Empty Files, Beautiful Reports: The Most Expensive Silent Failure in Modern Football Recruitment

March 2026, 9:20 in the morning, a second-floor meeting room at a training centre in southern Germany. On the table lay a 22-page scouting report on a 17-year-old midfielder. Bold headings. Eleven sections, from "Technical Foundation" to "Development Recommendation". Bar charts built in dedicated software, axes balanced, brand colours correct. Page eleven concluded: monitor for another six months before deciding.

I turned the document over and asked one question. Where is the raw data.

No one could answer. The player's GPS log for the past three months was empty. The video timeline was untagged. The under-19 match reports were absent from the internal archive. That 22-page report had been built on a layer of data that never existed, and it still looked beautiful. It was still persuasive enough for a technical director under pressure to finalise a squad list before the winter break.

What chilled me was not the emptiness. It was that the emptiness had put on a complete ceremonial uniform and walked into the meeting room as evidence. In 39 years in this trade I have seen hundreds of reports that looked this good. Only four times did I sit down long enough to trace them back to source. All four times, the source was empty.

The pipeline and its unguarded break points

A young player's file, from the moment a 15-year-old enters the radar to the day he signs his first professional contract, passes through at least five layers. Collection. Classification. Tagging. Metric calculation. Report writing. Each layer has an owner, and each layer can break without making a sound.

Collection is the most fragile. A source can sit behind a paywall. A federation website can change structure after a user-interface upgrade. An API credential can expire during a holiday week. A match video can download with the first three frames corrupted while the editing software still exports a file that appears valid. When collection breaks, the layers downstream receive no alarm. They receive an empty input, and they keep running.

Classification assigns positions. A box-to-box midfielder gets tagged as a defensive midfielder because he played in a double pivot for the youth team. One wrong label, and every metric downstream tilts off its axis. Tagging depends on human endurance. The third tagger on an eight-hour shift starts missing off-ball duels. Those are the duels that define a defensive midfielder's value, and they disappear from the data without leaving a trace.

Metric calculation is where belief is manufactured industrially. Opta, StatsBomb, Wyscout, Hudl, Transfermarkt — these names have become the shared reference frame of the industry. They are useful. They are also black boxes with very wide entrances and very narrow exits. A metric computed from a 180-minute sample of under-19 football carries an error margin larger than the gap between the two players it is being used to compare.

And the final layer, report writing, is the most dangerous. Because it is the only layer the decision-maker actually sees. A report with no spelling mistakes is always read as a report with no gaps.

I have sat in two very different meeting rooms. One in Munich, where the academy employed four full-time analysts, two video specialists and its own server room. One in Vietnam, where a single person filmed matches, took notes, booked the team's flights and wrote the report for the coaching staff. Both environments share the same blind spot: nobody is paid to sit and check whether the input is real.

When the source is blocked, the template survives

There is one kind of error I encounter more than any other, and it has nothing to do with football.

When a system is pre-programmed to a template and its input vanishes, it does not stop. It keeps running and emits the shell. I once read an analytical document in which, under "entities involved", the line still read as an internal instruction along the lines of "identify from the information points above" — while above it there were no information points at all. That instruction had survived the entire processing chain and travelled straight into the finished product.

This is not the writer's fault. It is the architecture's fault. A system with no mechanism to refuse output will always output. It will fill the blanks with template phrases, with dashes, with the words "insufficient information" — but it will still ship.

For a newsroom, the consequence is a bland article. For a football academy, the consequence is a player.

Based on my own experience watching Bundesliga under-19 matches and regional youth competitions in southern Germany across seven consecutive seasons, I have arrived at one rule: a report with a section left blank is more trustworthy than a report with every section filled. A blank field is a confession. A field filled with a template phrase is an unconscious lie.

A gap does not mean safety

This is the most expensive mistake, and it is common enough that I am almost no longer surprised.

When a player's medical section is empty, the reader understands "no issues". When a disciplinary section is empty, the reader understands "well behaved". When the financial data on a transfer is empty, the reader understands "clean deal".

None of that is logically true. A file's silence is a neutral state. It is not a certificate. But in a meeting room, at minute 40 of a three-hour meeting, nobody is lucid enough to distinguish between "no data" and "no problem".

I have seen the consequences of this false equivalence at scale, and it does not happen only in youth football.

In February 2026, the Premier League announced 115 charges against Manchester City relating to financial rules across the 2026-2026 period. In November 2026, Everton were deducted 10 points for breaching Profit and Sustainability Rules, later reduced to 6 on appeal in February 2026. In March 2026, Nottingham Forest were deducted 4 points for the same category of breach.

In all three cases, the central question was never whether a club spent too much. The central question was: at the moment the spending happened, how many people in the room could actually see the full number. And of those who could, how many believed that an empty cell on a spreadsheet meant the item did not exist.

A gap in a file never announces itself. It waits to be read as whatever is most convenient for the person who needs an answer.

The price of a decision built on an empty stratum

Here is a simple calculation I use with young trainees.

A bad transfer at first-team level for a mid-tier Bundesliga club costs roughly 8 to 15 million euros in fee, plus 2 to 4 million euros in wages across a four-year contract. If the player leaves on a free, the loss is booked into amortisation and drags across three more seasons. But that is not the most expensive part.

The most expensive part is a slot in the squad. A 19-year-old academy player who should have been promoted to the first team that season is pushed down because his position has been filled by the signing. He loses 18 months without top-level football at the fastest point of his development curve. No wage ledger records this loss. It sits in silence, like the empty data layer.

I know this from a file I signed myself.

In 2026, at 46, working as a player development consultant, I assessed a 16-year-old midfielder. Traditional data showed 78% pass accuracy in the under-17 Bundesliga — an entirely healthy figure. But GPS metrics had just been introduced, and they gave me something else: his top speed was only 28 km/h, below the team average. I withheld my recommendation and declined to propose his promotion to the under-19 side.

The coaching staff objected. He moved to another academy that summer.

Six years later, re-analysing the list of 19 players under 20 who started in the knockout rounds of a World Cup, I found that 14 of them had previously been rejected by German academies on physical or athletic grounds. A 19-year-old who scored four goals and threw Argentina's defence into disorder in the 2026 round of 16 carried the same data profile as the boy I had waved away: good technique, high speed, and a physical foundation undervalued by traditional assessment.

I wrote a 40-page internal memo. It openly admitted the limits of the method I was using. Nobody asked me to write it. I wrote it because after the storm, I counted 47 pieces of debris, and together they assembled a picture nobody wanted to look at.

But stopping there would have trapped me in the opposite error: blaming the eye and worshipping the number. Both are wrong in the same way.

The eye is also silent when it is wrong

Nine years after that episode, I still repeat this to younger colleagues: the eye is not an honest instrument. It is an instrument with a probability attached.

A scout watching a match can miss up to a third of the off-ball duels. Not through laziness, but because the human visual field is finite and the game does not wait. Yet the difference between the eye and a file lies here: the eye deceives without ever keeping minutes, while a file deceives and leaves a trail.

Because it leaves a trail, a file can be audited. An eye cannot.

That is why I still tell my trainees: I do not trust my eyes; I trust what the file leaves behind. This is usually misread as contempt for direct observation. Wrong. I watch every match with my eyes first, write it down immediately afterwards, then cross-check against GPS and video within 24 hours. Three independent sources. If the three disagree, I do not conclude. If only one remains, I do not conclude. If none remain, I say plainly: insufficient data.

This sounds slow. In a six-week transfer window, it is slow. But youth is an unexcavated geological layer; do not rush to pour concrete over it.

One more layer, one more place to break

The industry's default reaction to a data failure is to buy another tool.

A new analytics system. A machine-learning model predicting potential. A real-time physical monitoring platform. An AI layer that summarises scouting reports.

I do not oppose these tools. I oppose the assumption behind them: that adding a processing layer adds accuracy.

In operational reality, adding a layer means adding a silent failure point. If layer four does not check layer three, and layer three does not check layer two, then layer five only makes the error flow more smoothly. A crude error smells like an error. A beautiful report generated from a crude error has no smell at all.

This is the central paradox of modern recruitment. The more layers of tooling, the harder the final product is to challenge — and the harder it is to verify.

I have seen this in both football cultures I have lived in. In Germany, the data infrastructure is built so tightly that a report missing its raw data can still look fully compliant, because every field has the right format. In Vietnam, where sources are thinner, the gap usually surfaces earlier and is therefore sometimes less dangerous. This is one of the few cases where scarcity works as a protective fence.

The German model is not a universal yardstick. Every stratum has its own rules. Before any judgement, I ask myself which layer my own sediment was built from. If the answer is "I don't know", I do not judge.

The gate, not the model

After four times tracing back to source and four times finding emptiness, I changed how I work. I did not change the tools. I added a gate.

The rule is simple enough to sound trivial. No raw data, no report. Not a report with a footnote. Not a report with a blank field. No report.

In the four years since, the number of reports I sign has fallen by nearly half. The quality of judgement in the remainder has risen noticeably, and I need no model to demonstrate that. I simply count how many players get promoted to the first team and are then sent back down. That number fell from seven to two in two seasons.

Three sentences I write on the whiteboard in the analysis room and never erase.

Every stratum tells a story; the question is whether we bother to dig.

Every transfer is an excavation; rush it and you break the artefact.

And the third: a conservative decision can bury talent, but it keeps the foundation from collapsing.

I know the third is the one young people hate most. It sounds like a justification for the slowness of a 55-year-old man who has sat in one place too long. I have asked myself that question for years. If someone more conservative than me had occupied my seat in 2026, would that 16-year-old have been waved away once more.

So I keep one open question in every report I sign. If someone else looks at the same data and sees the opposite, what would they need to add to prove it.

What I am still counting

Old files never die; they simply wait for someone patient enough to read them again.

I have reopened 47 files that I or colleagues had discarded. It became a private index I built myself. Every season I cross-check that list against what those players actually achieved at major tournaments. The index does not prove I was right or wrong. It forces me to answer one question each time I open it: was the layer of data I relied on real.

World Cups are not won on the pitch; they are won from forgotten files.

That sounds like a slogan, and I dislike slogans. Yet every time a transfer window opens, I see it proven in a very concrete way. The winning club is not the one with the most reports. The winning club is the one that knows exactly which of its reports has raw data behind it. Tracing back to source before deciding takes about two hours per file, and in a transfer window those two hours are the first thing to be cut.

The question I carry into next season is not which player will shine.

It is how many of the reports sitting on meeting-room tables across Europe this week actually have a line of raw data behind them, and how many are a very handsome skeleton waiting to be believed.

People usually only check after something goes wrong. But with a scouting file, by the time something goes wrong the player has already signed — and that stratum has long since been paved over.