What does an implied probability calculator show?
An implied probability calculator turns the odds on every outcome of a market into the chance each price suggests, adds those chances up, and shows how far the total is above 100%. That excess is the bookmaker margin, also called the overround or vig. It does not tell you what will happen.
Odds are a price, and a price can be read as a chance. A decimal price of 2.00 suggests 50%, a price of 4.00 suggests 25%, and so on. The tool above does this for two to 12 outcomes at once. With the default prices of 1.91 on each side of a two-way market, each side implies 52.36%, the total is 104.71%, and the margin is 4.71%.
The tool does not remove the margin to give fair probabilities. Later on this page you will see how to do that step by hand, with a worked example, so you can see exactly what the adjustment does and where it is weak.
How do you calculate implied probability from odds?
Implied probability is the chance a price suggests before any adjustment. For decimal odds it is 1 divided by the odds. For fractional odds a/b it is b divided by (a plus b). For American odds it is 100 divided by (odds plus 100) when positive, and the odds divided by (odds plus 100) when negative, ignoring the minus sign.
| Odds format | Formula | Example | Implied probability |
|---|---|---|---|
| Decimal | 1 / d | 2.50 | 1 / 2.50 = 0.40 = 40.00% |
| Fractional | b / (a + b) | 3/2 | 2 / (3 + 2) = 0.40 = 40.00% |
| American, positive (+A) | 100 / (A + 100) | +150 | 100 / 250 = 0.40 = 40.00% |
| American, negative (-A) | A / (A + 100) | -200 | 200 / 300 = 0.6667 = 66.67% |
The three positive examples are one price. Decimal 2.50, fractional 3/2 and American +150 all mean a winning stake of 100 returns 250 in total, so all three imply 40%. You can type any of the three formats into the tool and it converts them for you. If you only need to switch formats, an odds converter does that job on its own.
The reverse also works. To turn a chance into decimal odds, divide 1 by the chance: 25% is 1 / 0.25 = 4.00. This is useful for checking a price against your own estimate, but remember that an estimate of a real-world chance is much harder to get right than the arithmetic.
What is the implied probability of common odds?
The table below lists the implied probability of common prices in all three formats. Every figure was recomputed from the formulas above and rounded to two decimals. Short prices (decimal below 2.00, negative American) imply more than 50%. Long prices (decimal above 2.00, positive American) imply less than 50%.
| Decimal | Fractional | American | Implied probability |
|---|---|---|---|
| 1.25 | 1/4 | -400 | 80.00% |
| 1.50 | 1/2 | -200 | 66.67% |
| 1.67 (exactly 5/3) | 2/3 | -150 | 60.00% |
| 1.91 (rounded; exactly 1.9091) | 10/11 | -110 | 52.38% |
| 2.00 | 1/1 | +100 | 50.00% |
| 2.50 | 3/2 | +150 | 40.00% |
| 3.00 | 2/1 | +200 | 33.33% |
| 4.00 | 3/1 | +300 | 25.00% |
| 6.00 | 5/1 | +500 | 16.67% |
| 11.00 | 10/1 | +1000 | 9.09% |
| 21.00 | 20/1 | +2000 | 4.76% |
| 101.00 | 100/1 | +10000 | 0.99% |
Notice how the figures bunch together at the long end. Going from 21.00 to 101.00 changes the price enormously but moves the implied probability by only about 4 percentage points. Long prices are easy to misjudge because a small change in the chance makes a large change in the odds.
How do you use this implied probability calculator?
Enter the odds of every possible outcome in one market, one per row, and read the margin and the total. Use the add button for a third outcome or more, up to 12. The tool accepts decimal, fractional and American odds, and the result updates as you type.
- List every outcome of one market. A two-way market has two outcomes. A football match result usually has three: home win, draw, away win. A race has one per runner.
- Type the odds on each row. Use decimal (1.91), fractional (10/11) or American (-110). Decimal odds must be above 1.00 and American odds must be +100 or higher, or -100 or lower.
- Read the margin. The main result is the total of the implied chances minus 100, in percentage points.
- Read the total. The second result is the sum of the implied chances. A fair, margin-free market would total exactly 100%.
- Open the formula line. It shows each implied chance added up, so you can check the arithmetic yourself.
Example: Leave the two defaults at 1.91. Each side implies 100 / 1.91 = 52.36%. The total is 52.36 + 52.36 = 104.71%, so the margin is 104.71 – 100 = 4.71%. If you type -110 on both sides instead, each implies 52.38%, the total is 104.76% and the margin is 4.76%. The tiny gap exists because 1.91 is a rounded version of -110, which is exactly 1.9091 in decimal.
Why do implied probabilities add up to more than 100%?
They add up to more than 100% because the bookmaker builds a margin into every price. If the chances of all outcomes sum to 100%, fair prices would pay out exactly what is staked over time. Shortening each price slightly lifts the implied chances, and the surplus is the bookmaker’s expected cut.
The tool reports that surplus directly: total minus 100. There is a second way to express the same cut, and the two figures differ slightly. In the default example the total is 104.71%. If you spread stakes across both sides so that the same amount comes back whichever side wins, you stake 1.0471 to get 1 back. That is a loss of 1 – 1/1.0471 = 4.50% of the amount staked.
| Measure | How it is worked out | Default example (1.91 / 1.91) |
|---|---|---|
| Overround (what the tool shows as margin) | Total implied chances minus 100 | 104.71 – 100 = 4.71% |
| Share of the amount staked that is kept | 1 – 100 / total | 1 – 100/104.71 = 4.50% |
Both measures are close at small margins, which is why guides use them loosely. Hegarty and Whelan’s 2023 paper on the bookmaker’s margin points out that the second figure assumes the margin is the same on every outcome. They found that when bookmakers charge more on unlikely outcomes, realized loss rates run higher: in their soccer data about 7.8% against 6.5% predicted, and in tennis about 7.4% against 5.4%. A margin read from the overround can therefore understate what a typical bet costs.
What is a normal bookmaker margin?
There is no single normal figure, because margins vary by sport, league, bookmaker, year and number of outcomes. Published research finds football match-result margins averaging around 8% in one large sample, and totals that climb sharply as the number of runners grows, so two-way markets tend to be the tightest.
On football, a study of closing football odds from 14 bookmakers (2006 to 2017, English and German leagues, match-result odds) found an average margin of about 7.94%, with individual bookmaker averages ranging from 2.46% to 18.50%. On field size, University College Dublin lecture notes on field size and margins report that in a horse-racing dataset of 396,572 races the average overround rose from 105% in two-horse races to 167% in races of 30 horses. The notes do not name the data source, so treat the horse-racing figures as indicative.
| Market type | Outcomes | Typical total of implied chances (from sources above) |
|---|---|---|
| Two-way | 2 | Example prices 1.91 / 1.91 total 104.71% |
| Three-way (football result) | 3 | Sample average margin about 7.94%, range of bookmaker averages 2.46% to 18.50% (2006 to 2017) |
| Horse racing win market | 2 to 30 | Average 105% with 2 runners, 167% with 30 runners (one dataset) |
To see the same pattern with your own hands, compare these invented markets, computed with the tool’s formula. They are illustrations, not real prices.
| Illustrative market | Decimal prices | Total of implied chances | Margin |
|---|---|---|---|
| Two-way | 1.91 / 1.91 | 104.71% | 4.71% |
| Two-way, tighter | 1.95 / 1.95 | 102.56% | 2.56% |
| Three-way | 2.10 / 3.40 / 3.60 | 104.81% | 4.81% |
| Eight runners | 2.50 / 4.00 / 5.00 / 7.00 / 9.00 / 12.00 / 15.00 / 21.00 | 130.16% | 30.16% |
The eight-runner total of 130.16% means that, on average, roughly 23% of money staked across all runners is kept (1 – 100/130.16 = 23.17%), if the margin is spread evenly. More outcomes give a margin more places to hide.
How do you remove the margin from implied probabilities by hand?
The simplest way is proportional (multiplicative) normalization: divide each implied probability by the total of all of them. This scales the chances so they add to exactly 100%. The tool does not do this step, so here is the full method with numbers you can check.
- Convert each price to an implied probability. Divide 100 by the decimal odds.
- Add them up. This is the total, the second result in the tool.
- Divide each implied probability by the total, then multiply by 100. Each answer is its share of the total.
- Check that the results add to 100%. Small gaps are rounding.
- Turn them back into odds if you want. Fair decimal odds are 100 divided by the adjusted chance.
Example: Team A is priced at 1.60 and Team B at 2.40 (invented teams). Team A implies 100 / 1.60 = 62.50%. Team B implies 100 / 2.40 = 41.67%. The total is 104.17%, so the margin is 4.17%. Dividing by the total gives Team A 62.50 / 104.17 = 60.00% and Team B 41.67 / 104.17 = 40.00%. Fair decimal odds would be 100 / 60 = 1.67 and 100 / 40 = 2.50.
| Step | Team A | Team B | Total |
|---|---|---|---|
| Decimal odds | 1.60 | 2.40 | – |
| Implied probability | 62.50% | 41.67% | 104.17% |
| After dividing by the total | 60.00% | 40.00% | 100.00% |
| Fair decimal odds | 1.67 | 2.50 | – |
The same steps work for any number of outcomes. For a three-way market at 2.10, 3.40 and 3.60 the implied chances are 47.62%, 29.41% and 27.78%, which total 104.81%. Dividing each by the total gives 45.43%, 28.06% and 26.50% (rounded, so they sum to 99.99%).
Are there other ways to remove the margin?
Yes, and they give different answers. The proportional method assumes the margin is shared out equally among outcomes. Other methods, such as Shin’s model and regression-based approaches, allow longer prices to carry more of the margin. a paper in the Journal of Sports Economics by Å trumbelj found that basic normalization can produce biased probabilities, and that the method chosen can be enough to reach opposite conclusions. Berkowitz, Depken and Gandar reviewed seven ways of converting money lines to win probabilities, and recommend the standard normalization for its simplicity unless a favourite-longshot bias is suspected.
The practical lesson is that the adjusted figures are still estimates built on an assumption. The differences between methods are usually small on balanced two-way markets and larger on lopsided prices or long fields.
Is implied probability the same as the real chance of an outcome?
No. Implied probability is what a price suggests, and it includes the margin, so the raw numbers overstate the chances of every outcome. Even after the margin is removed, the result is a market estimate of the chance, not a measured fact. Real outcomes depend on things the price may get wrong.
Three separate things get mixed up: the implied probability (from the price), the adjusted probability (after margin removal), and the true probability (unknown). A price cannot tell you whether a bet is good or bad on its own. Doing that needs an independent estimate of the true chance, which is hard to get right. For casino games the situation is simpler because the chances are fixed by the rules; see the roulette odds and payouts guide for how the house edge arises there.
Over many bets, the margin is a running cost. At a 4.71% margin, staking 10 units on each of 100 two-way bets means 1,000 units staked, and the expected cost is around 45 units if the margin is spread evenly (1,000 x 4.50%). Individual results will swing well above and below that. If you want to keep track of what betting costs you, set limits first with the Safer play toolkit.
What mistakes do people make with implied probability?
The most common mistakes are treating the raw implied probability as the true chance, entering only some of the outcomes, mixing markets, and rounding too early. Each one gives a misleading margin or an overconfident reading of a price, and most can be avoided by entering a complete market from a single source.
- Leaving out an outcome. If you enter two prices from a three-way market, the total will look too low and the margin will be wrong. Enter every outcome, including the draw.
- Mixing prices from different sources. The margin belongs to one bookmaker’s market. Combining prices from several places gives a total that describes no real market.
- Treating implied probability as the true chance. It includes the margin and the market’s own errors.
- Reading a total below 100% as a good sign. It almost always means an outcome is missing or a price was typed wrongly.
- Mixing up the two margin figures. Total minus 100 is not exactly the share of stakes that is kept. At 4.71% the two figures differ by 0.21 percentage points.
- Rounding odds before adding. Decimal 1.91 and American -110 are close but not identical, so totals differ slightly.
What should you try next?
Try entering the prices from a market you already follow, and compare the margin with the markets in the tables above. Then repeat the by-hand margin removal on paper once, so the steps make sense. If you are working out what betting costs you over time, the Safer play toolkit is a good place to set a spending limit.
Related Chancepedia tools and guides are planned, including an odds calculator, a parlay calculator, an expected loss calculator and an explainer on what the overround means. They will sit alongside this page as they go live.
How to read the result
Bookmaker margin is the total of the implied chances minus 100, in percentage points. It is also called the overround or vig. The total of the implied chances is the sum of 100 divided by each decimal price. A margin-free market would total exactly 100%. Typical two-way markets total a little over 100% (for example 102% to 105%), while markets with many outcomes total much more. The line under the result lists the implied chance of each outcome. Open Formula to see the addition. A total below 100% usually means an outcome is missing or a price was mistyped.
How it is calculated
Each odds entry is first converted to decimal odds d: fractional a/b becomes 1 + a/b, positive American +A becomes 1 + A/100, and negative American -A becomes 1 + 100/A. The implied chance of each outcome is 100 / d, in percent. The total is the sum of the implied chances. Bookmaker margin = total - 100. At least two valid odds are needed. Up to 12 outcomes can be entered. Results are shown to two decimals for the margin and total and one decimal for each implied chance.
Worked example
Worked example
With the default odds of 1.91 on each of two sides, each side implies 100 / 1.91 = 52.36%, shown as 52.4%. The total is 52.36 + 52.36 = 104.71%, shown as 104.7% on the bar. The bookmaker margin is 104.71 - 100 = 4.71%. To remove the margin by hand, divide each implied chance by the total: 52.36 / 104.71 = 50.00% for each side.
Limits
The tool does not remove the margin or give fair probabilities. It does not tell you whether a price is good value or what will happen. It assumes you entered every outcome of one market from a single source, and it cannot check that. Decimal odds must be above 1.00 and American odds must be +100 or higher, or -100 or lower. Fractions are read exactly, but a rounded decimal such as 1.91 differs slightly from -110 (1.9091). The margin shown is the total minus 100, which is slightly larger than the share of stakes kept (4.50% for the default). If margins differ between outcomes, the cost of a typical bet can be higher than the margin suggests. Results are estimates, not predictions.