Hollow Reports: Selling Zero as Insider Intelligence in the Transfer Window

Hollow Reports: Selling Zero as Insider Intelligence in the Transfer Window

At three in the morning on 14 July 2026, while the North American World Cup still had two matches left to play, a forty-seven-page PDF landed in my encrypted inbox. The subject line read: "Nine-dimension deep analysis — midfield candidate, summer transfer window." I opened it and counted. Nine sections. Seventy-two tables. Each table neatly ruled, numbered, with bolded column headers, exactly like a genuine scouting report. But most cells contained three characters: N/A.

No minutes played. No advanced metrics. No age, no sprint distance, no touches in the box. No fee, no wage, no contract length. Not a single line on injury history — in a sport where injury history now sets a transfer price more than goals do.

Thirty-six hours later, a rumour surfaced across three sports outlets, carrying three very specific numbers: a fee, a wage, a contract term. I traced back through the PDF. Those three numbers existed nowhere in it.

A document containing no information was traded at the price of a document containing everything. I found it in a data table nobody bothered to look at.

The transfer window: where noise has a price

The transfer market is the only market in sport where the product is sold before it exists. People don't buy a player; they buy a possibility. And that possibility must be packaged in the language of certainty before it can change hands.

In seventeen years covering this industry, I have never seen the volume of "analysis" this high. Every summer window, the number of scouting reports circulating in closed groups multiplies, while the number traceable to a single live viewing stays flat. The gap between those two figures is the market.

A report doesn't need to be true to hold value. It needs to look professional enough that the recipient won't discard it, and vague enough that nobody can verify it. Nine sections, seventy-two tables, immaculate rulings — that is industrial design, not the output of analysis.

Summer 2026 is stranger than any window before it, because the World Cup finished exactly as the window opened. A seven-match tournament generates a small data sample projected onto an enormous market. Any midfielder who plays four good knockout rounds enters July with his price re-set twice over. And anyone who wants to inflate an asset has a ready excuse to publish a document.

Anatomy of a hollow report

The document's structure is worth dissecting, because it reproduces almost perfectly every scouting template I have read in a decade.

Section one assessed competitive value. Section two, player data. Section three, club operations and salary cap. Section four, league landscape. Section five, rules and governance. Section six, coaching staff and locker room. Section seven, risk. Section eight, media narrative and expectations. Section nine, industry ripple effects.

Those nine headings are entirely reasonable. A real coach needs all nine. But filling all nine requires at least three independent sources: verified match data, contract records, and internal information on physical condition. The document had none of the three.

What makes it dangerous isn't the N/A cells. It's that every N/A cell sat next to a rating scale. A cell reading "insufficient information" still carried a confidence label, still got ranked, still got scored. The frame was completed before the content was written. That inverts the process: the conclusion comes first, data comes second, and when data never arrives the conclusion stands alone.

I counted two hundred and fourteen empty cells formatted identically to populated ones. Same font size. Same background. Same borders. To a reader skimming on a phone, an N/A and a "17.4" look similar enough that the brain files them as the same category.

That is the whole trick. You don't have to invent numbers. You only have to make empty space look like occupied space.

Fingerprints in the metadata

I trust physical traces over testimony. In a PDF, the physical trace is metadata.

File creation and last modification were eleven minutes apart. Eleven minutes for forty-seven pages, nine sections, seventy-two tables. Nobody writes a scouting report in eleven minutes unless it was generated from a template waiting for a name.

The authoring software was a generic office suite, not any sports-analytics platform. The author field held the manufacturer's default string, not a person's name.

I cross-referenced this file against nineteen others collected across three recent windows. Eleven shared the nine-section structure. Seven shared the table count. Four shared the same default author string.

Four files, four different windows, four different players, one fingerprint.

I don't believe testimony. I believe fingerprints on contracts and shoe prints in corridors.

The amplification chain: three layers

A hollow report is harmless on its own. It becomes fact by passing through three layers.

Layer one is the closed group, where the document is dropped with a single line: "internal, don't spread it." That protocol manufactures value by itself. What is forbidden to share gets shared three times as fast.

Layer two is the aggregator. A mid-sized site receives thirty to fifty reports a day during the window and cannot verify a tenth of them. It does the rational thing under its constraints: republishes with the phrase "according to sources close to the situation." The phrase isn't grammatically wrong. It simply doesn't tell the reader whether those sources can read a spreadsheet.

Layer three is the market. Once a fee appears on three sites, it starts being cited backward. A fourth site repeats the number and credits the first. The first site reads the fourth and believes it has been confirmed. The loop closes within forty-eight hours, and the number then lives independently of its origin.

People watch the scoreline. I watch who gets paid after it.

The mathematics of a swapped denominator

There is a subtler form of hollow report, harder to catch because every number in it is correct.

In March 2026 I received a twenty-two-page dossier on a twenty-two-year-old midfielder pursued by three clubs. It was dense with data: sprint percentiles, pass completion by zone, recoveries per ninety, percentile ranking against midfielders in Europe's top five leagues.

Every metric checked out. I verified them one by one. The problem lay elsewhere: the sample was two hundred and fourteen minutes across nine matches, seven of them substitute appearances after the seventieth minute.

A percentile ranking computed on two hundred and fourteen minutes is not mathematically wrong. It is statistically meaningless. At that sample size, variance is wide enough that one long-range goal lifts an entire index into the top group. And buyers don't read variance. They read rankings.

This is the most dangerous variant: not empty in content, empty in denominator. The report doesn't lie. It just fails to mention it is describing very little football.

Based on my experience tracking matches, there is one pragmatic rule: for any dossier with advanced metrics, the first thing to find is not the metric but the minutes line. No minutes, no metric.

On my own match-tracking log, I write that number in the top-right corner of the first page, before the player's name. Names carry bias. Minutes don't.

Where the money runs

A hollow report doesn't generate itself. Someone pays for it, and someone earns from it.

There are three payer groups. First, fans paying subscriptions for "insider" content, a few dollars to a few dozen a month. Second, bookmakers and data platforms buying information flow as raw material for pricing. Third, the agents themselves — paying not to learn the truth but to manufacture a version of it that helps their client at the negotiating table.

The earning side is broader. Some sell subscriptions. Some sell positions. Some need only a name in print to move a deal in which they hold a percentage.

That is why I never assess a report by its content first. I assess it by one question: if this spreads, who profits in the next forty-eight hours?

Every contract has two pages: a public one and a real one.

Injury files: the legally sanitised hollow report

One category of hollow report is never called that, because it is protected by a beautiful word: privacy.

Throughout my career I have kept touching the industry's largest void — medical data. Clubs publish the injuries that suit them and withhold the ones that don't. A hamstring tear announced as "tightness, two weeks" may in truth be a grade-two strain and eight weeks out. Fans buy tickets. Sponsors renew. The share price of a listed club doesn't move.

At the 2026 World Cup I was assigned to review match footage and found a midfielder recording eleven sprints above thirty-two kilometres per hour, while his club injury file logged a hamstring tear four months earlier. Cross-checking GPS data from qualifying, his distance covered was up twenty-three percent against his two-year average. No doping evidence. Only a gap between match data and medical records, and nobody has answered for it.

I logged it and kept watching. That is all you may do when you hold one data point and one question.

My protocol for receiving a report

You cannot verify a hollow report unless you have a procedure ready before it arrives. This is mine.

One: read the footnotes before the headline. The data source sits at the end, and it is the only part pretty formatting cannot fake.

Two: find the denominator. Minutes, matches, seasons. Without a denominator, every ratio is decoration.

Three: check time consistency. How long between creation and last modification. A forty-page document built in ten minutes is a document with its conclusion pre-installed.

Four: identify the beneficiary — not to accuse, but to rank credibility. Information from an interested party isn't false; it just needs a second source.

Five: cross-check against an unrelated source. If two sources share an origin, you have one source counted twice.

Six: timestamp everything. During a window, a true number today can be false by Friday.

Seven: say "I don't know" when you genuinely don't. This is the hardest step and the only one that gives the other six any value.

The reasonable case on the other side

Fairness requires this section, because otherwise my own analysis becomes another single-sided document.

Templates are not fraud. A small second-division club without a full analytics department still needs a frame for note-taking. The frame prevents omissions. The problem is when the frame is sold as a finished product rather than a tool.

Medical confidentiality is partly reasonable too. An athlete has the right not to broadcast his physiological weaknesses. A club has the right not to hand opponents information about where it hurts. I am not asking for privacy to be abolished. I am asking for honesty about the existence of a gap — instead of filling it with a diagnosis that sounds milder than reality.

And the buyer shares responsibility. The market pays for certainty, so the market receives counterfeit certainty. If a platform told readers it did not know, subscriptions would fall. Readers themselves subsidise hollow reports by refusing to pay for not knowing.

The sports analytics industry has produced genuinely excellent work over fifteen years. Hollow reports are a tax on that work. People who do it properly must compete with people who do it fast, and speed always wins the first forty-eight hours.

Who answers for it

I make no accusation against that PDF, because I have no evidence of intent. I have evidence of consequence: a document with no data produced three numbers cited across three outlets, and those numbers are still circulating as I write.

Scandals do not fall from the sky. They are initialled, scheduled, and staged step by step.

What I want in the next window is not fewer reports. I want every number to carry a provenance label, like an ingredient list on a milk carton. Where did this number come from. Over how many minutes was it measured. What does the person handing it to you gain if you believe it.

Those three questions cost nothing. They only require a habit.

And if a forty-seven-page document cannot answer the first one, the valuable thing isn't the document. It is our willingness to say to each other that it is empty.