new season, first-time members on a weekly or monthly Silver or Gold plan pay half price for their first 2 months after the 7-day free trial, until 30 November 2026. See the NBA season page.
18+ only. Hit rates describe the past, not what happens next, and no tool makes an outcome certain. Please gamble responsibly: if it stops being fun, talk to Gambling Help Online on 1800 858 858, see our responsible gambling page, or self-exclude through BetStop.