When the Match Data Returns Zero

In October 2026, at John Smith's Stadium, Huddersfield Town beat Manchester United 1-0. That night I sat in a first-year dorm room in Chicago, rewinding the footage and writing two numbers side by side on the same sheet of paper: Huddersfield generated 0.35 xG, United generated 1.82. Every model I knew said the away side should have won. They lost. A week later I found the third number that no major outlet printed: 27 tackles by Huddersfield in front of their own penalty area. It appeared in no standard statistical table. It was the entire story of the match. I started a small site called I Have A Number, and promised myself I would only write about the data everyone else left behind. When xG lies in a match, every number has to be interrogated from scratch. Eight years later, on a Monday morning, the screen in front of me returned a completely empty table. No competition name. No team. No player. Not a single information point. Every cell carried the same line: no data. I sat there for a while, because this profession trains you to fear wrong numbers, and almost nobody teaches you how to read an empty one. I work as a data consultant for a football club in the United States. The job began with scanning GPS data during training sessions, then expanded into match-metric reconciliation and transfer valuation. Most of it is not glamorous. It is hours spent with files nobody opens: distance covered, sprint intensity, touches taken under pressure, decisive defensive actions in front of the box. Somewhere in that work I learned a distinction that looks small and decides everything. Zero data and missing data are two different things. A midfielder who covers 9.4 kilometres has a low but real figure. A midfielder whose distance cell is blank because the device failed has no figure at all. Beginners mix the two, fill the gap with the squad average, and carry the result into a meeting. That error travels all the way down into transfer reports, and by then nobody can trace its origin. The empty table that Monday belonged to the second category, and I had to establish the cause before doing anything else. There are three possibilities. The source broke: a dead link, a deleted page, a region block. The extraction pipeline failed: the system received text but could not read anything out of it. Or the source genuinely contained no analytical content, an image-only page, for instance. Three causes, three different responses, and none of them is guessing. The principle I have kept for years is simple: when data goes silent, my job is to record that silence, not to fill it with a plausible story. In sport, a conclusion invented from an empty dataset outlives the person who made it, because it gets printed, quoted, and eventually written into a contract. Back to Huddersfield. What kept that match with me for years is how it exposed the limits of a single metric. xG measures chance quality based on shot location and situation. It does not measure a defence actively dismantling an attack before the shot is ever taken. United shot often, from good positions, and still did not score, because most of those chances were broken a step earlier. The 27 tackles sit exactly in the gap xG cannot reach. Summer 2026 was the first World Cup I watched as an analyst. After the group stage I collected data from 48 matches and built a comparison of distance covered. Croatia averaged 116.2 kilometres per match, second highest in the tournament, while their average xG was only 1.08. American coverage called them old and slow. I wrote a long piece, used a model of opponent speed decay in the final 30 minutes, and predicted Croatia would reach the final on the strength of extra-time endurance. The road to a final is not run by feet; it is run by the distance a team is willing to cover. When Croatia beat England in the semi-final, a Spanish analytics site translated the piece. I earned my first freelance fee of 120 dollars, and the name DataMonk started circulating in data-analysis circles. The bigger lesson was not the money: data only has value when it sits next to an argument nobody else has made. In mid-2026 world football stopped. I was studying for a master's in sociology and assumed my analytical career was over. When the Bundesliga returned to empty stadiums, I pulled 26 matches before the restart and 26 after it and compared them. The result startled me: the home win rate fell to 34.6 percent, down 10.4 percentage points, while the draw rate jumped to 31 percent. When the stands are empty, I watch the winning formula break into a thousand pieces and reassemble in a different shape. I wrote a long essay on the death of home advantage. Three days later the sporting director of Chicago Fire wrote to offer me an assistant analyst role, starting with exactly the GPS scanning work I mentioned earlier. Without that disrupted season I would probably still be a blogger with a small site and a few hundred readers a month. In January 2026, after the 2026 World Cup, I sent club leadership a 14-page analysis of Sofyan Amrabat. He had recorded 24 ball recoveries across five matches at that tournament, and I argued that 18 million euros to trigger his release clause at Fiorentina was sound business. The sporting director dismissed it flatly: Amrabat has no commercial value, nobody buys his shirt. By summer 2026 Amrabat had joined Manchester United on loan. My analysis circulated through professional front offices, and a European club approached me for remote consultancy. The transfer market is only a mirror reflecting the fears of the people who run it. Around the same period I received statistical packages from a Gulf league, where late-career European stars arrive on large contracts. Read the numbers alone and everything looks reasonable: high goal counts, stable minutes, assist rates holding up. Place them next to sprint intensity and the distances involved in decisive duels and the picture changes completely. Those attractive numbers are mostly produced in an environment with far less intensity in contact, and the real value of the deal sits elsewhere: image revenue, regional broadcast rights, the commercial pull of a name global audiences already recognise. Over those same years I gradually lost faith in the tool used most often in presentation rooms: the heat map. It is elegant, it is intuitive, and it convinces an audience that they understand a player. The problem is that a heat map only answers where a player stood, never what he was responsible for inside the tactical system. A full-back whose heat map stretches across half the pitch may have been instructed to hold the touchline all match, or may have been freed because a teammate covered for him. Two entirely different situations, one image, one meeting room. Esports took me to a harsher version of the same problem. Sample sizes are smaller, noisier, and dependent on game patches in a way football never is. A team can win because a patch favours their comfort picks, and a three-match win streak says nothing about genuine level. In esports I hear the echo of football before the data era: plenty of inspiration, very little probability, and a transfer market that runs on belief. Every match is a confession; my job is to read between the lines of code. The irony is that a complete dataset is the dangerous one. An empty table forces people to admit they know nothing. A full table delivers a false sense of safety, and that safety flows straight into investment decisions. In football it produces 20-million-euro deals built on ten good matches. In esports it produces contracts signed with players who shone for exactly one season in exactly the right meta. Data is never in a hurry; it waits until you are clear-headed enough to ask the right question. The problem is that people are always in a hurry. Transfer windows close, boards face results pressure, fans have very short memories. When everything pushes at once, publicly stating that the data is not yet ripe becomes a decision as hard as publishing a wrong conclusion. I have stayed quiet many times, and each time it cost me a byline. I still believe it was the right call. Another trap sits in the distance between correlation and causation. Small esports samples almost always yield a few beautiful perfect correlations: teams that win the first teamfight have a high win rate, teams with strong objective control post higher damage numbers. But when the next patch shifts the balance, those correlations vanish before anyone can defend them in front of the board that paid for them. My experience following matches has taught me that data should only persuade other people after it has persuaded me, across repeated cycles, in different contexts, against different opponents. What I carried away from that empty Monday table is not a data-handling technique. It is a habit of asking one more question. Every flawless dashboard has a column that was never recorded, and that column usually sits exactly where the match was decided: the defensive actions nobody counts, the training sessions nobody films, the objections raised in meetings and never written into the minutes. The major tournament cycle is approaching. When national teams reach the knockout rounds, I will again read the emotional surges in the stands before opening the data file, because that order matters. I do not believe in luck, but I believe in the probability of the shots nobody remembers. And I believe in the teams that run further than anyone thought necessary.