VAR Intervention Threshold at Euro 2026: Three Incidents, Three Outcomes, One Process Failure

On 10 July 2026, at Signal Iduna Park in Dortmund, I sat in Manchester with two windows open on the same screen: the live feed of the Netherlands–England semi-final on one side, and the VAR protocol checklist I use whenever I follow a major tournament on the other. In the 15th minute, Harry Kane struck a shot inside the box, Denzel Dumfries came across, and there was contact. Referee Felix Zwayer did not award a penalty immediately. I pressed record. About thirty seconds later, Zwayer's earpiece crackled. He stopped play, drew the television screen in the air, and jogged to the touchline.

By then I no longer needed to rewatch the incident. What I needed was the sequence: who in the VAR room proposed opening the review, at what second, and how long Zwayer stood at the monitor before changing his decision. Those details never make it to broadcast. They live in the match report, and most viewers never touch the match report.

That same night, both England and the Netherlands had reason to believe they had been treated unfairly. What interested me was not that feeling. What interested me was that nine days earlier, in Stuttgart, an incident with a similar structure had ended in the opposite direction; and forty-one days before that, also in Dortmund, a third incident had ended in a third direction.

Football has no VAR, only shadows waiting to be exposed. And the three cracks of Euro 2026 opened exactly where nobody wanted to look: in the referee appointment process and in how the intervention threshold is calibrated.

Context: one protocol, four categories, one vague threshold

VAR first appeared at a World Cup on 16 June 2026, in France–Australia in Kazan. Antoine Griezmann went down inside the box, the referee did not award a penalty, the VAR team invited him to the monitor, and after the review the decision was overturned to a penalty. I was twenty-seven that year, working as a data editor for a football site in England, and I stayed behind after hours to log every moment a referee left the pitch and walked towards the screen. I recorded forty-seven VAR interventions across that tournament, and what I realised earlier than my colleagues was simple: the intervention threshold has never been stable.

IFAB, the body that writes the laws, defines VAR as usable in only four categories: goals, penalties, direct red cards, and mistaken identity. The condition for intervention is a clear and obvious error, or a serious missed incident. The interesting words are "clear" and "obvious". They describe a level of human perception, not a measurable quantity. There is no millimetre count, no percentage, no numeric threshold written into the laws.

That means every match runs on a scale decided by two people: the referee and the VAR. That scale is calibrated by experience, by pre-tournament training, and by technical meetings the public never hears. At Euro 2026, held from 14 June to 14 July 2026 across ten German cities, with twenty-four teams and fifty-one matches, that scale was placed under the heaviest pressure European football has ever carried.

Three technical factors made things harder. First, the tournament debuted connected ball technology with a sensor transmitting at 500 Hz, which can pinpoint when the ball touches an object. Second, semi-automated offside technology was deployed alongside it. Third, image standards were raised sharply, meaning every contact incident had at least five crisp camera angles.

Those three factors sound like progress. They create a paradox: when observational capacity rises, the boundary between "foul" and "no foul" becomes blurrier, because people see more of what the law never quantified. I wrote in my Anatomy of VAR series that VAR does not make referees more accurate; it makes them more transparent at some stages and more opaque at others. Euro 2026 is the fullest evidence of that claim.

Analysis: three incidents, three outputs

Incident one – Denmark and Germany, 29 June 2026, Dortmund

This is the incident with the highest level of technical transparency of the three. In the second half, Joachim Andersen of Denmark was penalised for a handball inside the box. Connected ball technology confirmed contact, and referee Michael Oliver awarded Germany a penalty.

The mechanism matters. With sensor technology, this incident leaves the zone of judgement and enters the zone of measurement. There is no longer a question of whether the ball touched the hand, only whether the arm position fell outside the body line in an unnatural way. And that second question is still judgement. The result is a decision half measured, half sensed, presented to the public as though it were entirely data-driven.

What few viewers noticed: minutes earlier, Andersen himself had scored and the goal was disallowed for offside detected by semi-automated technology, by a margin of a few centimetres. I logged the sequence: goal disallowed, then a penalty awarded to the opponent. Within that short span, one team was stripped of a goal by data and then punished by a judgement. That is the psychological structure that made the Danes unable to accept the result, however correct the process may have been.

Incident two – Spain and Germany, 5 July 2026, Stuttgart

This was the most contested incident of the tournament. In extra time, Jamal Musiala's shot struck the hand of Marc Cucurella inside the Spain box. Referee Anthony Taylor did not award a penalty. The VAR checked and decided not to invite the referee to the monitor.

In my checklist, this is the fracture point. If we take the standard applied to Andersen in Dortmund as the benchmark, Cucurella's arm in a high-speed, close-range incident sits in a similar grey zone. But the standard was not applied. The usual technical explanation is that reaction distance was too short, and the arm movement originated from a balance-keeping effort rather than a deliberate expansion of the body's area. That explanation has a basis in law. The problem is that the same explanation was not used to protect Andersen.

I spent nearly four hours rebuilding both incidents on a single timeline at identical frame rates. Both fell below the reaction-time threshold that referee coaches often call the zone of exempt liability. The difference was that one incident had a measuring tool intervening, and the other had only human eyes. And when two incidents are identical in nature but different in tooling, the public reads the difference as bias, not as a difference in method.

Every foul is a question about intent; data only answers questions about consequence. That is precisely what happened in Stuttgart. Technology answered the consequence question, while the intent question remained inside the head of a man standing fifteen metres from the ball.

Incident three – Netherlands and England, 10 July 2026, Dortmund

Back to the incident I recorded at the start. Zwayer did not give the penalty immediately. VAR invited him to the monitor. He reviewed it, then awarded England a penalty. Kane equalised in the 18th minute, and England won 2-1 through Ollie Watkins in the 90th.

What matters here is not the conclusion but the duration. In my notes, Zwayer stood at the monitor longer than the tournament average. That does not prove he hesitated under pressure, but it shows the uncertainty level of the incident was high. For a foot-to-foot contact inside the box, the deciding factors are contact force and direction of movement, two quantities a 24-frame-per-second image cannot render accurately.

This is the crux the media skipped. Every post-match debate circled the question "was it a foul". The correct question is whether the intervention threshold in the first half of a semi-final matches the intervention threshold in extra time of a quarter-final. From the three incidents above, the answer is no, or at least it was never demonstrated.

I checked a similar incident I had followed since 2026: Joshua Risdon's contact with Griezmann in Kazan. The structure almost mirrors the Dumfries–Kane incident, yet the intervention threshold then was markedly lower. Six years later, the same kind of contact, the same protocol, two different standards. For someone who reads match reports for a living, this is more important data than any argument about whether the decision was right.

What the data shows

In my Euro 2026 log, the handball-in-the-box category accounted for the largest share of VAR reviews, and also the highest overturn rate. That is a predictable result. Handball is the only category where the law depends on a geometric definition drawn by humans — the body line — while both the ball and the arm are in motion. Foot-to-foot contact incidents have a lower overturn rate but higher uncertainty, because contact force cannot be seen.

In other words, VAR's risk map is not randomly distributed. It distributes according to how far a physical event can be converted into a number.

That is why I refuse to join debates of the "this referee favours that team" kind. Those debates have no output, because they have no yardstick. What does have a yardstick: decision speed, frequency of monitor visits, the number of times the VAR overruled itself, and the gap between standards applied to the same category at different stages of a tournament.

The contrarian view: the problem is not at the corner flag, it is in the meeting room

There is a reflex I consider structurally wrong: after every contested decision, the public demands that referees explain, apologise, and be disciplined. That approach handles individuals and ignores systems.

At Euro 2026, the story should have been placed somewhere else. In the process of appointing referees to major matches.

In Dortmund on 10 July, Felix Zwayer was assigned the semi-final. He is a referee who was banned for six months by the German Football Association in 2026, in connection with accepting money from Robert Hoyzer during the 2026 match-fixing scandal. He was the key whistleblower in that affair, but he was also the man who took money he should have declared immediately. In December 2026, Jude Bellingham was fined forty thousand euros for comments alluding to that past after Dortmund–Bayern.

My question is not whether Zwayer is competent. My professional conclusion is that he is, and there is no evidence his Dortmund decision was shaped by anything other than reading the incident. My question belongs to process: why a background file with an obvious sensitivity did not trigger any additional review step before he was handed a European Championship semi-final.

This is the kind of question I ask daily in my work: who is authorised to trigger the process. In VAR, the answer is the VAR, with two layers of control. In referee appointments, the answer is not published. UEFA's referees committee has criteria, rankings, and post-match evaluations. But those criteria are not public, which means any error automatically becomes suspicion.

The referee is the only person on the pitch not allowed to be guided by emotion. But the referee is also someone who should not be suspected simply because the appointment process lacks transparency. Those two statements do not contradict each other. They complement each other.

Here is a counterintuitive point worth making: if UEFA had published full appointment criteria and the reasoning behind choosing Zwayer for the semi-final, there is a strong chance Zwayer would have been removed from that assignment before the match. A transparent process does not only protect referees. It also protects the tournament from putting itself into an unnecessary controversy.

A second counterintuitive point: people often say VAR destroys the emotional continuity of football. But looking at my Euro 2026 log, VAR does not remove emotion. It compresses emotion into a waiting period, and during that wait, viewers write their own script. When the final decision contradicts that script, the reaction is no longer disappointment about football, but outrage about fairness. This is a reaction data cannot fix, because data cannot answer the question of fairness.

Let me be clear: I do not believe Zwayer was wrong. I believe the system placed him in a position where any decision would be read through the lens of his record. When a technical decision is read through a record, the process has failed before the referee blows the first whistle. VAR does not correct mistakes; it only changes who carries the blame. And at Euro 2026, the blame carrier was selected in advance by an unpublished process.

The appointment process under an investigator's eye

I once mispronounced Kyle Walker's name three times in a row during England–Slovakia at Wembley in 2026, and that night I received an email from my editor-in-chief. To calm myself, I reopened the recording and rewatched every phase of play to understand how a referee precisely determines offside. What I learned that night had nothing to do with offside. It had to do with how a small error in a process can be turned into a story about the character of the person who made it.

Referee appointment works on the same logic. A list of dozens of referees is ranked by performance, fitness, big-match experience, and post-match evaluation. Within that list, each person has a background file containing events unrelated to officiating competence. The problem is: which department assesses the media risk of that file, and which criteria are used.

I have asked this question at many press-conference workshops. The answer is usually "we assess all factors". That is a reasonable and useless answer, because it does not tell you which check was performed, who signed it, and who holds veto power.

Viewers see the incident, referees see the moment, I see the whole process. The process here includes the period before the ball rolls: who was chosen, for what reason, and how much was disclosed.

Process lessons from three decisions

In my study of Liverpool's tactical fouls under Jurgen Klopp, conducted during the 2026 Premier League suspension, I found the side committed an average of 10.2 fouls per match, mostly in midfield to stop counter-attacks. When I sent that analysis to an editor, he rejected it because nobody reads about football when football is not being played. I still wrote twenty pages of notes, and those notes became the basis for the e-book on law and tactics I published in 2026.

What I took from Liverpool then applies to Euro 2026. A tactical foul is not a player's failure. It is a language of spatial control, repeated deliberately. How a team fouls tells you which zone it is protecting. And how an organisation handles its own errors tells you what it is protecting.

At Euro 2026, three handball and contact incidents inside the box produced three different conclusions. Read as three individual mistakes, we go hunting for three referees to criticise. Read as three data points inside one system, we go looking for inconsistent threshold calibration and for background files that were never risk-assessed.

I lean towards the second reading, because it produces work to do.

First, the intervention threshold should be codified into specific categories with observable criteria, similar to how offside law was made concrete through semi-automated technology. There is no need to turn every contact into a number. There is only a need to publish that category A has criterion B, and that when criterion B appears, VAR must invite the referee to the monitor.

Second, the audio between referee and VAR should be released in full after the match, not selectively per incident. Selective release only prolongs the argument.

Third, the appointment process should be published at a level that lets outsiders verify the steps: rankings, background risk assessment, and the approval mechanism.

The law never stands outside the match; it is the second match played in parallel. At Euro 2026, that second match was more contested than the real one.

What I want to see at the next tournament

In my analysis of the penalty shootout at Euro 2026, I wrote three thousand words on Gianluigi Donnarumma choosing the right direction five times out of seven, based on studying opponents' standing-foot habits, with eleven metres from the spot to goal and an average ball flight time of about 0.3 seconds. The editor cut it to five hundred words because it was too technical. The full piece I kept said one thing: in a window shorter than a breath, decisions are made on the basis of prior preparation, not on-site observation.

That is true of goalkeepers, and it is true of referees.

A referee in the 15th minute of a semi-final does not decide based on what he sees in the moment. He decides based on the trained intervention threshold, on how he was instructed to handle this category, and on how much trust he places in the person sitting in the VAR room. If those three variables are not published, every post-match debate is a debate about belief.

I want to see the next tournament publish those three variables before the ball rolls, not after a controversy. I want to see calibration criteria placed side by side across referees within the same tournament, so viewers know whether the scale is shared or individual. And I want to see the appointment process written down, so a sensitive record is not turned into one person's morality tale.

When three structurally similar incidents produce three different conclusions, the problem is not in the referee's eyes. It is in the fact that those responsible for calibration never disclosed how they calibrated. Football will always have arguments. But an argument about a decision is different from an argument about a process. Euro 2026 mixed the two, and the referee is the one who paid for the mixture.