eFG%, True Shooting and Net Rating: Advanced Stats for Bettors
Available in US
The shift in my betting that produced the biggest profitability jump was not a strategy change. It was an information change. I stopped reading box scores and started reading the four or five advanced metrics that NBA analytics shops have been using for fifteen years to evaluate teams. Once eFG%, true shooting percentage, and net rating became the variables I checked first, my pre-game projections got noticeably sharper and my bet selection got more disciplined.
Standard box-score statistics – points per game, field goal percentage, rebounds – are the metrics that sportsbook marketing pages still reference because they are intuitive. They are also misleading in specific, predictable ways. The advanced metrics correct for those distortions, and the bettors who use them are reading the game on the same wavelength as the operators’ trading teams.
Effective Field Goal Percentage
Standard field goal percentage treats a two-pointer and a three-pointer as equivalent makes. A team that shoots 50 percent from two-point range and 35 percent from three-point range gets a blended field goal percentage that looks unimpressive next to a team shooting 52 percent overall, even though the three-point shooting team is producing significantly more points per shot attempted.
Effective field goal percentage corrects for this. The formula adds half a credit to each made three-pointer, reflecting that a three is worth 1.5 times as much as a two. A team shooting 40 percent from three is actually producing 60 percent equivalent two-point efficiency – substantially better than a team shooting 50 percent only from two. eFG% surfaces that reality in a single number.
For betting purposes, eFG% matters because it is the most accurate predictor of offensive efficiency available in a single metric. Teams with high eFG% generate more points per possession, which means they cover spreads more consistently and contribute to total overs more reliably. Teams with low eFG% – typically defensive specialists or rebuilding rosters with limited shooting – produce the opposite pattern.
The trap with eFG% is treating it as a fixed property when it is in fact heavily matchup-dependent. A team with a great eFG% against the league average may shoot dramatically worse against an elite defence, particularly one that contests three-point attempts aggressively. The matchup-adjusted version – the team’s eFG% against opponents with the relevant defensive profile – is the input I trust most for game-by-game projections.
One concrete pattern: when a team that ranks in the top five for eFG% plays a team that ranks in the bottom ten for opponent three-point percentage allowed, the over on the high-eFG team’s total points has historically been a reliable bet. The defensive weakness is exactly the kind that the offensive team is built to exploit, and the operator’s total line – based on broader season averages – often underprices the specific matchup advantage.
True Shooting and Free-Throw Inclusion
True shooting percentage takes eFG% one step further by including free throws. The reason this matters is that free throws are pure-points shots without the variance of contested attempts, and players or teams that get to the line frequently are generating efficient scoring that eFG% misses entirely.
The formula is points divided by twice the sum of field goal attempts and 0.44 times free throw attempts. The 0.44 multiplier accounts for the fact that most free-throw trips deliver two shots, but some deliver one (and-ones) and some deliver three (fouled on three-point attempts). The blended result is a per-shot efficiency metric that captures every scoring opportunity equally.
For team analysis, true shooting percentage typically tracks within a few percentage points of eFG%, with the gap reflecting how often the team gets to the line. A team with a 56 percent TS% and a 53 percent eFG% is generating an extra 3 percentage points of efficiency through free throws – a meaningful chunk that puts them in the top tier of offensive efficiency.
For individual players the gap can be much larger. A player who attacks the rim aggressively and draws fouls frequently can have a TS% 7 to 10 percentage points above his eFG%. That difference matters for player props – a high-foul-draw player generates points from the line that the eFG% measure cannot see, and the prop line on his points total is often built off the eFG-equivalent projection.
The defensive corollary is opponent TS%. Teams that allow high opponent TS% are giving up efficient scoring, often through a combination of poor rim defence and high foul rates. Their games tend to feature elevated totals because the offence is getting cheap points. Conversely, teams with elite opponent TS% – typically the league’s top defensive units – produce reliable unders against offences that depend on rim attacks.
NBA participation in the UK has been a quiet driver of betting volume too. Adult basketball participation in the UK has grown roughly 50 percent since 2021, reaching about 344,400 adults playing at least twice a month, with 1.2 million children playing weekly. That bottom-of-the-funnel growth feeds the top-of-the-funnel betting interest, and advanced metrics are increasingly the common vocabulary for UK punters who are betting NBA as a primary sport rather than a casual diversion.
Net Rating as Team-Level Edge
If I could only have one advanced metric to project NBA games, it would be net rating. The formula is simple: offensive rating (points per 100 possessions) minus defensive rating (opponent points per 100 possessions). The result is the team’s per-100-possession scoring margin, the cleanest single-number summary of how good a team actually is.
The reason net rating matters more than win-loss record is that net rating describes the underlying performance, while win-loss record describes the outcome. A team with a great net rating that has lost close games is unlucky in the short term but likely to revert; a team with a mediocre net rating that has won close games is fortunate but likely to regress. Net rating is the predictive metric. Win-loss record is the descriptive one.
For betting purposes the most valuable application is comparing the projected matchup net rating to the implied probabilities of the spread and moneyline. If two teams have net ratings of +6 and -3 respectively, the matchup implies the +6 team should be favoured by roughly 9 points in a neutral-court game. Home court is worth around 2.5 to 3 points, so on the road the favourite would be expected to win by 6 to 7 points. If the operator’s line is the favourite -4.5, there is potential value on the favourite covering.
The complications come from lineup variance and short-sample volatility. A team’s net rating includes minutes from rotation players who may or may not be available for tonight’s game. The lineup-specific net rating – the team’s net rating with the actual five players who will be on the court – is the version that matters for the bet, and it requires more granular research to produce.
There is also a rolling-window dimension. Season-long net rating averages over 50 or 60 games are stable but can hide recent trends. A team that ranks +3 over the season but +9 over the last 15 games is a different team than the headline number suggests. I rely on 10-game rolling net rating for live betting decisions and full-season net rating only as a baseline reference point.
The other dimension of net rating that matters is the offence-defence split. Two teams can have the same +5 net rating with very different profiles – one through an elite offence and middling defence, the other through an elite defence and middling offence. The two profiles produce different totals: the offence-heavy team generates higher totals; the defence-heavy team generates lower totals. Net rating alone tells you the matchup outcome; the split tells you the totals expectation.
How to Combine the Three in a Bet Decision
The practical workflow that I have built combines all three metrics into a single pre-game read on each game on my radar. The process takes about ten minutes per game once you are familiar with where the numbers live.
Step one: pull each team’s 10-game rolling net rating, including the offence-defence split. This tells me which side has the underlying edge in this matchup and roughly how many points it should translate into on the spread.
Step two: pull each team’s rolling eFG% and TS%, with the opponent-adjusted versions if available. This tells me whether the offence is efficient enough to support the implied net rating, and whether the defence is genuinely suppressing opponent efficiency rather than just benefiting from a soft schedule.
Step three: combine the two to produce a projected total. The offence-defence interaction – Team A’s offensive efficiency against Team B’s defensive efficiency, plus the inverse – multiplied by the projected pace gives me a defensible total projection. I compare it to the operator’s line and the gap tells me whether the over or under has value.
The mental model is that net rating tells me who wins; eFG% and TS% tell me how they win; pace tells me how high the total goes. Each metric corrects a blind spot in the others. Using them together gives me a multi-dimensional read on the game that single metrics cannot produce.
This is what professional NBA bettors have been doing for two decades, and it is what the operator trading models do too. The basketball share of US sports-betting handle hit roughly 28 percent in 2024, and the operators competing for that handle have all converged on similar analytical frameworks. The bettor who reads the same metrics the operator reads is fighting on equal terms; the bettor who reads box scores is fighting with a worse map.
The next layer down – turning team-level advanced metrics into individual player projections, particularly for props – is its own discipline. The single most useful metric for that work is usage rate, which I cover in detail in my usage rate guide.
The Habit of Trusting the Numbers Over the Eye
The hardest part of advanced-metric betting is learning to trust the numbers when they contradict what the eye tells you. A team that is on a five-game winning streak feels good. The same team can have a deteriorating net rating, a declining eFG%, and a defence that has been propped up by lucky opponent shooting variance. The numbers say the regression is coming; the streak says the team is hot. The numbers are usually right.
The discipline is to look at the metric every time, write down what it says, and bet accordingly even when the storyline argues otherwise. The bettors who outperform over a season are not the ones who are smarter; they are the ones who are more consistent in trusting the right inputs. Advanced metrics are the right inputs. Box scores are the wrong ones.
Basketball Reference and NBA Stats both publish all three metrics for every team and player. Cleaning the Glass is the more sophisticated version, with garbage-time filtering and opponent adjustments that produce sharper numbers, but it requires a subscription. For most betting purposes the free sources are sufficient. Both, but the application is different. Pregame, advanced metrics produce projections you bet against the operator’s line. In-play, they help you assess whether unusual game flow reflects matchup advantages that will persist or random variance that will revert. A team shooting 60 percent eFG% in the first half against a defence that allows 50 percent typically has a real shooting edge; that read informs in-play totals and second-half wagers.Where can I find current eFG%, TS%, and net rating numbers?
Are advanced metrics useful for in-play decisions or just pregame?
Creado por la redacción de «nba Betting Expert».
