The Empty Data File and Football's Most Expensive Mistake: Reading 'No Findings' as 'No Risk'

Marseille, 2:47 a.m. I open the analysis file that the system sent back after the weekend round, and the only thing on screen is nine analytical frames with every content field reading two words: nothing. No team name. No player name. Not a single minute played. Not a passing figure, not a revenue line, not a timestamp. Nine dimensions — tactics, finance, results, league landscape, rules and governance, dressing room, risk matrix, media narrative, industry transmission — and all nine return the same sentence: insufficient information to assess.

A lay reader would breathe a sigh of relief. A report that found no problems — what could be better. That is precisely the most dangerous moment in my profession.

An empty report is not a clean report. It is an incident that has not yet been named. In football, those two states get read into each other every single day: a club reads a blank scouting assessment as "nothing wrong with this player"; an editor reads a white statistics table as "nothing to say about this match"; a supporter reads the medical room's silence as "the kid is fine". All three are standing on the same void, and a void never announces that it is empty.

Thirty-one years in this trade, I have written thousands of pieces off numbers. But the biggest lesson I ever learned did not come from a correct number. It came from a wrong file — specifically, an empty one.

Modern football runs on a pipeline nobody looks at

A single Ligue 1 match today generates thousands of recorded events: every pass, every duel, every off-ball run. Those events flow through three layers. Collection — data providers, cameras, sensors, on-site coders. Transformation — models computing xG, xGA, PPDA, physical-load indices. Interpretation — scouts, coaches, journalists, and the people placing bets.

Those three layers rarely sit in the same room. That is where the accidents happen.

I once watched an analyst at a mid-table French club present an opponent briefing to the coaching staff. The opponent had played a back three in four consecutive matches. His chart said back four. Nobody in the room challenged it, because the chart had colour, had graphs, had a vendor logo. By the tenth minute of the match, the head coach discovered that his left flank was being carved open by a player who, according to the data, should not have existed in that position.

A small error. But it reveals the mechanism: a filled table carries a power an empty one does not — the power to be believed without verification.

An empty file is the opposite. It persuades nobody. It simply gets ignored, filed under "pending", or worse, read as a conclusion. In the report that reached my machine that night, one line was written in administrative tone: this is an "unassessed" state, not a "low risk" verdict. That line was correct. And it was the only sentence across nine pages with genuine analytical value.

The cause of an empty file usually has nothing to do with football. The source article sits behind a paywall. The page is JavaScript-rendered, so the scraper receives an empty body while the status code still reports success. A response gets truncated between two processing layers. Or the server simply blocks bots and politely returns a blank page. Four causes, four different fixes, and none of them visible on the end reader's screen.

What keeps me awake is not a single failure. It is the possibility of silent failure repeating across an entire batch — thirty articles, thirty empty files, thirty times nobody raised an alarm, because the system does not treat "no content" as an error.

The tactical layer: when xG, xGA and PPDA all fall silent

This is the layer most vulnerable to fabrication, because it is the layer readers believe they understand.

xG estimates chance quality by summing the probability that each shot becomes a goal. xGA applies the same measure to chances conceded. PPDA counts the passes an opponent is allowed before each defensive action — an inverse metric, lower meaning more aggressive pressure. Together, those three numbers usually tell a tactical story in forty seconds.

But when all three are blank, the only honest thing a writer can say is: I do not know whether this team presses high or sits deep.

I have sat in the Velodrome stands on many afternoons and seen the gap between those two things. Based on my experience watching matches, there are games where the home side presses like mad for twenty minutes — PPDA can dip below 7 — then abruptly drops deep and lets the opponent hold 65% of the ball. The full-time table records an average that looks entirely ordinary. Someone who was not at the ground writes: "the home side pressed moderately." Someone who was there writes: "the home side ran out of legs after minute 25, and that is the whole story of the match."

Numbers walk me to the stadium gate; eyes take me into the dressing room.

The problem is that when data is missing, many writers refuse to drop the sentence. They substitute feel for xG, adjectives for PPDA, and deliver all of it in the same confident register. That is the moment a tactical analysis becomes a descriptive essay with player names attached.

There is a subtler failure mode: the data exists, but only half of it. A match where the provider captured the home side's events but lost the away side's. The final table shows the visitors with zero shots, zero chances. An automated model concludes the visitors defended superbly. In reality they have just lost 0-3 and the coding system crashed at half-time.

The only way to block that failure is a single question asked before any table: is this complete data, or data that survived an incident?

The financial layer: where a blank cell can become a sanction

UEFA's FFP arrived in 2026 to tie club spending to revenue. The Premier League's PSR is the stricter descendant, and it has real precedent: Everton were docked 10 points in November 2026, reduced to 6 on appeal in February 2026, then docked a further 2 points in April 2026 for a separate breach period. Nottingham Forest were docked 4 points in March 2026. Those numbers are real, dated, documented.

And precisely because they are real, the financial layer is where fabrication is most dangerous.

A serious FFP analysis needs at minimum four inputs: broadcast revenue, commercial revenue, wage bill, net debt. Missing one of four, any conclusion about "breach or no breach" is a guess dressed in terminology. Missing all four, the only honest conclusion is: sustainability cannot be assessed.

I have seen transfer reports written like this: "Club X cannot spend €60m because of PSR." No balance sheet, no amortisation, no instalment structure. Just a preconception about the club's ambition, phrased in accounting language.

Do not look at the number on the price tag; look at the team after the player leaves. A transfer does not end at a figure. It ends in structure: how many years of instalments, performance add-ons, sell-on clauses, and where the new wage sits inside the dressing-room hierarchy. Those four factors decide whether a club slides into an internal crisis. None of them lives in an empty file.

The solidarity mechanism is a fine example of scattered data. FIFA allocates 5% of a transfer fee to clubs that trained a player within a defined age band. A small sum, but it depends on registration records held by individual clubs in individual countries across ten years. One small academy failing to keep paperwork, and the money vanishes from the system with nobody able to trace why.

Results and the opinion cycle: when the denominator is zero

Football is a sport of small sample sizes presented as large ones. Four recent matches is a sample size. Three home games is a sample size. A four-match winning run may be form, or it may be a fixture list containing four bottom-half opponents.

When the file is empty, the question "is this team rising or falling" has no answer. But deadline pressure does. The piece still has to be on the page by 7 a.m. So the writer chooses to describe a feeling instead of measuring a phenomenon.

I have made that mistake. In 2026, when I wrote that Monaco would collapse after selling Mbappé, I leaned on a very specific set of numbers: 107 Ligue 1 goals in 2026-17, with Mbappé scoring 15 of them and creating a string of decisive moments. But I ignored something the data never records: the runs that opened space in the 70th minute, when opponents had lost their shape. Three months later, when the €180m deal closed, the piece was shared more than 50,000 times and I was declared right. But I knew I had been right for an insufficient reason. From then on I changed the rule: go to the ground first, write second.

Matches are decided where the crowd is not looking. The bench. The way a striker reties his laces before coming on. The referee's breathing after a controversial call. None of that lives in any model, and it is often the only thing left when the model has gone blank.

That is why I do not panic when I receive an empty file. I only panic when someone intends to keep writing from it.

The league landscape: assigning a tier without a table

To say a team belongs among title contenders, European places, mid-table or the relegation fight, you need four inputs: which league, which division, whether there is continental football, and the points gap. Four inputs. An empty file contains none of them.

Football analysis has a damaging habit: assigning tiers by reputation. A club that once won the title gets called a giant for years after selling its squad. A newly promoted club gets called a phenomenon until it proves otherwise. Those labels are memory, not data.

The real way to locate a team in the food chain is not the league table. It is the direction of transfer money. A club that sells its key players upward two seasons running is a mid-tier club, whatever position it currently occupies. A club buying players from its own domestic rivals is a club climbing tiers. The direction players travel predicts the table roughly eighteen months ahead.

Rules and governance: the wrong framework produces a confident, meaningless answer

A case at FIFA level, UEFA level, national association level or competition-organiser level has its own file, its own limitation periods, its own sanction scale. Assessing a case against the wrong framework produces a conclusion that sounds highly professional and means nothing.

The clearest example is how sanctions get compared. A points deduction in the Premier League cannot serve as precedent for a club licensing breach in another league, because the financial mechanics, the accounting period and the revenue definition all differ. Online, every sanction gets placed side by side in the same screenshot.

In an empty file, this layer returns exactly one status: the applicable rule system cannot be identified. That is the correct status. It is also why I never accept a piece about a sanction that lacks a date and the name of the deciding body.

The dressing room: the thing data never touches

This is where I believe I hold an edge, and also where I am most likely to fool myself.

A file can tell you a 31-year-old has played 2,870 minutes this season. It cannot tell you whether he still believes in the dressing room. It can tell you a coach has used four different shapes in six matches. It cannot tell you whether he still controls the team.

Those things are read with the eye: who speaks last in an argument on the training pitch; who sits next to whom at meals; who is first out of the tunnel when the team has lost three in a row.

Two concepts I use often, neither of which appears in any model. The "new manager bounce" — the short results upturn after a change — is usually explained psychologically, but in my observation it tracks closely with a squad escaping an overloaded training regime. The "FIFA virus" — players returning to their clubs fatigued or injured after international duty — is something anyone who has stood at a training ground gate at 8 a.m. on the Wednesday after a break recognises instantly, while the database still lists the player as "available".

A shocking opinion is only worth something when it stands on a detail everyone else skipped. And the most valuable detail in modern football is often a negative one: a player not running, a coach not looking, an assistant not speaking.

The risk matrix: "unassessed" is not "low risk"

This is the most important line in that whole report, and the most misread.

Every risk matrix has six channels: sporting, financial, personnel, rules, public opinion, systemic. An empty file cannot score any of them. But the tabular presentation creates an illusion: a table with six rows looks like a table that has been assessed.

For me, the biggest risk in an empty file is not a football risk. It is decision risk inside an information vacuum: the chance that a decision-maker treats "no findings" as "no problems".

The fix is simple and almost never applied: label that file explicitly as a null result, keep it out of the same folder as genuine reports, and block downstream decisions until fresh data arrives. A null result labelled correctly is an asset. A null result blended into real data is a landmine.

Media narrative and transfer-rumour credibility

January is when this layer runs at full capacity. It is also when source classification matters most.

Three basic tiers. Tier one: an official club or league statement, or a journalist with a multi-year track record of confirmed calls on that specific club. Tier two: mainstream sports media with editing, usually relaying tier one. Tier three: aggregators, social accounts, and anything built to generate clicks.

The problem is that tier three moves six to forty-eight hours faster than tier one. In that window, a rumour can travel the world, spawn hundreds of articles, and become "consensus" before anyone checks where it started.

Agent motive is the second variable. A transfer rumour is not always released to make a deal happen. It can create negotiating pressure, inflate a price, or simply prove to a client that the agent is working. A rumour about a player in his final contract year usually has a different purpose from a rumour about a player who signed a five-year deal last summer.

In an empty file, the source tier cannot be graded. That is when I choose to say it plainly: I do not know how reliable this is, and therefore I will not write about it.

Industry transmission: an event that never happened cannot transmit

Transmission analysis needs a trigger: a transfer, a sanction, a broadcast deal, a format change. No trigger, no chain.

But this layer deserves attention because it is where Vietnamese football and European football meet most often — and where I am most easily lured into the "outsider" trap.

I was born in Vietnam and work in France. I can view a deal from two sides: the buying club and the football economy that sells players. But if I use that angle only to colour a piece, I am selling a brand rather than an analysis. I use it only when it genuinely reopens the question.

Three structures matter here, and all three are lessons in missing data. First, third-party ownership — TPO — banned by FIFA in 2026, yet its traces survive in complex loan structures. Second, multi-club ownership networks, where a player can move between three countries with nobody controlling the whole path. Third, the academy system, where small institutes develop players for big clubs and are routinely under-recorded.

All three are fragmented-data zones, and therefore where the most value leaks away.

Where I might be wrong

I have spent most of this piece arguing that an empty file is an incident. There is another possibility, and I have to state it.

Possibility one: the pipeline did not fail. The source article genuinely contained no information. In an era when every club issues dozens of statements a week and every outlet publishes hundreds of pieces a day, an article containing no event is entirely plausible. In that case the empty file is a correct result, and the only action required is not to publish.

Possibility two, and this is where I distrust myself most: perhaps I am using "data quality" as a shield to avoid making hard judgements. A hot-take writer can hide behind a number. A hot-take writer can equally hide behind the absence of a number. Both are hiding.

Possibility three: my faith in the "dressing-room eye" may have hardened into a small religion. I have dismissed uncomfortable numbers more than once because they did not match what I saw from the stand. Sometimes I was right. Sometimes I was defending an impression formed in the twelfth minute of a match I watched in less than optimal condition.

My correction: use data as a reverse check. When a number contradicts what my eyes saw, the first job is to name the anomaly, not wave it away. The model may be wrong. I may be wrong. Both possibilities have to be written down.

And if I am entirely wrong — if the real lesson of that night was a trivial technical fault undeserving of a piece this long — I will publish the correction as soon as new data arrives. For someone who works by judgement, a sharp reversal is a weapon, not a defeat.

What I am betting on

Three checkable predictions.

One: within the next two transfer windows, at least one transfer decision at a top European league will be postponed or reversed just before signing, citing incomplete medical records or physical-load data. The tell: a deal announced as "almost done" that disappears without explanation.

Two: at least one tactical analysis in a major European newspaper will be written on unverified match data, and someone will catch it because the numbers contradict the footage. The tell: a small correction, published quietly at the weekend.

Three: the number of clubs publicly saying "we do not have enough data to conclude" will rise, slowly but genuinely, because the financial cases have taught them that a wrong conclusion costs more than an empty admission.

People laughed at me for three months, but laughter never scores. And an empty data file never scores either — literally or figuratively. The only thing it can do is make us believe the match ended safely.

Next time you open a dashboard and every cell is blank, ask yourself: who collected this data, what did they actually capture, and who decided that the gap was not worth alarming anyone about?

When the world pauses, I return to where the audience is still listening. And there I still carry a notebook, a pen, and one unchanging rule: better to write "I do not know" than to publish a sentence that is syntactically correct and factually false.