Referee Home Bias 2026: Away Teams Take 54% of Cards

Every card market you bet is quoted as one number, and the two teams in the fixture are not facing the same one. Referee home bias is worth about 0.30 of a yellow card a match to the away side, which across a 19 game away schedule is roughly eight extra bookings a Premier League club collects on the road and not at home. Price a team card line off a league average and you pay the wrong price twice, once in each direction.

This page measures the effect rather than argues about it. We recounted every yellow and red card in 25,298 matches across the nine European leagues behind our referee pages, eight seasons from 2018/19 to 2025/26, from football-data.co.uk records downloaded on 1 November 2026. That window contains the only clean natural experiment football has run on this question: two seasons played in front of nobody.

The bias is real, the crowd causes it, it disappeared when the crowds did, and it is now larger than it was before the pandemic. There is also one European league where it has gone entirely, and a 13 point spread between individual Premier League referees that dwarfs every league level difference here.

Referee home bias, measured across 25,298 matches

Start with last season, because that is the baseline a 2026/27 card market is priced off. The nine leagues we track handed out 1.80 yellows a game to home sides and 2.10 to away sides in 2025/26. Pool the reds in and the away card share is 53.7 per cent. That is 0.302 of a yellow every match, in the same direction, in eight leagues of nine.

The spread between leagues is wider than the headline. The Eredivisie sent 56.8 per cent of its cards to the visiting side and the Primeira Liga 49.7 per cent, and those two sit at opposite ends of the card volume table as well: Portugal is the most carded league in Europe and the Netherlands the least. That is the first clue that bias and strictness are separate properties. Our league card rate guide ranks the volume side.

Hold on to the Premier League row. At 1.67 home yellows against 2.08 away, a club takes about eight more yellows across its 19 away fixtures than across its 19 home ones, before the opposition or the referee is considered at all.

The empty stadiums settled the argument

For decades the honest answer to whether referees favour home teams was that nobody could separate the referee from the team. Home sides have more of the ball, attack more, and commit fewer of the fouls that get carded, so a card gap proves nothing by itself. In March 2020 football removed the crowd and left everything else in place.

Read the third row twice. In the 639 matches played after the 8 March 2020 shutdown, home sides took more yellows than away sides, 2.041 against 1.879, an away card share of 48.01 per cent. The direction of a bias that had held in every season on record flipped inside a month.

The full closed doors season is stronger evidence because it is 3,218 matches rather than 639. There the gap did not invert, it collapsed to +0.026 of a yellow card, which is nothing. Eight of the nine leagues moved toward parity, La Liga the lone exception at a fifth of a point the other way.

One caveat on the 639 match cut: it is six leagues, not nine. Ligue 1, the Eredivisie and the Scottish Premiership abandoned 2019/20 and contribute nine matches between them. Of the six that restarted, four show home sides clearly taking more yellows, one is level and one is marginally the other way.

What the published research found, and where it stops

We are not first to this. It is a well established result in economics and sports psychology, and the popular version of it is wrong in one specific way.

Scoppa's paper is the one to read if you read one. Same underlying source as this page, ten seasons, 917 closed doors matches, and a result that lines up with ours: 1.87 home yellows against 2.18 away with a crowd present, against our 1.79 and 2.08 for 2018/19. Independent samples, same answer.

What most summaries get wrong is where the bias sits. The research consistently finds no reliable difference in the fouls referees award against home and away sides. The difference is in what happens after the foul, in the discretionary act of reaching for a card. A foul count is therefore not a proxy for this, and a referee's fouls per card figure, which our card betting method guide leans on, measures something else.

Where the literature stops is 2021. Every widely cited study ends with the empty stadium seasons, and whether the effect survived crowds returning has not been answered anywhere we can find.

The bias came back wider than it was before

It survived, and then some. Here is the whole eight season window pooled across the nine leagues.

The shape is a V with a higher right arm. The gap was +0.290 in 2018/19, bottomed at +0.026 in the closed doors season, then climbed in every season since: +0.265, +0.278, +0.279, +0.295 and now +0.302, the largest figure in the table. Away card share went 53.94 per cent, down to 50.60, back to 53.73.

We would not make much of 2025/26 edging 2018/19 on its own, because a hundredth of a card sits inside season to season noise. What is not noise is the recovery: five consecutive seasons of widening from a base of effectively zero, back to where it started. Anyone hoping the pandemic permanently civilised refereeing has eight seasons saying otherwise.

The practical reading is simple. Treat the pre 2020 research as current. The effect it describes is the effect operating in 2026/27, at full strength, which is not something you can usually say about a six year old football number.

Red card bias is the one that genuinely closed

Split the reds out and a different story appears, and this one is a real change rather than a return to normal.

In 2018/19 away sides were sent off 1.38 times as often as home sides and took 58 per cent of all reds. In 2025/26 the ratio is 1.03 and the away share is 50.8 per cent. Home reds per game barely moved across the window, 0.0845 to 0.0875. The whole convergence came from away dismissals falling, 0.1165 a game to 0.0903.

That is a 22 per cent drop in away reds over seven seasons while the yellow gap went the other way, so the cause is specific to dismissals. The obvious candidate is video review, which applies to red card decisions and not to second yellows, and which arrived across these leagues over exactly this period. Our VAR guide sets out what it does and does not touch.

The consequence for a bettor is narrow but clean. A red card market should not carry a home or away adjustment any more. A yellow card or booking points market still should.

Portugal has none left, the Eredivisie has the most

League averages hide more than they show, so here is the same three season comparison league by league.

The Primeira Liga is the outlier that matters. In 2025/26 it gave 761 yellows to home sides and 760 to away sides across 306 matches, and sent off 41 home players against 33 away. One card apart over a full season, with the reds pointing the other way. Portugal has trended this way since 2018/19, when it sat at 53.8 per cent, and is the only league in the set that did not rebound after the closed doors season.

At the other end the Eredivisie gave away sides 558 yellows against 428, which is 130 more cards over 306 matches, on an away share of 56.8 per cent. It is simultaneously the softest card league in Europe at 3.22 yellows a game. Our Eredivisie referee guide covers that pool official by official.

The Bundesliga is the quiet mover: largest away share of the nine in 2018/19 at 55.7 per cent, second lowest now at 51.4, a fall of 4.3 points while most of the set rose. Germany is the one big five league where a home or away card adjustment is close to not worth making.

Premier League referee home bias, 41 officials

Everything above is a league effect, and league effects are the small part. The referee is the large part, and it is the number no competitor publishes, because it needs match level records joined to a resolved referee identity rather than a league table.

Our Premier League set runs from 2005/06 to 2025/26, 7,980 matches. Among the 41 officials with 50 or more of them, away card share runs from 47.7 to 60.7 per cent, a spread of 12.9 points, four times the widest gap between any two leagues on this page. 38 of the 41 sit above 50 per cent, so this is a near universal tendency of very uneven size.

Two things there are worth separating. Mark Halsey is third most away leaning at 59.7 per cent on the lowest yellow rate in the group, 2.04 a game, and John Brooks is the only official in the set who cards home sides more often, 47.7 per cent, on the highest rate in the group at 4.32. Bias and volume are uncorrelated, and a lenient referee can be the most lopsided one in the division.

Among the active group the range is narrower and still useful: Robert Jones and Peter Bankes at 57.8 per cent against Anthony Taylor at 51.0. Our Premier League referee card stats page carries the full 2026/27 group.

One season of a referee's split is not a signal

Here is the trap, and it is the one that costs money. The splits above are built on 50 to 476 matches each. A current season split is built on ten to twenty, and at that sample a home or away share is mostly noise.

Martin Petersen's 66.7 per cent in the Bundesliga rests on ten matches, Florian Badstubner's 38.7 per cent on thirteen. The 28 point gap between them is more than twice the spread across 41 Premier League officials with full careers behind them, and almost all of it is sampling variation. One ten card derby lands in one column rather than the other and a referee moves ten points.

The working rule: under 50 matches, regress hard toward the league figure; under 20, use the league figure and ignore the referee's split. A career number from a 150 match official is a property of the referee. A single season number from a fifteen match official is a property of those fifteen matches.

That table names three leagues and not nine because those are the competitions where we hold the minute and the recipient of every card rather than a season end total, as our methodology page sets out.

Turning away card bias into a priced bet

The effect is small, consistent and almost always absent from the posted price. That combination is worth something, provided you put it in the right market. Four steps, using numbers from this page.

1. Only touch team card markets. Home bias redistributes cards between the sides, it does not create them, so a total cards or over 3.5 cards line is untouched by any of this. Team cards, team booking points and player to be booked are where it lives, and our booking points guide sets out how those settle.
2. Start from the league split, not 50 / 50. In the Premier League that is 55.4 per cent of cards to the away side for 2025/26, which on a 3.75 yellow game is 2.08 away against 1.67 home. That is your anchor before any referee adjustment.
3. Apply the referee as a shift, sized by sample. Robert Jones at 57.8 per cent over 110 matches is 3.2 points above the division's 54.6 career figure. Take the gap, regress it toward the league in proportion to how thin the sample is, and move the anchor by what is left. Do not replace the league figure with the referee's.
4. Check the reds separately, which now means not adjusting them. At an away to home ratio of 1.03 there is no edge left in a red card tilt, whatever the market offers.

None of it starts until you know who has the fixture, which is why our appointments page comes before this one in any real workflow. An unappointed match has a league baseline and nothing else.

What these numbers cannot tell you

Five limits, stated plainly, because a number you cannot audit is worth nothing.

A card share is not proof of a biased referee. Home sides have more possession and attack more, so they commit fewer of the fouls that attract cards. The published work removes that objection by holding the teams fixed and removing only the crowd, which is why the closed doors rows carry the argument and the 2025/26 rows only describe the state of play.

Game state is inside every figure here. A side that is losing commits more desperate fouls, and away sides lose more often. Some unknown share of the gap is that, not the official.

The 2019/20 closed doors cut is six leagues and 639 matches. Three leagues abandoned the season. Treat it as corroboration of the 3,218 match 2020/21 row, not a standalone result.

Referee splits come from source name strings. Two names in the Premier League table arrive abbreviated and are marked, and our Spanish, Portuguese and American pools carry enough unresolved name variants that we have deliberately kept them out of the referee level tables here. Anything ambiguous is flagged rather than merged, as our methodology page explains.

Nothing from 2026/27 is in these rates, nor are the IFAB law changes in force this season, two of which remove cautions referees used to issue regularly. Early season card rates are the least reliable numbers in football.

Used with the appointment sheet and a sense of sample size, the home and away split of a referee's cards is one of the few inputs in a card market nobody else is pricing.