The Night the Data Sheet Went Blank: When Football Asked for Human Eyes Again

At the 71st minute, the screen in front of me went blank. The possession column showed a dash. The completed-passes column showed a dash. Distance covered, sprint count, expected goals, duel success rate — every one of them a dash, as if the match had never been played. In my ear, the Nagoya stand was still roaring. On the pitch, two teams were still playing. Only the spreadsheet was dead.

That night the broadcaster's data feed failed. I had to call a professional football match with exactly two tools: my eyes and my memory. No heat maps, no sprint data, not even a possession figure to give the audience at the fifteenth minute. There was a silence long enough that I could hear my own breathing inside the commentary box.

I remember saying one thing that night, and I still find it truer than any table I have ever read: the pitch does not need a poet, but the poet needs the pitch. What I lacked that night was not data. What I lacked was a different way of seeing.

I began writing about football in 2026, after graduating from the Journalism Academy, working both for a domestic football paper and as a correspondent in Madrid for an international sports daily. Back then a match reporter's kit was a notebook, a pencil and legs that had to walk. Nobody handed you statistics. You counted yourself. You remembered yourself. You got things wrong yourself, and you owned the error.

Nearly fifty years later I sit in a box in Japan with three screens in front of me, and within two seconds I can know that the away team's left-back covered 9.7 kilometres, sprinted twenty-three times, touched the ball forty-one times and passed at eighty-four percent accuracy. I can know everything. And I realised something: knowing everything does not mean understanding more.

My trade has changed. People no longer buy a match for what you can tell them, but for what you can measure. Broadcasters build graphics that update in real time. Clubs run whole data departments with dozens of staff. Scouts no longer travel to watch players with their own eyes; they download profiles and filter. An eighteen-year-old in a provincial town can be undervalued simply because his club has no tracking cameras.

I am not against data. I am against letting data think on my behalf.

In 2026, in the knockout round on Russian soil, I sat in the box for Japan against Belgium. It was 2 July 2026, at the stadium in Rostov-on-Don. In the 48th and 52nd minutes, Haraguchi and Inui scored in quick succession. The box erupted. Then in the 69th minute Vertonghen pulled one back with a clever header. In the 74th, Fellaini equalised. And in the 90th plus fourth, Chadli ran into space after a counter-attack that in my head I could only count as fourteen seconds. The score stopped at 2-3.

I keep those fourteen seconds the way I keep a shattered gemstone.

Across decades behind a microphone, I have believed that every passage of play has a heart beating inside it. That belief never came from a spreadsheet. It came from nights when I had to count the breathing of a stand to know where a match was heading. And precisely because of that, I want to speak plainly about the three metrics this industry worships like idols.

The first is possession. Of all the numbers I have ever read, this is the most deceptive. It does not measure control. It measures time on the ball, and time on the ball is a quantity any midfield capable of sideways passing can inflate. A team with sixty percent possession is not stronger than a team with forty percent; it simply talks to the ball more.

Picture a side that completes seven hundred passes in a match, four hundred of them in its own half, between two centre-backs and a holding midfielder. Its opponent completes only three hundred, but puts the ball into the box twelve times and takes nine shots. The graphic tells the audience the first team had seventeen percent more of the ball. The graphic does not tell the audience that the second team played its football where matches are decided.

To escape that trap, possession has to be split by zone. What share of the time is the ball in the opponent's final third? How many passes genuinely move the ball towards goal rather than sideways? That is the tactical story. Total possession is only the headline, and headlines are usually more seductive than the substance.

The second is distance covered. This is the metric I see abused most often in bulletins. A midfielder who runs 12.4 kilometres is praised as tireless. But run where, run for what, and what does running after you have already been played past actually mean?

Distance covered does not reward intelligence; it rewards exhaustion. A player constantly dragged out of position will run enormous distances, because he is forever chasing what he has already lost. A centre-back who reads the game well may cover only 9.8 kilometres and still control his entire defensive corridor.

Modern tracking splits distance into layers: jogging, running, high-speed running, sprinting. It sounds sophisticated. Yet even the sprint layer cannot distinguish a sprint to escape a marker from a sprint back towards your own goal after losing the ball. The same number, two opposite meanings. The reader of the spreadsheet does not see the difference. The person in the stand sees it clearly.

The third is expected goals. Here I must be fair: this metric has rescued many analyses from laziness. It forces people to look at chance quality rather than only the scoreline. A team that wins 1-0 through a lucky long-range strike should not be praised like a team that wins 1-0 after ten one-on-ones with the goalkeeper.

But its blind spot lies elsewhere. Expected goals is an average value drawn from a population of players, not a verdict on one individual in one specific situation. A team can accumulate 2.5 expected goals from twenty-five low-value shots, while another manages 1.2 from four high-value ones. The spreadsheet concludes the first team attacked better. The human eye concludes the opposite.

And there is one more thing every expected-goals model cuts away: the pass before the shot, and the run that created the space for that pass. The metric stops at the shot, because the shot is the only thing that can be counted. But football is created two beats earlier.

I still remember a derby in Japan that I hosted for a podcast with two young journalists. That match contained forty-seven duels, and it took us nearly two hours to agree that the number forty-seven said nothing at all, because it lumped a decisive tackle inside the box together with a harmless collision in midfield. That episode reached twelve thousand listens in its first week. The lesson was clear: audiences do not need more numbers, they need more meaning.

There is another metric worth mentioning, because it is the perfect example of the data trap. It is called the number of passes a team allows the opponent before each defensive action. A low figure is read as a sign of ferocious pressing. But a bad pressing team also has a low figure, because its players run everywhere, dive into hopeless challenges and get played through with two passes. The same number, one side a system, the other chaos. To tell them apart you must ask a further question: where on the pitch does that team win the ball back, and what does it do next?

At this point I must return to the fourteen seconds in Rostov-on-Don, because it is the finest illustration of what I am trying to say.

In the 90th plus third minute, Japan won a corner. The score was 2-2. Extra time was waiting. Japan pushed players forward. They wanted to finish the match that night rather than drag out another thirty minutes on heavy legs.

The corner was cleared. The Belgian goalkeeper collected the ball and released it immediately down the right. De Bruyne carried it through the gap in midfield where a Japanese midfielder should have been standing. The ball went wide right to Meunier. Meunier drove a low cross across the face of goal. Chadli arrived at the far post. The score closed at 2-3.

During those fourteen seconds, not one metric on my screen changed quickly enough for me to tell the audience what was happening. All I had were my eyes, and a cold feeling at the back of my neck as I realised how wide the space in midfield had become.

The post-match spreadsheet would tell me how many more kilometres Japan ran than Belgium. It would tell me the possession share. It would tell me the passing accuracy. And not one cell in that spreadsheet could answer the only question worth asking: why, in the 90th plus third minute, did Japan leave only two players behind for a corner?

The most important metric of that night was a blank cell — the number of players left behind for a set piece, in the final minute of stoppage time, in a knockout match. No broadcaster put that line on screen. No bulletin read it out. And so millions of viewers wept without knowing precisely what they were weeping for.

I am not saying that decision was wrong. A team willing to push players forward in the last minute of stoppage time in a knockout match is a team willing to live. But I am saying that the way we retell that story has been distorted. We tell it as an accident. It was not an accident. It was the consequence of a choice, and that choice is measurable — we simply chose not to measure it.

Nearly seven decades of watching football, and I am still learning how the ball rolls every single day.

In 2026, when stadiums stood empty through the pandemic, I opened a slot on local radio in Nagoya to read supporters' letters on air. Some weeks I received three hundred letters. I read fourteen of them across three unbroken hours, without a break, because each letter was a real person sitting in a real house, waiting to hear their own name.

Around the same period, Nagoya Grampus lost its main sponsor, a contract worth 2.5 million yen. I did not appeal to anyone in particular. I simply broadcast the club's commitment on air, so the community could hear it and decide for itself. Two weeks later, donations passed forty million yen.

The spreadsheet will record: forty million yen, sixteen times the value of the contract that was lost. But the spreadsheet cannot record the thing I remember most: the voice of an elderly woman who called the station at eleven at night, saying she did not have much money, but she wanted to read out her husband's name, the man who first took her to a stadium in 2026.

A letter addressed to an empty stand, and football reads every line.

That is why I have kept an archive of tapes and a professional diary for decades. Not as evidence for anyone. But to remind myself that every match I ever commentated was once the present, and every one of those presents contained details no spreadsheet ever bothered to keep.

Nagoya still echoes back to me every time I touch the old microphone.

So where does the greatest blind spot of collective memory lie?

It lies in the fact that we archive what happened and forget what did not happen. Football is decided by both. The pass not played. The run not made. The player left behind. The corner not taken the way it should have been taken.

Modern data is excellent at recording what exists. It is almost entirely blind to what is missing. And here is the paradox: the more precise it becomes, the more readily we trust it, and the more we trust it, the less we see the blank cells.

I once visited the data department of a professional club. A large screen covered the wall, data streaming in real time, each player a moving point of light on a map of the pitch. They told me the system could predict a passage of play before it happened. I asked in return: can it tell me which player is afraid?

Nobody answered.

That is the limit. Not a limit of technology, but a limit of what we choose to measure. We can measure heart rate, but not fear. We can measure distance, but not hesitation. We can measure passing accuracy, but not the loneliness of a young player who has just moved to a foreign country, sitting in the dressing room and understanding not a single sentence of the coach's tactical talk.

Football is a game made of things that cannot be measured, then recorded by things that can. That mismatch is exactly where a commentator has to stand. Stand on the side of the spreadsheet and you will be right and meaningless. Stand on the side of pure emotion and you will be eloquent and wrong. The correct position is in between, and that position is very narrow.

Back to the night of the blank data sheet in Nagoya. I called that match with my eyes for the final seventy minutes. I talked about how the away team's right-back dropped three metres deeper after the first goal conceded. I talked about how the home team's central midfielder began receiving the ball with his back to goal more often, a sign his confidence had gone. I talked about how the stand fell silent for about seven minutes, then lifted itself again without anyone prompting it.

After the match, a young colleague asked whether I had found it frustrating without data. I said no. I said that for the first time in years, I had been allowed to do my actual job.

Based on my experience of watching matches across nearly fifty years, I can state something I know will irritate many people in this industry: most post-match statistical tables are produced to protect the writer, not to explain the match. The writer includes numbers to prove he did the work. The reader consumes numbers to believe he understood. Both sides are satisfied, and the match itself remains untold.

What I want to leave to those who come after me is not a set of metrics. It is a habit: leave one column blank in every report you write. That column is for what you saw but could not count, what you felt but could not prove, and what you know matters most but no system will ever record.

Whenever I open my professional diary and read the lines I scribbled hastily inside a commentary box, I notice that the most valuable lines are always the ones containing no number at all. They contain only a person's name, the name of a stand, and a question I asked myself.

Football will keep producing new metrics. There will be better prediction models, more detailed tracking systems, fuller data sheets. I am not fighting that. I only hope that while chasing the filled cells, nobody forgets that what decides a match sometimes lives in the blank one.