The VAR Model, Kim Min-jae, and the Limits of Numbers in the Transfer Market

Late in 2026, in a windowless meeting room in Gangnam, I presented a risk ranking to the leadership of a transfer consultancy. I had spent half a year building that ranking. Atop the red group sat Kim Min-jae. My model, running on more than a thousand referee decisions from five European top-flight leagues, produced a single number: 0.73 fouls per match in Serie A, which it rated as high card risk. I advised the firm not to recommend signing him. Napoli signed him anyway. The season ended with Napoli's first Serie A title in 33 years, and Kim Min-jae was in the team of the season. For many nights afterward I rewound old footage until dawn, not to find fault with Italian referees, but to find where I had misread a number. The gap between a number and a truth has rarely been wider. And it did not begin with Kim Min-jae. It began with the belief that referee data can measure the quality of a defender. CONTEXT: WHEN THE WHISTLE IS NO LONGER THE ONLY SOUND ON THE PITCH Over the past decade, player analysis has undergone a quiet but total shift. Where the 1990s measured players through scouts' impressions and VHS tapes, and the 2000s added running data, the 2010s and 2020s saw referee data enter the boardroom. Cards, fouls, penalty probabilities, contact locations, the interval between incident and whistle — all digitized. VAR did not just change how matches are governed; it created a new data layer the transfer market had never possessed. I entered this profession from a different direction. In 2026, at 23, I worked as a VAR assistant for a broadcaster in Incheon. FC Seoul against Jeonbuk Hyundai Motors, round 29, minute 67, Lee Dong-gook put the ball in the net. I detected he was offside by 0.3 meters. But absorbed in the rear camera angle, I sent the alert fourteen seconds late, against FIFA's seven-second standard. The main referee could not intervene in time. The goal stood. The executive director scolded me in front of the entire editorial room. For three nights I could not sleep, rewinding the footage, asking how to optimize the decision process. From that shock I built an automated log called VAR Decision Analysis, recording response times and camera angles for every incident. My writing became precise to the second, yet dry as a technical memo, missing the emotional thread of the match. I did not yet understand that what I was building was not a ledger but a way of seeing. In 2026, I was sent to Russia as a VAR analysis assistant for a Korean broadcaster. In the group stage match between Spain and Iran, I began collecting every handball incident of the tournament. The final count was 27, and only 31 percent were handled consistently under IFAB's new rule. I wrote a 40-page report and sent it to my editors. They published only a small chart. Frustrated, I started a personal blog and posted the entire dataset without asking permission. The post drew 50,000 reads from referees, sports lawyers, and fervent fans alike. From there I abandoned the memo style and shifted to investigative prose: cite the data, quote the law, ask open questions. I used percentages in place of feeling in every analysis. But I quietly forgot one thing — that behind every number is a person, and behind every person is a circumstance the ruler never touches. In March 2026, global football stopped. The broadcaster cut my contract for budget reasons. Instead of worrying, I retreated into research as an escape. For six months I analyzed 1,247 VAR decisions from five European leagues. The finding startled me: with no crowd, referees' VAR consultation time fell 22 percent, but the rate of upholding the on-field decision rose 15 percent. The noise of the stadium is not written into the law, yet it carries legal weight. I wrote a 60-page report and posted it on an academic network. A director of the Asian Football Confederation reached out and invited me to serve as a data analyst for the referees' committee. In that period I also learned to present hypotheses, methods, and limitations. The writing was rigorous in structure, but I used technical terms without explaining them, making it readable only to specialists. Once again I stood between two shores: too strict to be read widely, too dry to be loved. By 2026, as a mid-level staffer, I built a player-evaluation model from VAR data for a transfer consultancy. And Kim Min-jae appeared in the red group. CORE: DISSECTING A MODEL THAT MISREAD A DEFENDER My model was no impulsive product. It was built on 1,247 VAR decisions collected during the pandemic, plus tracking data from European domestic leagues. For a defender, it measured four main indicators: fouls per match, probability of being penalized inside the box, frequency of contact, and number of reviews. Kim Min-jae, while at a Serie A club, produced 0.73 fouls per match. On my scale, that signaled a center-back prone to fouling, likely to pick up cards, and therefore a defensive risk. But data is never as neutral as we think. It is only valid when the context around it stays fixed. The first thing I overlooked was teammates' cover. A center-back who steps up to intercept, who proactively cuts out the ball, generates contact situations that someone without cover never would. He fouls more not because he is weak, but because he stands where his teammates do not. The data records the collision, never the gap he filled. The second, more important thing I overlooked: Italian referees interpret the law differently from Korean referees. On the same challenge, Serie A referees tend to permit more physical contact, which inflated the fouls recorded in my dataset. I measured fouls with a Korean ruler, then applied it to a match played under an Italian ruler. The third thing, and the one that cost me the most sleep, was the very concept I was measuring. I called it a foul. But a foul is not a physical constant. It is the product of an interaction between a player and a referee, inside a rule that has a certain elasticity. Every VAR error is a crack in the mirror that reflects the law. And my mirror, turned on Kim Min-jae, cracked exactly where I had built it on belief rather than definition. To see this trap clearly, return to the 2026 World Cup. IFAB had just revised the handball rule, and the football world threw itself into arguing over the threshold of a hand. I collected 27 incidents across the tournament and only 31 percent were handled consistently. The trap of 2026 was not in the hand, but in the belief in a definition that does not exist. The community believed there was a correct threshold — that if the ball touched a hand in a certain posture, it was certainly a foul. There is no such threshold. There are only camera angles, conventions, and different decision-makers. The same thing haunted my model on Kim Min-jae. I believed there was a per-match foul threshold that reflected defensive quality. There is no such threshold. There are only incidents, covering teammates, and referees with their own conventions. In 2026, I learned this lesson the hardest way. Fourteen seconds late against a seven-second standard. In those fourteen seconds, an offside goal became a valid one. But looking closer, the fault was not entirely mine. It lay in the limits of the observation system. I had only a few camera angles, and I chose the rear angle because I believed it was most reliable. My error was the error of a tool that was not yet enough. VAR was born from the fear of error, but it nurtures the fear of late truth — the fear that by the time the tool sees the truth, the match has moved too far on. A wrong decision does not ruin a match; the silence after it ruins trust. When my editors in Russia published only a small chart instead of the 40-page report, what they sowed was not anger but the sense that the truth had been seen and then put away. Fans can forgive a referee who errs. They do not forgive an organization that stays silent. Back to Kim Min-jae. When I built the model, I had no ruler to measure the courage of a center-back who leaves his comfort zone to hold the line. I had no ruler to measure his move from a low-contact league to a high-contact one, or how long adaptation takes. Nor did I have a ruler to measure how a referee talks with a player. None of that ever appears in VAR data. When Napoli signed him and Kim Min-jae became a pillar, nothing happened as a shock. It happened as a public correction of the tool I had worshiped. At the end of that year, I wrote a 10-page self-review and deleted the model. Since then, every piece I write carries a section titled limitations of the data. I began interviewing referees and coaches to clarify context, no longer trusting numbers absolutely. But the story does not stop at one defender. It opens a larger problem for the industry. THE TRANSFER MARKET: NUMBERS PUSHED UP BY BELIEF, NOT BY DEFINITION Looking at today's transfer market, we see something close to the trap of 2026. Young players' prices are being pushed up by belief, not by definition. A player with fewer than 50 top-flight appearances can be valued near 100 million euros. That number reflects not the minutes played, but the expectations of those sitting in the boardroom. The problem is not talent. The problem is how we measure talent. When my model misread Kim Min-jae, I did not fail for lack of data. I failed because I believed data could replace definition. The transfer market is the same. Clubs buy potential, but they price that potential as if it were a finished asset. The bubble in young-player prices did not erupt from nothing. It erupted from a collective belief that youth plus a few numbers guarantees success. But if Kim Min-jae taught me anything, it is that context weighs more than numbers. A good center-back in a zonal system will show different indicators from a good center-back in a man-marking system. A young player in a healthy development environment differs from one pushed into a patched-up squad. How referees treat play in one league differs from another. A player's natural position is never in the passing data or the foul data. It lies in the relationship between him and the system around him. I use natural position here in a broad sense: the place where a player stands most reasonably, where he does not have to strain to become someone else. When a club buys a young player for 100 million euros, it buys a number, not a natural position. It bets that the number will create the position. History shows the opposite happens more often. Meanwhile, in esports, the problem is more severe. An esports pro's career is far shorter than a footballer's. If a footballer can play to 34 or 35, an esports pro usually faces the big question at an age when his football counterpart is still peaking. Yet youth development and post-retirement support are close to zero. They are valued by metrics, bought on belief, but when the metrics fall, no one holds a ruler to measure the person left behind. This is the intersection of VAR and the transfer market. Both believe there is a perfect ruler. Both keep discovering that no such ruler exists. VAR did not create justice on the pitch. It created an excuse to stop arguing. And in many cases, it merely moved the argument from the whistle to the screen. CONTRARIAN ANGLE: THE PROBLEM IS NOT THE DATA The easiest way to tell my story is to turn it into a personal confession: an arrogant analyst fooled by data, who then learned humility. That story sounds good, but it hides the real problem. The real problem is neither in the data nor in me. It is in our craving for fixed definitions so intense that we invent them, then defend them as if they were natural facts. The handball of 2026 had no single definition. Perfect defending has no single number. A young player has no single objective price. What is counterintuitive is this: data transparency does not reduce controversy; it increases it. When an organization publishes all its data, people assume trust will rise. In reality, each person picks the metric that suits their existing view. We do not seek justice on the pitch. We seek an excuse to stop arguing — and data merely supplies another place to keep arguing. If my model was not technically wrong, only wrong in its definition, deleting it is not the solution. The solution is to state clearly what it cannot measure. A model that does not declare its limits is a model lying through silence. And perhaps that holds for organizations too. I grew up in Vietnam and work in Korea, enough to recognize that the same rule can be explained two different ways by two football cultures, each convinced it is right. Not because one is dumber. Because each culture naturalizes what it has grown used to. I allow myself one parenthetical at the end: is there something important that even the best data never touches? I think the answer is yes, and I do not know what it is. TAKEAWAY: THE BORDER BETWEEN THE RULER AND THE ONE HOLDING IT In this season's matches, whenever a controversial VAR incident appears, I no longer ask who is right or wrong. I ask three questions: the truth was seen through how many camera angles, the decision was made in how many seconds, and what did the organization say afterward. Those three questions give me no justice, but they give me a measure of the quality of trust. Every VAR error is a crack in the mirror that reflects the law, and the right question when peering into the crack is not who broke it, but whether we dare to look at it. If Kim Min-jae could be misread by a model built on more than a thousand decisions, then any young player bought for a hundred million euros is being read by a ruler with the same limits. That does not mean we should stop measuring. It means we should print the ruler's limits right on the ruler. An honest model is not a correct model. It is a model brave enough to say what it cannot measure. The gap between the number and the truth is not a mistake. It is the space where people step in.