This season just concluded, we became the first side in the Premier League’s 34 editions to go through a campaign without conceding a penalty or receiving a red card. We did, of course, concede a penalty of consequence in the Champions League final, and two others on the way there. However, that is beyond the scope of this disciplinary review.
The editors requested an examination of our record both in a historical context and in comparison with the teams we have competed against for the PL title in recent seasons.
They may have hoped that this would reveal a deep conspiracy among referees to do us down. Not to bury the lede, as they used to say when journalism was still honest and thoughtful, it is really, really difficult to find coherence in how the employees of the Professional Game Match Officials Limited brandish the two cards they carry in their pockets.
Perhaps ‘Limited’ should be at the start of their name, not the end.
There was a foul roughly every four and a half minutes in the league last season, an average of 21 per game. Our average of 10.3 per game was slightly below our fair share. Wolves committed the most, at 13.0 per game, and City the fewest, at 9.7. Yet thirteen of the 20 PL clubs were roughly within half a foul per game of our figure.
The blue line in the chart below shows that our number last season was close to the midpoint of the narrow range of our foul rate over the past decade. Myth no. 1 to dispel: Arteta has got his team fouling less. He hasn’t. Where the fouls are conceded is a different matter, and that has affected our yellow-card count. More on that later.
Next, let’s consider red and yellow cards, represented in the chart by, er, the red and yellow lines. Last season’s yellow card count was relatively low, but not the lowest over the past 10 seasons. Nor has there been a steady decline during the Arteta years. This term was the first since 2022-23 when we received fewer yellow cards than the previous season.
In general, in the Premier League, the correlation between fouls committed and yellow cards issued is very strong, as one would expect. However, for Arsenal, considered in isolation, the correlation is weaker than in the league as a whole, though it remains reasonably strong.
That, I shall argue later, is not a coincidence.
The red line shows no correlation with either yellows or fouls conceded. We should not expect one. Red cards are relatively rare, and VAR puts its oar in, not always for the best. The Key Match Incidents panel reckoned we should have had four reds last season, not zero. The previous season, the panel had said we should not have had as many as we did.
However, the tally is what it is, making last season the second of the past four seasons without any red cards. For the record, the most red cards we’ve had in a season in the Premier League era is seven in 1998-99.
Where we ranked for fouls, yellow cards and red cards among the other PL clubs over the past ten seasons underlines the randomness of red cards.
Last season we ranked joint 18th for red cards (City and Brighton also had no red cards last season) but 16th for fouls. The previous season we were 17th for fouls but top for reds. In 2020-21, we were second for reds but 20th for fouls. Go figure.
It would be easy to see a conspiracy afoot. Yet, as the next pair of charts show, the same erraticity in yellow and red card issuance applies to the other members of the Big Five and the neighbours, over the past ten seasons.
The only reasonable conclusion the man on the Clapham omnibus could draw is that there is no coherence or consistency among Premier League referees.
No surprises there, then.
The final chart shows our yellow card count by player for last season. Two-thirds were for fouls, roughly one-sixth for time-wasting or delaying restarts, and the remainder for a miscellany of infractions.
I draw your attention to the numbers for Gabriel and Saliba, four and two, respectively, of which only half were for fouls.
I would suggest that these figures hold the key to why our yellow card count has been relatively low compared with other clubs in recent seasons, and that this season our red card count has evaporated, even though we are still fouling as much as ever.
Gabriel ranked 33rd and Saliba 62nd, respectively, for yellow cards last season, not among all PL players but among just PL centre-backs. The top 10 included Brighton’s Dunk and van Hecke, with 10 and nine yellows, respectively. Newcastle’s Burn had nine. Everton’s Tarkowski had eight, as did Bournemouth’s Senesi. Wolves’ Mosquera led the way with 11. There was a sprinkling of red cards among that cohort, too.
I would argue that those extraordinarily low figures for Gabriel and Saliba are a consequence of the discipline and organisation Arteta has instilled in the team, rather than the result of specific measures to reduce cards.
I will acknowledge that after Saliba was sent off at Bournemouth in October 2024, Arteta said, “we cannot continue to play with 10 men at this level”, and that the team needed to eliminate red cards. I will pair this with Arteta’s oft-stated maxim that one quality he values in a player is availability, which includes avoiding suspensions.
Yet, to my mind, it is the defensive organisation built on strict positioning principles and the discipline to hold that structure for long periods in games, allied to a game plan based on control and on making the opposition play where we want them to play, i.e., away from yellow and red card-prone areas in front of our goal, that has made the difference to the card counts.
Our defenders have fewer desperate last-ditch tackles to make — and in Saliba’s case, as I mentioned in a previous drinks, while his numbers for the dirty work of defending are anything but impressive, his positioning and anticipation stats are off the charts compared with any centre-back in Europe’s top five leagues, let alone the PL. He rarely needs to make last-ditch challenges, and when he does, he has the Rolls-Royce qualities to make them cleanly.
I would ask you to guess — without first looking at the answer in the next paragraph — how often Saliba and Gabriel foul on average.
The answer is 0.8 of a foul every 90 minutes. That is remarkable given the intensity and physicality of the Premier League. Even our most carded players, Jurrien Timber, Ricardo Calafiori and Viktor Gyökeres, all with five yellows last season, were committing only 1.4 fouls every 90 minutes on average.
That further suggests a systemic effect from team discipline and organisation. Control of the ball makes players less likely to commit fouls, and even less likely to commit the desperate fouls that result in cards. We got a yellow for every 7.66 fouls last season, down from every 5.7 fouls in 2024-25
There are second-order dimensions to this. Our out-of-possession 4-4-2 midblock rarely exposes individual players. A study of all 20 PL teams, conducted after 10 games last season and dividing the defensive third into ten zones, found that we were the only team to foul below the league average in all ten.
Another is our high defensive line when in possession. Most of our fouls are committed near the halfway line. In the Palace home game, all six of our fouls were committed in the Palace half. One benefit of fouling high up the pitch is that referees view such fouls as less serious offences and are less likely to reach for their cards. Another point is that free kicks conceded in the middle third are not in direct shooting range.
Paolo Maldini, who would feature in many all-time XIs at centre-back, once said that if he had to make a tackle, it was because he had already made a mistake. Artetaball is built on avoiding mistakes, in the sense of avoiding jeopardy.
We now have a mature, disciplined team that is confident in its tactical organisation. It no longer fouls in the most dangerous areas out of desperation, but rather where it is tactically most effective. The foul count hasn’t declined, but the card count has as a result.
