- No penalty rate is published here: the downloadable data we use does not include penalties or VAR decisions.
- Fouls can be measured: 25.2 per match across 16 leagues, from 20.9 in the Netherlands to 30.1 in Greece.
- A rare event needs many matches: with 30 matches for a referee, one penalty more or less moves the average a lot.
Save this article: you will want to open it before the next match you care about, along with the prompt for requesting a referee's record.
Your progress is saved in this browser.
Can you measure a referee's penalties?
This is the shortest article in data and the longest in honesty.
We wanted to give you "the referee who awards the most penalties."
We cannot, and we explain why.
The public football-data.co.uk files we use include goals, cards, fouls, corners and odds.
They do not include penalties.
They do not include VAR decisions either.
Honest limit: we will not publish a penalty or VAR rate per referee.
We do not have that data, and inventing it would be worse than not having it.
What follows is a framework for thinking, plus fouls, which can be measured.
This is not betting advice and is for adults 18+ only.
What can we measure: fouls per match?
Fouls are in the files.
We took 39,376 matches from 16 European leagues, from the 2018/19 season to 2026/27.
The average is 25.2 fouls per match.
The home team commits 12.4 fouls and the away team 12.7.
For each foul there are, on average, 0.17 yellow cards.
That is one every six fouls.
| League | Matches | Fouls per match | Yellows per foul |
|---|---|---|---|
| Belgium (B1) | 1,853 | 23.2 | 0.17 |
| Bundesliga (D1) | 2,204 | 22.6 | 0.17 |
| Bundesliga 2 (D2) | 2,223 | 24.4 | 0.18 |
| Premier League (E0) | 2,740 | 21.4 | 0.17 |
| Championship (E1) | 4,007 | 22.5 | 0.16 |
| Ligue 1 (F1) | 2,409 | 24.8 | 0.15 |
| Ligue 2 (F2) | 2,498 | 27.3 | 0.14 |
| Greece (G1) | 1,483 | 30.1 | 0.17 |
| Serie A (I1) | 2,750 | 25.5 | 0.17 |
| Serie B (I2) | 2,710 | 29.4 | 0.17 |
| Netherlands (N1) | 2,158 | 20.9 | 0.15 |
| Portugal (P1) | 2,223 | 29.1 | 0.18 |
| Scotland (SC0) | 1,613 | 23.3 | 0.16 |
| LaLiga (SP1) | 2,759 | 26.0 | 0.19 |
| LaLiga 2 (SP2) | 3,334 | 27.7 | 0.18 |
| Turkey (T1) | 2,412 | 25.9 | 0.18 |
The Netherlands has 20.9 fouls per match.
Greece has 30.1.
That is a difference of more than nine fouls.
The league remains the biggest factor.
But watch out: a foul in midfield is not a foul in the box.
The fouls we count do not say where they happened.
We cannot use them to estimate penalties.
Do fouls change from one referee to another?
Yes, according to data from the English and Scottish leagues, the only ones with a referee in the files.
We took the 86 referees with 30 matches or more.
The difference from their league average ranges from −3.3 to +3.2 fouls per match.
In the Premier League, among referees with 30 matches or more, the range runs from 18.1 to 23.4 fouls per match.
The league average is 21.4.
The full story on yellow cards is in referees and cards.
Verdict on "some referees whistle more fouls": confirmed.
Verdict on "that predicts penalties": it cannot be known with this data.
Why is a rare event so hard to measure?
Because it happens rarely.
Fouls happen dozens of times per match.
Penalties happen, at best, once every few matches.
Let us look at an example with numbers invented on purpose to understand the math.
Imagine a referee with 30 matches and 8 penalties awarded.
Their average is 0.27 penalties per match.
| Case (example) | Penalties | Per match | Change |
|---|---|---|---|
| Base: 30 matches | 8 | 0.27 | Reference |
| One more penalty | 9 | 0.30 | +12% |
| One fewer penalty | 7 | 0.23 | −13% |
| Estimated standard error | ±2.8 | ±0.09 | ±35% |
With a single penalty of difference, the average changes by more than 10%.
The standard error of a count like this is the square root of the number of events.
With 8 penalties it is about 2.8, which is 35% of the total.
This whole table is an illustrative example.
These are not real data for any referee.
It shows why a penalty ranking built on just a few seasons is very unreliable.
What should you check about a referee before a match?
If you want to build your own framework, this is the list of what is worth having.
Each item includes how much can be measured today.
| What to check | Public data? | Difficulty |
|---|---|---|
| Fouls per match | Yes, in football-data | Low |
| Yellow cards per match | Yes, in football-data | Low |
| Penalties per match | You have to look for it | High |
| VAR reviews | Partial, depends on the league | High |
| Style: lets play go on or stops it often | Impression, not data | Medium |
- Sample: how many matches they have officiated in that league and over how many seasons.
- Context: whether the matches were even, knockout ties or high-stakes.
- Source: where each figure comes from and whether you can open it.
- Comparison: how it looks against the league average.
- Odds: what the bookmaker pays and whether that price already includes what you know.
How do I ask an AI for a referee's record?
An AI with web search can gather sources faster than you can.
But it carries two risks: inventing figures and mixing up seasons.
The prompt has to forbid both.
Act as a football data analyst. I will give you a referee and a league: [REFEREE] in [LEAGUE]. Search the web for their record over the last three seasons. Give me penalties awarded per match, cards per match and fouls per match. For each figure, state the source with a link, the period and how many matches it includes. If a figure does not appear in a source you can open, write "no data". Do not estimate or fill in gaps. At the end, list which figures are worth verifying by hand. Do not make any betting recommendation.
What are the risks of betting on penalties and VAR?
- High variance: a rare event may not show up in twenty matches and show up twice in one.
- Market rules: some bookmakers count only penalties in regular time, others include shootouts or penalties retaken after a VAR review.
- Late information: the referee is confirmed a few days before and the odds have already moved.
- Unclear data: each source counts differently, and AIs mix up seasons.
- Recall bias: you remember the controversial penalties, not the thousands of matches with none.
If you are going to try it, use small stakes.
Write down each bet and each source.
Keep in mind that betting carries risk and is for adults 18+ only.
What to read next
If you want to see what can be measured about a referee, start with referees and cards: how much they influence the numbers.
It is the part with real data in this pair of articles.
For card frequency league by league, there is card stats from 39,856 matches.
For corners, corners: myths put to the test.
And to build your own assistant, see how to create your AI for betting.
The full route is in the guide to AI for betting.
To see how we measure our own picks, there is the Betting Journal.
Betting carries risk and is for adults 18+ only.
If you feel you are losing control, ask for help.
Past data guarantees nothing going forward.
Frequently asked questions
Which referee gives the most penalties?
With the data we use, it cannot be known.
The football-data.co.uk files do not include penalties per match.
You need a source that records them, and a large sample per referee.
Distrust any ranking without a source.
Did VAR change penalty betting?
It cannot be measured with this data.
We have no penalties or VAR decisions per match.
What did change is that there is more dead time and more reviews, something worth checking in each market's rules.
How many fouls are there per match?
25.2 on average, across 39,376 matches in 16 leagues between 2018/19 and 2026/27.
The home team commits 12.4 and the away team 12.7.
It varies a lot by league: 20.9 in the Netherlands and 30.1 in Greece.
Do fouls predict penalties?
There is no data to claim that.
The fouls in our files do not distinguish where on the pitch they happened.
A foul in the box is a different story from one in midfield.
Can an AI give me a referee's penalty record?
It can try if it has web search.
Ask for the source, seasons and number of matches behind each figure.
If it cannot give them, do not use the number.
Always verify at least one figure by hand.
Is it a good idea to bet on penalties?
It is a rare-event market with a lot of variance.
A short streak tells you nothing.
Only bet what you can afford to lose, and only if you are 18 or older.