NBA Points, Rebounds and Assists Combo Props for UK Punters
Available in US
The PRA line as a single number with three drivers
The points-rebounds-assists prop, or PRA in shorthand, is the prop bettor’s compound bet. One line, three statistical drivers, settled on a single arithmetic total. UK books quote it on most starters every night, and the casual punter often treats it as a “soft” version of the points line – same idea, slightly bigger number. That reading underprices the product. A PRA prop is not a points line plus padding. It is three independent probability distributions added together, and the combined distribution behaves differently from any of its components.
The simple version of why it matters: a points-only prop has a known floor and ceiling driven by usage, minutes and shooting percentage. A PRA prop adds two more distributions on top, each with its own floor, ceiling and variance. The aggregate distribution has wider tails than the points distribution alone, but it also has a more stable middle, because rebounds and assists are partially independent of points. A scorer having a cold shooting night may still rack up assists and rebounds, padding his PRA to the line even when his points fall short.
This is why PRA is the cleanest prop in many UK punters’ books. It absorbs single-stat variance better than any individual category, and the price reflects the combined floor rather than any specific stat. The trade-off is that PRA volatility on the over side is higher than punters expect, because three positive tails can stack into 50-plus performances that would be unthinkable from any single-stat line. SportsLine’s NBA projection model entered the second round of the 2026 NBA Playoffs at 26-9 (74 percent) on top-rated NBA spread picks for the season, and that kind of model-driven approach is exactly what is needed to read PRA distributions accurately, because eyeball pricing struggles with three-variable products.
How books build a PRA prop from minutes and usage
The trading desk does not invent a PRA number from scratch. It builds it from three component projections – points, rebounds, assists – each of which comes from a model output, then sums them with an adjustment for correlation. Knowing the build helps you reverse-engineer the line and spot the components that have been mispriced.
The points projection starts with usage rate and minutes. A starter playing 32 minutes a night with a usage of 24 percent has a points projection that follows a roughly known distribution centred around his season average, with shot quality and opponent defence as adjustments. The rebound projection comes from his per-minute rebounding rate, scaled to projected minutes, then adjusted for opponent rebounding tendencies. The assist projection follows the same pattern: per-minute assist rate, scaled to minutes, adjusted for teammate shooting percentage.
The correlation adjustment is the subtle part. Points and assists correlate negatively – a guard taking more shots usually creates fewer for teammates. Points and rebounds correlate weakly. Rebounds and assists are largely independent. Books fold a small negative-correlation adjustment into the PRA total because of the points-assists relationship, typically 0.5 to 1.0 points off the simple sum of components. A punter who calculates the simple sum and gets, say, 38.5, while the book has the line at 37.5, is not necessarily seeing a misprice – the difference is the correlation adjustment.
The 130-plus international players in the 2025-26 NBA season have expanded the projection complexity. Players moving between systems mid-season – a trade, a coaching change, a role redefinition – have stale per-minute rates that the book’s model takes 5 to 10 games to reset. PRA lines on recently moved or recently re-roled players carry the highest pricing error rates of any prop on the board, and that is the cleanest hunting ground for a punter who tracks role changes carefully.
Why a PRA over is more volatile than a points-only over
The naive read is that more variables mean more variance. The mathematical read is that more variables mean less variance per stat but more aggregate spread. The practical read for a UK punter is that PRA overs hit at the high end harder and miss at the low end more steadily, and that asymmetry has real implications for sizing and selection.
Consider a starter with a points line at 18.5, a rebounds line at 7.5, and an assists line at 4.5, summing to 30.5 – call it a 31.5 PRA line after correlation adjustment. The points-only over hits roughly 50 percent of the time at a fair line. The PRA over also hits roughly 50 percent at a fair line, but the distribution behind that 50 percent is different. A 25-point single-stat night cleared the points by 6.5. A 25-point night plus 6 rebounds and 4 assists clears the PRA by 3.5. A 12-point night with 9 rebounds and 7 assists fails the points by 6.5 but clears the PRA by 1.5. Same player, same night, different bet outcomes depending on how the stats distributed.
The volatility shows up most on big-tail nights. A 35-point single-stat performance clears the points line by 16.5. A 35-point night plus 12 rebounds and 8 assists clears the PRA by 23.5. The over wins bigger when it wins, and the size of the over-win matters for any compound calculation – for instance, if you are using PRA legs in bet builders, the implied combined probability matters more than the individual leg.
What all this means in practice: PRA overs are a higher-volatility play than they look. A run of cold games can break a PRA-heavy punter’s bankroll faster than a run of cold points-only nights, because the PRA bet was sized as if the variance was lower than it actually is. Position-sizing PRA at slightly smaller units than your equivalent points-only bet is the unglamorous correction.
Defensive context: how matchups change PRA splits
The opponent’s defensive scheme is the third invisible variable in every PRA prop, and it shapes the bet more than most UK punters realise. A team that switches every screen punishes off-the-dribble guards on assists. A team that hedges hard on the pick-and-roll bottles up scoring guards on points but leaves rolling big men open for rebounds and finishes. A team that drops conservatively gives shooters open looks but limits transition baskets.
Each scheme changes the component-stat distribution differently. A point guard projecting to 21 points and 7 assists against a switching defence might land 18 and 5 – the points-only line dies, the assists line dies, the PRA over dies. Against a drop defence, the same player projects to 24 and 8 – points line cashes comfortably, assists line cashes, PRA over cashes by a wide margin. Same player, two different opponents, two different PRA worlds.
UK books incorporate matchup data into PRA pricing, but unevenly. The big books have model integrations that adjust for opponent scheme; smaller books may run a flat adjustment by opponent points-allowed without breaking down the scheme details. That uneven application is where line-shopping pays off: the same PRA prop priced 32.5 at one book and 31.5 at another reflects different scheme adjustments, and the over at 31.5 against an exploitable matchup is meaningfully better value than the over at 32.5. A wider read on how possessions per 48 minutes shape the offensive opportunity behind every prop is the supporting framework that makes scheme reads actionable.
Pace is the other matchup variable. A high-pace game produces more possessions, which means more shots, more rebounds, more assists for everyone on the floor. A pace-heavy projection lifts every PRA component upward, and the over price often does not move enough to compensate. Pace-heavy matchups against weak defences are the cleanest PRA over spots on most nights, and the cleanest PRA under spots are pace-light playoff games or matchups against elite defences.
A worked £10 PRA bet at 5/6
Let me walk through a concrete example, because the numbers anchor everything we have just discussed. Imagine a starting forward with a points projection of 18.5, rebounds 7.0, assists 3.5. Simple sum: 29.0. With correlation adjustment, the book posts the line at 28.5, and the over is priced at 5/6.
The 5/6 quote is decimal 1.83. Implied probability after margin: 54.6 percent. A £10 stake returns £8.33 plus stake on a win, loses £10 on a loss. Break-even probability is 54.6 percent. So the bet is positive expectation only if the true probability of clearing 28.5 is higher than 54.6 percent.
Now run the matchup check. The opponent allows 14 percent more rebounds to forwards than the league average. The pace projects 4 possessions higher than the player’s season average. The defensive scheme switches everything, which marginally suppresses assists but boosts rebound chances on the back side. Adjusted projection: 19 points, 8 rebounds, 3 assists. New simple sum: 30.0. With correlation: 29.5. The over of 28.5 looks comfortable.
The implied true probability of clearing 28.5, given the adjusted projection of 29.5 and historical variance, is around 60 percent. Expected return per £10 stake: (£8.33 × 0.6) – (£10 × 0.4) = £5.00 – £4.00 = £1.00 of expected profit per £10. That is a 10 percent edge, which is excellent over a long sample.
The exercise looks simple in retrospect. The work is in actually doing the matchup check and the variance read consistently – not in finding the line, but in deciding which lines deserve the analytical effort and which to skip. PRA props at fair prices are everywhere; PRA props with genuine edge are the ones where the matchup and pace adjustments push the projection past the line by 1.0 or more, and those are perhaps one in eight on a normal night’s slate.
Does a PRA prop settle on regulation or include overtime on UK books?
UK books almost universally settle PRA props on the final box-score numbers, including overtime. That is worth knowing on close games where a starter playing extra minutes in OT can push a PRA over the line that would have failed in regulation. The exception is a small handful of edge-case markets that specify regulation only, and the settlement rule sits in the prop terms before you place the bet.
Which is the better value: a single-stat over or the same number split into a PRA combo?
It depends on the player’s stat distribution. A balanced player whose three stats all contribute regularly tends to offer better PRA-over value because the combined floor is more reliable than any single category. A specialist whose production concentrates in one stat – a high-scoring guard with low rebounds, for example – usually offers better single-stat value, because the PRA combo line gets dragged down by his quieter categories.
Written by the editors at nba Betting Discussion.
