Away teams are booked more per foul — unless the ground is empty

Your club’s card count is not just about your club. Two things it cannot control move it as alongside how it plays: whether it is at home or away, and which referee it draws. Here is what more than 10,000 EFL matches — every yellow, every red and every foul the provider holds across seven completed seasons from 2019/20 and this one so far (League Two joins from 2022/23) — say about both, and about the decisions that change a scoreline outright. It is the Championship, League One and League Two; the Women’s Super League is left out because its fouls are not recorded far enough back to join in.

First, the ground rules

Card counts are not stable across these seven seasons, and two breaks matter before any of the rest makes sense. In 2020/21, played behind closed doors, bookings fell sharply everywhere — the Championship from 3.39 yellow cards a match to 2.91, League One from 3.61 to 3.18. Then in 2023/24 the EFL’s participant-behaviour charter pushed them back up and added a wave of bookings for dissent and time-wasting. So where this piece compares referees, it does so only on matches from 2023/24 onward, and where it compares home and away, it shows every season, because that empty-stadium year turns out to be the most revealing one in the set.

Away sides are booked more — for the same fouling

Across every season with a crowd, away teams were shown more bookings per foul than home teams: 11.3% more overall. That is not because away sides foul more and so collect more cards — this is bookings divided by fouls, so the fouling is already taken out. Give away the same number of free kicks and they were carded harder for them.

The obvious question is whether that is the laws or the atmosphere, and 2020/21 answers it as cleanly as football ever will — with the caveat that it is one season, and one in which more than the crowd changed. With the stands empty, the away premium was -0.9% — within a point of zero. Away sides were booked at essentially the home rate. In every other season it sat between 9.4% and 15.3%. It came back the moment the crowds did. Whatever the mechanism — pressure on the official, or away players behaving differently in a hostile ground — it travels with the noise, not with the laws of the game. One caveat worth stating plainly: we hold no record of where a foul happened. If away sides commit more of the cynical, transition-stopping kind that are more cardable — which this data cannot see — some of the premium would be that. Not the part that vanished in 2020/21.

And it depends who you get, and on what for

Referees are not interchangeable. Among the 49 who have taken 51-plus EFL matches since the charter, the rate at which they book a foul runs from 0.104 per foul (A. Chilowicz) to 0.139 (M. Corlett) — the strictest gives about a third more cards for the same fouling than the most lenient. These rates are shrunk toward the average to keep a light-sample official from topping the list by chance, and even after that the spread is 1.55 times what pure luck would produce. It is a real difference between people, not noise — though which matches a referee is given (the senior names take the bigger, hotter games) is inside that number, and no single-referee dataset can pull it out.

About 25% of bookings since 2020/21, when reasons were first recorded, are for no foul at all — dissent, time-wasting, simulation — and this is where officials differ most, because it is the closest thing to a free choice the laws give them. Non-foul bookings run from 0.82 a match for the most tolerant referee to 1.73 for the strictest. But that count says as much about the matches as the official, which is the next section.

What the whistle does to the scoreline

A penalty is the one refereeing decision that scores by itself. There were 1,058 awarded in the 6,624 matches since 2022/23, about one every six games, and home sides got 1.33 of them for every one an away side received — a raw count ratio that includes how much more home sides attack, since we hold no shot-location data to separate the two. What they do to a result is stark, once you split by the score when they were given. A side level when awarded a penalty went on to win 69% of the time — against 29% for a side simply level at the same stage of the game, the like-for-like comparison the red-card section below also uses. A side trailing salvaged at least a draw 56% of the time. The honest caveat: a penalty is not random. The side winning one was attacking the box, which usually means it was the better side that day — so some of that 69% is the team, not the spot-kick. But unlike a red card, a penalty award is far less tangled with the state the game had drifted into than a red card’s timing is.

The same score-state lens explains why a club’s booking record is not simply its discipline. Time-wasting bookings go 74% to the side that is winning and just 2% to the side that is losing — you waste time when you are ahead. Bookings for dissent lean the other way, 38% to the side that is behind. So a club that spends a season in front learns one kind of card, and a club chasing games learns another, and neither is really about how cleanly it plays. And heavily-refereed matches finish closer: games with five or more cards ended within a goal 72% of the time, against 62% for games with two or fewer — with almost identical goals per match. That is a tight game producing cards, not cards producing a tight game, most likely — a close, fractious match invites cards — but with nothing here to time-order the two, that is the plausible reading rather than a proven one.

A fair-play table that accounts for who you drew

Put those two corrections together — home or away, and which referee — and you can ask a fairer question than a raw card count allows: did a club pick up more bookings than the referees it actually faced would normally give, home and away, for the fouls it actually committed? The table below is each club’s real bookings against that expectation. A figure in red is a club carded above what its fixtures would predict; green, below. The bold clubs are the only ones whose gap is bigger than the margin of error — and there are strikingly few of them.

That is the real finding, and it is a deflating one for anyone hoping to prove their club is victimised. Of 72 club-seasons across the three divisions, 6 have a booking count whose 95% interval clears zero — and seventy-two intervals at that confidence will throw out about 3.6 by luck alone even if every club is perfectly average. So 6 is barely above what chance predicts, and even those names carry that warning. The honest summary is that almost nothing here rises clearly above scatter, and everyone else sits within a normal season’s noise of what the fixtures would give them. For the record, the clubs that clear the line are, in the Championship, Sheffield Wednesday (-19), Derby (+20), Hull City (+22); in League One, AFC Wimbledon (-24), Plymouth (+28); in League Two, Cambridge United (-25). Which side of the line a club falls on is partly the luck of the draw: the interval is wider for clubs expected to collect more bookings, so a big number does not always clear it and a smaller one sometimes does.

Championship, 2025/26

League One, 2025/26

League Two, 2025/26

What a red card costs — once you allow for the score

“Down to ten and the game’s gone” is football’s most confident cliché, and the raw numbers seem to back it. But a late red is often the symptom of a game already lost — a frustrated lunge at 2-0 down — so counting what happens to ten-man sides tells you as much about the score as the sending-off. The fair test is to hold the score fixed: take sides that were level when reduced to ten, and compare them with every side that was level at the same stage and venue and kept eleven men.

Done that way, timing is everything. A home side sent off while level inside the first half-hour (34 such matches) took 0.41 points a game from there, against 1.54 for the nearly six thousand that stayed eleven — a cost of about 1.1 points. The same dismissal after 80 minutes (50 matches) cost only 0.2: there is barely any game left to punish it. The cells are small, so read the figures as the shape, not the decimal. The cliché is true for an early red and close to empty for a late one. And the sides sent off were not, on the evidence, weaker teams having a bad day — their pre-match form was within 0.2 of a point a game of the elevens they are measured against, so the damage looks to be the red itself, not a gap in quality dressed up as one. This rests on 1,134 sending-offs across the seven completed seasons, with matches that saw a red to both sides, or a goal in the same recorded minute, set aside.

The short version

The whistle leans home, and the crowd is the likeliest reason why: away sides are booked 11.3% more per foul with a crowd in and not at all without one, and penalties — the decisions that swing games hardest — go to home sides a third more often, though how much of that is the whistle and how much is home sides attacking more, this data cannot say. Referees genuinely differ, most of all on the cards that need no foul. Penalties swing games hard and go to the home side far more often. A red card early is close to decisive and late is close to nothing. And the club that feels hard done by is, 6-in-72 times aside, getting about what the fixtures gave it.

All figures are regular-season EFL, from match data supplied by API-Football. Referee rates are estimated on 2023/24 onward; League Two before 2022/23 is excluded for want of card data; expected-bookings intervals use a standard count model. Where a foul happened is not recorded, and the piece says so where it matters.