How to read the probability of a football match
A percentage does not tell you who will win. It tells you how often each team would win if the same match were played many times. This guide explains how to read it, using our own data.
What 60% means
If we give the home side a 60% chance, it means that in 10 matches played under the same conditions the home team would win about 6. In the other 4 it would draw or lose.
favourite wins draw loses
That is why 60% is not a sure thing: 4 times out of 10 the result will be different, and that does not mean the probability was wrong. A probability can only be judged over many matches, not one.
Why a favourite can lose
Football is low-scoring. A penalty, a red card or a mistake in the 80th minute can change a whole match, and that happens more often than it seems. Even the clearest favourites in a league drop points several times a season.
When a favourite loses, the useful question is not "was the prediction wrong?" but "do favourites at that level lose as often as we said?". If we say 65% and the favourite wins 65 times out of 100, the figure was right even if you saw the defeat.
Real example · Tue 10 Nov Trival Valderas 0–1 Leganes We gave Trival Valderas a 60 % chance of winning. They did not win: this is the kind of match that falls in the remaining percentage.The draw, the most overlooked result
In Europe's top leagues, roughly one match in four ends in a draw. It is almost never the single most likely outcome, but added up it weighs a lot.
19% of the matches we analysed this year ended in a draw (19 of 99).
When the draw reaches 30% or more, no outcome clearly stands out. We flag it with this label: Open match
How to tell if a probability is good: calibration
A source is well calibrated when its figures hold up in practice: outcomes it rates at 60% happen close to 60% of the time. This table groups our 1X2 predictions from the last year by the probability we gave the chosen outcome.
| Probability given | Matches | Average given | Actual hit rate |
|---|---|---|---|
| < 45 % | 48 | 41% | 38% (18/48) |
| 45–55 % | 30 | 49% | 63% (19/30) |
| 55–65 % | 13 | 58% | 77% (10/13) |
| ≥ 65 % | 8 | 73% | 88% (7/8) |
With few matches per group the actual figure can drift by chance; it settles as more matches are played. We publish every hit and miss and never delete any. See full performance →
Probability and odds are not the same
Odds can be turned into a probability with 1 ÷ odds: odds of 2.00 equal 50%. But bookmakers add a margin, so the three probabilities of a match add up to more than 100%.
Our figures are clean probabilities that add up to 100% and come from our own statistical model, which we compare with the market. They help you understand a match, not tell you what to bet on.
PronosLab is not a bookmaker and does not recommend betting. If you bet, do so responsibly and only if you are over 18. Responsible gambling
Frequently asked questions
What does a 60% probability mean in a match?
That if the match were played 10 times under the same conditions, that outcome would happen about 6 times. The other 4 times something else would happen, so it is not a certainty.
If the favourite loses, was the prediction wrong?
Not necessarily. A probability is judged over many matches: if favourites we rate at 65% win close to 65% of the time, the figure is good even if they lose a particular match.
Why is a draw almost never predicted?
Because the draw is rarely the single most likely outcome, even though it happens in about one match in four. That is why we flag matches where the draw is at 30% or more as an "Open match".
How do you convert odds into a probability?
Divide 1 by the odds: 1 ÷ 2.00 = 50%. Since bookmakers include a margin, the three probabilities add up to slightly more than 100% and must be normalised before comparing them.
