Empty Data and the Temptation to Invent: Notes from La Commanderie to World Cup 2026

The night before a Ligue 1 match, I received a data pack. Forty-two files, one per match, each containing GPS positioning, ball touches, pass coordinates, pressure metrics. I opened the first file and found it empty. I opened the second. Empty. By the twelfth file, I understood: this was a pipeline failure, the data had been cut somewhere between the sensor and the server. But I still had seven hours before the report had to go live.

In those seven hours, there was a very specific temptation. I knew the lineup. I knew the score. I knew who scored. And I had a ready-made analytical template in my hands, with nine dimensions and dozens of cells waiting to be filled. The temptation was to fill them with what I thought I had seen, rather than with what had been recorded. The temptation was to write "a defense lacking cohesion" because the team lost, not because I had evidence of a defense lacking cohesion. The temptation was to turn an empty template into an article that looked full.

I tell this story not to talk about a specific match, but to talk about the moment the football analysis profession faces every week: when the template is already there, when the story is hot, and when the data — the very thing that should fill that template — does not arrive. The danger of this profession is not the lack of numbers. The danger is that an empty template is always ready to be filled with imagination, and imagination is never checked.

Recently, when I took part in a two-stage analysis process — stage one breaking the source article into information points, stage two applying a nine-dimension professional template to those points — I met that same moment again. The stage-one input was empty: no title, no source, no information points, no entities identified. Technically, the stage-two report still ran through all nine dimensions, but every analytical position was marked "insufficient information." Not a single football conclusion was drawn. To an outsider, that is failure. To me, it is the most correct behavior a system of analysis can perform: refusing to invent.

Because in this industry, the worst thing is not an empty table. The worst thing is a table already filled in, looking perfect, and nobody able to check the origin of each cell. Numbers do not lie, but they hide the most important thing. And when there is no number at all, the only thing a decent professional can say is: I do not have enough data to judge.

Context: when football is taught to viewers through templates

Over the past twenty years, the way the public understands football has changed at the root. Previously, fans absorbed a match through feeling: who ran more, who was tougher, who "played with fire." Today, they absorb it through a template. They hear about possession, about pass counts, about heat maps, about xG, about formations. Every match is now presented with a set of metrics like a citizen's identity card.

There is a good side to this. It forces writers like me to answer harder questions: why did team A control sixty percent of possession and still lose? But it also carries a new disease. When the analytical template becomes something everyone knows, the greatest temptation is to fill the template without needing data. People have grown accustomed to the phrase "this team deployed the wrong shape," so they will write it even when they have never watched the recording of the in-game formation, having only seen the paper diagram on television before kickoff.

I learned this lesson in a very specific place.

In the summer of 2026, when I was twenty-three, I worked as a research assistant at Olympique de Marseille's La Commanderie training center. My job over three consecutive weeks was monotonous: processing the GPS positioning data of right-back Hiroki Sakai. I recorded high-speed running distance, average receiving position, sprint counts, and compared each matchweek to the early season.

Those three weeks gave me a number that forces you to pause. Sakai's high-speed running distance had fallen eighteen percent compared to the start of the season. And his average receiving position had dropped seven meters deeper than before.

The standard reading of a newsroom is: the player's form has declined. Sakai has lost speed, has slowed down, is no longer himself. That reading is smooth, easy to write, and easy to please readers with. It is also wrong. Because when I placed that number beside another — how the right flank was left open after coach Rudi Garcia switched the shape from a four-two-three-one to a four-one-four-one — the picture changed entirely.

In the four-two-three-one, the right-back has a wide midfielder mapping runs alongside him, someone to cover, someone to dare to push up with. When Garcia switched to a four-one-four-one, the four-man midfield lay flat across but with only one holding midfielder left, and the right channel became no man's land. Sakai was still himself. What differed was the structure. People called it a dip in form. The misalignment is not the machine's fault, but the thing people choose not to see.

Seven meters deeper is not a sign of decline, but the consequence of the channel above having lost its cover. I wrote a twelve-page report concluding not that Sakai had a technical error, but that the system change had distorted the very data people used to convict him. The report sat for two weeks. When Marseille lost to Monaco with nothing left to hide behind, the coaching staff finally pulled up all of my data again.

The lesson is not that I was right. The lesson is that if I had filled the ready-made template — "player in decline" — I would have had a faster article, a tidier article, and a worthless one.

Core: when data is empty, the template generates its own story

I want to devote most of this article to the actual mechanism of the problem, because this is where few analyses dig deep enough. When an analytical system processes empty input, it faces two choices. The first is to mark insufficient information across every dimension, draw no conclusion, make no inference. The second is to fill the template with default assumptions that look plausible.

The second choice is dangerous for three technical reasons, and all three are familiar to anyone who has worked in sports data.

First, an analytical template has a pull. A dimension with a ready-made name — "tactical analysis," "club financial analysis," "industry transmission analysis" — always carries an expectation. When you have an empty cell titled "degree of tactical sophistication," your brain instinctively wants to put an assessment in it. The mere existence of the cell creates pressure to have content for the cell. This is the same psychological mechanism as an editor setting a headline first and then going out to find evidence for it.

Second, missing data always leaves a causal gap, and the human brain is intensely uncomfortable with causal gaps. When you do not know why the team lost, you are forced to create a cause. In football, the default causes are always available, and they follow an almost programmed order of priority: first player form, then the coach, then the referee, then psychology. Very rarely is the first default cause "the coach's structural system change." Yet that is precisely the variable with the greatest explanatory power in most of the cases I have handled.

Third, and this is the most subtle point, the scoreline always creates an illusion of evidence. When a team loses one to three, the writer typically assumes the scoreline has confirmed every hypothesis about that team. But the scoreline is only the final outcome of a chain of events, not the explanation of that chain. A team playing well can still lose one to three. A team playing badly can still win one to nil. Using the scoreline to fill analytical cells is the most common form of intellectual laziness in football writing, and it is worse than leaving the cell blank.

The nine-dimension template I work with — tactics and technique, club finance and the transfer market, results and the opinion cycle, league landscape and team positioning, rules and compliance, management and the dressing room, risk profile, media narrative and expectation, football industry transmission — is a very powerful machine. But a powerful machine does not mean a machine that knows what it is processing. It is only powerful when it has the right material. Without material, its power converts into the power of fabrication: each dimension produces a fluent paragraph, no dimension is cross-checked, and the whole report looks like a professional analysis while in reality it is nothing but a coloring page.

When I compare this with the stage-two case above, I notice something professionally interesting. Stage one was empty — no title, no source, no information, no entities. In theory, this is the worst situation for a process. Yet the handling was precisely the standard I want to spread: marking insufficient information across all nine dimensions, making no inference, inventing nothing. The report closed with a procedural conclusion rather than a football conclusion: the input is defective, re-supply it before issuing any judgment.

That may sound trivial, but it stands in direct opposition to the reality of football media. If you replaced that analytical system with a newsroom, and replaced the empty cell with a marquee match, you would not receive that humble silence. You would immediately receive a three-thousand-word commentary, fluent, gripping, with full diagrams and full conclusions, about a match for which the writer never had a single number to rely on.

Magic is only a name for what we have not yet measured

To see this problem more clearly, I want to give a second example, one more familiar to the public, and also the example that made colleagues laugh at me a few years ago.

In July 2026, while working as a young researcher at a sports data company in Paris, I was assigned to write the daily tactical briefing on Croatia at the World Cup in Russia. After the semifinal against England, I wrote a two-thousand-word piece arguing that Luka Modric was not a wizard, but the product of a system. Specifically, a three-center-back system with two deep-lying midfielders created the spaces for Modric to receive the ball in the zone around the center circle. The figure I used then was 9.4 receptions in the central circle area per match.

My office colleagues laughed at me. In the middle of the whole world worshipping Modric as a god, my daring to "de-sacralize" him was seen as a betrayal of a beautiful emotion. But three months later, when I compared Croatia's transition map with France's pressing data in the final, that same colleague asked for my analysis file as reference material.

I retell this not to praise myself. I retell it to point out that football writing has something called "magic," and that magic is almost always measurable — people are simply too lazy to measure it. The sublime long shot, the impossible solo run, the destined goal — each can be traced back to a mechanism: the starting position, the space the opponent exposed, the defender's reaction, the ball trajectory trained in advance, the breathing rhythm of a whole system rotating around one player. Magic is only a name for what we have not yet measured. And when we have not measured it, instead of admitting we have not measured it, we call it magic — a way of avoiding responsibility for our own ignorance.

Connecting back to the moment of the empty data pack, I realize something: my reaction that night and my French colleague's reaction to Modric are two sides of the same problem. When data is missing, the human default reaction is to fill the gap with something already in memory — and what is already in memory about football is always an emotional story, not a mechanism. Both times, my choice to go against that default reflex gave me a conclusion that was neither fast nor charming, but correct.

Contrarian angle: the real enemy is not missing data, but hollow analysis dressed as professionalism

At this point I want to offer a contrarian view, because if I only said "use more data," this article would have no value. My view is this: in today's football media environment, the greatest enemy is not missing data, but hollow analysis dressed as professionalism.

In plain terms, hollow analysis is a piece with the perfect form of analysis — diagrams, named metrics, terminology, decisive conclusions — but beneath which no chain of reasoning can actually support the conclusion. It is a building with a beautiful facade, erected on ground the contractor never drilled to test. It is far more common than an emotional piece, because it is harder to accuse. An emotional piece everyone knows is emotional. A hollow analysis piece carries the prestige of science, and that very prestige lets it pass through the reader's mental filter with ease.

That is why, in working reality, I believe an empty nine-dimension table, with nine rows reading "insufficient information," is worth far more than a nine-dimension table filled to the brim where not one cell has been cross-verified. The empty one tells the reader a truth: I do not yet know. The full one seems to say: I already know. And when the full one is produced from imagination, the reader does not receive an analysis — they receive a fraud with a polished interface.

But — and this is where I must push back on myself, because a structural view always risks being taken to an extreme — it is not always right to leave the template empty. There are matches, there are moments, where the most worth-saying thing is precisely the human factor: the fear of a player taking a penalty, the trembling of a defense in stoppage time, the moment a person surpasses himself. An analysis that speaks only of structure, of misalignment, of pressure, with no room at all for fear and courage, is also a kind of hollow analysis in another sense — hollow in the human part. I have carelessly overlooked this part. I was wrong. A decent analyst must know where to stay silent before structure and where to listen to the human being.

I must add one more thing about the numbers themselves. There is a common misunderstanding that numbers are objective. Not true. Numbers are objective only in the collection stage; in the stage of selection and interpretation they are thoroughly subjective. My choice to include Sakai's eighteen-percent drop in high-speed distance in the article, rather than another number, is a choice. My choice to pair it with the seven meters deeper, rather than with sprint counts, is another choice. A trustworthy analysis is not one with many numbers, but one transparent about why these numbers were chosen and which others were left out. I do not believe in miracles. I believe in data collected properly — and "properly" includes honesty about the limits of what we have.

Third case: silent stadiums and the trap of emotional trends

In 2026, when European football was paralyzed by the pandemic, I was working as an analyst in Ligue 1. The editorial board asked me to write a series of nostalgic pieces about stadium atmosphere during the football shutdown. I refused. I offered an alternative: build a dataset comparing pass rates, match tempo, and sprint counts for teams with crowds versus without crowds, in lower-division matches that still had to be played behind closed doors. I wanted to answer the question "what actually changed," not the question "how are we feeling."

The results surprised many. Match tempo in Ligue 2 rose six percent without crowds. But the number of risky passes into the final third fell eleven percent. Two numbers moving in opposite directions, and that broke the beautiful story the editorial board wanted. The beautiful story was: football loses its soul when crowds are absent. The data said something different and more complex: the match ran faster but was more cautious, safer, less adventurous.

I wrote a four-thousand-five-hundred-word piece arguing that silence does not create cautious football; it exposes the caution already present in coaches' thinking. When crowds no longer roar for adventurous plays, when jeers no longer pressure players to push forward, what remains is the true instinct of a system trained not to make mistakes. Football did not die when stadiums emptied. It only revealed its true skeleton.

The lesson here is about emotional trends. When a major event occurs, the entire media industry is swept into a shared emotional current, and that current becomes the default template for filling every analytical cell. A pandemic occurs, the default template is "football lost its soul." A player shines, the default template is "exceptional genius." A team loses, the default template is "form has declined." That template is not entirely wrong — it is simply unverified. And my job, at least in principle, is to verify before writing, accepting that I am slower than the news, in exchange for something more durable than the news.

The key is in the question "what produced this number"

If there is one single principle I want the reader to carry away after finishing this article, it is this: every time you see a number, ask "what produced this number," not "what does this number prove."

The first question is a tool. The second is a trap. When we ask "what does this number prove," we have prepared a conclusion in advance and are merely looking for evidence to sign our name onto it. When we ask "what produced this number," we force ourselves back to the chain of causation, to structure, to the recording. That is the only way an analysis can be refutable, and that very refutability is what distinguishes an analysis from a belief.

I think about this every time I see a fluent commentary on a match the writer never watched on tape. I think about it every time a referee spends two minutes at the VAR monitor and the rhythm of the match cools, so that all that remains in the viewer's memory is two minutes of stillness rather than a passage of play. I think about it every time I hear someone say mid-table midfields are using physicality to turn football into athletics, and I wonder whether the speaker has data for that, or is merely repeating a shared feeling.

My profession is not the profession of delivering truth. No one owns the truth about football. My profession is the profession of delivering evidence, placing evidence side by side, and being responsible for the chain of reasoning that links them. When that chain breaks for lack of data, the most decent and also the hardest thing is to stop and say: I do not yet know. An honest empty template is always better than a fake full one. That is what the empty data pack that night, and a whole analytical process stopping just in time before emptiness, taught me once again.

Football will always have room for miracles, because people need to believe in something beyond their control. I understand that. But let miracles be for others to believe; I will keep my data. Next match, when I open a pack of files and find it empty, I know what I will do: shut the laptop, call the data officer, and wait. If I have to wait longer than the news, then wait. One day, my readers will understand why.