The Empty Report: Nine Dimensions of Analysis, Not a Single Piece of Information

At 2:14 in the morning on February 14, 2026, in a fourth-floor flat overlooking the Lach Tray pitch in Hai Phong, I opened a forty-seven-page file. The name of the file: Stage-2 Deep Professional Analysis, Football Domain. Inside there were nine analytical dimensions. A four-row tactical assessment table. A six-cell risk matrix. A three-tier transmission diagram with arrows running from youth academies through clubs down to broadcast and derivative markets. A five-star information-value rating. A priority-sorted risk warning list. A ten-entry glossary, including definitions of metrics the document itself admitted could not be applied.

Across all of those cells, there was not a single piece of information.

Not a player. Not a club. Not a competition. Not a match date. Not a line saying who played where, or why. Nine dimensions. Forty-seven pages. And an answer repeated like a litany: insufficient information, cannot assess.

I read it three times. The first time I laughed. The second time I opened a spreadsheet and counted the empty cells. Four hundred and nine. The third time I sat still for a long while, because I realised I was looking into a mirror held up to an entire industry.

This was the most honest document I have read in thirty-seven years in this trade. It was also the most frightening.

A report that knows it knows nothing is harmless. A report that does not know it knows nothing, yet still arrives properly formatted, with tables, with star ratings, with recommended actions, is the dangerous thing. And I have received hundreds of documents like that over the past decade. They differ in exactly one respect: they do not leave the cells empty.

They fill them in.

The machine that is never allowed to fall silent

In 2026 I started writing. The same year, the newspaper where I learned the trade was founded. Back then, the equipment of a football watcher consisted of four things: a notebook, a pencil, a stopwatch, and a pair of eyes capable of enduring a cold November stand.

In 2026, a company in England began paying people to sit and watch video and count every pass. They called it detailed statistics. By the late 1990s the first camera systems were installed in stadiums to track players by signal. In 2026 the world governing body approved goal-line technology. In the 2026-14 season the English Premier League installed it across every ground. In 2026 the two largest data companies merged into a single entity holding most of European football's event data. In 2026 FIFA began operating a central clearing house to control cross-border transfer money. In April 2026 UEFA replaced its old financial fair play rules with a new financial sustainability framework. And in June 2026 the World Cup expands to forty-eight teams, one hundred and four matches, running from June 11 to July 19 across three countries.

Thirty-five years. Four changes of measurement unit. And one consequence nobody wants to name: football moved from a shortage of data to a surplus of data, and from a surplus of data to a surplus of format.

Format is not data. Format is the mould. Data is the dough. You can own a thousand moulds and not a single gram of dough, and you will still produce a thousand hollow cakes, beautifully packaged, fully labelled, with nutritional information, a use-by date, and a barcode.

The forty-seven pages I read this morning were one of those cakes. Baked at the right temperature, for the right time, in the right mould, and empty inside.

The economy of the mould

I need to be precise here, because this concerns Vietnam more than people realise.

An average Vietnamese sports outlet publishes between eight and twenty-five articles a day. On a big match day, that figure doubles. Of those, the share containing genuine analysis, built on match observation or self-collected data, I would estimate at under ten percent.

The rest is mould.

Mould one: five reasons. Mould two: three talking points. Mould three: preview and score prediction. Mould four: betting angles. Mould five: predicted line-ups. Mould six: manager quote round-up. Mould seven: players to watch. Mould eight: head-to-head comparison across six criteria.

These moulds are not evil. I have used them my whole career. A mould helps a reader orient quickly. The problem lies elsewhere: when the speed of production exceeds the speed of observation, the mould becomes the only thing left. Writers begin filling the mould with whatever is cheapest to find: ready-made quotes, copied data, the general mood of the majority. And the mould, because it is always full, creates the illusion that something is always inside.

Ten years ago a young editor asked me the secret of writing fast. Write slower, I said, then hold back. He did not understand. This trade does not pay you for silence, I added. But it does not pay you for filling a gap with air either. It pays you for saying correctly what you saw. He still did not understand, and three years later he was running a team producing thirty articles a day with automated tools.

I do not blame him. Economics does not blame anyone.

Since 2026 a new class of content has appeared and spread very quickly: short answer blocks built for search engines and automated answering machines. The format is astonishingly tight: a core answer of no more than sixty words, three to five bullets of no more than twenty-five words each, with source and publication date, plus two or three follow-up questions. That is a good standard. It forces concision, sources, and full entity names instead of pronouns.

It also teaches an entire generation that the value of an article is measured by how easily it can be extracted, not by what it can see.

And no cell in that template asks: were you at the match?

Dissecting a void

The forty-seven pages are interesting because they are not the product of laziness. They are the product of a serious, disciplined, clearly layered process that failed at exactly one link.

I spent two days reconstructing its path. There are three layers.

Layer one: the pipeline

Every automated content system runs through four steps: fetch, render, parse, extract.

Fetch opens the link and pulls the content. Render builds the page as a user would see it. Parse locates the body, the headline, the byline, the timestamp. Extract turns prose into structure: who, where, when, how much, and to what conclusion.

The forty-seven pages show a beautiful paradox. The classification step succeeded: the system knew this was football content. The extraction step returned zero.

Which means the pipeline did not fail at the point of decision. It failed at the point of reading.

Three causes are common for this kind of blockage, and I have met all three in daily work.

The first is a paywall. The content sits behind a login. The system retrieves the shell but not the substance. The page still returns a success code, still has a headline, still has a cover image, but the text is only an invitation to subscribe.

The second is geo-blocking. The server refuses to serve by network address. I have met this while following Japanese outlets from Hai Phong.

The third, and the most common in Vietnam, is content built by code that runs live in the browser. The page returns an empty frame, then calls data and injects the text. A machine reader that cannot run that code sees only the frame.

All three lead to the same outcome: the system receives a document with a valid shape and an empty body. And because the downstream process is designed never to fall silent, it must do something with that emptiness.

It produces forty-seven pages.

This is the point I want read carefully. The failure is not that the pipeline returned zero. The failure is that somebody decided zero should be packaged as a document that looks complete. Those forty-seven pages are not an accident. They are a design decision.

And every major sporting organisation is making exactly that design decision, except they have not named it.

Layer two: substitution

The information-value table in that document has five star levels. Four of the five categories were rated one star, with a note that there was no sporting content to evaluate. The fifth was rated two stars with a very good line: reference value exists only as a negative exemplar.

I looked at that table for a long time. It is formally perfect. It has a scale, criteria, notes, a conclusion. A hurried reader would skim it and think: fine, this document has been quality-assessed.

But it is scoring something that does not exist.

This is the error I call false precision. False precision occurs when you measure an entirely empty thing with great care, and the result of the measurement looks identical to the result of a real measurement. In football, false precision is everywhere.

It appears when a report uses possession share to conclude something about a team's quality without anyone having watched the match. A side with sixty-eight percent possession can still lose by three goals to three counter-attacks. I have watched those matches. I have seen a back line push up beautifully, textbook-correct, and get torn apart by one through ball in four seconds because a centre-back rotated his hips half a beat late. No data table records that half beat.

It appears when a preview concludes something about a team's mentality from three press-conference quotes. A team's mentality lives in the tunnel, in who stands apart from the group, in the captain talking to a young defender twelve seconds before kick-off. Press conferences have never given me any of that.

It appears when a transfer analysis rates a player on data from one league and concludes he will fit a league where the tracking systems operate entirely differently.

And false precision is why I never trust a prediction merely because tables accompany it. I trust a prediction when the writer describes what they saw well enough that I can picture the scene inside my own head.

Tables cannot do that. Description can.

There is a paradox I must admit: I am myself a product of the mould. The Quang Nam affair of 2026, which I will come to, was packaged as five reasons. Five reasons. A neat, easily extractable format.

Was my analysis right because it was right, or because it was presented as if it were right?

That question has cost me many nights of sleep, and the honest answer is: I am not certain.

Layer three: the standard for emptiness

There is one concept in that document I believe should be mandatory in this trade. Handling an empty input.

That document did one thing correctly: it did not invent. It printed the entire framework and filled every cell with a line stating that there was insufficient information to conclude. It labelled the confidence level of each inference. It stated plainly at the top that no conclusion in the document should be treated as an analytical output.

I reread that section four times. Four times, because I wanted to be sure I was reading it correctly.

In thirty-seven years I have never seen a football organisation do anything similar. No club has said: we do not have enough data to conclude. No federation has said: we have not yet assessed this. No coaching staff has said: we signed this player on a sample too small to justify.

Our industry has no word for I do not know.

It has words for we are disappointed with the result. It has words for we need more time. It has words for the referee was wrong. It has words for the development process is on the right track. But it has no word at all for admitting that the evidentiary basis of the decision just announced does not exist.

I once sat in a press conference in Vietnam where a coach was asked why a young player started seven consecutive matches and then vanished from the squad for twelve. The answer: he needs time.

I answered that question in my own head for three years. Perhaps he was injured. Perhaps he lost form. Perhaps there was a disciplinary issue. Perhaps he was playing well but being played out of position. Perhaps there was a clause in his contract whereby a certain number of appearances would trigger a bonus the club did not want to pay. That last possibility exists in Vietnamese football more than anyone cares to admit.

None of us knows. And an entire football nation runs on gaps like that, filled with answers that have a mould.

How I watch a match

People often ask what I look at from the stand. I give a very short list.

For the first ten minutes I do not watch the ball. I watch the posture of both back lines when their team does not have possession. The distance between a centre-back and his full-back on the same side tells me which zone they have been coached to protect. The direction a centre-back turns tells me whether he fears the through ball or the cross.

For the next ten minutes I watch the midfield when the ball is lost. Who drops first. Who stands still. Who turns his head toward the bench. Those three behaviours repeat all match and describe the team more accurately than any formation diagram.

Then I watch the coach. Standing or sitting. Which hand he signals with. Whom he calls. By the thirtieth minute, if a team is behind and the coach is still sitting, that signals a plan different from the one I am watching on the pitch.

And by the seventieth minute I watch the players without the ball. This is my favourite part and the hardest to record. A good midfielder is always in the right place before the ball arrives: he moves one beat ahead of the pass. You will not see that in a statistics table, because the table records the action after the ball has reached the foot.

If I had to compress it, a match I watch properly gives me twenty to thirty usable observations. The rest of the evening is noise, emotion, cold coffee, and people around me asking which team will win.

I usually answer that I do not yet know. Not to appear modest. Because I genuinely do not.

The Quang Nam bet, and what it does not prove

In January 2026 I wrote an article with a fairly loud headline: Quang Nam will win the title, and here are five reasons.

Let me set the context properly. At the time I was a writer of long standing in Hai Phong, produced by that city's own academy. Most experts placed Quang Nam in the relegation group. Some colleagues called me an idiot. I did not argue, because I understood why they said it: Quang Nam had no big-club budget, no deep pool of national team players, and no championship tradition.

I wrote that article because I had gone to watch three of their pre-season friendlies. Three. Not ten. Not a full season. Three.

What I saw was in no statistics table. I saw a team transitioning from defence to counter-attack faster than the league average. I saw their midfield hold horizontal spacing almost unchanged through the second half, even when behind. I saw captain Dinh Thanh Trung reposition his two wide midfielders four times in one friendly without ever looking at the bench. That is the signature of an internally stable structure.

At the end of the season Quang Nam were champions with thirty-eight points, one point ahead of Ha Noi. Dinh Thanh Trung later told the press that the article had lit a fire under the whole squad.

I was right. And for years I let that correctness underpin everything I wrote afterwards.

Now I want to say something I have never said publicly.

Three friendlies is a sample far too small. I saw real things: the stability of the spacing, the capacity for self-correction, the speed of transition. But I inferred a conclusion far larger than my evidence. I did not merely see a well-structured team. I inferred that this team would win a twenty-six-round league in which injuries, suspensions, managerial changes and plain luck across twelve decisive matches all carry weight.

Had Quang Nam finished second on thirty-seven points, my article would be a textbook case of over-reading. Had they finished fifth, I would have been the idiot my colleague called me, with nothing to fall back on, because in that article I had not stated my confidence level at all.

I did not write a single line about the probability of being wrong. Not one.

Before Quang Nam became a story, I read them. That is true. But a good reader of the game is not someone who is always right. A good reader of the game is someone who states clearly where they might be wrong, before the result arrives and turns words into history.

Where I was wrong

In March 2026 I wrote a long piece predicting Becamex Binh Duong would win that season's V-League. I had reasons. I had observations. I had data on the average age of rival back lines, on minutes played by key players in the first half of the season. I assumed the season would be completed.

In August 2026 the season was cancelled because of the pandemic. There was no champion.

That is the worst possible outcome for someone who bets on judgement. I was not right. I was not wrong. I was voided.

And I noticed something about how we handle such cases. We do not publish them. We wait for public memory to shorten, then issue a new prediction, and if the new one is right, the new correctness replaces the old one that has faded. Nobody ever audits.

In thirty-seven years I have never published my hit rate. I have never published my miss rate either. I am an analyst with no audit report.

Think about that the next time you read a betting preview containing eleven predictions, of which at least three contradict each other about the same match.

My industry sells probabilities and never returns the test results.

What if the void is the correct answer

Here I have to raise my own deepest doubt.

There is a possibility I have considered and cannot dismiss: those forty-seven pages are not a failure. They are a correct answer to a question football is not ready to hear.

Suppose most of the football content we consume daily contains, in truth, insufficient information to analyse. Suppose the real figure is seventy or eighty percent. In that case an honest system would return seventy or eighty percent of cells as insufficient information. And such a system would be judged broken, sent for repair, and replaced by a system that knows how to fill the gaps.

I am not certain I am right here. Perhaps I am rationalising. Perhaps I am trying to turn a technical fault into a philosophical thesis because I prefer large theses to error reports.

But one fact tilts me the other way. When I write a genuine deep analysis, it costs me twenty to forty hours of work: rewatching footage, taking notes, cross-checking figures, calling two or three people in the trade, then writing. Such a piece contains perhaps fifteen to twenty genuinely valuable observations.

The rest of the article is frame, guidance, context, smoothing.

Which means that even at its best, the ratio of real information to total words is about fifteen percent. If an automated process read such an article and extracted only the frame, it would obtain exactly what those forty-seven pages obtained.

Nothing at all.

Perhaps the pipeline is not broken. Perhaps it is simply reflecting honestly that most of the words in a football article are filler. And we have built an entire industry around packaging filler attractively.

My eye is also a model

This is the hardest part. I write it because I have postponed it too long.

I am known for sitting in the stand and reading things others cannot read. I call it the field eye. I have used it to explain Quang Nam, to explain the German national team, and dozens of smaller calls.

But the field eye is a forecasting model.

It has parameters. It was trained on data. It was trained over four years at the Lach Tray academy, over afternoons with a Japanese coach who said very little, over eight Olympic Games, eight World Cups, and journeys following the great European cycling tours, where I learned that a three-week race is decided not by the strongest rider but by the one who distributes his strength most wisely.

It has biases. It leans toward structure over individual talent. It leans toward stability of spacing over moments of brilliance. It undervalues teams with outstanding individuals but loose organisation, and I have paid for that bias many times.

It has blind spots. I do not read pure short-distance speed well, because my eye was trained in a slower era. I am also poor at judging teams built around set pieces, because I get absorbed in the rhythm of open play.

And it has an error rate I have never published.

The only difference between me and the machine that produced those forty-seven pages is that the machine printed a confidence level for every inference. I did not.

Champions are born to shine. Critics are born to see the darkness before they do. But the darkness inside the critic's own eye is invisible to everyone, including its owner.

Three at the back and the art of dodging blame

Over the past two seasons, the number of elite teams lining up with three centre-backs has risen again. Many analyses call it tactical evolution. I think most of it is caution dressed in progressive language.

A back three has one very convenient property for a coach. It absorbs failure better. When a team plays a back four and is cut open on the flank, the fault is systemic: the flank was vacated, the wide midfielder did not drop in time, the outside centre-back was dragged out. Analysts can see it, point at it, write about it.

When a team plays a back three and is cut open, the fault is usually assigned to an individual: loss of concentration, poor judgement, weak duels. Because the system has a spare man in the middle, almost every goal conceded can be framed as a residual individual error.

That is the difference between a system that owns errors and a system that transfers them.

At club level this shift comes with a specific economic benefit. A back three lets a team list more attackers on paper without buying quality full-backs. You push two wide players high, call them attacking full-backs, and sell the crowd a line-up that looks adventurous.

Then, when the match reaches the second half level, that team shrinks, both flanks drop, and on television you see exactly what you saw twenty years ago: a five-man line in front of the box.

For Vietnamese football I have a specific note. A back three demands higher situational reading and spacing coordination than a back four, because the gaps between centre-backs are larger and there is no one covering vertically. Domestic teams rarely get enough training time to automate that system. The result is that they have the shape of a back three and the behaviour of an uncoached back line.

I have watched many such matches. In the first half you see three centre-backs in a beautiful straight line. By the sixtieth minute the gap between the middle and right defender has stretched to nearly twenty metres, and neither central midfielder drops in to fill it. The goal comes from exactly that gap, and the next day's report will say the right centre-back played badly.

Signing fees for free agents: a loophole with a name

UEFA's financial sustainability framework, introduced in April 2026 to replace the old financial fair play rules, caps squad cost as a share of revenue and limits losses over a three-year cycle. The new framework is tighter in many respects. It still leaves one wide opening.

That opening is the signing fee for a free agent. The money paid to a player when his contract expires and he moves on a free transfer. In the accounts, it is usually amortised over the contract length, just like a transfer fee. But it does not pass through the central clearing system between two clubs, so it is scrutinised far less.

Compare two situations. Club A pays twenty million euros in transfer fees to Club B. That deal appears in the international transfer report, in the central clearing records, in every media statistics table. Everyone knows. Everyone analyses. Everyone compares it with market value.

Club A pays twenty million euros in a signing fee to a free agent. There is no club at the other end. There is no international clearing record. The sum appears as one line in the annual accounts, labelled prepaid costs or signing remuneration, and is usually revealed only by a handful of investigative journalists with their own sources.

The same money. The same effect on sporting competitiveness. A gulf in transparency.

As free transfers grow more common, because players and agents understand they can capture most of the saving from the selling club, this gap widens every year. Europe's richest clubs increasingly favour the method, not because they like free agents, but because it is the only way to increase squad spending without damaging their compliance ratios.

If you think this has nothing to do with Vietnamese football, think again. Domestic clubs are shifting in the same direction: less in transfer fees, more in direct remuneration to players and agents. That makes money flows harder to track, harder to value, harder to audit. A large expenditure quietly becomes a private negotiation between two people.

Expected goals and the limits of a good metric

I use expected goals in most of my analysis. It is a good tool. It forces an analyst to separate chance from outcome, a six-metre shot into the top corner from a thirty-metre shot into the stand.

But I want to state three of its limits clearly, because I see it cited as a final verdict.

The first limit is the data source. An expected goals model is built from a set of past shots with variables for position, distance, angle, body part used, number of defenders nearby, and ball speed. But those rows were collected by a specific organisation under a specific definition. Two providers will give you two different values for the same shot. Nothing is absolute.

The second limit is the concept of a chance. A shot with a very low scoring probability can still be the signature of a good combination. A shot with a high probability can be the product of luck after a defensive error. The metric measures the outcome of a process, not the process itself.

The third limit is Vietnamese football. Domestic leagues lack full event-tracking data for every match. In many competitions, data exists only for a selected group of fixtures, or only in basic form: shots, shots on target, corners. In those conditions, any advanced analysis of a domestic league is in truth an analysis of an unrepresentative sample.

I have made this error. I once cited a team's metrics from just seven matches in a twenty-six-match season. Those seven were the fully broadcast fixtures. Fully broadcast fixtures are usually the fixtures of the most-followed clubs. Which means I analysed the most-followed club using its own data, then generalised the conclusion to the whole league.

That is a selection error. It has a name. I still commit it.

The analyst in the dressing room

Over the past fifteen years the number of analysts at professional clubs has grown very quickly. That is good. But there is a problem nobody wants to write about: most analysts at smaller clubs have no access to training. They receive data from a supplier, build a report on a laptop, and hand the coaching staff a document that has never been verified by anyone who was on the grass.

Which means we have an intermediary layer between the pitch and the decision, and that layer cannot see the pitch.

The result is a very specific paradox. The report says the opponent is weak at defending counter-attacks after losing the ball in their own half. The coach reads it, believes it, pushes the line high. The match ends with two goals conceded from counters. The post-match report says the line was too high.

Both reports are correct according to the data. And both are useless, because both ignore a question only someone present could answer: is this opponent willing to sit deep and counter from the first minute today?

I have watched such matches. You can see the answer in the first ten minutes, in the way the opposing midfield does not push up when they have the ball in their own half, in the way the striker does not contest but drops to screen the gap. Those signals are in no event dataset. They live in the posture of players before the ball moves.

A good analyst in a dressing room is not the one who reads the most data. It is the one who walks out to the training pitch and asks the coach a single question: how do you intend them to play today?

Vietnamese football and the ritual of silence

I have lived in Hai Phong a long time. I follow the domestic game with the patience of a man long accustomed to waiting for changes that do not come.

What I see most clearly over the past decade sits in a communication standard, not in tactics.

In Vietnam, when a club announces a signing, the announcement never contains a data assessment. It never says: we watched two hundred and forty minutes of footage of this player, and the sample is still small. It never says: we accept a high injury risk because the transfer fee is low. It never says: we do not yet know whether this player fits our system.

I understand why. An announcement admitting uncertainty is read as weakness. An announcement admitting uncertainty makes supporters lose faith. An announcement admitting uncertainty gives a sponsor a pretext to walk away.

But consider the price on the other side. We have a system in which every decision is presented as the output of deep analysis, while most are made on one person's instinct, under one week's pressure, with less information than a betting preview written by strangers on the internet.

Media attention concentrates on a very small group of players, names like Nguyen Quang Hai, Nguyen Tien Linh, Do Hung Dung, while most of the league's remaining players do not have a single line of complete data recorded about them. A football nation that analyses thirty names and forgets the other three hundred has a sampling problem, not an enthusiasm problem.

The empty-input standard applied in that document may be exactly what Vietnamese football needs most. Print a sheet, write clearly: insufficient information to conclude. Sign it. Stamp it.

I know nobody will. But I want it written down, so that later, when somebody does, we can say the idea existed beforehand.

One hundred and four matches, and the sample-size problem

In June 2026 the World Cup expands to forty-eight teams, one hundred and four matches, running five and a half weeks across three countries. It is the largest expansion in the tournament's history.

I have heard two arguments. The first says more matches means more data, more analysis, more stories. The second says quality will dilute.

Both miss what I consider the most important point.

A tournament of one hundred and four matches will generate more analytical content than audiences can consume. When supply exceeds demand, value shifts from analysis to speed. Whoever publishes first wins. And once speed becomes the criterion, producers are forced to use more mould, observe less, and fill gaps with whatever is available.

One hundred and four matches is a vast machine for producing gaps.

And here I return to those forty-seven pages. In a season like that, any system that dares return insufficient information for most matches will be considered out of date. Any system that knows how to fill in will be bought. And we will get a month and a half of content that is very smooth, very full, very well formatted, and empty.

What I take away from those forty-seven pages

Collapse is not an endpoint. It is a test that sorts practitioners from performers.

The forty-seven pages I read at 2:14 in the morning on February 14, 2026 sorted me.

I laughed on the first read. That is the reaction of a man who thinks he stands outside the story. But I do not stand outside it. I am part of that pipeline. I have written for thirty-seven years, produced an enormous volume of words, and I have never published my hit rate, never published a confidence level for each conclusion, never stated in any article how much of it was observation and how much was filler.

The machine managed it. It printed cell by cell: insufficient information.

What I want to leave behind, instead of a summary, is a test anyone can run right now.

Take the last football analysis you read. Count the cells you can verify yourself. Not the cells with numbers. The cells with numbers that you can source and confirm within ten minutes. For most current writing in Vietnam, I would guess you count fewer than three.

Then do the same with something you wrote yourself.

And if you are a coach, a technical director, a decision-maker, write it down once in your life, in black and white, that you do not know.

I will bet that within two years, at least one club in Southeast Asia will publicly announce a personnel decision accompanied by an acknowledgement that the data was incomplete. Not because they are better. But because the cost of always knowing will become too high.

And if I am wrong about that prediction, I will be the first to write that I was wrong, with dates, and with a list of the times I was wrong before.

That is the only thing a person in this trade can promise without needing any more data.