International football scores exactly as many goals as club football. It just shares them out differently, and the difference is big enough to move a price.
Across every match in our database, the Nations League averages 2.63 goals a game and international friendlies 2.61. English league football - the Premier League, Championship, League One and League Two together - averages 2.64. On volume there is nothing in it.
But both teams score in 52.2 per cent of those English league games, and in only 46.6 per cent of Nations League matches and 43.8 per cent of friendlies. Same goals, fewer shared.
| Competition | Matches | Goals per game | Both teams scored | Over 2.5 goals |
|---|---|---|---|---|
| English league football | 10,562 | 2.64 | 52.2% | 49.7% |
| UEFA Nations League | 363 | 2.63 | 46.6% | 48.5% |
| International friendlies | 1,501 | 2.61 | 43.8% | 47.4% |
Both gaps clear a significance test against the club-football baseline. The Nations League difference of 5.6 points returns p = 0.035; the friendlies difference of 8.3 points returns p < 0.0001 on a sample of 1,501 matches. The over 2.5 column does not: 48.5 and 47.4 per cent against 49.7 are inside the noise.
That combination is the whole finding. The goals are there. They are just not being shared. A 3-0 and a 2-1 both clear over 2.5; only one of them pays a both-teams-to-score line, and international football produces more of the former.
The reason is not mysterious. International fixture lists put mismatches on the same card as even contests in a way a domestic league never does. Armenia beat Latvia 2-0 and Liechtenstein lost 2-0 to Lithuania on the same night that Netherlands drew 1-1 with Germany. A league of twenty roughly comparable clubs does not generate that spread.
We found this the expensive way. On Thursday and Friday this column took three both-teams-to-score legs across the Nations League and one won. Portugal beat Wales 1-0, Kosovo beat the Republic of Ireland 1-0, Poland drew 0-0 with Bosnia and Herzegovina.
The model had those fixtures in the low-to-mid fifties, because it carries a goals correction measured on club football, where it is known to under-rate goals. On internationals that correction points the wrong way. Adding roughly five points where it should be subtracting five or six is an eleven-point error on every international goals line, which is more than enough to turn a losing price into an apparent edge.
Worth being precise about what did and did not constitute evidence here. Matchday one alone was sixteen games and 44 per cent - and on its own that proves nothing at all, returning p = 0.50 against the club baseline. Sixteen games is an anecdote. It was only worth acting on once the 363-match and 1,501-match samples said the same thing.
Three practical reads.
Treat over 2.5 and both teams to score as different questions in international football. In a domestic league they move together closely enough that you can reason from one to the other. Here they come apart: the totals market is priced about right and the both-teams market is not.
The effect is strongest in friendlies, at 43.8 per cent - which is a second, independent reason to leave them alone on top of the rotation and the experimental elevens.
It cuts both ways. If both teams to score is over-priced by the market in internationals, the no side of the same market is under-priced. We are not tipping that on the strength of one weekend's work, but it is the half of this worth testing next.
Our own model correction is being reviewed rather than patched mid-break. Until it is, this column is not taking goals markets in international football, and the Daily Acca has gone back to the EFL.
Do international matches have fewer goals than club football? No. The Nations League averages 2.63 goals a game and international friendlies 2.61, against 2.64 across the Premier League, Championship, League One and League Two. The volume is the same.
Why is both teams to score less likely in international football? Because the goals are shared less evenly. Both teams score in 46.6 per cent of Nations League games and 43.8 per cent of friendlies, against 52.2 per cent in English league football. International fixture lists put heavy mismatches alongside even contests in a way a domestic league does not, producing more one-sided scorelines.
Is the difference statistically significant? Yes for both teams to score. The Nations League gap of 5.6 points returns p = 0.035 over 363 matches and the friendlies gap of 8.3 points returns p below 0.0001 over 1,501. The over 2.5 difference is not significant.
Does this change how Statz prices international matches? Yes. Our goals correction is measured on club football, where the model under-rates goals, so it pushes both teams to score upwards. On internationals it should push down, an error of around eleven points. Until the correction is reviewed we are not taking goals markets in international football.
18+. Please gamble responsibly - begambleaware.org. Match data from our own database, all completed fixtures held for each competition.