An NBA winning margin bet asks you to pick the winning team and the band its winning margin falls in, such as 1 to 5 or 6 to 10 points. Each band is priced from the same thing that sets the line: how many points the bookmaker expects the favourite to win by, and how widely real results spread around that. A typical NBA margin lands within about 12 points of the line either way, so the bands near the line are the most likely and the far ones pay long odds. The catch is the bookmaker's cut: with a dozen outcomes, it is usually far bigger than on head to head.
This guide works through one invented game from the line to fair odds for every band, so you can read a winning margin market rather than guess at it. No selections here: the numbers are made up to show the maths.
How NBA winning margin markets work
Australian bookmakers list NBA winning margin as a set of outcomes covering both teams. Layouts differ between bookmakers, so check the one you use. For this guide we use five-point bands:
- Team A by 1-5, 6-10, 11-15, 16-20, 21-25, 26+
- Team B by 1-5, 6-10, 11-15, 16-20, 21-25, 26+
That is 12 outcomes, and exactly one of them wins. There is no draw: under NBA Rule 5, a game level after four quarters goes to five-minute overtime periods until someone wins. Most bookmakers settle on the result including overtime unless the market says otherwise, which is the general rule Action Network describes for spreads, totals and head to head. Read the rules on the market before you bet.
If you are newer to the NBA board, our NBA odds explained guide covers the head to head, line and totals markets first.
Where the band prices come from
A bookmaker doesn't price 12 bands from scratch. It starts with the line, which is its best guess of the expected margin, and then needs one more number: how far real results stray from it.
The expected margin
Say Team A is -6.5 on the line. That means the market expects Team A to win by about 6.5 points on average. Half the time Team A covers, half the time it doesn't, give or take the bookmaker's margin.
The spread around it
NBA results scatter widely around the line. Wayne Winston, the sports analytics author, puts the standard deviation of the margin relative to the point spread at near 12 points, and explains it from the randomness of about 86 shots a team per game. Hal Stern's 1994 paper in the Journal of the American Statistical Association modelled the score difference in an NBA game as a random walk, which ends in a bell-shaped spread of final margins, and found it fitted 493 games from the 1991-92 season well.
So a simple model is: final margin follows a bell curve centred on 6.5 with a standard deviation of 12. Roughly two games in three land within 12 points of the line, either way.
Worked example: fair odds for every band
Using that model (centre 6.5, spread 12), here is the chance of each band and the fair odds, which are 1 divided by the chance. We split the small chance of a tie after four quarters evenly between the two 1-5 bands, on the assumption that overtime games are usually decided by a few points.
| Band | Team A chance | Team A fair odds | Team B chance | Team B fair odds |
|---|---|---|---|---|
| 1-5 | 17.3% | $5.79 | 13.6% | $7.38 |
| 6-10 | 16.4% | $6.11 | 8.0% | $12.44 |
| 11-15 | 14.3% | $7.00 | 4.5% | $22.26 |
| 16-20 | 10.5% | $9.53 | 2.1% | $47.28 |
| 21-25 | 6.5% | $15.38 | 0.8% | $119.13 |
| 26+ | 5.7% | $17.65 | 0.4% | $261.07 |
Three things stand out:
- The favourite's 1-5 and 6-10 bands are close to even, because the line of 6.5 sits right on the boundary between them.
- Team A wins about 71% of the time in this model, so all six Team A bands together are worth about $1.42 fair.
- The underdog's bands fall away fast. Team B by 16 or more is about 3% across three bands.
Where the simple model is wrong
The bell curve puts a tie after four quarters at about 2.9%. Real games go to overtime more often than that: RotoGrinders counted overtime in 6.2% to 6.4% of NBA games in each of the 2013-14, 2014-15 and 2015-16 seasons. Close games cluster more than a smooth curve suggests, likely because late-game tactics such as fouling and slowing the game change how the last few minutes play out. The practical upshot is that the true chance of the two 1-5 bands is probably a little higher than the table shows, and that a model needs checking against real results before you trust it.
The spread number matters too. Change it from 12 to 13 and Team A by 26+ moves from $17.65 to $13.90 fair. A small change in the assumption makes a big difference in the tails, so treat long-shot fair odds as rough.
How big is the bookmaker's margin?
Now compare those fair odds with a bookmaker's. These prices are invented but shaped like a real winning margin board:
| Band | Team A odds | Team B odds |
|---|---|---|
| 1-5 | $5.00 | $6.25 |
| 6-10 | $5.25 | $10.00 |
| 11-15 | $6.00 | $17.00 |
| 16-20 | $8.00 | $34.00 |
| 21-25 | $12.00 | $67.00 |
| 26+ | $13.00 | $101.00 |
Add up 1 divided by each price and you get 121.6%. That 21.6% above 100% is the overround, the bookmaker's built-in margin. For comparison, an invented head to head market on the same game at $1.36 and $3.15 adds up to about 105.3%.
That gap is normal for markets with many outcomes. A public Bookmaker Margin Panel on Mendeley Data, which measured 52 bookmakers from June to August 2026, found Australian-licensed bookmakers averaged a 5.25% margin overall, and player props averaged 7.93% against 4.77% for head to head. More outcomes, and less attention on each, mean more room for margin.
The margin isn't spread evenly
Compare each invented price with our fair odds and the cost per dollar bet goes like this:
- Team A 1-5 at $5.00 against $5.79 fair: about 14 cents in the dollar.
- Team A 26+ at $13.00 against $17.65 fair: about 26 cents.
- Team B 26+ at $101.00 against $261.07 fair: about 61 cents.
The long shots carry most of the margin. This is why the quick method of removing the margin, dividing every implied chance by 121.6% evenly, misleads here. It gives Team B by 26+ a "fair" price of about $123, when our model says $261. The even method is fine for a two-way market (our No-Vig calculator does it in one step), but on a 12-way board it understates how much you are paying on the long shots. For more on how margin builds into a price, see what vig is and how it works.
How a line move reprices the bands
Winning margin boards are big, and bookmakers don't always reprice every band at once. Watch what one piece of news does to the fair odds. Say Team A's star is ruled out and the line moves from -6.5 to -2.5:
| Band | Fair odds at -6.5 | Fair odds at -2.5 |
|---|---|---|
| Team A 6-10 | $6.11 | $6.72 |
| Team A 26+ | $17.65 | $36.18 |
| Team B 1-5 | $7.38 | $6.06 |
| Team B 16-20 | $47.28 | $25.53 |
The bands near the middle barely move. The tails roughly double or halve. That is where a stale price is most likely to sit after news, and also where your own fair odds are least certain. An odds screen that shows the line moving across bookmakers is the quickest way to see when the market's expected margin has changed, before you look at any band.
Bands, alternate lines and middles
A band is two alternate lines in one. Team A by 6-10 wins when Team A wins by 6 or more but not by 11 or more. So its chance is the chance of "Team A by 6+" minus the chance of "Team A by 11+". If you already price NBA alternate lines, you can check a band against them. If the band's price is well away from what the difference implies, one of the prices is out of line with the rest of the board.
The same idea links to middle betting: a middle wins when the margin lands between two lines, which is exactly the space a band covers, though a middle uses two separate bets at two prices.
A quick checklist before you read a margin board
- Start from the line: it tells you where the market thinks the centre is.
- Check which band the line sits in. Those bands, and the one next to them, are the most likely.
- Add up the book. Anything well above 110% means you are paying a lot for the choice of bands.
- Look hardest at the tails, where the margin is biggest and fair odds are least certain.
- Check the overtime rule on the market itself.
- Track your results by band so you can see whether your own estimates hold up over hundreds of bets, not a handful.
Winning margin bets are high-variance: even a well-priced band loses most of the time, and no model removes that.
See the NBA board in one place
BetBank puts NBA lines, totals and prop prices from Australian bookmakers side by side, with the No-Vig and Hold calculators to check a market's margin yourself. If you haven't had a paid plan before, the NBA season special gives first-time members half price for their first 2 months on a weekly or monthly plan, until 30 November 2026.
Betting is for adults 18+. Odds and models show the price of a bet, not its result. Set a limit, bet only what you can afford to lose and gamble responsibly. Free, confidential help is available 24/7 from Gambling Help Online on 1800 858 858.