When the Analysis Room Returns a Null Result: The Integrity of Football Data

A morning in Liverpool. I opened the data export from my analysis pipeline and found a blank page. That blank page was not a font error, nor had I forgotten to switch on the machine. It was blank because every field carried the same phrase: insufficient information. Title empty. Source empty. The list of information points empty. No club, no player, no competition identified. I sat still for a long time, hands on the keyboard without typing, and the first thing that came into my head had nothing to do with an error message.

What came into my head was a January evening in 2026, when Liverpool beat Manchester City 4-3 at Anfield, and I sat in a small room a few streets from the ground, watching the pressing figures of the two sides diverge like railway tracks that never meet. The silence on that screen taught me something that more than thirty years in the trade had never taught me fully: data does not lie, but data also has the right to fall silent, and an honest professional must respect that silence. In my industry, a blank table is usually read as failure. I began to think otherwise.

The age of the data pipeline

Over the past two decades, football has shifted from a game of the eye to a game of the eye plus the spreadsheet. Clubs such as Brentford built an entire recruitment model on data to climb from the lower divisions to the Premier League. Liverpool expanded its analysis department into a unit with a voice in transfer deals. Manchester City built a data system stretching from the academy to the first team. Behind the million-pound signings sits a chain of work few people see: collection, cleaning, coding, then analysis. I call that chain the pipeline, and my job is to stand at the end of it and read what flows out.

That pipeline has two stages. The first breaks a match or a file down into discrete information points: who passed, where, when, under how much pressure. The second is where I build the story: setting context, comparing, finding patterns, asking questions. When the first stage returns an empty list, the second has nothing to build on. And this is what I want to say to those who read football through data: a null result is not an error to be hidden, but a signal to be read. It tells you that somewhere in the pipeline, a link has broken.

I once thought analysis was a profession of answers. The longer I work, the more I see it is a profession of the right questions. A null result answers very little but asks a great deal: was the source blocked, was the text corrupted, did the document type fall outside the model, or did the pipeline simply swallow a page with no words on it? The right question is worth more than a hasty answer. And in an industry where the pressure is to have an opinion every day, holding the right silence is a form of discipline.

I learned that discipline very early. In 2026 I joined the sports desk of Belgrade Television. There was no xG then, no PPDA, no passing maps. We had notebooks, pencils, and one rule: what you cannot see, do not write. My old editor called it the honesty of the blank page. He said a sports journalist may be wrong about a prediction, but must never invent a fact. Thirty years later, that rule still holds; the only difference is that the blank page now lives inside a data file.

A null result is a finding

When all nine analytical dimensions of a file read insufficient information, outsiders usually conclude that the analysis failed. I read it the other way. A null result has high diagnostic value because it localises the fault. If every dimension is empty, the error is not in the tactical stage, not in the financial stage, not in the governance stage. The error is in the extraction of the input. A null result is not the absence of a conclusion; it is a conclusion about where the fault lies.

In medicine, a negative test is still information. In statistics, a null hypothesis that is not rejected is still a result. In football, a goalless match is still a match worth analysing, sometimes more so than a seven-goal one. So why is a file with no data treated as worthless? Because we have grown used to measuring value by quantity, and have forgotten that emptiness has weight too.

I once stood before a data table and felt I was witnessing a miracle at Anfield. That is a moment I keep to myself, and I mention it here only once. But that moment taught me that the miracle of data is not that it gives us more numbers, but that it tells us when a number does not exist. The good reader of numbers is not the one who reads the most, but the one who notices at once when a blank cell appears in a full table.

In recruitment practice, this matters more than we think. A young player in a league that data systems do not cover will have an almost empty file. If the scout rushes to fill the gap with instinct, he produces a report that sounds very convincing but has no basis. If he is honest, he writes clearly: insufficient data, three more live viewings needed. That honesty may cost him a deal in the short run. But it keeps him in the trade over the long run.

PPDA and that January evening

To see why a blank cell matters, look at one specific metric. PPDA, the passes allowed per defensive action, measures pressing intensity. The lower the figure, the fiercer the press. In the 2026-18 season, I calculated Liverpool's average PPDA at 8.2, the lowest in the league, while a conservative side such as Manchester United sat at 15.7. A gap of nearly double spoke of a philosophy: Liverpool turned the opponent's half into a trap, while the conservative side turned its own half into a fortress of waiting.

I wrote a long piece on gegenpressing and published it on my personal blog. The response came fast and hard: too mechanical, soulless, turning football into a maths problem. I read all of it, did not argue, and noted one thing: people do not object to data, they object to the feeling of being judged by data. That was a lesson in psychology, not in mathematics. The 4-3 win over Manchester City on 19 January 2026 strengthened my faith in metrics, but also taught me that faith in metrics must come with humility in how it is presented.

What I did not write that day, and now regret, was the limitation section. A low PPDA does not automatically mean quality. A team that presses hard without holding its shape burns its own energy and leaves space behind. The figure of 8.2 only means something beside the number of recoveries in the opponent's third, beside the goals conceded from counter-attacks. A metric standing alone is half a truth. And half a truth, in my trade, is more dangerous than ignorance, because it looks so much like the truth.

From then on I added a section to the end of every piece: the limitations of the analysis. That section did not make the piece shorter; it made it longer. But my readers, the ones who stayed, began to trust me more. They realised that when I say something, I have questioned myself three times before writing. In a news market where everyone wants to be fast, slowing down is a competitive advantage.

World Cup 2026 and the forgotten model

In 2026, at forty-three, I was commissioned by a sports site to write a special series on the World Cup in Russia. I built a homemade xG model and ran it across all sixty-four matches. The model gave me an early conclusion: France would be champions, because their chance-creation was the highest in the tournament, averaging about 2.4 xG per match. I wrote that from the group stage. The piece was mocked. I had described Croatia as a team with low xG but good fortune, and that angered many people.

Croatia reached the final. I was emotionally exhausted. I hid in a library for two weeks, reviewed all the data, and found the error: my model ignored set pieces, especially corners. A significant share of Croatia's attacking power came from dead-ball situations, and I had let them fall outside the model. The error was not in the data. The error was in the person who built the model, that is, me.

The biggest lesson of World Cup 2026 was not that I predicted the champion. The lesson was that I predicted correctly for an incomplete reason, and was criticised for a legitimate one. Being right about the outcome while wrong about the path is a dangerous kind of luck, because it reinforces confidence in the wrong place. If I had looked only at the final result, I would have rewarded myself with an undeserved belief.

xG is a revolution, but every revolution needs time before people accept it. I say this not to excuse slowness, but to remind that those who go first are often misunderstood in the early phase. In 2026, most viewers still measured chance quality by eye. A beautiful long-range shot is remembered longer than an efficient close-range finish. xG reverses that emotional order, and reversing emotion always causes discomfort before it is accepted.

Empty stadiums and an empty truth

In March 2026, world football stopped because of the pandemic. Liverpool were twenty-five points clear of Manchester City and all but certain to win the Premier League, but the season was suspended. I lost my faith deeply. If data could not predict a pandemic, what did data mean? I wrote three drafts and deleted all three. That was the period in which I understood that an analyst can lose faith not because the data is wrong, but because the data cannot cover what is really happening to people.

When football returned to empty stands in June, I realised the data had changed shape. The home-win rate fell from around 46 per cent to around 39 per cent. Home advantage, which had seemed immutable, turned out to depend on a variable I had never put into the model: the crowd. Empty stadiums did not distort the data, but they made the truth empty. The numbers remained technically correct, yet they lacked the thing that made a number mean something.

I had to be alone for weeks to redefine my model. I added a new layer I call data context: every metric must be read alongside the environmental conditions around it. The same shot, taken before twelve thousand fans and taken in silence, is two different events psychologically, though identical in coordinates. This is a lesson no spreadsheet can teach itself.

I learned at Anfield that belief is also a variable. A player who runs one extra metre in the eighty-ninth minute does so not only out of fitness, but because he believes the stand behind him will push him on. When the stand falls silent, that belief leaves the equation. A good data analyst must bring even invisible variables into the model, or at least admit that they exist outside it.

The transfer market and fates waiting to be written

I work as a transfer market administrator, so I read data through the lens of price. Every number in the transfer table is a fate waiting to be written. A fee of one hundred million euros for a player who has not yet played fifty top-level matches is a naked gamble, dressed up in grand language. The youth price bubble is not a story about talent; it is a story about cash flow, about expectation, and about clubs' fear of missing out.

What is striking is that valuation models often ignore adaptation risk. A player who shines in league A may struggle in league B because of tempo, space, culture. His data in league A looks superb, but that data does not travel with him across the border. When a transfer file is full of numbers yet lacks data on the new context, the sober reader must see a large blank. And that blank is usually where the deal fails.

I have seen deals built on a single season. The sample is too small, the variance too large. A striker who scores fifteen goals in half a season may be just a lucky run, and a model that does not check sample size is deceiving itself. In statistics we call it regression to the mean. In the transfer room we call it a one-season wonder. The same phenomenon, two names, one consequence: money lost.

The trap of filling in

My trade has one great temptation: filling the gaps. When data is missing, people want to fill it with instinct, with rumour, with a plausible story. That temptation is dangerous because it is rewarded. A complete report is always welcomed more than an honest report of deficiency. But a complete report built on invented data is a time bomb under a club's feet.

I have seen recruitment models assign metrics to players no one has ever watched live. Those metrics are interpolated from similar leagues, similar players, and gradually they become truth in the user's eyes. This is what I call second-order truth: a thing repeated often enough that no one checks its origin any more. And when the origin is a blank filled with guesswork, the whole building stands on sand.

Refereeing and VAR give me a parallel example. The space for subjective judgement in VAR is larger than people think. The phrase clear and obvious error is, in itself, a vague clause. Clear to whom, obvious to what degree, from which camera, at what speed? Every time a video referee reviews an incident, he is not only reading images; he is filling a gap of interpretation. That gap cannot be erased by technology; it only changes place.

What I take from VAR controversies is this: when data is insufficient, people default to what they already believe. That is instinct, and instinct is not always wrong. But a data professional must be able to stand between instinct and evidence, and say clearly where he is standing.

The counter-intuitive angle: silence is punished

This is the part I want to state most plainly. In today's football media, silence is punished. A writer with no opinion every day is pushed down by the algorithm. A club that announces nothing is assumed to be hiding something. An analyst who says I do not have enough data is assumed to be incompetent. So the system creates a quiet pressure: say something, regardless of whether it has any basis.

My counter-intuitive angle is this: the greatest value of data is not that it gives us answers, but that it gives us the right to say we do not yet know. In an environment where everyone must know, the one who dares to say I do not know is protecting the truth. It is a losing position, even a lonely one, but it is the foundation of every trustworthy analysis that follows.

Those who are right before their time always pay in loneliness. I say this not to flatter myself. I say it because I have seen analysts who were right but pushed to the margins, good models shelved because they did not suit the taste of the meeting room. Being right about the numbers does not mean being recognised. But that loneliness, if we can bear it, is the most honest part of the trade.

The biggest blind spot in my industry is the confusion between correlation and causation. A team that presses hard wins a lot, so we conclude the pressing is the cause. But perhaps the team wins because its squad is better, and the pressing is merely a consequence of having more of the ball. A player runs a lot, so we conclude he is diligent. But perhaps he runs a lot because he reads the game slowly and is always arriving late. Data rarely states the cause on its own. People assign cause to data, and that is where errors are born.

I once thought that more data meant fewer mistakes. I now think otherwise. More data means more opportunities to fool ourselves, because we can pick out the numbers that support what we want to believe. The truth is not in the volume of data. The truth is in the discipline of asking questions.

Ending with a signal for the next round

When the analysis room returns a null result, I no longer panic. I read it as I read a sign on a health monitor: somewhere a link has broken, and my job is to find that link before I rush to invent a replacement story. In a major-tournament season, when emotion is compressed and everyone wants a quick conclusion, holding an honest blank cell is how I protect both my trade and my readers.

Data whispers, and those who know how to listen will hear miracles. But I have learned that before hearing miracles, we must learn to hear silence. The next round of this story does not begin with a new number. It begins with a question I will carry into every file: if I had to write a piece today, would I have enough evidence not to invent anything at all? As long as the answer is yes, I can sit still before a blank page and smile.