Full Report, Empty Information: Silent Failure in Football Analysis

In October 2026, at Olympique de Marseille's La Commanderie training centre, I spent three straight weeks processing the GPS positioning data of right-back Hiroki Sakai. His high-speed running distance had fallen 18% from the start of the season. His average receiving position had dropped seven metres deeper. I wrote a twelve-page report, blaming not his individual form but the fact that head coach Rudi Garcia had shifted the shape from 4-2-3-1 to 4-1-4-1, leaving the right flank unprotected so that Sakai had to cover space the midfield no longer shielded. The report sat on a desk for two weeks. Only after Marseille lost 0-3 to Monaco did the coaching staff dig my files back out.

That was a failure, but at least it was a visible one. I saw the number, I traced the cause, I wrote it down. Today I want to talk about the other side of the mirror: the failure nobody sees. The report still prints with full pages, full tables, full charts, yet is hollow inside. No sound tells the reader.

Magic is simply the name we give to what we have not yet measured. Football has spent two decades measuring everything. Now it is time to talk about the day we measured it — and the machine forgot to write it down.

Modern football runs on a layer of data the audience never sees. A single match in a top European league generates between 1.5 and three million coordinate data points. GPS systems in training vests log heart rate, distance, accelerations. Optical systems mounted around the pitch record the ball and 22 players at 25 frames per second. Event providers such as Opta or StatsBomb resell every pass, every duel, every shot with coordinates and expected-goal value.

At the other end of the pipeline there is a reader. A coach reads the pre-match report. A sporting director reads a transfer dossier. A journalist reads the stats table to write. A fan reads the graphic on television. All of them believe what they are reading is the output of a process already completed.

Between those two ends lies a chain of steps nobody audits: collection, transmission, analysis, verification, publication. A typical football data pipeline has at least five stages. Collection pulls raw data from sensors or a vendor. Transmission sends it to the server. Analysis computes the metrics. Verification cross-checks plausibility. Publication packages it into a report for the end user.

What matters is that verification is usually the thinnest stage in manpower. At many mid-tier clubs, one or two analysts must compute metrics, write the report and answer the coach. Nobody has time to confirm that the raw input data actually exists. When collection fails, the later stages keep running. Machines do not know how to stop. They just print a table with blank cells.

In data science this phenomenon has a name: silent failure. A process returns a success status while producing no usable information. No error message. No red light. Just an empty object drifting quietly through the stages until it reaches the reader wearing the shape of a finished report.

Axis deviation is not a fault of the machinery, but of what people choose not to see. Yet here, even axis deviation does not exist. There is nothing to deviate. The table still has column headers, still has rows, only the cells hold no values. To a glancing eye it looks like an ordinary report. To a careful eye it is an empty room labelled "verified".

Silent failure lives even in the field we assume is most subjective: player evaluation. In July 2026, writing the daily tactical bulletin on Croatia at the World Cup in Russia, I argued that Luka Modric was no wizard but the product of a three-man defence plus two deep midfielders, giving him an average of 9.4 receptions inside the centre circle per match. A colleague in the office laughed. Three months later, that same colleague asked for my file back to cross-check against France's pressing data in the final.

What I learned from that episode was not that Modric is not great. He is great. Numbers do not lie, but they conceal the most important thing. A cell filled in the right place can replace a legend, and a cell left empty can create another one. When we cannot measure something, we tend to call it magic. When we can measure it but refuse to check, we tend to call it truth.

In May 2026, when European football was paralysed by the pandemic, my editors asked me to write a nostalgic series on stadium atmosphere. I refused and proposed another route: building a dataset comparing matches with crowds against closed-door matches in the lower divisions. The result: match tempo in Ligue 2 rose 6% without fans, but risky passes into the final third fell 11%. Silence does not create cautious football. It exposes the caution the coach already carried.

Football did not die when the stands emptied. It simply exposed its real skeleton. The trouble with a real skeleton is that it is not appealing to sell. Nobody wants to buy a ticket to watch a skeleton. So people keep selling classic atmosphere, and ignore the data that has argued otherwise for a long time.

Football is especially vulnerable to silent failure for three structural reasons.

First, the decision cycle is brutally short. A coach has a few days between matches to digest a report. He has no time to query the data source. He trusts the format. If the report looks professional, he uses it.

Second, the pressure of results turns data into a political weapon. A selectively chosen metric can protect a player or indict a coach. When data is both evidence and weapon, the incentive to check it becomes complicated. Sometimes people do not want to know the cell is blank.

Third, football culture still prizes intuition. That sounds like a strength, and sometimes it is. But it also means that when data and the eye conflict, people pick whichever is more comfortable to believe. A report with gaps that matches the prevailing feeling is accepted more easily than a complete report that says the opposite.

What makes silent failure more dangerous than loud failure is how it gets filled. A data gap is rarely left empty for long. It is filled with three substitute materials.

The first is the default value. The system automatically inserts an industry average into the missing cell. A midfielder with no pressing data is assigned the league's average pressing level. That error is not flagged.

The second is human interpretation. The analyst sees the blank, remembers the previous match, and fills the gap with memory. Memory is dominated by the strongest impression, usually a goal or a mistake, not by repeated behaviour.

The third is silence. The blank is skipped, and that skipping itself becomes a false signal: nothing to worry about. In risk governance this is the gravest error. The absence of evidence is read as evidence of absence.

I have told young colleagues many times that I do not believe in miracles. I believe in properly collected data. But I must add the other half of that sentence: badly collected data is more dangerous than no data at all, because it carries the shape of truth without its content.

A coach with no report knows he is blind. He will rely on his eyes, on the opponent, on conscious intuition. A coach with an empty report thinks he can see. He makes decisions with the confidence of a man who has been informed, when in fact he has only been formatted.

In my analysis I call these gaps dark zones. Not the dark zones of ignorance, but the dark zones of understanding presumed complete. They sit exactly where the data pipeline is most confident: standardised metrics, automated tables, reports published on schedule.

Dark zones do not discriminate by club size. A big club with ten analysts can have more dark zones than a small one with two, because the big club produces more data to overlook. At Marseille in 2026, Sakai's data was not missing. It was so abundant nobody read all of it. A good analyst is not the one with the most numbers, but the one who knows which numbers are missing.

Here is the contrarian point I want to put on the table. The football analytics industry has spent a decade improving the quality of input data — more sensors, more vendors, more metrics. It has spent almost no effort improving the detection of missing input data. We optimise measurement; we do not optimise knowing what we have failed to measure.

The result is a paradox: the more complex the system, the harder silent failure is to detect. A handwritten report by a single analyst may contain errors, but they are the errors of a person we can interrogate. An automated five-stage pipeline can emit an empty report for which nobody is specifically responsible. Responsibility is diffused until it dissolves.

People trust format over content. A table with gridded cells looks more credible than a scrawled note, regardless of what is inside. Clubs have built entire analytics departments on the assumption that professional form means professional substance. That assumption holds in most cases, and fails disastrously in the rest.

A loud error in football has its uses. When the TV feed drops, viewers see a blue screen and know something is wrong. When VAR loses connection, the referee is told and play stops. These failures annoy, but they are honest. They admit the break.

Silent failure has no such quality. It pretends to be seamless. In a decision problem, a loud error costs us time; a silent one costs us direction. Time can be recovered. Lost direction accumulates quietly, match after match, until a run of bad results forces people to dig out all the old data — exactly what happened at La Commanderie.

If I had to extract one practical test, it would be this. For every number on a football report, ask two questions. First: what question does this number answer? Second: if this cell were blank, would the reader notice? If the answer to the second is no, that report carries a dark zone, however complete it looks.

Magic lies where we have not measured. Dark zones lie where we think we have finished measuring. An empty stats table can deceive, and it deceives better than any lie, because it does not need to say anything at all.

The next match you watch, try a small exercise. Pick one player, find one metric about him, and ask yourself how that metric was measured, from what source, at what moment. If you cannot answer, you have just touched a dark zone. No need to panic. Just remember that modern football does not lack data. It lacks people to check whether the data is actually there.

And if one day you read a tactical report so smooth it has no gaps at all, be suspicious. An honest report always has room for the unknown. A perfect report is usually just a report that has not yet been questioned.