What percentage of matches end in each scoreline
Dardan Hasku4 min read
Football has a very short list of likely results and a very long tail of everything else. Knowing the percentages will not win you a round on its own, but it tells you which of your calls are ordinary and which ones you are gambling on.
These are long-run shares across Europe's big five leagues, rounded. Any single season moves them by a point or so in either direction, and a low-scoring league like Serie A in a defensive spell will shift more than that.
The percentages
| Score | Share of matches | Running total |
|---|---|---|
| 1:1 | about 12 percent | 12 percent |
| 1:0 | about 10 percent | 22 percent |
| 2:1 | about 9 percent | 31 percent |
| 0:0 | about 8 percent | 39 percent |
| 2:0 | about 8 percent | 47 percent |
| 0:1 | about 5 percent | 52 percent |
| 1:2 | about 5 percent | 57 percent |
| 2:2 | about 4 percent | 61 percent |
| 3:1 | about 4 percent | 65 percent |
| 3:0 | about 4 percent | 69 percent |
Five scorelines cover roughly half of all football. Ten cover about seven matches in ten. Everything else, every 4:3 and 5:1 and 0:4 you remember for years, shares the remaining third between dozens of results that each turn up once or twice a season.
The same numbers as outcomes
Group them and the picture is simpler:
- Home win: around 44 percent. Down from the high forties a decade ago.
- Draw: around 25 percent. Remarkably stable across leagues and decades. More on why.
- Away win: around 31 percent. The number that has grown as home advantage shrank.
- Both teams scoring: around 50 percent. Close to a coin flip, which is why it makes such a good bonus question.
- Over 2.5 goals: around 50 percent. Same story, and the reason 2.5 is the line everyone uses.
So what percentage of correct scores do people actually get?
If you always picked the single most likely scoreline for every match, you would land somewhere around 11 or 12 percent of them. Real predictors do a little worse than that on the easy games and a little better on the ones they know something about, so a good season sits in the 10 to 15 percent range.
In a ten match round, that is one exact score, sometimes two, occasionally none. If that sounds low, it is: correct score is a hard ask, and the full arithmetic is here. A prediction league that only rewarded exact scores would be a lottery. That is why ours pays for the outcome and the goal difference and each team's goals separately, so a near miss still counts for something.
Reading the table the useful way
The percentages matter less as targets than as a measure of how alone you are. Two calls that are both wrong cost the same. Two calls that are both right do not pay the same, because a correct 1:1 will be shared with half your league and a correct 3:2 will not.
Our scoring takes that seriously. Nail an exact score and you get a rarity bonus scaled to how few people picked it, so the value of a scoreline depends on both its chance of landing and its popularity. How rarity works has the formula.
- share is the fraction of everyone who predicted that exact score.
- It is frozen at kickoff, so nothing you do after the lock changes it.
Worked example
0:0 is the value pick of this table. It happens about as often as 2:0 and gets predicted a fraction as much, because nobody enjoys calling nothing.
Where these numbers stop applying
League averages describe a season, not a fixture. A title chaser at home to a relegated side has a completely different distribution: the draw share collapses, the 3:0 and 4:0 shares climb. A tight derby pulls the other way, with the low-scoring results all growing.
Use the table to sanity check yourself, then adjust for the actual match. If your coupon has five different scores on it that each happen 2 percent of the time, you have not been brave, you have been careless.
Put the percentages to work on a real coupon. Ten matches, your bankers to place, and the field to beat.
Make your calls