When the Data Goes Silent: The Craft of Football Analysis and the Price of a Rushed Verdict

The Moment the Screen Went Blank

There is a moment that anyone who has ever sat in a football analysis room knows well: the screen is bright, the data has not arrived, the deadline is knocking, and a very plausible conclusion is already sitting in your head. All you have to do is write it down. Nobody can verify it right away. The audience will nod. The editor will be pleased. And that is precisely the most dangerous moment in my profession.

I call it the trap of the empty space. When a dataset is empty, a scouting report is missing, a statistical table fails to load, the natural human reflex is not to stop, but to fill. The brain hates a void. It will automatically weave a coherent story out of a few fragments, then present that story in the tone of someone who saw it with their own eyes.

In modern football, voids appear more and more often. A single match contains thousands of events, but only a fraction are recorded. A player runs eleven kilometres, but only a few dozen ball touches are digitised. A referee's decision lasts three seconds, yet is dissected for three days. Between those facts, there is always a dark zone that no camera and no algorithm can touch.

A poor football writer fills the dark zone with guesswork. A good football writer names the dark zone and leaves it alone. The difference between the two is not knowledge, but something far harder to train: the capacity to tolerate not knowing.

That night, my screen displayed a line that many would consider a failure: “Insufficient information to assess.” I could have deleted it and written a very persuasive analysis of a match I had not fully watched. I chose to keep it. This article explains why that choice, in the football of 2026, matters more than any goal.

Football in a Flood of Data

Football has never had so much data. Every matchday in the Premier League, La Liga or K League 1 generates millions of data points: player positions recorded every ten seconds, ball trajectories, goal probabilities, pressing metrics, passing maps, sprint speeds. A single match in the English Championship today can produce more data than an entire season did in the 1990s.

But quantity of data does not equal quality of understanding. This is the central paradox of contemporary football: the more metrics there are, the more voids are concealed beneath a coat of paint called manufactured precision. People assumed that once every footstep was measured, every argument would end. The opposite happened. Each new metric opens a new battleground, because each metric rests on a chain of assumptions that few bother to trace to its root.

Take a simple example. When you read that a team controlled 65% of possession, you believe you are looking at a fact. But behind that number is a specific definition of what counts as “possession,” set by a private data company, applied across thousands of matches, according to standards no spectator ever sees. You are not reading the truth. You are reading one way of reading the truth, chosen by someone else.

In Vietnam, this gap is even wider. Fans access most data through translations, roundups and leaderboards shared onward without any source. A metric that passes through three reposts can change meaning entirely. I once spent an evening tracing the origin of a widely circulated statistic about a Vietnamese club's passing accuracy, and found it came from a friendly match, calculated by a different standard, then labelled as a whole season. No one lied on purpose. The gap between origin and reader was simply too wide for the truth to survive.

The data flood also carries a darker undercurrent: live data supplied to betting companies. This is the most sinister side effect of the digitisation of sport, and it is rarely stated plainly. The same dataset that helps a coach understand an opponent also helps a bookmaker price risk and helps a gambler place a bet in an instant. The line between sports analysis and financial speculation grows ever blurrier, and the football writer stands right on that border, often without realising whom they are serving.

I do not write these lines to condemn technology. I write to remind us that whenever a new tool appears, the first question is not “how powerful is it,” but “what is it used for, and who pays for it.” In football, the second question is almost always skipped.

So what should an analyst do amid that flood? My answer comes from a principle I learned while sitting in the press stand, watching the assistant referee's flags: before you assert, verify; when you cannot verify, say you do not know. It sounds simple, but it is the hardest discipline to keep in an industry that rewards speed and treats hesitation as weakness.

Anatomy of a Conclusion

Expected Goals: A Double-Edged Blade

Expected goals, or xG, is the perfect example of both the power and the trap of football data. In essence, xG assigns each shot a probability of becoming a goal, based on position, angle, shot type, the number of players in front and a range of other variables. The idea is beautiful: instead of merely counting goals, we measure chance quality.

But who builds the xG model, and on which dataset, is another matter. Every major data provider has its own xG model, and they can differ substantially for the same shot. A shot from the edge of the box might be valued at 0.08 xG by one model and 0.12 xG by another. That difference, multiplied across hundreds of shots per season, is enough to change an entire conclusion about a team.

When a player like Erling Haaland repeatedly scores more than his xG, people call him a “finisher who beats the model.” That may be true. But it may also be that the model simply has not learned enough about a striker with rare positional instincts. Outscoring xG does not automatically prove talent, nor does it automatically prove luck. It only proves there is a gap between model and reality, and that gap needs explaining, not celebrating.

This is where I see many analyses collapse. They take one metric, treat it as truth, and build an entire tactical story on top. The conclusion sounds very clever. But it stands on a foundation the author never checked. A metric is not the truth; it is a way of seeing, and every way of seeing has blind spots.

Pressing Metrics and the Illusion of Aggression

If xG measures chance quality, then a metric like PPDA — passes allowed per defensive action — measures pressing intensity. The lower the PPDA, the higher the press. This is a favourite metric among modern analysts, because it allows teams' styles to be compared on a single scale.

But PPDA has a fatal blind spot: it measures action, not effectiveness. A team that presses frantically but is constantly played through will have a very low PPDA and be rated highly. A team that deliberately sits deep, concedes territory and then counter-attacks in a flash will have a high PPDA and be called passive. The metric cannot distinguish intelligent pressing from foolish pressing, proactive defending from cowardly defending. It only counts.

I once spent four hours with slow-motion video to verify a PPDA-based claim about a K League 2 side. The initial conclusion was neat: this team presses high, therefore controls the match. Reviewing every phase, I found the opposite: they pressed high because they kept losing the ball in midfield and were forced to chase. A low metric was not a sign of strength, but a symptom of chaos.

That lesson has stayed with me throughout my career: when a metric tells a story that is too beautiful, go looking for the story it is hiding. Data does not lie, but neither does it speak for itself. There is always an interpreter, and the interpreter can be wrong.

Transfer Rumours and the Hierarchy of Evidence

During the transfer window, the gap between truth and rumour becomes an industry. Every day, hundreds of pieces of information about deals are released, and most have no basis beyond the wish of an agent, a club or a journalist needing clicks.

My job in this period is to build a hierarchy of evidence. At the top are items with official club confirmation, disclosed contract structures, specific release clauses. In the middle are items from reputable journalists with a track record verified over time. At the bottom are unsourced items, usually originating from anonymous accounts and spread by the very people who do not believe them.

What is striking is that in the transfer window, the most traded commodity is not players, but belief. A club can sign no one and still achieve its communication goals by letting a rumour travel far. An agent can negotiate nothing and still apply pressure on a parent club by leaking information about a third party. The structure of the release clause and the wage bill is the real story; the name is only the tip of the iceberg.

I learned to read rumours through the times I was fooled. Once I published a transfer story from a source that sounded very plausible, complete with a specific transfer fee. Three weeks later, the deal collapsed, and my source vanished. No one remembers my wrong article except me. It took me three months to believe I was right, and two years to understand that being right is never enough. Being right must come with the responsibility of explaining why you are right, and admitting when you are wrong.

VAR: A Machine That Cannot Fix Fear

No field exposes the gap between data and truth more clearly than the video assistant referee, or VAR. When VAR arrived, many believed the arguments would end. There are cameras at every angle, offside lines, handball-detection technology. The truth would be laid bare.

Reality is far more complicated. VAR does not replace human judgement; it merely supplies additional data for that judgement. And judgement remains judgement. An offside line can be accurate to the millimetre, but the moment the ball leaves the foot is still determined by a human, and the error at that moment can be larger than the offside margin itself. A shoulder contact can be handball or not, depending on how the law is read regarding where the arm begins.

This is where I always return to my core principle: the law is never wrong; only the reading of the law is wrong. When a decision is controversial, I do not hunt for “the referee's mistake.” I hunt for which reading of the law led the referee to that decision. Because the referee is only the reader, while the law is an inviolable text.

And this is what I believe deeply: VAR does not correct referees' mistakes; it merely exposes their fear. When a referee stands before the monitor, he faces not only the phase of play. He faces tens of thousands of spectators, the pressure of public opinion, the memory of past criticism. That fear seeps into every decision, and no camera records it.

I once sat in a control room, watching a VAR team handle a situation in the 85th minute. The assistant referee raised his flag about three-tenths of a second off, enough to reverse a decision. On the screen, everything appeared clear. But in the mind of the decision-maker, time did not flow by frames. It flowed by fear. The silent whistle at 23:47 is a verdict, and that verdict is not only for the player, but for the one who blows the whistle.

The Art of Handling Null Values

Back to my blank screen. In data science there is a concept called null handling. When a data field has no value, you have two choices: fill it with an estimated value, or mark clearly that it is empty. The first choice gives you a beautiful table. The second gives you an honest one.

Most football analysis systems choose the first, because a beautiful table sells. But every estimated value inserted is a small lie, and thousands of small lies add up to a completely false conclusion. A model trained on fake data will produce fake predictions, and those fake predictions will be presented with the confidence of a truth.

When I receive an empty dataset, the correct professional response is not to invent a story. The correct response is to stop, report that the data is insufficient, and request that the collection process be re-run. It sounds like an act of failure. But it is the only act that protects the integrity of the entire analysis chain downstream.

A conclusion built on empty data is not analysis; it is a delusion dressed up in terminology. And in football, where each decision can change a club's fate, that delusion is worth an entire season.

The Counter-Intuitive Angle

Here I want to offer a view many colleagues will oppose: in modern football, the best analyst is not the one who delivers the most conclusions, but the one who knows exactly when to stay silent.

The entire industry is racing according to the opposite logic. More articles, more predictions, more decisive assertions is better. Hesitation is read as a sign of insecurity. Saying “I do not know” is treated as professional failure. And so an entire generation of analysis is being trained to always have an answer, even when no answer exists.

But look at the referees themselves. On the pitch there are twenty-two players and one person who is not allowed to be wrong. That person cannot say “I do not know” when the ball has entered the net. That pressure creates a special psychology: a tendency to cling to the first decision, a tendency to avoid reversal, a tendency to stay silent when the whistle should be blown. Those tendencies are not personal faults; they are the product of a system that places a human in a position where they cannot be wrong.

Football analysts, by contrast, have the privilege of being wrong. We have time, tools, the ability to watch a phase ten times. Yet we routinely choose manufactured certainty over honest uncertainty. We voluntarily surrender our greatest privilege.

The view from the substitutes' bench shows you how the system erodes the truth. A coach watching from the touchline can see things the stand cannot, but also misses things the stand sees clearly. No viewpoint is comprehensive. Every position on the pitch, in the stand, in the control room has its own blind spot. The analyst's task is not to pretend they have no blind spot, but to state clearly where their blind spot lies.

This leads me to a counter-intuitive conclusion about the industry's development. As technology grows stronger, the pressure grows to believe that everything can be measured, predicted, controlled. But football remains a game of humans, played by tired bodies, tense minds, racing hearts. No algorithm models the moment a defender hesitates for half a second out of fear of making a mistake. And that very moment often decides the match.

So I propose a new standard for our profession: judge an analysis not by the decisiveness of its conclusion, but by the transparency of its process. An article that states clearly “I have this data, I lack that data, and therefore my conclusion holds only within these limits” is more credible than one that asserts everything with absolute confidence. Transparency about limits is not weakness. It is evidence of honesty.

There is an investigative journalist I always learn from, one who dares to say what no one wants to hear and accepts the cost. From him I understood that in sports journalism, courage lies not in attacking, but in refusing to join the game of unsupported assertions. And there is one thing I always remind myself: the rules were written to protect the match, but some people use them to protect themselves.

What to Take Away

I did not write this piece to convict data. I wrote it to remind us that every tool mirrors the person who uses it. A metric in the hands of an honest person is light; in the hands of someone who needs a fast conclusion, it is a weapon.

What I want the reader to carry away is not advice on how to read expected goals or pressing metrics. It is a small habit: whenever you read a football analysis, ask yourself whether the author is telling you what they know, or what they want you to believe. The gap between those two things is where the truth lives, or is buried.

Football will keep digitising, keep generating more metrics, more models, more cameras. But there will always be a dark zone technology cannot touch: the dark zone of human intention. Whether a player truly meant to play that pass. Whether a referee truly believes in his decision. Whether a journalist truly believes what he writes. No data can answer those questions. Only honesty can.

Perhaps in the coming years, when artificial intelligence models can write a structurally perfect football analysis in seconds, the value of a football writer will no longer lie in producing text, but in deciding when not to produce it. In a world where every answer is available, the rarest thing is the courage to say: I do not have enough information.

I leave that question open. If tomorrow you read an analysis that is perfect, coherent, full of numbers, with not a single admission of what it does not know, will you trust it more, or less? Your answer says more about the future of football than any prediction model. And I may have missed a detail somewhere in that night's stream of data. If so, I hope you will point it out, because an analyst who knows he can be wrong is more trustworthy than one who has never considered the possibility.