When the Slow-Motion Replay Never Arrives: Refereeing and the Limits of Evidence

Minute 56 at San Siro, November 2026, Milan hosting Juventus on matchday 12 of Serie A. Gonzalo Higuaín put the ball across the line and made it 2-0 to the visitors. In the VAR operations room, my monitor carried a camera angle from behind the goal: the tip of Higuaín's boot was roughly 0.2 metres beyond the shoulder of the last Milan defender. I hesitated for exactly three seconds, and in those three seconds the fear of being wrong outweighed the probability that the frame was right. I did not recommend an on-field review. Milan lost 0-2. After the match, the referee supervisor stood in front of the whole officiating team, rolled the clip back and read out my name.

The story does not end with that incident. Four months later, during a routine audit, we found something worse: on certain matchdays the VAR room received an empty feed — no camera angle, no data, no log — and the system recorded that outcome as "no incident". A phase of play with no evidence was processed as a clean phase of play. That was the moment I understood that the job of the person sitting in the dark room is not to look very hard, but to know precisely what is missing.

Serie A brought VAR into operation from the 2026-18 season, and the early phase was an experiment in intervention thresholds. The protocol states clearly that the video referee intervenes only for a "clear and obvious error". The protocol says nothing about the worst case: when the evidence never arrives. Over the first twelve months I re-watched 47 offside situations similar to the Higuaín one, logged every parameter and built a 37-criterion checklist — distance, camera angle, frame rate, the moment the ball leaves the foot, the tilt of the shoulder, the running speed of the last defender.

That checklist produced a professional conclusion: most VAR controversies live not in the law but in the quality of the input. Referees err when the frame is blurred. Referees err far more when the frame does not exist and nobody dares to say so. From the 2026-19 season, UEFA added a requirement to log even non-interventions. A small procedural change, a large cultural one: a decision to do nothing must also be accounted for.

Inside the VAR room, every incident is graded on two columns: intervention correct or incorrect, and non-intervention correct or incorrect. There is no third column for missing data. An officiating team that lost its camera feed midway through a half was still recorded as "no incident requiring action", and that line travelled straight into the weekly report without challenge. For months, the system awarded itself credit for its own silence.

In statistics, an empty sample does not mean the phenomenon is absent; it means the measuring instrument is not working. Football commits this error at far greater scale. A match with no shot on target is logged as "a solid defence". A side that ran 8 km less than its opponent is logged as "managing the workload". A player who did not touch the ball in the final 20 minutes is logged as "having done his job". An empty conclusion is a signal that the measurement system has broken, and it certifies safety for nobody. This is the point the data departments now pushing into dressing rooms tend to skip: they are trained to read what is on the chart, and almost none of them are trained to read what has vanished from it.

In June 2026, at the World Cup in Russia, Sky Sport Italia invited me to commentate on VAR for the round-of-16 tie between France and Argentina. Before kick-off I pulled the data from Kylian Mbappé's last 14 matches in Ligue 1 and the Champions League: a peak sprint speed of 36.5 km/h, 2.8 km/h above the Argentina defensive average. A 1,200-word analysis argued that Argentina's defensive structure would fracture between minutes 60 and 70, the window in which Mbappé accelerates into the space between full-back and left centre-back. France won 4-3; Mbappé won a penalty and scored twice. I do not trust my eyes, I trust the slow-motion replay — and that night the slow-motion replay was a spreadsheet assembled before the ball moved.

In October 2026 Serie A returned to empty stadiums because of COVID-19. Milan went seven matches without a win, and the convenient explanation was that selling André Silva to Monaco for 35 million euros had removed the goals. I did not see it that way. I analysed transition data across the previous 14 matches: Milan's back line lost 42% of its counter-attacking defensive capacity in the phases where crowd noise normally drives the front line to press. A 30-page report went to the editors at La Gazzetta dello Sport and was published in full. That was the first time I wrote to a structure of symptom, root cause and a recovery roadmap with dated milestones.

The counter-intuitive point sits here. The public blames a referee for one bad whistle and blames a coach for one substitution made ten minutes late. What actually ruins a season is the line reading "nothing unusual" in the weekly report, repeated 38 times, read by nobody, questioned by nobody. Emotion can always find an individual to point at; the laws can only ever find a system to repair. A referee has no right to invent a camera angle. An analyst has no right to invent a number. Both of them have the right, and the duty, to say: I do not have enough evidence. Technology does not kill football, it kills blind faith. And the most dangerous blind faith of all is the belief that silence means cleanliness.

The 2026 major tournament cycle has brought semi-automated offside and ball sensors into every matchday. The thicker the technology, the harder data gaps are to hide and the easier they are to ignore, because nobody wants to admit that an expensive system can return an empty result. My proposal fits into a single procedure: publish the non-interventions, publish the empty logs, publish the moments the VAR room lost its signal. Then supporters no longer have to trust the credibility of one person; they can trust a process that can be checked. Every judgement deserves a review, including the judgement of the data. Before I blow the whistle, I review myself. A sport that wants to be believed should start with that same gesture.