When the Data Sheet Is Empty: The Discipline of a Tennis Analyst Who Refuses to Invent

At 2:17 in the morning, I reopened a tennis analysis file and found it empty. No title, no source, not a single information point. Nine analysis frames sat waiting, each carrying the line “insufficient information.” I sat there for a while, hands on the keyboard, wondering whether to fill it with a story that sounded plausible. For someone who works with data, that moment is the real test: when you have nothing in hand and must choose between silence and invention. I chose silence. Here is why. Some years ago, I built a two-stage process for my analysis work. Stage one deconstructs the source article: title, source, core viewpoint, and most importantly the information points. Stage two takes those points and runs them through nine analytical dimensions, from technique and form to rules and media. The rule is unbreakable: every conclusion at stage two must anchor to a specific information point at stage one. If stage one returns a void, stage two must stop. The machine is not lazy; it was built to halt when there is nothing to analyze. This annual season taught me that the principle applies to people too. Tennis media lives on stories, and stories are always ready to fill any gap in the data. A player loses early in the third round, and within hours there is a piece about a “psychological crisis.” Few check how many sets he played in seven days, on what surface, or how his opponent served. The gap gets filled with feeling, and feeling is always available. Based on my experience following matches, most such “stories” dissolve once you place a few numbers beside them — schedule load and second-serve points won. This is where I learned the most important lesson, from my own mistake. In the summer of 2026, when Liverpool paid 42 million euros for Mohamed Salah from Roma, I published a long analysis built on Serie A data. Salah's numbers sat in Europe's top 5% for finishing and box penetration, and I concluded he would score more than 30 goals. He scored 32. But in the same piece, I predicted that Gylfi Sigurdsson, at 45 million pounds, would dominate Everton's midfield — and he faded all season. The data told the truth, but I had ignored one variable: the new role the manager asked him to play. When the market mocked Salah, the data nodded silently. When I trusted Sigurdsson, the data did not nod — I was the one nodding. Since then, I never write on a single metric. Every analysis must include a “role variable” section: the team's tactical system, how the player is used, schedule pressure, and surface. In tennis, this is even stricter. A player with a high return-points-won rate on hard courts may not hold that number on clay, where each point stretches a few beats longer. A 68% first-serve-in rate sounds great, until you learn the tour's best returner is standing across the net. Numbers do not lie, but they only speak when you place them correctly, and at the right moment. The 2026 World Cup taught me a second, more painful lesson. After the Croatia–England semifinal, I used xG to argue that Croatia created only 0.8 xG while England had 2.1, then wrote that Croatia “did not deserve” the final. The community pushed back at once: football is not a computer simulation, and Modrić's mentality and stamina were what carried the team. I withdrew into video study for a month and found that Croatia's goalkeeper dived to his right 2.3 times more often than to his left. Croatia was not luck. xG had recorded the story before the ball rolled. But xG could not record a goalkeeper's reflex in a single instant, and that was the part I missed. I stopped using the phrases “deserved” and “did not deserve.” I replaced them with probability descriptions: Croatia won within a sequence of events with roughly an 18% probability, and that is what the data still cannot explain. Every piece now ends with a “data limits” section. That is why, when the analysis file is empty, I do not fill it with a story. No information points, no named player, no tournament, no source, no date. Inventing a conclusion from that would betray the very method I built to protect myself from costly errors. But there is a flip side I must admit. The discipline of refusing to speculate, pushed too far, becomes avoidance. I have seen analyses that are nothing but data and dare not offer a single judgment — a way of hiding behind numbers to avoid responsibility. That is another form of intellectual cowardice, just dressed more neatly. Readers do not need a table; they need a judgment with a probability attached, clear enough to argue with. And here is the counterintuitive point: data is not neutral. The way you collect it is already a choice. A PPDA metric only means something when you know over how many minutes it was calculated. Fans look with their eyes, I look with a probability distribution, but both have blind spots. Mine is believing that what can be measured matters more than what cannot. So when stage one returns a void, I treat it as a signal, not a failure. It tells me the input has a problem, that someone handed me a page with no words. The work is to go back and check whether the original article actually exists, not to fill the page with imagination. In a long season, where each week brings dozens of matches and hundreds of headlines, the ability to say “I do not have enough data to conclude” is perhaps the most underrated skill. The truth lies deep beneath the numbers, where headlines never reach. And sometimes, the simplest truth is that the sheet is empty.