Who Is Allowed to Trigger VAR? Anatomy of the Decision Process at Major Tournaments

On 16 June 2026, at the Kazan Arena, Uruguayan referee Andrés Cunha allowed play to continue after a challenge between Antoine Griezmann and Joshua Risdon. No whistle. No card. Griezmann lay on the grass, the stands behind the Australian goal roared, and nobody on the pitch knew that a process had just been triggered in a sealed room a few hundred metres away. Ninety seconds later, Cunha left his position, ran to the touchline and bent over a monitor. He straightened, pointed to the penalty spot. Antoine Griezmann scored, and France won 2-1.

That was the first penalty in World Cup finals history awarded through VAR. For most viewers the story ends there. For me it begins the moment Cunha stood up. The decisive moment of the whole incident did not happen on the Kazan turf; it happened in a room where four people sat in front of a bank of screens, and where a video assistant referee had to answer one question: is this challenge serious enough to ask the on-field referee to take another look?

Football does not have VAR. It only has blind spots that are waiting to be exposed.

A protocol written in deliberately vague language

IFAB approved VAR trials in 2026, after years of argument over whether judgement calls should be handed to machines. The system first appeared in a competitive match in the Netherlands, then spread to competitions in North America, Australia, Germany, Italy and Spain. By the 2026 World Cup in Russia, all sixty-four matches were covered, and more decisions were overturned than in every previous World Cup combined.

The Premier League adopted the system later, from the 2026-20 season, and quickly became the most VAR-contested league in Europe. The cause was not the technology. Different competitions operate the same IFAB protocol but interpret the intervention threshold differently, and it is that divergence in interpretation that produces the enduring sense of injustice among supporters.

The IFAB protocol limits intervention to four match-changing categories: goal or no goal, penalty or no penalty, direct red card — second yellows excluded — and mistaken identity. Outside those four, the video assistant referee may not speak, even when a clear offence is right in front of him. A foul in midfield that leads to a second yellow and then a red, whether inside or outside the box, is not within VAR's authority.

The threshold is phrased as "clear and obvious error" or "serious missed incident". That wording is deliberate, not laziness on the part of the drafters. IFAB wanted to preserve the principle that the on-field referee is the final decision-maker, and that VAR intervenes only when the error is large enough for an entire stadium to spot without a replay. The problem is that "large enough" has no unit. No centimetres, no percentages, no frame count.

Who actually presses the button?

Across three consecutive seasons watching Premier League and European matches, I logged the sequence of referee actions in every overturned incident. One detail is almost always misreported: the power to trigger the process belongs neither to the on-field referee nor to the coaching staff. Neither has a button. The on-field referee can only request a review; opening the process is the VAR team's decision.

A standard VAR team comprises a lead video assistant referee, a second assistant, an offside-camera operator, and in some competitions a support official. They sit in a central control room monitoring multiple feeds in parallel. When one of the four categories occurs, the lead assistant checks slow-motion angles, compares the first and last frame of the incident, and decides whether to invite the on-field referee to the monitor.

There are two outcomes. One is a silent check that nobody on the pitch notices, and play continues — the vast majority of cases. The other is the on-field referee being called to the touchline screen, at which point the formal process begins. Once the referee is at the monitor, aggregated figures from major competitions show that most reviews end in the original decision being overturned. In other words, once the process opens, the probability of reversal is high enough that the outcome can be predicted before the referee touches the screen.

That is the detail I consider most important in the entire system, and it is rarely discussed. A video assistant referee only recommends a review when he already believes the on-field decision is wrong. The touchline monitor is not a place of discovery; it is a place of confirmation. The referee walks out there with a recommendation in his earpiece, and in most cases he walks out there to agree.

VAR does not correct mistakes. It only changes who carries the blame.

The threshold applies one rule to two different scales

Most of the inconsistency supporters perceive at major tournaments comes from the protocol applying a single threshold to two categories of incident that are fundamentally different in nature.

The first category is objective: whether the ball crossed the line, whether it went out of play, whether a player was offside, whether the ball struck an arm inside the box. Here the answer exists independently of the observer. An offside line is right or wrong regardless of who sits in the VAR room.

The second category is subjective: whether the intensity of a challenge constitutes a foul, at what point an arm makes the body unnaturally bigger, whether a tackle from behind is serious enough to merit a red card. Here two good referees sitting side by side can reach opposite conclusions and both can justify them under the laws.

In the first category, the "clear and obvious" threshold works reasonably well because it can be measured. In the second, the threshold cannot function, because there is nothing to measure. What the protocol calls "clear" is in practice the VAR team's opinion aligning with the majority opinion of television viewers. That is a social standard, not a technical one.

The 2026-19 Champions League quarter-final between Manchester City and Tottenham is a textbook collision of these two scales. Raheem Sterling scored in stoppage time, the Etihad erupted, and the goal was disallowed for an offside in the build-up, established by a line measuring a few centimetres. Nobody in the stadium could see that gap with the naked eye. The goal was taken away by a measurement, not by a judgement.

Semi-automated offside and the relocation of power

The 2026 World Cup in Qatar marked the first official deployment of semi-automated offside technology at a finals. The system uses twelve tracking cameras and dozens of body points per player, including points attached to the parts of the body the laws permit to score, and relays a signal to the VAR room within seconds.

The group-stage match between Japan and Spain is the most cited example. Kaoru Mitoma kept the ball in before cutting it back for Ao Tanaka to score, and the system determined the ball still had a portion over the touchline. FIFA published the measurement: roughly 1.88 millimetres. The goal stood, Spain lost top spot in the group, and an entire bracket shifted because of a margin thinner than a thread.

What stands out is not the measurement. What stands out is the public reaction. When an offside decision is produced by a measuring system, supporters attack the law rather than the referee. When a similar decision is produced by the human eye and a hand-drawn line, supporters attack the referee. Technology has redirected discontent from individuals to institutions. That is progress in governance, even if it has not made anyone happier.

Viewers see the incident, referees see the moment, I see the whole process.

The handball law and the moving-target problem

Over the same period, IFAB amended the handball law three times in four years. From the 2026-20 season, goals scored directly or created by hand or arm were disallowed even when the contact was entirely accidental, and an "unnaturally bigger body" criterion was introduced to define the offence.

The following season, IFAB had to add a definition nobody had previously thought necessary: the boundary between shoulder and arm is the bottom of the armpit. Such a definition only appears when a series of incidents exists in which neither referees nor the public can agree which body part the ball struck. The "unnatural" criterion was then refined by requiring officials to consider whether the arm position was a consequence of, or justifiable by, the player's body movement in that specific situation.

Each amendment resolves one group of incidents and creates another. This is the familiar mechanism of any legal system: the more detailed the definition, the more boundaries exist, and every boundary is a potential point of dispute. A law written as "abnormally extended arm" gives referees broad but vague discretion. A law written as "bottom of the armpit" gives referees a precise marker that means nothing to supporters. Neither version fully satisfies both needs.

Laws never stand outside the match; they are a second match running in parallel.

At major tournaments, pressure compresses emotion and compresses error

The pandemic-delayed Euro 2026 finals remain the clearest illustration of how a major tournament turns a penalty shootout into a process test. The final between Italy and England at Wembley on 11 July 2026 finished 1-1 after extra time, and Italy won the shootout 3-2. Three England substitutes, Marcus Rashford, Jadon Sancho and Bukayo Saka, all failed to score. Rashford hit the post; Sancho and Saka were denied by Gianluigi Donnarumma.

Based on my experience watching matches, I spent many evenings reconstructing every kick of that shootout — not to assess player psychology, but to measure the time a goalkeeper actually has. The distance from the penalty spot to the goal line is eleven metres, plus the distance a keeper is permitted to advance before the ball is struck. For a shot at the average speed of professional football, that interval is under a second, and most of it is insufficient for reaction — it is only enough for a pre-decision. A goalkeeper does not read the kick after the ball leaves the foot; he places a bet before the ball leaves the foot, based on a dossier of the taker's habits.

This is why the law forbidding a keeper from leaving the line before the kick is so hard to enforce. The boundary between a legitimate bet and an early infringement is a few hundredths of a second thick. At a major tournament, where every shootout kick is replayed in slow motion from multiple angles, penalising a keeper for advancing becomes a decision almost impossible to make without generating controversy. The referee must choose between ignoring a real infringement and denying a team its chance over an error nobody can see.

The referee is the only person on the pitch who is not allowed to be guided by emotion.

The contrarian view: the protocol does not fix error, it redistributes responsibility

My disagreement with the majority concerns expectations, not VAR's effectiveness.

The "clear and obvious error" threshold is not a technical flaw waiting to be corrected. It is a political compromise between two sets of interests. One side wants every major decision verified by imagery. The other wants to preserve the continuity of the match and the final authority of the on-field referee. A high threshold sits between them, and its vagueness is not a defect — it is the condition that allows the compromise to exist.

The same applies to speed. Viewers tend to read the slowness of a review as a sign of weak competence. In most cases, a long review is a sign of a process running correctly: cross-checking angles, fixing the frame, locating the point of contact, verifying the preceding frame. A fast process on a difficult incident is a process that skipped a step. The impatience of the crowd and the precision demanded by the laws pull in opposite directions, and no technical fix satisfies both.

There is one more point rarely stated in these debates. VAR has not made referees more powerful; it has made them more exposed on every decision, because every decision is now examined from twelve angles and compared against a standard its own authors cannot quantify. The number of wrong calls has fallen, but the number of contested calls has risen, because the new standard makes every decision procedurally challengeable.

What comes next

The sensible next step is not making VAR faster, but separating the two categories of incident currently lumped together. For objective incidents, measurement technology should be pushed to full automation, so that whether a goal stands is answered by the system rather than by human judgement. For subjective incidents, the intervention threshold should be published as more specific criteria with reference examples, so that the public at least knows what the system is trying to achieve.

Another possibility, rarely discussed but emerging in other sports, is a limited challenge system for teams. If each side could request one review per match, the "clear and obvious" threshold would depend less on the VAR team's judgement and more on the accountability of the party requesting it. That is a systemic change, not a technical one, which is exactly why it is hard to pass.

I will keep logging every decision, every frame, every time a referee touches the touchline monitor. After years of this work, the thing I believe most firmly is that a transparent system is not a system that makes no mistakes. It is a system in which, after a mistake, people can still point to exactly which step went wrong. The only remaining question is which step.