NBA Efficiency Metrics for Betting: Reading the Numbers Behind the Numbers
Available in US
Table of Contents
- The Bet I Lost Trusting Points Per Game
- Why Per-Game Stats Lie In The Modern NBA
- The Four Numbers That Do Most Of The Work
- Effective Field Goal Percentage And Why It Matters For Props
- Defensive Rating: The Number Books Hide Behind
- The Four Factors And How To Use Them Quickly
- Pace-Adjusted Player Stats For Prop Markets
- How Efficiency Metrics Move Pre-Game Lines
- The Books Aren’t Hiding Anything
- From Stats To A Repeatable Process
The Bet I Lost Trusting Points Per Game
Two seasons back, I was burning a Saturday on a Bucks-Heat over. Vegas had it at 224.5. Both teams averaged 117 points per game on the season. Felt like an over. Smashed it. Final: 207. Loss.
The number that would have saved me wasn’t on the casual stats line. It was offensive rating – points scored per 100 possessions – and the matchup version of it was telling me Miami’s defence drops the Bucks’ offensive rating by about 5 points per 100. With pace, that’s a 6-7 point haircut on team total expectations. Adjusting for that, the line was actually long. Points per game is the surface. Efficiency metrics are what’s underneath.
Eight years on this desk, and the bettors who upgrade from PPG to per-100 numbers are the ones who survive. Here’s the toolkit I use, in language that doesn’t require a stats degree.
Why Per-Game Stats Lie In The Modern NBA
The NBA’s pace this season is 101.9 possessions per 48 minutes – highest in 30 years. League offensive efficiency is 114.3 points per 100 possessions, near the all-time peak of 114.5. The first 10 days averaged 117.7 points per game, third-highest in NBA history and the highest in 64 seasons.
What this means: two teams averaging 117 points per game can be wildly different. A team running 105 pace with a 111 offensive rating produces 117 points. A team running 98 pace with a 119 offensive rating also produces 117. They look identical in the box score, they’re not remotely the same team. Their points-allowed numbers diverge similarly. Per-game stats blend pace and quality together and you can’t unscramble them once you’ve combined them.
The fix is using per-100 metrics. Offensive rating, defensive rating, net rating, and a handful of derivatives. Strip out pace, and what’s left is quality.
The Four Numbers That Do Most Of The Work
Offensive rating is points scored per 100 possessions. Defensive rating is points allowed per 100 possessions. Net rating is the difference. Pace is possessions per 48 minutes.
Multiply pace by offensive rating and divide by 100, you get expected points scored per game in a vacuum. Multiply pace by defensive rating and divide by 100, you get expected points allowed. Two teams meet, you average the offence of one with the defence of the other, repeat for the other side, sum the two – you’ve built a crude total expectation in five minutes of arithmetic.
The crude version is more accurate than most bettors’ instinct. The refined version – adjusting for opponent quality, recent form, and home/road splits – is more accurate still. But you don’t need the refined version to start beating bad lines. You need the discipline to actually do the maths instead of trusting points per game.
Net rating is the closest thing to a “real strength” number a team has. A team at +6 net rating over a 30-game sample is functionally a contender regardless of their record (records can mislead due to clutch luck). A team at -3 net rating with a winning record is a regression candidate. Books know this. Sharp money respects net rating more than win-loss columns.
Effective Field Goal Percentage And Why It Matters For Props
Standard field goal percentage treats a three-pointer as a two-pointer that just happens to be worth more. That’s mathematically wrong. Effective field goal percentage (eFG%) adjusts for the value of threes by giving 1.5 credit for each made three.
A player shooting 45% from the field who takes 6 threes per game and makes 35% of them has a higher eFG% than a player shooting 50% from the field who takes 1 three per game and makes 30%. The first player is more efficient even though he looks worse on the surface stat. For points props, the first player is the safer bet because his expected points per shot is higher.
True shooting percentage goes a step further by including free throws. TS% accounts for the fact that some players draw fouls at high rates and convert from the line. For prop bettors, TS% is the single best volume-adjusted scoring efficiency stat. A player at 60% TS is producing roughly 1.20 points per shooting possession. A player at 52% TS is producing roughly 1.04. The difference is 16% per shot, and it shows up in points props every game.
Defensive Rating: The Number Books Hide Behind
Casual bettors look at points allowed per game and think they know which teams are good defensively. The Bucks allow 110 PPG, the Pacers allow 118 PPG, ergo Bucks defence > Pacers defence. Wrong. The Pacers play at 103 pace, the Bucks at 99. Pace-adjusted, the Pacers might actually have the better defence.
Use defensive rating. Always. The 19% of NBA games decided in the fourth quarter – based on analysis of 2,295 games – are heavily weighted toward defence: clutch stops matter more than blowout sequences. Teams with elite defensive ratings (108-110 range in the modern game) close games out. Teams above 115 defensively get scored on in the clutch and lose games they shouldn’t.
Match-up wise, defensive rating splits matter more than overall numbers. A team can be average defensively but elite at defending pick and roll, terrible against isolation, and decent against post-ups. If you’re betting a Luka Dončić over points, you want to know not just the opponent’s defensive rating but how their defence performs against high-usage iso ball-handlers specifically. Cleaning the Glass and Synergy break this down. Free tier of Cleaning the Glass is enough for most needs.
The Four Factors And How To Use Them Quickly
Dean Oliver’s four factors framework still works. Effective FG%, turnover rate, offensive rebound rate, free throw rate. Both ends. Eight numbers per team. They explain roughly 90% of which team wins a basketball game.
The quickest betting use is for matchup edges. If you bet a team that’s elite on the offensive boards against a team that’s poor on the defensive boards, you’ve identified a likely possession edge. If a team takes care of the ball (low turnover rate offensively) against a team that doesn’t force turnovers, the offensive team gets more shooting possessions. Stack two or three of these factor edges and you’ve found the line.
| Factor | What it measures | Typical league average |
|---|---|---|
| eFG% | Shooting efficiency, weighted for threes | ~54% |
| Turnover rate | Turnovers per 100 possessions | ~13% |
| Offensive rebound rate | Share of own missed shots rebounded | ~26% |
| Free throw rate | Free throws attempted per field goal attempt | ~24% |
The factor that moves totals lines hardest is eFG%. A 2 percentage point swing in expected eFG% between two teams translates to roughly 4-5 points of total. The factor that moves moneylines hardest is turnover rate, because turnovers create transition opportunities and asymmetric possession counts.
Pace-Adjusted Player Stats For Prop Markets
Player counting stats have the same pace problem as team scoring numbers. A player averaging 25 PPG on a 105-pace team isn’t the same as a player averaging 25 PPG on a 96-pace team. The second player is meaningfully better – he’s getting his 25 in fewer possessions.
Per-100 player stats answer this. Points per 100 possessions, rebounds per 100, assists per 100. The numbers are bigger and harder to read at first – a per-100 scoring rate of 35 is “normal star” territory rather than “video game” – but they’re the right currency for projecting future games where pace might be different.
Stats Perform’s Andrew Skweres has talked about how books integrate context – he noted “a sportsbook saw a 1.3× increase in bet count and double the average stake size after they included a contextual stat next to markets”. Per-possession context is exactly the kind of thing books are now showing alongside prop lines, and the bettors who think in those units rather than per-game ones are tracking the same numbers as the books.
How Efficiency Metrics Move Pre-Game Lines
Lines aren’t built from PPG. They’re built from a combination of pace estimates and efficiency estimates, run through a model, then adjusted for closing line value patterns and bookmaker edge. When you see a total of 224.5 for a Suns-Pacers game, the book has run something like: Suns offence (114) × Suns own pace (101) blended with Pacers defence (113) × Pacers own pace (104), reverse for the other side, sum, add bookmaker margin.
Your job as a UK bettor is to do roughly the same exercise and see if your number is far enough from theirs to bet. The 22 December 2025 NBA injury report rules – every 15 minutes during the 11:00-13:00 window for evening tips – give you the inputs reliably enough that the projection becomes worth doing. If a star is out, his minutes get reallocated to teammates with different efficiency profiles, and the team’s overall offensive rating moves.
The Senate Commerce Committee letter from Senators Cruz and Cantwell to Adam Silver in late 2025 pushed the league on integrity issues, and Adam Silver has spoken multiple times about how regulated betting markets help monitor anomalies. UK regulated books are running tighter modelling now than three years ago, which means edge against them is harder to find on basics – but the gap between casual bettors using PPG and the books’ per-100 models is bigger than ever.
The Books Aren’t Hiding Anything
Top UK books – bet365, William Hill, BetVictor, Unibet, Betfred – list 50+ markets per NBA fixture. The Gambling Commission has issued over 770 cease-and-desist notices since April 2024 (262 to operators, 205 to advertisers, with 64,000 URLs removed via Google), and Andrew Rhodes, the UKGC chief executive, has spoken at industry briefings about how operators are widening sports offerings into basketball, NFL, and cricket. Basketball is now mainstream UK product. The numbers behind it are too.
You don’t need access to NBA’s S2 data feed or Genius Sports’ tracking. The free public stats – pace, offensive and defensive ratings, eFG%, the four factors – are sufficient for the questions a UK NBA bettor needs to answer. Cleaning the Glass, Basketball Reference, NBA.com. Free, fast, daily updates.
For a deeper look at how the same per-100 framework applies to setting up totals expectations, my piece on how pace shapes NBA totals lines walks through the same maths from the totals angle.
From Stats To A Repeatable Process
Efficiency metrics aren’t a magic shortcut. They’re a discipline. The bettors who use them well aren’t pulling huge numbers from huge moves – they’re pulling small consistent edges over hundreds of bets across a season. Pace adjusts your expectations. Net rating tells you who’s actually good. eFG% and TS% tell you who scores efficiently. The four factors tell you why. Layer those together with injury news and matchup context, and the prop and totals markets stop looking like noise and start looking like a graph with a few obvious mispricings on it. That’s the goal. That’s the work.
What’s the difference between offensive rating and points per game?
Offensive rating is points scored per 100 possessions, which strips out pace. Points per game blends pace and scoring quality together. For betting, offensive rating is much more useful because it isolates how good a team’s offence actually is.
Is true shooting percentage better than field goal percentage for player props?
Yes. True shooting percentage accounts for the value of three-pointers and free throws, giving you a much better picture of scoring efficiency. A player at 60% TS is producing roughly 1.20 points per shooting possession.
Where can I find pace and efficiency stats for free?
Basketball Reference, NBA.com Stats, and Cleaning the Glass all provide pace, offensive rating, defensive rating, and the four factors at no cost. Cleaning the Glass is the most useful for matchup-specific splits.
Written by the editors at bet of the day nba.
