Odds are a probability written in another way. To know how far to trust them, though, you need a second number: the volume. The data from 36,211 matches in our archive.
When you look at a match, next to each outcome you find a number: 1.55, 4.40, 7.00. It is easy to mistake it for a price. In fact it is something more interesting: it is the way the market tells you how likely it thinks a given result is.
In this article we show you how to read that number, where it comes from and why, on its own, it tells only half the story. The other half is told by the volumes. And, as always, we let the data speak.
Odds are a probability
Turning odds into a probability is simpler than it looks: just divide 1 by the odds.
Implied probability = 1 ÷ odds
Odds of 1.85 therefore correspond to 54.1%: according to the market, the outcome happens a little more than once in two. The calculation also works the other way round: an outcome with a 60% probability has fair odds of 1.67, that is 1 divided by 0.60.
Odds and probability
| Odds | Implied probability | In words |
|---|---|---|
| 1.25 | 80.0% | 4 times in 5 |
| 1.50 | 66.7% | 2 times in 3 |
| 2.00 | 50.0% | 1 time in 2 |
| 3.00 | 33.3% | 1 time in 3 |
| 5.00 | 20.0% | 1 time in 5 |
| 10.00 | 10.0% | 1 time in 10 |
Looking at the table you will notice something many people miss: the relationship between odds and probability is not linear. Going from 1.25 to 1.50 moves the probability by more than 13 points. Going from 5.00 to 10.00 moves it by 10, even though the odds double. At low odds, even a few hundredths matter a lot.
How odds are formed
The odds in our archive are Exchange odds: they are not set by an operator, but form like the price of a share on the stock market, where supply meets demand.
For each outcome you always see two prices side by side: the best odds on the Back side and the best odds on the Lay side. When the two sides agree, the position is matched and those odds become the last traded price.
From that moment the odds update continuously: they shorten when confidence in the outcome grows and lengthen when it falls. What moves them is the information reaching the market, such as the official line-ups, an injury, a goal or a sending-off, but also simply the minutes ticking by at 0-0.
That is why it is better to think of odds not as a forecast but as a snapshot: they tell you what the market thinks at that precise moment.
Do the odds reflect reality? How accurate are they?
If odds express a probability, their accuracy can be checked. We took 36,211 matches from 40 leagues, played between 1 January 2024 and 20 November 2026, using the home team’s odds at kick-off. We grouped the matches into odds bands and, for each band, compared the average implied probability with the actual share of home wins.
For each band, the difference between the two percentages is the gap: it tells you how far the odds strayed from reality. In the table it is the last column, and right after it there is a step-by-step example.
Home team odds bands
| Odds band (home) | Matches | Average implied probability | Actual home wins | Gap (points) |
|---|---|---|---|---|
| 1.01–1.40 | 2,330 | 79.1% | 79.7% | +0.5 |
| 1.41–1.70 | 4,354 | 64.1% | 63.7% | −0.3 |
| 1.71–2.00 | 5,901 | 53.9% | 52.6% | −1.4 |
| 2.01–2.50 | 8,537 | 44.6% | 44.1% | −0.5 |
| 2.51–3.50 | 8,758 | 34.4% | 34.0% | −0.4 |
| 3.51–5.00 | 3,788 | 24.5% | 23.9% | −0.6 |
| over 5.00 | 2,543 | 13.7% | 13.0% | −0.7 |
Odds and reality compared
In the chart, the diagonal represents a perfect market: the closer the dots are to it, the more accurate the odds. With an average gap of 0.6 points, the odds proved very reliable. They stray furthest in the 1.71–2.00 band, with a gap of −1.4 points, beyond the margin of error.
This does not mean that every single price is right. It means that, over thousands of matches, the market as a whole rarely makes big mistakes. The more interesting question then becomes: under what conditions does the error grow? To answer it we need to bring in volumes.
Volumes: how much the market believes it
The odds tell you what the market thinks. The volume tells you how solid that opinion is.
Volume is the amount of positions matched on a market. Every matched position always has two sides, Back and Lay, so volume does not tell you which way the market is leaning: it tells you how much activity there has been at that price.
A useful comparison is with opinion polls. With ten respondents, two people changing their mind is enough to shift the result noticeably; with ten thousand, the figure is far more stable. The same principle applies to odds:
- With high volumes, the comparison of many opinions makes the odds more stable.
- With low volumes, a single order can move them significantly.
To compare leagues of very different sizes, we express volumes as an index: the average of the whole sample is 100. A league with an index of 300 has three times the average volume, one with an index of 20 a fifth of it.
Does more volume mean more accurate odds?
We repeated the check league by league, giving each league its own volume index on the full-time result market. But what we saw in the example still applies: a league has far fewer matches than the whole archive, so even with perfect odds its gap would not be zero. That is why next to each league we show the chance limit: how far the gap can go by pure chance, with perfect odds and the same number of matches (chance only goes beyond it one time in twenty). If the gap stays below that limit, the league’s odds are as accurate as they can be.
Volume and accuracy by league
High volume 13 leagues · index 50 to 770 · average gap 3.0 points
Medium volume 14 leagues · index 21 to 42 · average gap 3.5 points
Low volume 13 leagues · index 5 to 18 · average gap 3.6 points
All the numbers
| League | Country | Volume index | Matches | Gap (points) | Chance limit (points) |
|---|---|---|---|---|---|
| England | 770 | 1,043 | 2.6 | 4.3 | |
| Spain | 456 | 1,075 | 2.7 | 4.3 | |
| Germany | 420 | 850 | 3.8 | 4.8 | |
| Italy | 417 | 1,052 | 3.3 | 4.3 | |
| France | 193 | 843 | 3.1 | 4.8 | |
| Brazil | 170 | 1,102 | 2.9 | 4.2 | |
| Portugal | 140 | 875 | 1.7 | 4.5 | |
| Germany | 130 | 855 | 2.0 | 4.7 | |
| Turkey | 98 | 951 | 3.7 | 4.5 | |
| England | 88 | 1,546 | 1.8 | 3.5 | |
| Australia | 80 | 466 | 7.3 | 6.5 | |
| Netherlands | 69 | 888 | 2.9 | 4.6 | |
| Spain | 50 | 1,316 | 1.0 | 3.8 | |
| Denmark | 42 | 558 | 4.4 | 6.0 | |
| Saudi Arabia | 42 | 841 | 2.8 | 4.6 | |
| Argentina | 40 | 1,346 | 3.1 | 3.7 | |
| Greece | 38 | 671 | 4.0 | 5.2 | |
| Belgium | 38 | 775 | 3.9 | 5.1 | |
| Sweden | 38 | 703 | 5.9 | 5.3 | |
| United States | 34 | 1,550 | 2.9 | 3.5 | |
| China | 34 | 706 | 3.0 | 5.2 | |
| Romania | 33 | 894 | 2.5 | 4.7 | |
| Norway | 31 | 679 | 4.1 | 5.4 | |
| Italy | 31 | 1,085 | 2.7 | 4.1 | |
| Switzerland | 23 | 649 | 3.8 | 5.5 | |
| Finland | 23 | 518 | 3.3 | 6.2 | |
| England | 21 | 1,544 | 2.7 | 3.5 | |
| Austria | 18 | 555 | 4.4 | 6.0 | |
| France | 18 | 912 | 1.7 | 4.6 | |
| Poland | 17 | 870 | 2.6 | 4.7 | |
| South Korea | 16 | 660 | 2.9 | 5.2 | |
| Bulgaria | 15 | 799 | 5.2 | 4.9 | |
| Portugal | 14 | 873 | 5.1 | 4.6 | |
| Netherlands | 14 | 1,091 | 2.7 | 4.3 | |
| Croatia | 13 | 370 | 6.0 | 7.1 | |
| England | 12 | 1,538 | 3.1 | 3.5 | |
| Turkey | 11 | 1,013 | 1.8 | 4.3 | |
| Serbia | 10 | 834 | 2.5 | 4.7 | |
| Hungary | 5 | 558 | 4.7 | 5.9 | |
| South Korea | 5 | 757 | 4.6 | 5.0 |
In the leagues with the highest volumes the average gap is 3.0 points, while in the least followed ones it rises to 3.6. But it is a small difference, and chance is enough to explain it: 36 leagues out of 40 stay within their chance limit. Volume tells you how solid a price is, not how right it is: at kick-off the odds are accurate even where volume is low.
When the volume comes in
Volume is not spread evenly over time. A significant part is concentrated close to kick-off, in the hour when the line-ups are usually announced, and another during the match.
When the volume is matched
58% of the pre-match volume is concentrated in the last hour before kick-off. It follows that odds read the day before reflect the opinion of a limited number of participants, while the odds at kick-off sum up a much wider audience.
A match, minute by minute
To see odds and volume in action we look at a real match: Lazio – Cagliari, Serie A, 3 November 2025, which ended 2-0.
Lazio – Cagliari: odds and volume over time
Before kick-off the home team’s odds move from 1.67 (59.9%) three hours before to 1.69 (59.2%) at kick-off, while in the last hour the volume doubles compared with the hour before. During the first half, with the score stuck at 0-0, the odds gradually lengthen to 2.28: every minute without a goal reduces the favourite’s chances, and the market takes it into account.
Then, at 65’, the home team scores. Within seconds the odds collapse to 1.24 (80.6%) and the volume hits the peak of the match, 5.6 times the average.
The odds record the event; the volume measures the activity that event generated. Read together, they give a much fuller picture than either can give on its own.
In short
- Odds are a probability: you get it by dividing 1 by the odds. 2.00 equals 50%, 1.50 equals 66.7%.
- On the Exchange, the odds are set by the market and update with every new piece of information.
- In our sample the odds proved very accurate, with an average gap of 0.6 points.
- Volume measures how solid that opinion is: it tells you how solid a price is, not how right it is.
- Odds and volume should be read together: a move in the odds backed by high volume carries more weight than one backed by little.
Want to go deeper? FG Academy, FoxesGoal’s free course inside the platform, is coming soon: it starts from how an Exchange works and goes all the way to reading odds and volumes on the matches in our archive.
FG AcademyComing soon
The data updates automatically with the matches played. Last update: 20 November 2026. FoxesGoal is a football analysis and statistics tool.
The sample includes the matches in our archive from 40 leagues that ended with an official result and have the Exchange full-time result odds available at kick-off. The home team’s odds are the last price traded before kick-off; the implied probability is 1 divided by the odds. Each band shows the 95% margin of error (Wilson interval). The gap is the difference, in percentage points, between actual home wins and the implied probability; the average gap is the mean of the absolute values, weighted by the number of matches. The chance limit is the gap a league can show even with perfect odds and the same number of matches: chance goes beyond it once in twenty. Volume is the amount matched on the full-time result market, before and during the match, shown as an index (sample average = 100) or as a percentage share, never as an amount; the half-time volume is counted with the second half. Leagues with fewer than 300 matches are not shown individually. The data is updated every 10 minutes with the newly played matches.