When the Data Has Nothing Left to Say: A Lesson from the Tennis Circuit

One winter night in Liverpool, I opened my laptop at two in the morning and found an empty table in front of me. Not empty because I hadn't entered any data. Empty because the system had finished running, had returned a result, and the result was nothing at all. Every cell carried the same line of text: insufficient information. I sat looking at it for a long time, and suddenly realised this was the best tennis lesson I had been handed in years. In the data-consulting trade, people are taught that numbers are the answer. More metrics are better, more charts are more professional, more models are more credible. But that empty table taught me the opposite: the frightening thing is not the absence of data, it is having data and not knowing that it is hollow. A report that looks complete, with a title, a structure, nine sections of analysis, tables drawn with careful borders, but with a decorated blank space inside every cell. That is the kind of report the tennis world receives every week, from sources nobody verifies. I once read an analysis of a match that was written beautifully. It had an opening, arguments, a conclusion. But when I traced it back, the original article it was built on was a short news brief with no tournament name, no match date, and no player named. The author of the analysis had not invented facts. He had simply filled the blank space with sentences that sounded very reasonable. That is exactly where the danger lives. In tennis we live on numbers that move in cycles. Ranking points are defended on a rolling 52-week basis, meaning each player must protect the points earned at the same event the previous season. Without a date anchor, without the tournament week, the entire analysis of points-defence pressure becomes meaningless. I have written about this before, and I hold to it: a piece of tennis analysis that cannot identify the tournament week is not analysis, it is sports prose. For the same reason, every judgement about form needs a time anchor. A player winning five straight matches in March is not the same as a player winning five straight in October, because the surfaces differ, the opponents differ, and the accumulated fatigue differs. Without a date, a winning streak is just a nice number to cite. It says nothing about whether that player is about to collapse or about to explode. But what troubled me most about that empty table was the section on identity. In tennis analysis there is a precondition we routinely skip: you have to know who you are talking about. It sounds obvious, yet in real content production it is skipped constantly. People write about a tactical trend without naming a single player who represents it. People write about the rise of a new generation without naming a single concrete name. Those pieces read smoothly and sound knowledgeable, but when you ask who, there is no answer. There is something I learned at fifty-four that I wish I had known at thirty: honesty about blank spaces is itself a form of expertise. When an analyst says they do not have enough data to conclude, that is not weakness. It is the mark of someone who has been fooled by numbers before and does not intend to be fooled again. Conversely, the person who always has an answer to every question, who always finds a metric to explain every phenomenon, is usually the person who has never audited their own model. I remember a season when I tracked a young player through data. Every metric looked good, every model predicted a top-30 breakthrough within a year. But I overlooked one detail the data did not display: he had never played a match lasting beyond three sets on hard court in hot conditions. When summer came he lost four straight and nobody understood why, except those who had read the notes on playing conditions. The data was not wrong. It simply did not say everything. That is why I believe a good analysis must contain an admission of its own limits. Not to please anyone, but so the reader knows which parts to trust and which parts to check themselves. To me, a piece without that section is like a contract without a refund clause: it may still be legal, but it is not worth trusting. That night I changed nothing in the table. I left the empty cells as they were and wrote a small note beside them: go back to the source, find the article title, find the publication date, find at least one player named. If those do not exist, write nothing at all. That was the hardest advice I ever gave myself, because my trade is a trade of speaking, and silence is always the harder option. But tennis taught me this too: in a match, not every moment is a moment to swing. Some balls you let pass, not because you cannot reach them, but because you know swinging now will cost you position for the next one. In data analysis, the blank space is that ball. Sometimes the right move is to let it go and wait for a real one. There are things data never touches, like the way a stadium breathes. And there are things data touches very clearly, but only when people agree to admit they do not yet have enough of it to speak. The line between those two is where this trade is actually measured, by kindness to the reader, not by the number of charts in a report. The following days I started collecting from scratch. I looked for the title, the publication date, the tournament name, the player. Once I had all four, the story opened itself in a way an empty table never could. Like a match that only truly begins once the umpire knows who is standing on either side of the net.