When the Tennis Data Sheet Is Empty: A Decent Analyst Must Know How to Say Insufficient Data

Late at night in New York, I reopen the recording of an ATP Masters 1000 quarter-final and stare at the data sheet. The first-serve points won column is full. The return points won column is full too. But when I scroll down to the fitness and scheduling context, every cell is empty: no rest days, no count of sets played that week, no trace of injury. In that moment, eighteen years in the trade remind me of one thing — the hardest part of tennis analysis is not finding the number, but knowing when to stop because there are not enough numbers yet. People usually picture my job as sitting in front of a screen and prying out hidden truths. In reality, most of my time goes to verifying what I do not know, rather than declaring what I do. Every tennis match is a complex system: hard court or clay, altitude above sea level, ball quality, the opponent's tempo, and the psychology of a player after three sets lasting four hours. Blaming a loss on the first-serve percentage alone is the kind of intellectual laziness I have always fought against. In the sports data industry there is a trap called the empty assumption. When the input material is insufficient, a careless writer fills the gap with imagination. They build a story that sounds very reasonable: Player A lost because of poor serving, Player B won because of iron mentality. No one verifies it, because no one wants to verify a story that is already too smooth. But to a defensive data person, a gap is not blank paper to draw on. It is a signal: not enough grounds to conclude. I once learned that lesson at a considerable price. In 2026, after the World Cup semi-final between Croatia and England, I used xG to argue that Croatia did not deserve a place in the final, because they created only 0.8 xG against England's 2.1 yet still won 2-1. The community pushed back hard. I had to retreat for a month, rewatch every penalty shootout, and discovered that the Croatian goalkeeper dived to his right 2.3 times more often than to his left. Since then I have dropped the phrase deserved or undeserved, replacing it with probabilistic description: Croatia won within a sequence of events with a probability of roughly 18%, and that is something the data still does not fully explain. In tennis, that principle applies even more strictly. Imagine a player winning 6-4, 6-4 while hitting only 54% of first serves. At a glance, that number looks unremarkable. But if I know that across those two sets he faced seven break points and saved all seven, the story flips entirely. The points-won rate at decisive moments is what shapes the result, not the match-wide average. A player can serve averagely all match yet reach 78% first-serve points won in the tense games — and that is the data worth discussing. This is why I built a multi-layer verification process. The first layer is raw statistics: serving, returning, points won, unforced errors. The second layer is context: surface, schedule, weather, minutes of rest between sets. The third layer is tactical role: whether the player is asked to attack down the line or defend deep, whether the opponent tends to rush the net or hug the baseline. Only when the three layers align do I allow myself a judgment — and even then, the judgment must carry a probability level. Fans look with their eyes; I look with a probability distribution. That is not the arrogance of someone holding a computer, but a form of self-defence. Because in tennis, the sample is always small. A player can win a Masters event thanks to three matches in which opponents served below form. If I take those three matches as evidence of a career turning point, I am reading noise as signal. Small sample size is the number-one enemy of the analyst, and the silent ally of exaggerated stories. The truth lies deep beneath the data sheet, where headlines never reach. A young player can be exalted by the media after reaching a Grand Slam semi-final, while his own numbers show a return points-won rate still below the tour average. The media frenzy comes first, the data comes later, and when the two drift apart, the market and the crowd pay the price for that gap. My job is not to extinguish the frenzy, but to record that gap honestly. There is a paradox I want to state plainly: in this trade, insufficient data is a valuable conclusion, yet it is the conclusion the market values least. Newsrooms want a decisive answer. Bookmakers want a number. Fans want a name to believe in. No one pays for an article that ends with I do not know. But it is precisely the articles brave enough to end that way that keep the analysis industry from collapsing into a chorus of empty prophecies. So when I receive a match summary with empty data cells, I do not fill them with speculation. I state clearly: information missing on the schedule, information missing on fitness, information missing on tactical role. I set a sufficiency threshold before writing: three independent sources, or two data layers that align. If the threshold is not met, I stop. Stopping is not a failure of analysis; it is analysis protecting itself from its own most common disease — confidence without grounds. Every number in a player's file is a confession of the market. Agent fees, prize money, sponsorship deals — they all speak to what the market believes, not necessarily what is true. When a player is priced high after a run of wins over weak opponents, the data sheet is disguising a lucky summer. The market forgets nothing; it merely disguises itself as a new season. And a decent data person must peel away that disguise, carefully, one layer at a time. In modern tennis, where every serve is measured by radar and every rally is recorded as a point, the greatest temptation is to believe everything can be quantified. But some things never appear in the sheet: the pressure of defending ranking points before a deadline, the fear of losing a tournament entry, and the silence of an empty stand. An empty stand does not make the result wrong; it only strips away our illusions. That is why I always add a data-limitations section at the end of every analysis, even the ones I am most attached to. What I want to leave for the next round is not a prediction of who will win, but a reminder of how to read a player. Let the experience on court speak first, let our eyes live inside the rally, let the feel of tempo and fitness touch us — and only then bring the numbers in to illuminate. If the data sheet is empty, let it stay empty. That honesty toward the gap, across many seasons, will be worth more than any confident prophecy. Fans may not remember who predicted correctly, but the market remembers who made things up.