The Breath of an Empty Feed

The Breath of an Empty Feed

11 p.m. in Saigon. I open the document the desk sent over. Thirty-two lines. Every line reads "N/A." No title. No source. No club. No player. The list of information points is empty. Only one field is filled in: the domain label — "football."

Beyond that label, there is nothing.

Thirty-nine years in this trade, and I have received my share of bad drafts, scribbled notes from young reporters, wire copy rushed out before kickoff. But a document labeled "football" with not a single ball inside it — that is a first.

The desk built a nine-dimension analytical template: tactics, club finance, results cycle, league landscape, rules and governance, the dressing room, risk profile, media narrative, and the industry's transmission chain. It is a beautiful template — designed to read a match, a transfer, a boardroom decision, and return what the scoreboard can never measure.

It runs on a data pipeline. Tonight, the pipeline delivered an empty packet.

The first thing to state clearly: an empty packet is not a thin packet. A thin document can still be mined — it has a name, a date, a figure, a quote. From those fragments a decent writer can build something. An empty packet has nothing. No subject to attach risk to, no event to trace, no timestamp to cross-check. Every conclusion drawn from it, however plausible it sounds, is invented.

When a nine-dimension template meets an empty packet, it has two choices. One: say plainly that there is not enough information to analyse. Two: fill the blank with whatever sounds plausible.

The second option is far more dangerous than it looks.

Picture a transfer-window analysis. It has a fee, a contract length, a wage, a sell-on clause. Those four figures are enough to build a story about value, risk, and a player's age curve. Readers trust those numbers. They do not check the source, because the analysis looks immaculate. But if those four figures were generated from an empty packet — invented to fill the blank — then the reader is trusting a building with no foundation.

I once saw something like this on a smaller scale. In 2026, sitting through nine home matches for one club, I recorded three goals conceded in the final ten minutes of each game. Nobody noticed. No statistical table captured that tenth-minute moment. But it was real. I counted it with my own eyes.

By contrast, some numbers look highly professional but do not exist. A single bad data line can run through hundreds of reports in hours. A mispronounced name can become "fact" in a single evening.

The greatest danger of an empty packet is not that it is empty. The danger is that the template receiving it does not permit blank space.

A nine-dimension template is designed to return answers. It has nine layers, each with tables, cells, conclusion lines. When the input layer dies, the template still wants to be filled. This is the instinct of any system built to answer. It is exactly the instinct of a commentator behind a microphone: when the name slips, he calls out another name — any name — because silence is unbearable.

I know that instinct better than most. In 2026, on live television, I called a Russian player by the wrong name three times in one half. I did not stay silent. I filled the gap with a wrong name. And that wrong name went into millions of ears.

That is why I spent the following forty-seven days reviewing every match tape of the Russian national team, from qualifying. Forty-seven days. Not for a name. For a habit.

Today, looking at the empty packet on my screen, I realise that habit is being multiplied many times over. If a machine can generate nine layers of analysis from zero, then my old habit of filling blanks — one man's isolated act — has become a systematised process.

The machine does not lie out of malice. It lies because it is required to have an answer.

Here, most readers — and most professionals — will get it backwards. They will think the machine is at fault. They will demand a new algorithm, a new tool. But the machine only reflects what it is asked to do. The fault lies in how we define "analysis" as a product that must ship nine full layers, rather than a process that can return an empty answer.

A template with no "I don't know" mode will inevitably fabricate. Not may — will. Because the blank must be filled, and if the data does not arrive, prose arrives in its place. The professional-sounding phrases, the balanced sentence structures, the decisive conclusion lines — all of it can be written without a single scrap of fact.

We who work the trade call that "fabrication." In the data pipeline, it is called "task completion."

Readers cannot tell the two apart. They see only an analysis that looks good. They do not see that beneath it lies nothing.

If I told you that yesterday there was a match, that Team A lost to Team B by two goals, that Team A's manager is under pressure — you would believe me, wouldn't you? You would believe me because I said it with such certainty. You would not ask whether I was at the ground. You would not ask whether I watched the match with my own eyes.

But if I was never at the ground, what right do I have to speak about the atmosphere in Team A's dressing room?

That question is the line between a beat keeper and a text-generating machine.

Within those nine layers, every layer has a place to say it does not know. An "insufficient information" cell in the risk section. A "not assessable" line in the media section. A "source unverified" column in the transfer section. Nine such places. And all of them were used correctly in tonight's document.

That is the only bright spot of tonight.

But I wonder: how many times is that place left unused? How many transfer analyses do I skim each morning, with fee figures nobody sources, with predictions about managerial pressure nobody says where they heard from? What share of those are nine layers built from another empty packet — an empty packet nobody flagged?

The difference between a newsroom and a data pipeline: a newsroom has people. People have memory. Memory sometimes has the courage to say "I don't know." A pipeline has no memory. It has only processing capacity, and a template that needs filling.

I am not writing this to accuse anyone. I am writing to remind myself, and the trade, that between two whistles there is a world the scoreboard cannot measure — but that world only has value if it is real.

An empty feed has no breath. It seeds nothing in the reader's mind — neither truth nor falsehood. A decent writer, in the end, needs only one very old requirement: if you do not know, say you do not know. And if you do know, check once more before writing a person's name. Because a name set right can hold a reader's trust, but a name set wrong takes forty-seven days to fix.

And a name invented — that can take a lifetime.