The Empty Payload: The Silent Failure Threatening Modern Football Analytics

2:47 AM, March 14. I was sitting in front of two screens in my Milan apartment, opening the data package my tracking system had just pushed through after a round of Serie A fixtures. The file name followed the convention. The timestamp matched the time zone. The field structure was identical to every previous delivery. But when I scrolled down to the content, every cell was empty. No player coordinates. No pass counts. No team names. No match date. Only a single surviving label remained in the entire package: football.

In this profession, an empty data package is more frightening than a wrong one. Wrong data leaves traces. An absurd figure, a coordinate outside the pitch, a percentage that cannot exist, and you can see at a glance that something is off. An empty package stays silent. It raises no error. It does not turn the screen red. It simply contains nothing, and if you are not fully alert at 2:47 in the morning, you will unconsciously read that nothing as a conclusion.

It took me nearly an hour just to confirm I was not misreading. Then came three days of tracing the problem back upstream. What I found was not a small overnight glitch, but something far larger: the way modern football builds its entire decision-making apparatus on data pipelines that can fail without anyone knowing. From the analytics room of a mid-table club to the negotiation table of a giant, from the xG graphics on television to the scouting files of an academy, everything rests on an assumption that has never been fully verified: that the data will always arrive.

When it does not arrive, and no one notices it did not arrive, we get a kind of failure more dangerous than a mistake. We get false confidence.

When football learned to trust the pipeline

Twenty years ago, analysing a football match was a job for the eyes and the memory. You watched the game, you took notes, you argued with colleagues in a coffee shop, and you wrote. I began my career right around that time, when the big European newspapers still sent reporters to stadiums with a notebook and a pencil. Data then was crude: goals, cards, possession rounded to the nearest whole number.

Then everything changed. GPS vests appeared in training sessions, logging every metre of movement, every acceleration, every sprint. Scouting platforms like Wyscout, InStat and later StatsBomb turned every player into a digital profile. xG, the expected goals metric, became common vocabulary on television. Clubs began hiring data scientists, software engineers, modelling specialists, into roles that once belonged to traditional scouts.

I remember the summer I realised this shift was complete. It was around 2026, when a sporting director called me to ask about a winger in a smaller national league. He did not ask for my feeling about the player. He asked whether I had data on touches inside the box, a model predicting attacking contribution, a comparison against positional peers. The call lasted forty minutes and not once did we discuss how the player performed when his team was trailing.

That was the moment I understood that power had changed hands. The decision-makers no longer sit in the stands with a sharp eye. They sit in an office with a data table. And in many respects, that is progress. It removes countless biases, countless decisions made on gut feeling, countless failed transfers caused by a scout misremembering a beautiful piece of play.

But every advance carries a price. When you hand decision-making power to data, you become dependent on the pipeline that delivers it. And pipelines can break. A GPS logging server with a faulty buffer. A scouting platform that changes its application interface and silently drops a data field. A match postponed by weather that throws a synchronisation schedule off by a day. A software update that erases a data layer nobody rechecks.

All these incidents share one feature: they make no sound. No alarm. No exception is thrown. There is only silence, and an empty package waiting on an analyst's screen at nearly three in the morning.

In the software industry, this is called a silent failure. An error occurs, but the system does not treat it as an error, so it allows defective data to pass through the processing layers without being stopped. In football, where decisions are worth tens of millions of euros, silent failure is a monster that has never been named.

Dissecting a silent failure

Let me describe exactly what happened in my package, because the detail here matters more than its appearance.

The package my system pushed through was not a corrupted file. It was structurally valid. It had all the fields, the correct formats, the correct data types, the correct naming conventions. If you ran an automated format checker on it, the checker would pass it. That is precisely what makes it dangerous. Correct structure conceals empty content.

This is the point I want to stress with all the weight of a man who has misread many things. A structurally valid empty package is not a technical fault. It is a cognitive vulnerability, because it looks exactly like a normal package to anyone who does not open the content and read it.

I distinguish three states that football routinely conflates. The first is missing data, where one piece of information is absent but the rest is enough to work with, and you know exactly what you are missing. The second is low-confidence data, where information exists but you are unsure how reliable it is, and you assign it a level of doubt. The third is the empty package, where there is no information at all, yet the system still presents it as a complete result.

The difference between these three states is not academic. It determines how a club behaves. With missing data, you seek additional sources. With low-confidence data, you combine multiple sources and decide cautiously. With an empty package, you risk deciding as though you have full information, because nothing in the interface tells you that information is absent.

Again, my unease that night did not come from a lack of data. It came from data being absent while the interface still looked normal. I had grown used to beautiful, complete, smooth data nights. That very familiarity nearly made me overlook the empty cell.

Had I been a younger analyst, working faster, trusting the system more, I might genuinely have overlooked it. I might have written a report with a few very plausible-sounding observations, based on a package containing not a single data point. And that report might have reached the desk of a coach, a sporting director, or an investor.

That is the scenario that chills me. In our industry, an empty package does not stop at the screen of the person who discovers it. It stops only if someone is disciplined enough to block it. Otherwise it drifts onward, gets dressed in analytical language, and eventually becomes part of a real decision.

I spent years telling younger colleagues that numbers do not lie. But they do not tell the whole story either. And that night taught me something harsher: sometimes the number is not even present, and yet we still behave as if it is.

Nine analytical layers and the price of evidence

To understand why an empty package paralyses an entire workflow, you need to see the structure of professional football analysis. A serious report is not just a paragraph about a match. It is a building with many floors, and each floor rests on its own foundation of evidence.

The first floor is tactical and technical analysis. To discuss a formation, how a team presses, how a defensive block shifts, you need at least the team name, the player names, and a few metrics such as tackles per defensive action, possession share, chances created. Without those, any tactical claim is guesswork in makeup.

The second floor is club finance and the transfer market. To assess a deal you need concrete numbers: transfer fee, contract structure, wages, duration, add-on clauses. Without those figures, you cannot say whether a deal is expensive or cheap, sensible or reckless.

The third floor is results and the public-opinion cycle. To discuss pressure on a manager, you need league position, recent form, the upcoming fixture list. Without those, you cannot know whether pressure is rising or easing.

The fourth floor is league landscape and club positioning. To place a team among title contenders, European spots, mid-table or relegation, you need to know which league, which rivals, which resources. Without those, the competitive map is entirely blank.

The fifth floor is rules and compliance. To discuss disciplinary or financial-fair-play risk, you need to know which governing body, which clause, which precedent. Without those, there is no risk to assess.

The sixth floor is management and the dressing room. To discuss the relationship between a coach and his players, you need quotes, contract status, ages, behavioural signals. Without those, any guess about the internal atmosphere is merely rumour.

The seventh floor is the risk profile. To list risks, you need an entity times an event times a probability times an impact. With no entities and no events, the risk matrix has no rows to fill.

The eighth floor is media narrative and expectation. To analyse a media story, you need a headline, a source, an author's stance, an article's purpose. Without those, you cannot separate reporting from advocacy.

The ninth floor is the football industry's transmission chain. This is a second-order floor, meaningful only once a first-order event exists to ripple outward. Without an originating event, there is no transmission path to draw.

Nine floors, and all nine stand on the same foundation: evidence. When the foundation is empty, the building collapses from the bottom up, even though its facade can still be assembled to look beautiful.

That is the trap I call the architect's trap. A skilled analyst builds systems quickly. They can erect a complete, coherent, compelling framework in a few hours. But if the evidence beneath it does not exist, that elegant frame is a building without a foundation. It stands until someone comes to inspect it, and then it falls.

That night, what saved me was not analytical intelligence. It was a disciplined habit: before writing anything, I check whether I have evidence. That habit formed after many occasions when I misread. And this time, it gave me back a lesson about the value of saying two simple words: not enough.

In the culture of this industry, saying there is not enough information is often seen as weakness. People want a conclusion, a prediction, a decisive angle. But an honest analyst must be able to say that with the available data, I cannot offer any assessment at all. That is not evasion. That is respect for the truth.

The biggest risk is not analysing incorrectly. The biggest risk is analysing correctly in form but empty in content, and having no one notice.

When football was fooled by data

If this story were merely a personal technical incident, I would not have written it. What turns that empty package into an industry-wide problem is that it belongs to the same family as the occasions when football was fooled by data in more public, more famous, and more expensive ways.

Let me begin with the thing I have tracked for years: the millimetre offside line. When referee-assistance technology entered the big leagues, we were promised that controversy would end. Football would have absolute fairness, measured by lines accurate to the centimetre. But after a few seasons, I saw something else happen. Goals disallowed for a toe, a shoulder, a shoelace crossing the line. Attacking players learning to slow down, to keep themselves within a safe margin, to abandon the instinct to burst forward.

I once thought technology would make the game more transparent. Three months of analysing match-tracking data for a personal project showed me the opposite. I spent three months realising I had been misreading this position. I kept thinking the problem lay in the technology's precision, in camera error, in processing speed. But the real problem lay elsewhere: technology had pushed accuracy to a level humans could no longer read by intuition, and at that point decision-making shifted from the referee to a system the audience cannot see or verify.

That is another form of silent failure. Not a system returning an empty result. But a system returning a result that appears absolutely precise, lulling us into no longer asking questions. This false confidence came from too much data, not too little.

Another example is the way we read xG. Expected goals is a wonderful tool for measuring chance quality. But I have seen it misused to the point of counter-productivity. A team that loses with a higher xG is called unlucky. A striker who scores fewer than his xG is deemed poor. But xG cannot measure what happens inside a player's head when he stands before the goal in the ninetieth minute. It cannot measure fatigue, fear, the confidence lost after three missed attempts. It measures the geometry of a chance, not the human being.

Emotion is not data noise; it is data not yet decoded. When we push emotion out of the model, we do not make the model more accurate. We only make it more confident about an incomplete picture.

I have also witnessed scouting errors born of incomplete data. Back in March 2026, I published a long analysis of Gian Piero Gasperini's Atalanta. I used tracking data from thirty-seven Serie A matches to show that Robin Gosens was not an ordinary full-back but a hybrid, a wide number ten, averaging more than twenty-one touches inside the box per match, more than the team's main striker. When an Italian football magazine republished the piece, I gained press credentials and attended the 2026 World Cup.

But the true story of that article is not its success. It is my earlier mistake. For months, I had read Gosens as a normal attacking full-back. My data system classified him by his starting position, and I trusted that label. Only when I set the positional label aside and asked the data a different question, what this player actually does when his team has the ball, did the real picture emerge.

The lesson lies here: ask what the system has hidden before judging a defender. The label a system attaches to a player may be right about position and wrong about function. And when you build a model on wrong labels, you do not get a slightly wrong model. You get a model that is systematically wrong, and systematically confident.

At the 2026 World Cup, I learned yet another lesson. That July, I was in Moscow for the semi-final between France and Belgium. I noted how manager Didier Deschamps dropped his defensive block to an average of only about twenty-five metres, while instructing Blaise Matuidi to tuck inside to cut off passes into Kevin De Bruyne's feet. I wrote in great detail about space, about defensive layers, about the gaps being sealed.

That piece sank. A colleague wrote only about Vincent Kompany's tears after the defeat, and his article was shared six times as much as mine. At first I was annoyed. Then I understood. My spatial analysis was correct, but it lacked the human being. I had described a perfect tactical machine and forgotten that the machine was operated by people who feel pain.

From then on, I learned to open with a concrete spatial image, for instance that the distance between two centre-backs was only seventeen metres, then weave in the player's story as a catalyst to hold the reader. My tone remains dry, still analytical. But it has rhythm, and it has emotional contact points.

By the pandemic summer, when football shut down in 2026, I entered a prolonged period of anxiety at thirty-nine. I wrote nothing for six months. Instead, I stayed in my room, rewatched four thousand five hundred wide-attacking situations from Serie A between 2026 and 2026, and drew thirty-eight pressure maps by hand. In June 2026, as the Euros began, I turned forty and suddenly spotted a pattern.

Three months of isolation, 4,500 wide plays, and an answer simple to the point of surprise. Italy's central midfielders, Nicolò Barella and Marco Verratti, were generating roughly fourteen to fifteen passes into dangerous areas per match through triangular movement, a template that had never appeared in my data store. That taught me that my data store, however large, always has a frontier. And that frontier is exactly where football keeps evolving.

The paradox: the more data, the more fragile

Here I want to go against the industry's common intuition.

The prevailing belief is that more data means more safety. More data points, more models, more metrics, and therefore less risk. I believed this for years. But that empty package, together with what I have witnessed in the industry, forced me to reverse that belief.

In truth, every data layer you add creates a new point of possible failure. Every pipeline you build creates a new place to leak. Every model you deploy creates a new assumption that can be silently violated. Complexity does not make a system more robust. It makes it harder to inspect. And a system that is hard to inspect is a system that hides faults more easily.

Imagine two clubs. The first has a simple analytical setup: a handful of basic metrics, one experienced scout, and a coach who watches footage each week. When something breaks, it shows immediately, because there are few layers to hide it. The second club has ten integrated systems, hundreds of metrics, thousands of data rows a day. When a fault occurs in one layer, it can be concealed by the nine others operating normally. The second club looks more professional. But it may be more fragile at the very moment fragility carries the greatest consequence.

There is a line I always keep in mind when working with data: ask what the system has hidden before trusting it. A heat map shows position; an intent map shows thought. And we have plenty of heat maps, but very few intent maps.

I am not writing this to oppose data. I am writing it because I believe in data, and precisely because I believe in it, I want it built to be trustworthy. An honest data system is one willing to say I do not know. One willing to display an empty cell instead of auto-filling a default value. One that blocks defective data before it reaches the decision-maker.

And above all, one with people alert enough to recognise that silence does not mean calm.

What I learned from an empty package

I do not know what the original analysis I awaited that night contained. I may never know, if that data source cannot be recovered. And perhaps that matters less than the lesson it left behind.

An empty package is not a full stop. It is a signal. It tells me that somewhere in the chain a link broke, and no one told me. It tells me that the professionalism of modern football lies not only in how much data we collect, but in how many times we detect that the data never arrived.

If you work in this industry, in any role, from scout to coach to journalist to director, I have one request. Next time, when you receive a data table, a report, a model, take thirty seconds to ask one single question: if this were empty, would I know?

If the answer is no, you are standing on the same ground I stood on that night.

And a larger question remains: we have built a football industry in which the most dangerous thing is not a wrong number, but a void disguised as completeness. When our systems learn to stay silent so skilfully, will we still be able to hear that silence?