Divide each implied probability by the market total. The total drops back to 100% and each outcome gets a normalised probability. On a 1X2 market priced 2.10 / 3.40 / 3.60, that gives estimated margin-free odds of 2.20 / 3.56 / 3.77. Use a different method to spread the margin and you'll get different numbers from the same prices.
What are margin-free odds?
Margin-free odds are the prices you'd be left with if the bookmaker handed back the excess it builds into a market. Their main job is comparison. No two operators run the same margin, so their raw prices don't line up directly.
You'll see a few names for this. Margin-free odds, no-vig odds, de-vigged prices. They all mean the same operation: bringing the market's implied probabilities back to a clean 100%.
How do you calculate them?
The method used here is proportional normalisation. Each implied probability gets divided by the sum of the probabilities in the market.
normalised probability = implied probability ÷ sum of the implied probabilities
margin-free odds = 1 ÷ normalised probability
So the excess is shared out pro rata. An outcome that weighs twice as much as another soaks up twice as much margin. It's the easiest method to explain and to check for yourself. It's also the one this site's calculators use, starting with the bookmaker margin calculator.
A 1X2 example, start to finish
Take a three-way market priced 2.10 / 3.40 / 3.60. They're the same prices as on the page about the margin, so you can follow along from one page to the other.
| Outcome | Displayed odds | Implied probability | Normalised probability | Margin-free odds |
|---|---|---|---|---|
| Home win | 2.10 | 47.62% | 45.43% | 2.20 |
| Draw | 3.40 | 29.41% | 28.06% | 3.56 |
| Away win | 3.60 | 27.78% | 26.50% | 3.77 |
| Total | n/a | 104.81% | 100.00% | n/a |
The book percentage comes to 104.81%, which is an overround of 4.81 points. Normalisation takes the same share off every probability, and the total goes back to 100%. As a result, every margin-free price ends up above the displayed one. That's expected. You've just removed the very thing that made them less generous.
The whole calculation lives in that one move from 104.81% to 100.00% on the Total row. The other three rows simply follow from it.
Bear in mind. You won't find margin-free odds at any operator. Nobody's offering 2.20 on that outcome. It's a yardstick for comparing prices, and that's all it is.
Now for a real case, Borussia Dortmund v Eintracht Frankfurt (Bundesliga). The second table starts from the raw prices of an unnamed reference operator and strips the margin out pro rata, just like the example above. The chart and the first table show OddScore's margin-free odds for the same match.
From 1 May to the start, the margin-free odds on “Borussia Dortmund win” moved from 1.60 to 1.78 (+11.40%), a margin-free probability of 62.62% then 56.21%. Lowest value seen: 1.57; highest: 1.79. This chart describes a past market movement: it says nothing about the match result and is not betting advice.
| Outcome | First tracked odds | Last odds before the start | Change | Margin-free probability (first, then last) |
|---|---|---|---|---|
| Borussia Dortmund win | 1.60 | 1.78 | +11.40% | 62.62% then 56.21% |
| Draw | 4.90 | 4.55 | −7.12% | 20.41% then 21.97% |
| Eintracht Frankfurt win | 5.89 | 4.58 | −22.17% | 16.99% then 21.82% |
Raw odds from a reference operator, not named, before the start: the implied probabilities add up to 103.42%, a margin of about 3.42%. Dividing each probability by that total gives the margin-free probability (proportional method, slightly different from the OddScore consensus in the chart).
| Outcome | Raw odds | Implied probability | Margin-free probability |
|---|---|---|---|
| Borussia Dortmund win | 1.70 | 58.82% | 56.88% |
| Draw | 4.44 | 22.52% | 21.78% |
| Eintracht Frankfurt win | 4.53 | 22.08% | 21.34% |
| Total | 103.42% | 100.00% |
Source: OddScore. Margin-free odds computed from the odds of the international reference bookmakers, snapshots from 1 May to 8 May 2026, match already played. Data frozen on 11 October 2026.
Where the proportional method falls short
Proportional normalisation rests on one assumption: that the margin is spread across the outcomes in proportion to their probability. It's handy. Nobody has proved it.
Research on betting markets keeps finding the same pattern, known as the favourite-longshot bias. Big prices tend to carry proportionally more margin than short ones. When that bias is at work, the proportional method underrates the favourite's probability and overrates the longshot's. And the wider the gap between the prices, the bigger the error.
There are four main families of alternatives, and none of them has won everyone over. The power method raises the probabilities to a calibrated exponent instead of dividing them all by the same factor. The logarithmic method spreads the excess on a log scale. The Shin model treats part of the margin as cover against better-informed punters. Then there are empirical corrections for the favourite-longshot bias, calibrated on past results.
On a tight market, they all land in roughly the same place. On a lopsided one, say a favourite at 1.15 against a longshot at 15.00, the gaps get obvious. So picking a method is a real modelling choice, and you should own it out loud. I'd take a simple method with known flaws over a sophisticated one nobody can explain.
How OddScore removes the margin
OddScore removes each bookmaker's margin before comparing anything, and it explains the method instead of passing it off as an objective result. The principle is the one on this page. Every market gets brought back to 100%, so that the gap between two operators reflects their estimate rather than their commercial policy.
What you see on the platform is still an estimate of the margin-free price, recalculated at every snapshot. It's there for comparing and for following how things move. It doesn't point you to a bet.
Analyse a full market with the calculator →
Margin-free probability or true probability?
A margin-free probability is a reading of prices. A true probability would be a property of the match itself. They don't live in the same place.
The margin-free probability comes from the odds on offer. It inherits everything the market has got right, and everything it has missed. If every operator underrates a team, the margin-free probability will underrate it just as much. No amount of margin removal fixes that.
The true probability is something nobody gets to see. A match is played once, and the result doesn't show you the distribution it came from. That's why this page talks about estimates and never about the "right" probability. The phrase wouldn't point to anything you could check.
If you want a reference point that doesn't come from market prices, you need an estimate from somewhere else. That's what fair odds are about, and it's the starting point for any value bet calculation.
In short. Removing the margin shows you how a market splits once it's back to 100%. It also helps you gauge how far two operators disagree. It never tells you that an outcome is more likely than its price suggests.
Put it into practice
Apply this guide to real odds.
See a market reading
A real match with its odds, movements and 0-100 signal.
Open the demoMargin calculator
Measure a bookmaker's margin on any market.
Calculate the marginDig into the market
Odds movements are only part of the story. Here are the next topics to read.
Odds and probability: the complete guide
The whole journey, from the price on screen to the closing line.
Read the guideBookmaker margin: definition and calculation
Book percentage, overround and theoretical margin: what the excess actually measures.
Understand the marginImplied probability: getting it from the odds
The step before this one: turning a price into a percentage.
Convert some oddsFair odds: from an estimated probability to a price
The other route to a reference price, starting from your own estimate.
Work out fair oddsFrequently asked questions
What are margin-free odds?
They're the odds you get once a market's overround has been handed back to its outcomes. For each outcome, they're the inverse of its normalised probability.
How do you remove the margin from a market?
The most common way is to divide each implied probability by the sum of all the probabilities in the market. The total then comes to exactly 100%.
Do margin-free odds give you the true probability of an outcome?
No. They give you a normalised estimate built from the prices on offer. Change the prices or the spreading method and the result changes too.
Are there other ways to remove the margin?
Yes. The best-known ones are the power method, the logarithmic method, the Shin model and corrections for the favourite-longshot bias. Each spreads the excess differently, mostly on the big prices.
Why remove the margin before comparing two bookmakers?
Because each operator applies its own margin. If you don't strip it out first, the gap between their prices mixes two things: a difference in their estimate and a difference in commercial policy.
What's the difference between a margin-free probability and a true probability?
The first comes from market prices, once the excess has been redistributed. Nobody can observe the second, and no price will ever give it to you.
Sources & methodology
This page draws on the standard methods for normalising probabilities taken from odds, on economic research into the overround and the favourite-longshot bias, on the player guidance published by the British regulator and support services, and on OddScore's own odds-analysis method.
- Convert each decimal price in the market into an implied probability (1 ÷ odds).
- Divide each implied probability by the market total. That's proportional normalisation, the method used here and in the site's calculators.
- Invert each normalised probability to get the estimated margin-free odds.
- Present the result as an estimate, never as a true probability. Other spreading methods give other values from the same prices.