Cards in football: the stat where the referee matters more than the teams
TL;DR. Card averages are among the most requested and worst understood statistics in football. The factor that explains them most is not how aggressive the teams are: it is who referees and in which competition. Two different referees can produce averages that differ by several cards a match with the same teams on the pitch.
Note. Informational and statistical content. It is not betting advice nor a promise of profit. No statistic predicts a single match: they are there to estimate probabilities over the long run. 18+ only. Play responsibly.
The referee is the dominant variable
Refereeing standards vary enormously and, unlike a team's form, they are fairly stable: a strict referee tends to stay strict. So if you are going to look at cards, the referee's name is the first data point, not the last.
The second factor is the competition. Every league has its own culture of permissiveness, and the differences between countries are larger than the differences between teams within one country. Comparing a team's card average with one from another country without correcting for that context is a basic error.
What the team does contribute
| Factor | Approximate weight on cards |
|---|---|
| Appointed referee | The single most decisive factor |
| Competition and country | Very high: it moves the bar for what gets punished |
| Rivalry or importance of the match | High in derbies and knockout ties |
| Team style (high press, tactical fouls) | Moderate |
| Scoreline during the match | Moderate: the side that is losing accumulates fouls |
The weights are qualitative, meant to order your reading: they are not coefficients we have measured.
The usual traps
- Adding yellows and reds without distinguishing them. Many counts weight a red as two yellows and others do not. Always check the convention before comparing figures.
- Ignoring the second yellow. Depending on the convention, a second-yellow dismissal can count as two cards, as three or as one red: it is the biggest source of discrepancy between sources.
- Extrapolating from few matches. Cards carry very high variance: one seven-yellow match inflates a team's average for weeks without describing anything stable.
- Confusing rough play with cards. A side committing many small, spread-out tactical fouls can pick up fewer cards than one committing half as many but more spectacular ones.
What we do with cards
We publish them as statistical context on the match page, alongside possession and shots, but the engine issues no card forecast. The reason is the one above: without the appointed referee as a variable, a card model would be ignoring its most decisive input, and we would rather not publish an estimate we know is lame.
It is the same policy we apply across the site: show the data and explain its limits rather than filling gaps with estimates.
Frequently asked questions
What influences a match's cards more, the teams or the referee?
The referee, by some distance, followed by the competition. Refereeing standards vary enormously and are fairly stable over time, so the referee's name is the first thing to check, not the last.
Why do card averages vary so much between sites?
Mostly because of how a second-yellow dismissal is counted and whether a straight red is weighted as two cards. It is the biggest source of discrepancy between sources.
Does BetsTalent forecast cards?
No. Without the appointed referee as a variable, a card model would ignore its most decisive input. We publish the data as context and explain its limits instead of estimating something we know is incomplete.