The Empty Spreadsheet and the Line Between a Sports Analyst and a Rumor Merchant

The screen showed a spreadsheet with 76 rows and dozens of empty cells. I was merging data from two seasons. On one side, 76 matches played without spectators inside the Dalian and Suzhou bubbles in 2026, when the pandemic shut every stadium in the world. On the other, 76 matches by those exact teams in the 2026 season, when the stands were full and the roar was part of the game. The home-team possession column was complete. The expected goals per shot column was missing nearly half its values. The phone rang. My statistician collaborator said, flatly: "The xG column is missing shot-location data for three matches. Do you want me to interpolate from the season average?" I stayed silent for a few seconds. Interpolation means taking the average of the matches I had and assigning it to the matches I did not have. The spreadsheet would be full. The article would be done before 9 a.m. And it would be a lie dressed up neatly, subtle enough that nobody would catch it — including me. I said no. That decision sounds like it belongs to ethics. In practice, it belongs to technique. A spreadsheet with empty cells is still honest. A spreadsheet filled with guesses stops being data — it becomes literature. And in sports analysis, literature disguised as numbers is the most dangerous thing there is, because it does not give itself away. It does not smell of fabrication. It smells of professionalism. That night I did not file. I called back and asked for two more days to pull the real shot-location data from the source. Two days later, the article appeared with a tidy hypothesis: with the crowd noise gone, home-team possession rose from 51.2% to 54.1%, but expected goals per shot fell from 0.11 to 0.08. An empty stadium gives us data, but it takes away the thing data cannot measure: noise. Stage one and stage two Every serious piece of sports analysis runs through two separate stages. Stage one gathers facts: who played, where, at what minute, how many passes, conversion rate, distance covered, number of presses. Stage two reads those facts to draw a conclusion. It sounds simple. But almost every major mistake in this profession comes from mixing the two stages together, or worse, letting stage two run before stage one is finished. When the two stages blur, the conclusion arrives before the data. The writer can feel the story they want to tell, then goes looking for the numbers that will back it, ignoring the numbers that will not. That is the moment analysis turns into advocacy. And advocacy, however decorated with charts, is still just a personal opinion wearing a data costume. I learned this at eighteen, writing my first piece on the AFC Champions League semi-final between SIPG and Urawa Red Diamonds. I argued that Hulk was SIPG's biggest weakness. It sounded like a line designed to provoke the crowd. But behind it lay a very heavy stage one: Hulk completed 8 dribbles but produced only 2 key passes, while Wu Lei registered 0.4 xG despite barely touching the ball inside the box. I spent five days on that piece, revising every figure, because I knew a single data error would be enough for people to dismiss it with one remark about the writer's gender. What I kept from that experience was not caution. It was a formula: a provocative headline, evidence in the body. Every article begins with a claim that runs against the consensus, then develops into three numbered arguments anchored in specific figures. Readers find it hard to bend an idea that is numbered 1-2-3 and where each number has data at its back. In 2026, at the World Cup in Russia, I applied that formula to one of the tournament's most debated matches. After France beat Argentina 4-3, I wrote that Deschamps was killing attacking football — and that it was the best thing about this team. France held only 42% possession but fired 15 shots, 8 on target. Mbappé's two goals did not come from improvisation. They came from intent: ceding the pitch and leaving space behind Argentina's back line. Deschamps was not wrong that year — the crowd's view of ugliness was. The piece reached 200,000 reads and drew hundreds of comments along the lines of "what does a woman know about tactics". I did not reply. I rewatched four France matches over two weeks and wrote a longer, data-driven rebuttal. The only way to answer an emotional accusation is to add evidence, not emotion. When attacked, I do not explain myself. I dig deeper and publish a new analysis. When stage one returns an empty result That is the comfortable part of the story. The hard part is when stage one returns nothing at all. In this profession there is a situation I call the "empty payload": you sit before a topic, a match, an analysis, and the underlying data layer is completely blank. No tournament name, no team name, no player, no time marker, no figure to hold on to. On the surface it is just a technical glitch. Inside, it is a test of professional character. There are two roads when the payload is empty. The first: stop and say plainly, "not enough information to conclude." The second: fill the gap with something that sounds plausible. The second road is always easier, always faster, and in the short term, always rewarded by the market. An article stuffed with numbers — even numbers that trace back to nowhere — is more attractive than one that admits it has nothing yet. Based on my experience following matches, I have seen this often enough to know it is not an exception. In many analysis pipelines, the gravest error is not a wrong figure but a data layer left blank and quietly filled. Once a gap is filled without a trace, every conclusion built on it carries that error forever, passed from article to article, season to season, until nobody remembers where it began. The heat map — a new form of fortune-telling The heat map is the clearest example of this kind of gap-filling. For years, the positional heat map of ball touches has become a familiar ornament in every analysis broadcast. People look at a dark patch on the right wing and immediately conclude that the player "roams widely." But a heat map does not tell you whether that player received the ball under pressure or in space, with his back to goal or facing it, after a counter-attack or after a long goal kick. It has become a new form of fortune-telling. It presents a pretty picture and hides the player's real role in the tactical system. That is why I always tell young writers: if all you have is a heat map, you do not have data. You have an illustration. And a pretty illustration, when nobody scrutinizes it, can fool both the writer and the reader for years. By the same logic, I reject any conclusion built on a single variable. People ask me constantly: "Is this team weak because of the defense or the midfield?" The question is broken at the frame. A defense is only weak in relation to the pressing structure above it and the way the midfield screens in front of it. Don't ask how good the player is; ask how the system protects him. I often isolate one variable for analysis — a press, a champion ban, a substitution — but I never conclude that everything revolves around it. Isolate to see clearly, not to reduce. The hidden variable: noise There is one variable that traditional data can barely measure, and it is the one I have chased for years: noise. In 2026, when the pandemic shut the stands, I had a rare chance to isolate it. Same teams, same league, same tactics, differing only in one condition: whether there was a crowd. The result was later cited by a doctoral student in a thesis on Chinese football. But the bigger lesson lay elsewhere. It lay in my having to admit that some important things escape my model. An empty stadium takes away the noise, and noise — it turns out — is part of the rules of the game. The hypothesis I put forward then was bold: home advantage did not vanish, it moved into the referee's head. With no pressure from the stands, decisions tend to be more neutral, and home advantage shifts from the players' legs to the referee's mind. It was a hypothesis that could be refuted, tested, and possibly wrong. That is exactly why it was worth writing. Looking back, I see that my best analysis stage was never the times I found one more number. It was the times I identified which number was still missing and refused to fill it. Underlap and the "female perspective" tag In 2026, at the Euros, I discovered that Mancini's Italy did not attack the flanks the traditional way. Spinazzola pushed high but cut inside instead of crossing — a form of underlap. I wrote that Italy would win through underlap, while most coaches thought they were merely chasing the ball. A male editor killed the piece with the reason "don't lecture coaches on how to play football." I sent the data along: 11 inside cuts but only 3 successful crosses in one stretch, and Italy producing 2,434 passes in the group stage. When Italy went deep, the piece was republished — but this time tagged "female perspective." I objected with a second, purely logical piece, asking for the tag to be removed. I put the question directly: if a man had written this line, would he be questioned? I learned that authority in this profession does not come from claiming it, but from your data holding up under scrutiny. Gender is not a variable in the professional equation. There is only data and the way it is read. Data becomes the defensive wall, and I never appeal to emotion to defend a view — only add more evidence. The counterintuitive angle Here I have to say the thing that unsettles me. If data matters so much, why does the market keep rewarding those who have none? The answer lies here: the market does not buy truth. The market buys certainty. Someone who makes a decisive prediction, even a wrong one, feels safer than someone who says "I don't have enough information." Decisiveness sells advertising. Caution does not. And in periods when information is scarce — like a transfer window, when everything is rumor piled on rumor — the data gap becomes fertile ground for rumor merchants. I call them sellers of fake certainty. They do not invent players; they invent certainty about players. They do not say "the source is unverified"; they say "the deal is 90% done." That 90% cannot be measured, cannot be verified, and is never held to account when it is wrong. It exists only to create the impression that the speaker is in the room, that the speaker knows something others do not. This is the paradox of the trade: the more noise there is, the more people need a filter, yet the less they are willing to pay the person who makes the filter. Readers in a transfer window are drowning in rumor, and what they need most is a reliability filter — but what grabs their attention is the latest, most sensational rumor. I may be wrong. Perhaps decisiveness has its own value, and someone who always says "not enough data" becomes useless in a world that needs fast decisions. I accept that possibility. But I have seen no evidence that disciplined fabrication beats disciplined honesty. What I see is bad datasets passed from article to article, until they become part of the shared memory, and nobody can tell data from anecdote anymore. The transfer window: noise drowning signal In a transfer window, everything turns upside down. Noise drowns the signal, and the real signal usually sits in the driest places. A transfer is a contest between three brains and one cheque. Three brains: the agent, the sporting director, the coach. One cheque: the budget. In that contest, what decides is not the tweets but the structure of the release clause and the remaining wage bill. Readers get swept up in the noise of rumor, while the real signal sits in dry figures few bother to read: contract length, automatic renewal clauses, sell-on percentages to the old club, and the gap in the wage bill. I am not saying rumor is worthless. I am saying rumor needs to be ranked by evidence. A story sourced from the agent is a different animal from one inferred from an airport photo. A deal with a triggered release clause is a different animal from one merely "revealed" by an anonymous account. The analyst's job is not to retell rumor but to label each item with a reliability rating, so readers know what they can lean on and what they should read only for fun. And the most important thing in a transfer window: do not let noise fill the gap where data should be. When you do not yet know the true value of a deal, the honest move is to say you do not know — not to invent a number that sounds precise. Why one variable can change the whole system I believe in the thinking of "change one variable, watch the whole." It is not a presentation trick. It is the only way to separate cause from coincidence. Take a small example. A coach decides to pull a winger inside. Just one variable. Immediately, the opposing full-back loses his marker, the center-back must step up, space opens behind the back line, and the striker gains another passing lane. One decision, the whole system shifts. If you only look at that winger's heat map, you see a patch of color. If you look at the system, you see a chain of dominoes. This is also why I hold a clear position on inverted wingers. The trend has homogenized football. Traditional wingers — players who live by dribbling down the line and crossing — are being erased wrongly, not because they are worse, but because the system no longer has a role for them. The best system does not create superstars; it creates the perfect role. And when the role disappears, a whole generation of a player type disappears with it. In esports, the rule is identical. The meta is not invented by anyone — it reveals itself when someone bothers to calculate. A small patch changes one champion's stats, and at once the entire pick-ban system rotates around it. The team that reads that variable first wins for the first two weeks. After that, when every team has read it, the edge vanishes and the match returns to execution. That is why I do not believe in "super rosters" on paper. Paper does not play. Systems play. What I want to leave behind That night, when I said no to interpolation, I lost two days. But I kept something far harder to build: the confidence that my spreadsheet can be scrutinized to the very bottom without collapsing. In this transfer window, as hundreds of rumors scroll across the screen each day, readers do not need one more engaging storyteller. They need someone willing to say "I have nothing here yet." I do not know whether the market rewards that honesty. But I know this: an analysis is only worth reading when it can be refuted with data, not with emotion. So next time you see a spreadsheet with empty cells, do not rush to fill them. Ask what they are trying to say. An empty cell is sometimes the most important discovery in the entire dataset — it is a reminder that the line between an analyst and a rumor merchant is not drawn by who speaks louder, but by who dares to stay silent when the evidence is not yet in.