← All articles

Penalties won are a refereeing stat

Penalties won sits in player profiles and team summaries alongside goals, assists, and shots, and it is usually read the same way: as something a player or a team did. A forward who wins a lot of them is described as clever, quick, or hard to defend in tight spaces. A team that wins a lot of them is described as dangerous in the box. Some of that is true. But a penalty is not an action in the way a shot or a pass is an action. It is a decision taken by an official, increasingly reviewed by a second official watching replays, and the number a player carries is a joint product of his own play, the defender’s choices, the referee’s threshold, and a great deal of chance.

A count of decisions, not events

A shot happens whether or not anyone approves of it. A penalty exists only when a referee decides that contact, or an arm, or a shirt pull, crossed a line. The same tangle of legs at the edge of the six-yard box can be given in one match and waved away in the next, and both outcomes can be defensible under the laws as written. Referees differ in how much contact they tolerate, in how they read a player who goes down easily, and in how game state shapes their willingness to award a decision that will probably settle a match.

The laws themselves move. Handball guidance in particular has been rewritten several times in recent years, with each version changing what counts as an unnatural position and when accidental contact is punished. Every one of those revisions changes how many penalties are awarded across the league without any attacker playing differently. Handball penalties also expose the oddity of the stat most clearly: the attacking side is credited with having “won” something that was entirely produced by a defender’s arm. The only attacking contribution was getting the ball into the area.

Video review moved the threshold, then moved it again

Video review was sold as a way to remove error, but it did not remove judgment. It added a second layer of judgment with its own standard. Under the “clear and obvious error” principle, the on-field decision is the default, and borderline calls tend to stand whichever way they were originally given. That means the referee’s first instinct still determines most of the marginal cases, which are exactly the cases that separate a player with many penalties from one with few.

Review also changed the mix. Offences that a referee could not see in real time, such as holding at set pieces or a trailing foot caught from behind, can now be spotted on replay, while some soft on-field awards get overturned. The net effect on totals has varied between competitions and between seasons as officiating bodies adjusted how often they want the video assistant to intervene. Comparing a player’s penalty count from before video review with one from after it mixes two different regimes, and comparing two seasons of the review era can mix two different interpretations of it.

Volume in the box drives the count

Nearly every penalty requires an attacker to be in the opposition area, usually with the ball or about to receive it. That makes penalties won heavily dependent on how much time a team or player spends there. A side that dominates possession and repeatedly works the ball into the box generates far more opportunities to be fouled than a side that defends deep and attacks in brief bursts. A winger who carries the ball inside and takes on defenders one against one invites contact that a player who crosses from wide never will.

This is why raw totals mostly measure volume. A more useful framing is a rate: penalties won per touch in the opposition box, or per carry into the area, or per dribble attempted inside it. A rate asks a better question — when this player is in the area with the ball, how often does it end in a spot kick? — and it separates the player who is simply on the ball near goal constantly from the one who converts his presence into fouls. Even the rate carries style effects. Direct dribblers who drive at defenders will draw more contact per touch than players who prefer cutbacks and one-touch finishes, and that difference reflects tactics as much as any special talent for being fouled.

Reputation complicates it further. Defenders may hold back against an attacker known for going down, which suppresses fouls, while referees may be less generous to a player they suspect of exaggeration. Neither effect appears in the count.

The sample is tiny

Even the teams that spend the most time in the opposition box win only a handful of penalties in a season, and individual players fewer still. At counts that low, randomness dominates. Imagine two forwards with exactly the same underlying chance of earning a penalty on any given touch in the area. Over a single season, ordinary variation can easily leave one with twice as many as the other, and the gap will be read as a difference in skill.

The practical consequence is that penalties won is weakly repeatable from one season to the next at the player level, and any single-season figure should be pulled heavily toward the league average before it is used to judge anyone. Rates built on box touches help, because the denominator is large, but the numerator is still small. A player who led the league in penalties won last season is a reasonable bet to win more than average this season if he keeps getting the ball in the box. He is not a good bet to lead the league again.

Why penalties distort goals and expected goals

Penalties matter disproportionately because they are so valuable. Roughly three in four are converted in top-flight football, with the exact figure varying by sample and by era, and expected-goals models typically assign a penalty a value in that range. Most open-play shots are worth a small fraction of a goal. A single penalty therefore adds as much expected value as a large number of ordinary chances.

That is the reason analysts separate non-penalty goals and non-penalty expected goals from the headline totals. A striker who is his team’s designated taker collects goals that reflect a coaching decision about who stands over the ball, not his ability to create or finish chances in open play. Two forwards with identical open-play output can finish a season several goals apart purely because one takes the penalties. At team level, an expected-goals difference that includes penalties partly reflects officiating, and stripping penalties out gives a cleaner read of which side actually controlled the quality of chances.

There is also a subtler problem in how penalty value is counted. A penalty is usually awarded for a foul that interrupted something: a dribble, a shot about to be taken, a cross about to be met. The expected value of that interrupted chance disappears from the record and is replaced by the value of the spot kick. The attacking teamThe attacking team’s figure dependsrsquo;s figure therefore depends on whether the foul happened, which is the defender’s and referee’s contribution rather than the attacker’s.

The honest read

Winning penalties is not a fiction. Getting the ball into the area, carrying it at defenders, and arriving in dangerous positions are real attacking skills, and players who do those things will earn more spot kicks over time. But the count itself is a blend of that skill with refereeing interpretation, shifting video- review thresholds, possession volume, and sample noise. Read it as a rate per touch in the box rather than a total, regress single seasons hard, keep the player who wins penalties separate from the one who takes them, and judge attacking quality on non-penalty goals and non-penalty expected goals. Treat the penalties-won column as part attacking output, part officiating record, and part luck, in proportions that no one can cleanly separate.