Huijsen, the 63rd-Minute Red Card, and the Data Skeleton of a Madrid Derby

At the 63rd minute, the score was 1-1. Dean Huijsen lunged in with both feet, arriving before the ball. Referee Ortiz Arias, aged 41, a man who had officiated matches nobody remembers, brandished a straight red card. No consultation. No pitch-side monitor. Just a decisive flourish, as though he had prepared himself psychologically for this moment long before kickoff.

Real Madrid were down to ten men for the final 27 minutes. Atlético Madrid scored the winning goal to make it 2-1. And later that evening, Jose Mourinho walked into the press conference carrying a stack of colour-printed photographs — not tactical diagrams, but images of uncalled fouls. He laid each one on the table, neatly aligned, like an auditor reconciling books in front of a court.

I sat ten thousand kilometres away, in Melbourne, rewinding the footage at 0.25x speed. And what I saw was not in the red card. It was thirty seconds before, when Huijsen had not yet decided to lunge. That is where this derby was truly decided.

When the whole world zooms in on the red card, I rewind thirty seconds and zoom in on the space Huijsen left behind him.

Context the data has not fully told

The Madrid derby between Atlético and Real always carries a characteristic that statistics struggle to capture: it is a match where both teams deliberately play differently from how they perform in ordinary league fixtures. Atlético in recent seasons have been renowned for a low defensive block, ceding possession and waiting to strike in transition. Real Madrid, conversely, operate as a possession machine built on individual quality in midfield and the movement of attacking outlets.

But a derby does not allow either side to maintain that identity intact. When the opponent shares your city, your pitch, your domestic calendar, every tactical model gets dragged onto a single psychological plane. That is why I always tell my editors: data from a derby only has value when placed alongside the data of those same two teams across the four preceding rounds. Otherwise, we are reading a momentary phenomenon rather than a structure.

Across the seven most recent Madrid derbies prior to this fixture, Atlético's PPDA was significantly lower than their seasonal average — meaning they pressed less, deliberately ceded the initiative more. Real Madrid's PPDA was higher than their seasonal average — meaning they pressed more, pushed higher, and accepted greater risk in the space behind their defence. That is a tactical paradox few viewers notice: in a derby, the stronger side is the one accepting more risk.

And the very space Real left behind is where Huijsen's story begins.

To understand why the 63rd-minute red card was not a cause but a consequence of a chain of events, I must rewind to the first half. Real Madrid controlled 61% of possession in the first half, firing nine shots for a total xG of 1.14. Atlético held 39%, firing five shots for a total xG of 0.38. Reading those two figures, any reader of broadcast statistics would conclude: Real dominated. But xG is not a scoreboard. It is an X-ray machine, and that X-ray revealed something else.

Real's shot count came mostly from outside the box — specifically, six of nine originated from beyond 18 metres. That is the signature of a team controlling possession but being blocked in the final layer, forced into long-range solutions. Meanwhile, three of Atlético's five shots came from inside the box, despite a lower total xG. Two entirely different attacking structures were hiding beneath a scoreline that felt level.

The data evidence chain: from the opener to the space at minute 62

To analyse a derby the way I want, I divide it into five-minute data windows. This approach is not new in professional sports analytics, but it is especially effective in derbies, where match tempo fluctuates sharply with each psychological cycle. I do not trust a viewer's feeling, including my own. I trust counting.

Window 10-15: Atlético raised their PPDA from 9.2 to 8.4 — meaning they began pushing higher. This was the window in which Valverde lost the ball twice in midfield, both leading to quick Atlético counterattacks. No goals, but the structure had formed: Atlético realised that if they pressed at exactly the point where Valverde lost possession, the space behind would open.

Window 25-30: Real's opening goal. Here I must state clearly what post-match reports usually omit. The goal came from a passage in which Bellingham ran 28 metres off the ball to drag Atlético's left full-back out of position. The ball never reached Bellingham's feet in that passage, but his run opened the space for the scorer.

When the whole world looks at the goal, I look at the off-ball run.

This is why I always request raw movement data rather than post-match box scores. A box score will record: Bellingham — 0 goals, 0 assists, 1 shot. But GPS data shows that across 90 minutes, Bellingham ran a total of 11.8 km, of which 2.3 km was at speeds above 20 km/h and 11 sprints above 25 km/h. That figure says what a box score never can: he was the unlocker, not the finisher.

Window 45-50: Atlético equalised from the penalty spot. The foul leading to that 12-yard kick was the first contentious moment of the match, and it was also the incident VAR referee Trujillo Suárez confirmed as the on-field decision. I do not have sufficient data to assert whether that decision was right or wrong — and I will not do what emotional commentators do: adjudicate a passage from a television angle. I simply note a fact: across the last three seasons, Atlético Madrid have been the team awarded the most penalties in derbies against Real Madrid. That is a statistical fact, not an accusation.

Window 55-63: This is the window I want to spend the most time on. Atlético shifted from medium pressing to high pressing across a broad area. Their PPDA in this period dropped below 8.0 — the figure I once used to decode Croatia at the 2026 World Cup. When a team drives PPDA below 8.0 for nearly ten consecutive minutes, it means they are imposing a tempo that forces the opponent to respond either by playing long or by lunging into tackles to escape the press.

In this window, Real Madrid lost possession four times in their own half. All four occurred when this team's centre-backs were forced to hold the ball longer than 2.5 seconds — a threshold I call the hesitation threshold. When a centre-back holds the ball beyond 2.5 seconds under high pressing, the probability of an error or foul spikes. And Huijsen was the defender holding the ball longest in that period, averaging 2.9 seconds per touch.

Low PPDA does not decode Atlético. It decodes the football Real were hiding inside a glamorous shell of possession.

By minute 62, Huijsen received the ball near the left touchline, roughly 35 metres from his own box. Ahead of him, two Atlético players had already moved into pressing positions. Normally, a centre-back would play back to the goalkeeper or look for a channel pass. Huijsen chose to hold. He held for 3.1 seconds. Then he lunged.

That lunge was not a moment of madness. It was the result of a chain of decisions distorted by pressure Atlético had built across the preceding seven minutes. When a team drives PPDA below 8.0 for ten minutes, they are not merely applying physical pressure — they are applying cognitive pressure. Opposing players start holding the ball longer, hesitating more, and when they hesitate, the natural reflex is to choose a decisive solution — sometimes decisively wrong.

The 63rd-minute red card was not the cause. It was the symptom.

Referees, VAR, and the lesson from an old name

On referee Ortiz Arias. This is a figure whose data I have tracked across the last three seasons. He belongs to the group of referees with an above-average tendency to brandish straight red cards — roughly 0.28 direct reds per match, against a La Liga average of about 0.17. He also belongs to the group least likely to use the VAR monitor to review his own decisions — only about 12% of incidents reviewed at the pitch-side screen.

Those two figures combined reveal an officiating style: high decisiveness, high confidence in direct judgement, low dependence on assistive technology. With such a referee, a two-footed lunge in midfield in a derby would almost certainly produce a straight red card. That is not emotional prediction — it is a behaviour model built from data.

But here is a point I want my readers to note carefully. When a referee has an above-average tendency to issue straight reds, that does not automatically mean he is wrong. Nor does it automatically mean he is right. It simply means the referee variable must be built into the analytical model, much as one builds weather variables into a transfer forecast model.

The pandemic season did not erase data. It stripped away the glamour and left the skeleton of the game.

That is why I never write a post-match analysis without a refereeing data section. The referee is not a supporting character. He is a tactical variable.

Now to the hardest part: the name Negreira. In that evening's press conference, Mourinho referenced memories of Spanish football's refereeing controversies, including indirect allusions to past cases involving relationships between referees and clubs. I will not go deeper into specific accusations, because doing so means stepping into territory a data journalist should not enter: adjudicating cases for which he does not hold the complete raw evidence.

But I will say this. When a manager brings colour-printed photos to a press conference and lays them on the table, that is not an emotional act. It is a calculated communications act. It aims to shift the narrative from an on-field defeat to an off-field injustice. Technically, as media strategy, it is a clever move. As a result on the pitch, it changes nothing.

I reviewed the 14 photographs presented at that press conference. Some incidents genuinely warranted review. Others were routine collisions present in any match. Over one afternoon, I logged the file: 14 incidents, of which 6 showed clear signs of physical contact, 5 were at the level of ordinary ball contests, and 3 lacked sufficient data for assessment due to incomplete camera angles.

That is an exercise I recommend to any sports journalist: whenever someone presents a set of evidence, categorise it by confidence threshold. Never accept the whole package, and never reject the whole package.

After the red card: what actually changed

This is the section where I believe my data has the most value, because it runs counter to the intuitive conclusion of most post-match reports.

In the 27 minutes remaining after the red card, Atlético Madrid scored one goal and controlled 58% of possession. That figure seems to confirm the familiar story: the red card broke the match, the team with the extra man exploited the advantage. But look at another metric. Atlético's shot count in those 27 minutes was four, for a total xG of 0.47. In the preceding 63 minutes, they had five shots for a total xG of 0.38. In other words, in terms of the quality of chances created, Atlético were not much better after their opponent went down a man. They simply had more ball.

What is more interesting lies with Real Madrid. After the red card, Real dropped their PPDA to 11.2 — meaning they essentially abandoned high pressing and retreated into a medium defensive block. But their passing accuracy rose from 87% to 91%. This is a phenomenon I have logged in many matches involving red cards: the team down a man sometimes passes more accurately, not because they play better, but because they aim for safer, lower-risk passes.

The consequence: the red card did not shatter Real's structure. It simply pushed them into a different structure — one they could still operate, but one without the capacity to generate enough pressure to score. Atlético's winning goal came in the 81st minute, after a counterattack in which Real's defence, with one fewer centre-back, exposed a gap through the middle — the very gap they had controlled all match.

A small discovery in some match sounds like a whisper, but a few seasons later it becomes a roar in decisive fixtures.

What does that mean for a data tracker like me? It means: in post-match analysis, the right question is not whether the red card decided the match. The right question is: which structure pushed the team into a position where a red card became the deciding factor?

I do not need to see how many matches they play. I need to see how many metres they run in a situation nobody notices.

And in this derby, the unnoticed moment was the seven minutes before minute 63 — when Atlético drove PPDA below 8.0 and began controlling the cognition of Real's defence.

The possession shell and the data trap

There is a phenomenon I have tracked for years across European domestic leagues, especially in major derbies. It is what I call fake possession.

The team controls 60% of possession, completes several hundred more passes than the opponent, fires more shots, yet loses. Broadcast media will call it an injustice. Data will call it an exposed structure.

In this derby, Real Madrid controlled 61% of possession in the first half. But look at their shot count from inside the box: only two across the entire first half. Atlético had three. This is the figure I use to distinguish two entirely different types of possession:

Active possession: the team retains the ball heavily and creates many chances from dangerous areas. Passive possession: the team retains the ball heavily but can only create chances from outside dangerous areas.

Real Madrid in this match belonged to the second category. And the notable thing is they could still have won without the red card. Because sometimes a passive-possession team still wins through individual quality in a moment. But when stripped of a man, that individual quality is no longer enough to compensate for a deficient structure.

This is why I believe the 63rd-minute red card, though decisive for the scoreline, was not the tactically decisive moment. The tactically decisive moment was seven minutes earlier, when Atlético realised they could control Huijsen's cognition by driving PPDA below 8.0.

In that period, I logged a recurring behavioural pattern in Real's defence: after each loss of possession in midfield, the ball-carrying centre-back tended to hold the ball 0.5 seconds longer than his usual habit. This is what sports psychologists call increased cognitive load. Players under cognitive pressure tend to process the ball more slowly, and when processing more slowly, their decisions are made under higher tension.

No algorithm replaces observing human behaviour. But one metric can indicate when human behaviour begins to distort. In this match, that metric was PPDA — from an average of 9.2 in the first half, Atlético shifted to 7.6 in the 55-63 window. That is a PPDA level close to what Croatia achieved against Argentina at the 2026 World Cup.

Croatia 2026 did not only suppress opponents' play through PPDA. They suppressed through the patience of a side that knew it was counting every beat.

And Atlético Madrid, in those seven minutes, counted every beat in exactly the same way.

The contrarian angle: when the red card is not the turning point

Now I come to the section I know will irritate my colleagues in Melbourne and Sydney. I want to say that the 63rd-minute red card, viewed through data, was not the turning point media labels it.

I will use a hypothetical test I apply regularly. Had this match continued 11 versus 11 for the final 27 minutes, would the result differ? I cannot answer with certainty. But I can rely on one fact: across the 63 minutes before the red card, Atlético had created more chances from dangerous areas than Real, despite controlling less possession. They were on an upward PPDA trajectory. They had scored once from the penalty spot. The structure was tilting gradually toward them.

In other words: if I had to bet on an 11-versus-11 scenario for the remaining 27 minutes, I would give Atlético the greater chance. Not because they were better individually — but because they owned the structure better suited to the closing phase.

This is the kind of conclusion data permits me to reach, and it is also the kind emotional commentators struggle to produce. When you watch footage at normal speed, you see a red card. When you rewind at 0.25x, you see a structure. The red card did not create that structure. The red card was created by it.

I want to issue a caution. I am not saying the red card was unimportant. It was hugely important for the scoreline, because it reduced Real's ability to maintain a medium defensive block without gaps through the middle. But I want to distinguish two kinds of importance:

Scoreline importance: what changes the number on the board. Structural importance: what changes the probability of subsequent events.

The 63rd-minute red card belongs to the first. The seven minutes before it, when Atlético drove PPDA below 8.0, belong to the second.

Selective confrontation, as I still tell younger colleagues, does not mean attacking many targets. It means choosing the right target. In this derby, the right target is not the referee, not the red card, not VAR. The right target is the structure both teams operated in the seven minutes before minute 63.

Data never lies — but I needed ten years to know when it tells half a truth.

And the half-truth in this derby was this: broadcast statistics said Real dominated. But the shot-zone map said the opposite.

Hjulmand, Rudiger, and unread variables

One of the things I hate most when reading post-match analysis is the treatment of defensive players as extras. In this derby, there are two names my data says deserve closer reading: Hjulmand and Rudiger.

Hjulmand, Atlético's holding midfielder, possesses a metric I track all season: interceptions before the opponent receives the ball. In this match, he registered 11 such actions — the highest in the game. What few noticed is that 8 of those 11 occurred before minute 63, meaning Hjulmand was the man who established Atlético's pressing structure before the red card happened. He did not benefit from the red card. He was the enabler of it.

On Rudiger, the notable figure is not his duels but the number of times he had to cover space behind teammates. After Huijsen left the pitch, Rudiger had to run an average of 1.4 km over 27 minutes — equivalent to a midfielder's output in the same span. That is a workload no centre-back can sustain for 90 minutes, and it explains why the second goal came from a gap through the middle.

As for Romero and Lee Kang-in, two attacking variables introduced in the second half, they played different roles in the story. Romero came on when Atlético needed a fresh attacking outlet to exploit Real's tired defence, while Lee Kang-in came on when Real needed a ball-carrier to stretch time and generate set pieces. Neither scored, but their movement data reveals two different strategies being deployed.

In the 70-80 window, Lee Kang-in ran 1.8 km — the highest of any player on the pitch in that period. He ran not to score, but to drag Atlético's defence out of position, creating space for teammates. And precisely because he ran so much, Atlético's defence was forced to stretch, opening a gap they could not close in the 81st-minute counterattack.

The empty stadium of 2026 did not make players weaker. It exposed the artificial metrics that crowds had once shielded.

In this derby, the crowd had returned. And precisely because of that, some artificial metrics were shielded once more. That is why I always prioritise automatically collected movement data over manually recorded data.

What the final scoreboard does not say

The final score was 2-1 to Atlético Madrid. Three points stayed in the capital, and the refereeing story will be replayed for weeks. If you read this article and carry away a single conclusion, I want it to be this: a football match does not have a single turning point. It has a chain of turning points, and how we choose which one to tell the story around says more about us than about the match.

Across the last three seasons, I have logged more than two hundred derbies across European leagues. The most repeated behavioural pattern I have found is this: the teams that win derbies are always the ones that control their opponent's cognition in the ten-minute window before a decisive event. Not the team that scores first, not the team with more shots, not the team with more possession. The team that controls cognition.

I do not intend to turn this conclusion into a formula for every derby. In football, every behavioural pattern has five exceptions, and a behavioural pattern with five exceptions can still be a valuable behavioural pattern. It means: if you are looking for a signal to track in the next derbies, pay attention to PPDA in the ten-minute window before the match produces a major event — a red card, a penalty, or a goal. If a team drives PPDA below 8.0 in that period and holds it steady, the probability they generate the decisive event is significantly higher than a team with average PPDA.

But I want to close with a reminder about the limits of this method itself. Last season, I used a PPDA model to predict a major Serie A derby. My model suggested the away side would win. The away side lost 0-3. On review, I found the away side had driven PPDA below 8.0 as predicted, but they did so in the first half, when both teams were at full strength. In the second half, once the opponent adjusted their structure, their PPDA climbed back. My model was not wrong in mechanism, but it was wrong in timing.

That is why I always remind myself before every piece: the question is not what the data says. The question is when the data says it.

In this Madrid derby, if you gave me a single metric to summarise the match, I would not choose the 2-1 scoreline. I would choose the moment Atlético's PPDA dropped to 7.6 at minute 55. That is the moment this match was decided. Everything that followed, including the red card, was merely the consequence of a chain of decisions that began there.

At that moment, referee Ortiz Arias stood in the middle of the pitch, yet to make any significant call. Huijsen was still on the field. VAR Trujillo Suárez had not yet been summoned. Mourinho had not yet printed a single colour photo. And several journalists in the media area still believed Real Madrid were controlling the match.

Only the data said otherwise.

In modern professional football, the most frightening thing is not a stronger team. The most frightening thing is a team that controls your cognition for ten minutes, then lets everything else collapse through inertia. Atlético Madrid did that in this derby. And they will do it again, not because they were lucky, but because they own a repeatable model.

The empty stadium. Only the data knows who actually ran.

That was the line I wrote in Melbourne during the pandemic season. And this season, with crowds back, I still hold that line — only changing the subject. It is not the crowd that hides data. It is the storyteller's emotion.

I do not need the red card. I need the ten-minute window before it happened.