Data Corridor: When Referees Already Have Their Answer Before Walking to the VAR Monitor

In the last three matches of the team whose data I have been tracking in the V.League 2026/26, the PPDA index — the number of opponent passes allowed before each defensive action — dropped from 11.4 to 8.7. That signals higher, more aggressive pressing. But the points earned fell from 7 to 2. That number is not on any league table. It sits inside a tracking-data file that only three people in the club's analytics department know how to read. Last Sunday, after a derby that ended at 10 p.m., I stayed until 2 a.m. re-coding every phase of play. There were 17 passes that no mainstream stats page credits as a "chance created" — but if you look at how quickly the opposing back line rotated after each of those passes, a pattern emerges. Defenders turned 0.3 seconds slower. The holding midfielder drifted out of position. A goal was built from there. This is why I do this job. The numbers that are not on the stat sheet. They live between two touches. I work as a data consultant for football clubs. Most of the time I look at what the crowd does not see: the distance between two touches, a goalkeeper's reaction time to a shot, the rotation speed of a defender facing a forward running diagonally. That is the daily work. This season I took on an additional task: analyzing VAR decisions in the domestic league. Not to criticize referees. That is not my habit. But to answer a simpler question: when does VAR actually correct an error, and when does it merely create a new layer of error? I remember a coach once telling me: "You do not need to analyze the match. You need to tell me how the next match will unfold." That is the pressure of this job. No one pays you to describe the past. They pay you to predict the future. But football's future is not in what we know. It is in what we have not yet seen. The domestic league has now passed 94 matches this season. VAR intervened 31 times. Nineteen of those involved penalty or red-card decisions. I downloaded the full dataset, coded each incident, and cross-checked against the motion data from the league's eight-camera system. The results forced me to write this piece. The first finding: the average time for the VAR referee to call the on-field referee to the monitor is 47 seconds. But the average time for the on-field referee to reach a final decision after reviewing is only 22 seconds. That means in more than half of all incidents, the on-field referee already had the decision in his head before walking to the screen. The review only confirmed it, it did not change it. This is not necessarily bad. The on-field referee is closest to the incident. His instinct is usually right. But when VAR exists, people expect it to correct errors the naked eye cannot see. And when the referee already has an answer, the chance he changes it drops sharply — regardless of what the monitor shows. The second finding: of the 31 interventions, 12 times the team that benefited from VAR was the team leading — a rate of 38.7%. That number proves nothing about intent. But it reveals a clear psychological effect. When the leading team is reviewed by VAR, the referee tends to uphold the original decision. When the trailing team is reviewed, the referee tends to change it. This matters for smaller clubs. They are usually the team trailing. They need VAR to overturn decisions to have a chance. But the data shows that when they are reviewed, the probability of being penalized rises. When they need VAR to correct an error for them, that probability is lower. The third finding, and the one that kept me awake: in 12 reviewed penalty incidents, 9 showed via tracking data that the fouled player had already shifted his center of gravity before the contact occurred. They were actively seeking the contact. The eight-camera system does not capture this because it does not record center-of-gravity movement. But the sensors in boots and shirts do. If you read that number and think "diver", you are oversimplifying. A center-of-gravity shift does not mean pretending to fall. It is a natural reflex when a player senses contact is coming — the same way you pull your hand back when touching something hot. No one teaches them. It happens in 0.2 seconds. There is one match I will never forget. The home team conceded a penalty in the 89th minute. The score was 1-1. The referee signaled, VAR reviewed, and after three minutes the decision stood. The away team scored and won 2-1. The next day I downloaded the tracking data and found something no broadcast mentioned: the home defender had rotated 0.4 seconds slower than his season average. He was tired. That does not mean he did not foul. But it means the story we told about that incident — that this defender was too clumsy, that the home team deserved to lose — ignores a more important layer of data: fatigue. This is where I must say something plainly that I know will upset many people. "Clear and obvious error" — the clause IFAB wrote for VAR — is not a technical standard. It is a vague concept written to give referees room for interpretation. But that room for interpretation is not measured. No quantitative threshold. No scale. Every referee has a different "clear threshold." I tried to measure it. Across the 31 VAR incidents this season, the same type of contact — a challenge from behind — was ruled "clear and obvious" in 7 cases and not ruled "clear and obvious" in 5 others. Same season. Same league. Same clause. The difference is not in the incident. It is in the decision-maker. Last week I spoke with a female goalkeeper whose performance I once analyzed for the national U19 side. She told me how she reads the shooter's belly step before the ball leaves the foot — something that never appears in a data export. She said: "I do not look at their feet. I look at their hips." And I wondered: if a U19 goalkeeper can read that signal, why can a VAR system with eight cameras not? And here is what I want the VAR critics to understand properly. The problem is not that VAR is bad. The problem is that we are using a high-precision measuring tool to answer a question defined in vague language. No technology can fix that. I spent three months analyzing VAR decisions in the domestic league this season. What I learned is not whether VAR is right or wrong. What I learned is that we are looking in the wrong place. We argue about the contact. But the answer lies in the step before the contact — something no camera rotates fast enough to see. There is a line I often hear from coaches: "VAR kills the emotion of football." I do not think so. VAR does not kill emotion. It shifts emotion from the moment of the goal to the moment of waiting. Fans no longer scream when the ball hits the net. They scream when the screen shows the word "checking penalty." That is a change. But not a loss. It is simply different. What is more concerning is that VAR is creating a new layer of power that no one controls. The VAR referee — sitting in a closed room, no one seeing his face — has the authority to call the on-field referee to review an incident. But no one audits the VAR referee's decision. No public report. No evaluation score. We know who the on-field referee is. We do not know who the VAR referee is in most cases. And here is the question I cannot answer. If the VAR referee decides which incidents are reviewed, then who decides the VAR referee? We have created a system in which power is concentrated in a person nobody sees, and we call it fairness. The season has 12 rounds left. Over those 12 rounds, I will keep coding every VAR decision. But this time I will add another layer of data: the "pre-contact shift" index — the degree of center-of-gravity change in the 0.5 seconds before contact. If the pattern I saw in the first 31 incidents is confirmed, we may have to rewrite our understanding of the penalty. Not in the way coaches want. But in the way the data forces us to accept. And I wonder: if one day we can measure a player's belly step before the shot arrives — will we still need to argue about VAR? Or will we shift to arguing about who has the right to define "clear"?