NBA Prop Bet Hit Rates — How Often Each Market Cashes

Editorial Team «nba Best Player Prop Bets» June 8, 2026 Reading time: 18 min

Why one market cashes 70% and another barely clears 55%

The first time I ran a clean dataset on NBA prop hit rates, I expected modest spreads. Maybe two or three percentage points between markets. What I actually found stopped me mid-coffee. A graded set of 10,580 predictions from the 2025–2026 season showed blocks props cashing at 69.9% while points props limped in at 55.7%. That is a 14-point gap between two markets that, on the surface, look like the same kind of bet — pick over or under, see what happens.

Most punters in the UK never see that gap. They scroll the bet builder, grab a points line on whichever star is trending, and call it research. So this article is the opposite of a tip sheet. I will walk you through every major NBA prop market with a real strike rate attached, explain why those numbers settle where they do, and show you how to plug the data into your own approach without falling for the obvious traps. Hit rates are not a shortcut to value. They are a starting map.

Throughout the piece I am quoting decimal odds, UK terminology, and the kind of bookmaker behaviour you actually meet on a UKGC-licensed site. American prop content tends to assume FanDuel logic. I am writing for someone holding a phone with bet365 or Sky Bet open.

How the hit rate dataset is built

I want to be straightforward about what these numbers represent before we go any further, because half the prop content online quotes hit rates without ever explaining the engine behind them. The 10,580 graded predictions I keep returning to come from a model that ingests projections, line movement and final box score data, then settles every prop as a binary win or loss. No half-credits. No pushes muddying the picture. If a player projected for 22.5 points scored 23, the over wins; if they scored 22, the under wins.

That is the cleanest format for comparing markets, but it has limits worth knowing. A 69.9% strike rate on blocks does not mean every blocks prop you bet will cash seven times in ten. It means that across thousands of model-tagged value bets, the over-under outcome aligned with the model 69.9% of the time. Bet without a model and you are running a different distribution.

The other thing I always remind people: sample size matters more than hit rate. Five hundred graded blocks props is a meaningful base. Fifty is a coffee chat. When you are reading hit rate content elsewhere, the first question to ask is how many predictions sit behind the percentage. If the answer is hidden, treat the headline number as marketing.

One more frame before the market-by-market breakdown. Hit rate tells you how often a prop cashed in retrospect. It does not tell you whether the price was good. A 65% hit rate market priced at 1.50 is a winning system; the same hit rate priced at 1.40 is a slow leak. Pair every percentage in the next sections with the vig discussion later in the piece, otherwise you will draw the wrong conclusions.

I treat any market hit rate above 60% as worth investigating, and anything from 55 to 60% as a market where line shopping and projection accuracy do most of the work. Below 55%, you are in casino territory — the hit rate alone will not save you, and the pricing usually reflects that.

Blocks props — the highest hit rate on the board

Blocks props sit at the top of the hit rate table at 69.9% based on the 2025–2026 sample, and that figure surprises almost everyone I show it to. The instinctive read is that blocks are random, low-frequency events — a league average of about 4.5 per game per team — so why would the over-under settle so decisively?

The answer sits in distribution shape. Blocks props are typically set at 0.5 or 1.5 for most rotation players. That means the bet is essentially binary in the cleanest possible way: did the player record at least one block, or at least two? Players who block shots for a living — your Wembanyamas, your Mobleys — produce them in clusters tied to defensive scheme, opponent shot profile, and minutes. The prop line tends to lag behind the actual rate because books are conservative on counting stats with high variance, and that conservatism creates the value.

The trade-off is volume. You will not find blocks props on every player, every night. UK bookmakers offer the market for established rim protectors and a thin pool of switchable wings. Outside that pool, the market either does not exist or is priced so low it carries no edge. So 69.9% is real, but the addressable universe is narrow. I treat blocks as a precision tool, not a daily staple.

One more practical note. Blocks props are sensitive to opponent style in a way that points props are not. A team that takes 45 three-pointers a game generates fewer block opportunities than a team that drives to the rim 50 times. When you are shopping a blocks line, the opponent profile matters more than the player’s average — and that is where a lot of the residual edge lives.

Three-pointers — the second-highest cashing market

Threes props cashed at 63.2% in the same dataset, which puts them firmly in the high-strike-rate tier alongside blocks. The logic differs, though, and that difference is where research pays off.

Three-point attempts are far more democratic than blocks. Just about every NBA rotation player takes at least a couple per game, which means UK bookmakers post threes lines on 12 to 14 players a side. That is a much wider menu, and wider menus tend to carry softer prices on the back end of the depth chart. Stars get sharp lines. The eighth man on a 100-pace team going up against a defence that gives up volume from the corners — that is where lines stay stale.

I rate three-pointer props as the most actionable high-hit-rate market in the NBA, and the reason is sustained over-under separation. A player averaging 2.4 threes a game with a 2.5 line might look 50-50 on paper, but pace adjustments and matchup tilts can push the expected output 20 to 30% in either direction on any given night. A 2.4 average against a switching defence on a 105-pace night is not 2.4 — it is closer to 2.9. The line does not always move with that math, and the over wins more often than the headline average suggests.

The catch with threes is volatility. A player can take ten attempts and shoot 1-of-10. Hit rate of 63.2% across thousands of predictions is reliable; a five-bet sample on threes can swing wildly in either direction. If you are using this market to build a steady edge, sample size discipline matters as much as anything else.

Steals props — the third high-percentage tier

Steals come in at 61.9%, just below threes. The two markets share more DNA than people realise. Both are high-variance counting stats with relatively low daily totals, and both tend to be priced conservatively because the books cannot move limits the way they can on points or assists.

Where steals differ is the player pool. Top-of-the-line steals lines belong to a smaller crew — perimeter defenders with quick hands, point guards with ball-pressure instincts, a handful of forwards who anticipate passing lanes. Outside that pool, the lines are 0.5 or 1.5 with low odds on either side, and the margin gets eaten by the vig.

What the 61.9% number does well is reframe how you should think about steals as a market. It is not random noise. There is a real edge to be captured, but the addressable bet count per night is small. I usually find one or two steals plays a slate, never ten. If you are betting more than that, you have moved from finding edges to filling a quota.

Assists props — where research separates winners from losers

Assists props cashed at 57.6% in the dataset, which puts them in the middle tier. That is interesting, because assists feel like they should behave more like blocks: a stable counting stat, tied to specific roles, with predictable distribution shapes. The reality is messier.

The reason assists props sit lower than blocks or threes is that books have far more data on lead playmakers. The market for primary ball handlers is sharp. Brunson, Doncic, SGA, Haliburton — the lines on those players move within tight windows, and the vig eats most of the edge. Where assists props open up is the secondary playmaker tier: the wing who runs the second unit, the rookie point guard absorbing minutes when the star sits, the centre who has quietly become a hub passer.

Usage rate matters more here than for almost any other market. A player with USG% of 28 and a 7.5-assist line is materially different from a player with USG% of 22 and the same line. Minutes, usage and matchup combined can move expected assist output by 20 to 30% in either direction, and that swing is exactly the band where a 57.6% market becomes profitable for a researched bet and unprofitable for a coin-flip bet.

One trap I see often: punters bet assists under on a star whose team is missing key shooters. Logic is that fewer shooters means fewer easy assists. Sometimes true, sometimes not — when your best playmaker is the only creator left, his assist rate can spike rather than drop because he is touching the ball on every possession. Read the box score effect, not the headline injury news.

Rebounds props — narrower distribution than you expect

Rebounds props cashed at 57.3%, basically tied with assists. The interesting feature of the rebounds market is how tight the distribution is for established players. A starting centre averaging 11 rebounds a game will land between 8 and 14 in roughly 70% of his appearances. The line typically sits at 10.5 or 11.5, which means the bet usually comes down to whether the player is on the high or low end of his typical range.

That tightness is both why the hit rate is moderate and why the market has a research ceiling. There is less ground to make up than in three-pointers, where a hot shooting night can push a player far past his average. Rebounds tend to settle near the mean.

The exceptions are the situations that matter most for prop research. A team missing its starting centre creates a scramble for boards that benefits whichever forward absorbs the minutes. A pace differential can swing total rebound opportunities by ten or more in a single game. And opponent profile, particularly opponent shooting accuracy, has a bigger effect on rebounds than people credit — a team that shoots poorly creates more long boards, period.

I treat rebounds as a workmanlike market. Solid hit rate, low volatility, decent vig, and best used when you have a specific scenario in mind rather than as a daily go-to.

Points props — the lowest-cashing main market

Here is the line that usually gets the strongest reaction from people in this niche: points props cashed at 55.7% in the dataset. That is the lowest figure of the six main markets, and it sits there for the most predictable reason in sports betting — points are the most popular prop, and popularity tightens lines.

Points over-under is the single biggest prop market in the NBA, accounting for somewhere between 35 and 40% of all prop handle on major books. Books pour their sharpest pricing into the most-bet markets. They have to. A loose points line on a star bleeds money in five-figure tranches within an hour of opening. So the lines on stars are rigorously priced, the vig is competitive but not generous, and the residual edge is small.

What that means in practice: if your prop strategy is built around points props on top-tier names, your hit rate will struggle to clear break-even after vig. The 55.7% figure is the model’s hit rate against settled outcomes. Subtract bookmaker margin (4.76% on a standard –110/–110 line, which is 1.91/1.91 in decimal) and the buffer over a coin flip is razor-thin.

This does not mean points props are unbettable. It means the pool you should be working in is narrower than for blocks or threes. The edges sit on second-tier scorers in pace-up situations, on stars facing defensive matchups the public has not properly priced, and in alt-line points where the standard line and your projection diverge enough to justify a stretch. Bet points the same way you bet first basket props — selectively, with a specific theory, never as filler.

Why hit rates separate by 14 percentage points

Looking at the table — 69.9% on blocks down to 55.7% on points — the obvious question is what produces that spread. The answer involves three intertwined factors, and understanding them changes how you read every prop line.

The first is market depth. Books direct sharper pricing toward markets with higher handle. Points props get the surgical treatment; blocks props are priced more conservatively because the volume is lower and the variance harder to model. As one US-based handicapper observed about prop markets, books are often slower to react on player props than on sides and totals, partly because the limits are smaller and partly because the algorithms producing the lines can be beaten with proper information. That observation maps directly onto the hit rate table.

The second factor is data resolution. Points and assists have decades of granular tracking. Blocks and steals have tracking too, but the underlying events are rare enough that a single role change or scheme tweak can scramble a player’s true rate for weeks. Books cannot adjust as quickly as reality does, and that lag becomes the edge.

The third is line construction. A standard –110/–110 line in decimal is roughly 1.91 each side, with implied probability of 52.4% per side and a 4.76% overround. That overround is the bookmaker’s margin. On points, the books can run very tight overrounds because the volume justifies it. On blocks, overrounds are wider. Higher hit rate on blocks does not necessarily mean higher long-term return — wider vig eats some of the edge — but the combination of model alignment and bookmaker conservatism is real.

None of this is a free lunch. Higher hit rate markets have lower volume, smaller pools of viable players, and tighter daily slates. You cannot scale a blocks-only strategy to twenty bets a night. You can build a discipline around finding the three or four highest-conviction opportunities across all six markets, weighted by hit rate and pricing.

Putting hit rates to work in your own research

I want to close with the practical part, because the hit rate table is useless if you do not know how to apply it. I treat the percentages as a tilt mechanism, not a betting menu. They tell me where to look first when I open a slate, not what to bet.

The order I go in is roughly: blocks first, threes second, steals third, then assists, rebounds and points based on specific situations rather than market preference. The first three are where I expect to find soft pricing on a typical night. The bottom three are where I need a story — a pace differential, a defensive matchup, a usage spike — before I take a position.

Inside each market, I run the same checklist regardless of hit rate: what are the projected minutes, what is the player’s usage rate, what does the opponent’s pace and defence-vs-position number tell me, and what does the line look like compared to my own projection after stripping the vig. The five-factor frame is the same; the hit rates just point me at the most efficient places to start.

If you want a deeper walk through the projection-and-pricing framework — minutes, usage, pace, defence-vs-position and no-vig math — that is the next piece I would read after this one. I built it as a companion to the hit rate data, and it covers the math side that hit rates alone cannot answer. You can find it in the NBA player prop strategy guide.

One last thing. Hit rates change. The 2025–2026 figures will not be identical to the 2026–2027 figures, because rule changes, rotation patterns and rule-prompted market restrictions all shift the underlying distributions. Treat this article as a snapshot of where the markets sit right now, not as a permanent ranking. The relative order — blocks high, points low — has been stable for several seasons, but the magnitudes drift, and any researcher worth the name updates their priors at least once a season.

Created by the ”nba Best Player Prop Bets” editorial team.