Nine Analytical Dimensions, Forty-One Tables, Not a Single Player's Name

3:47 in the morning, Chicago. The machine in my study received a 41-page file from the analytics system of a newsroom I contribute to. The file name: Stage-2 Deep Professional Analysis Report. I read it once on screen, once more on the balcony, then made coffee and read it a third time to be sure I was awake. Inside: nine dimensions of professional analysis, forty-one tables, a six-row risk matrix, a five-star information-value scorecard, and even a glossary of industry terms. And not a single player's name. Not a team. Not a score. Not one minute of game action. Every data field read N/A or insufficient information to assess. A flawless report on a game that does not exist. In forty-four years of reading sports reports, this was the most honest document I have ever held, and the most useless.

It arrived during the trade deadline week of the 2026-26 season, when every outlet must push out fifteen to twenty pieces a day, each needing its own angle, each angle needing a table of numbers. Nobody has the staff to re-watch game tape. Nobody has the hours to call an assistant coach in Ohio at midnight. So they built a pipeline: one layer reads, one layer writes. A pipeline does not tire, does not ask for a raise, does not miss deadlines. And that week, it delivered on time a report about nothing at all.

The pipeline runs in two stages. Stage one reads the source article and extracts information points — atomic, verifiable factual claims: who, did what, when, how much. Stage two takes that payload and applies nine dimensions: tactics, player data, salary and the cap, league landscape, rules and governance, the locker room, risk, media narrative, industry ripple effects. That framework is not bad. It is what I wished I had while sitting in a press room in Kazan in 2026, hungry and sleepless, trying to remember how often Germany had switched play to the flanks. The framework was right. The data was empty.

Stage one died in silence. It returned an empty list, raised no error, blocked nothing. The report itself listed three possibilities: the extraction step failed and passed an empty payload downstream; the source sat behind a paywall or a JavaScript-rendered page, leaving no readable text; or a schema mismatch wrote the information points into a field stage two never reads. All three are ordinary technical faults. There is no conspiracy here. Just a clogged pipe, and a system that kept running because nobody ordered it to stop.

In fairness, the first thing to say is this: the writer of stage two did not fabricate. It did not assign anyone eighteen points, did not conjure a rising star out of nothing, did not invent a seventy-million-dollar transfer. It wrote N/A in every cell and listed precisely what it needed to work: player names, sample window, TS%, payroll relative to the apron, the coach's name, the reporter's name. In a week when hundreds of sports pages ran hundreds of unsourced trade scenarios, refusing to invent is a moral act. But it saved no one.

Because the report was still published. It was formatted, paginated, tagged, archived. Tomorrow it sits in a folder, next week it is cited in a meeting, next month someone opens it and sees nine dimensions of professional analysis, forty-one tables, a fully scored risk matrix. Nobody reads the N/A sections. What is frightening is not a model that is wrong, but a model that is perfectly formatted. The report itself prescribed the cure: a hard gate — if the information-point count is zero, halt the job and raise an alarm instead of printing paper.

People saw Manchester City win; I saw a man dozing on the other side of the pitch. This time the sleeping giant wore no kit. It was the editorial layer — the layer that should have noticed an analysis with no human names in it cannot be an analysis. For twenty years, this industry's editorial layer was taught to check sources, to call and verify, to demand two independent sources for a transfer story. Then automation arrived, and everyone forgot that a pipeline is also a source, and it needs checking like a person.

Based on my experience covering games — twenty-two consecutive NBA Finals spent inside the arena, eight years writing a column for VnExpress — there is one thing no model teaches: a game always holds a detail nobody logs. A guard glancing at the bench before receiving the ball. A centre breathing hard in the sixth minute of the fourth quarter. A lineup moving half a beat slower when the ball is far away. None of that lives in an information point. It lives in the eyes of someone sitting close enough.

In 2026, at fifty-one, I sat in a brand-new podcast studio in Chicago watching Liverpool 4-3 Manchester City at Anfield. While the room praised Kevin De Bruyne, I shouted into the live mic that Mohamed Salah would break the Premier League scoring record. At that point Salah had eleven goals in eighteen matches. The forums mocked me. I staked my reputation on two things: expected goals and his carrying speed. By season's end Salah had scored 32 in 38 games, breaking the record Luis Suárez set in 2026. The studio received two hundred collaboration offers.

Here is the part I rarely tell. The numbers did not tell me Salah would break the record. They told me he was shooting from better positions than his goal tally implied. The rest came from having watched him run. I had seen how he received on the right touchline, checked for half a beat so the full-back flew past, then cut inside. The model counted the shot. It could not count the check. A metric carries weight only when the writer has seen the rhythm behind it. Strip out the rhythm and metrics become decoration.

In the summer of 2026 I flew straight to Kazan, Russia, to cover Germany's 0-2 group-stage defeat to South Korea on 27 June. Kim Young-gwon opened the scoring in stoppage time after a VAR check, and Son Heung-min sealed it in the 96th minute with the German goalkeeper stranded upfield. The world was stunned that the holders were out. I did not write a lament. I went into a local beer hall, bought Korean reporters a round, and wrote a long piece on systemic failure: Germany played roughly 12 percent fewer vertical wing passes than in 2026. Three thousand furious comments. The piece was still shared across Europe.

Germany had probably lost before the first ball was kicked — people simply were not sharp-eyed enough to see it. That 12 percent was only evidence for something I had seen with my own eyes: their full-backs received, looked up, and passed backward. Nobody dared break the line. Their death lay in hesitation, and hesitation has no column in a box score.

That is why I do not trust the way analytics departments are walking into locker rooms. They bring lineup-performance tables, plus-minus, impact indices, and they present them to men who have just lost four straight and are arguing about who holds the ball. A table can be arithmetically right and humanly wrong. Its conclusion detaches from the actual rhythm of a group that is coming apart. That nine-dimension report is the purest proof: a system delivered a complete professional verdict without ever touching a minute of play.

The same logic runs through the transfer market. Clubs now buy profiles rather than people: a striker with a pretty long-range shooting number, a defender with a high long-ball completion rate, a free agent who costs no transfer fee so nobody scrutinises the wage and signing fee. Every fallen giant is a slap for those who collect names instead of collecting people. A sixty-million signing is not guaranteed to change more than a shy academy kid who knows how to watch.

I call what these pipelines are building ghost football, and in my sport, ghost basketball: a game with no stands, no smell of sweat, no rubber squeaking on hardwood, only data fields locking into place. In that game every player is a vector, every coach a function, every failure explainable. And in that game, a report with no names in it still counts as complete, because there is no one left to ask where the missing name went.

I could be wrong, and this is where I doubt myself. There is another reading: that empty report was the most dignified document of deadline week. Among hundreds of pieces asserting unconfirmed trades with total confidence, a file willing to write N/A forty-one times is the only voice that did not deceive the reader. If so, the fault is not in the model. The fault is that we built a machine to speak, when the only thing worth saying that week was that we knew nothing.

And here is where I have to examine myself. My trade is to make a call before the result arrives. A brand like that rots easily. Any day I write about a team whose games I have not watched in three weeks, I am publishing that same nine-dimension report — except mine carries no N/A labels, so nobody catches it. I keep a ledger of every prediction right and wrong, reconciled quarterly. It does not make me wrong less often. It only makes it impossible for me to pretend I was never wrong.

The real cause, most likely, is deeply unglamorous: a 401 error, a login wall, a JavaScript-rendered page. The pipe clogged on something trivial, and nobody in an entire newsroom opened the door to check. In basketball we call that a team losing because nobody called timeout. In this trade, we do not have a name for it yet.

So here is a verifiable prediction: within twelve months, at least one sports outlet will publish an analysis that nobody on that desk watched a single minute of tape for, and it will carry not one N/A field to warn the reader. When that happens, do not blame the model. Find the person who pressed send. And if one day you read a piece of mine with no detail from the floor in it, send me back those forty-one blank pages.