Empty Data: Lessons From an Analysis That Only Looks Complete

At two in the morning, the data file from the match that had just finished opened completely blank. No xG. No passes into the final third. Not even a starting line-up. In the top corner, exactly one line survived: football.

I sat looking at that frame far longer than an empty file deserves.

In May 2026, round 15 of the V-League, I wrote my first 900-word blog post about SHB Da Nang losing 0-3 to Hanoi right at home. All three goals came down the left side. Hanoi completed 134 more passes yet produced only 4 shots on target. Underneath, one account left a single line: “What does a girl know about football to talk like that.” I did not write a word in reply. I added an average-position map for every player to the end of the post and went to sleep.

Tonight there was nobody to answer, because there was nothing to answer. Only a gap, and a template waiting to be filled. The temptation lies in how professional that template looks.

Football analytics in Vietnam has lived off rented data for years. A blogger in Da Nang, a fanpage team in Saigon, a reporter in Hanoi all pull numbers from international providers such as Wyscout, WhoScored or SofaScore, or receive internal files shared by clubs after a match. The pipeline has four stages: fetch the document, parse the numbers, build the story, publish.

The odd part of this chain: when the first stage breaks, the other three still run smoothly.

A document may fail to arrive because the link is dead, because the page sits behind a paywall, because the content is rendered in JavaScript and the reader only receives a blank page. What reaches the writer then is a shell. The frame is complete: a title field, a source field, a summary field, a stance field, a field for the list of information points. Inside every field is a zero. No error flag is raised. No red line states that retrieval failed.

Data people call that state an empty payload: a perfectly shaped box with nothing inside.

What stands out is that the system never crashed. It ran exactly as designed. The design simply allows a document that does not exist to pass through the gate with nobody checking tickets.

A fully presented analysis is not the same thing as an analysis that contains information. Completeness of format and completeness of content are two different things, and merging them is the most dangerous mistake a writer or a reader can make.

I tried putting myself in the reader's seat. A nine-part document, every part with a table, a conclusion section, an evidence section, a risk matrix, and a glossary at the end. Skim it and it looks exactly like a finished report. The evidence section states plainly: evidence base empty. The conclusion section states plainly: no conclusion can be drawn. But the eye skips those lines, because the structure has already persuaded before the content can speak.

I call that the false authority of formatting.

For someone working in football, this trap has a nastier version. The more experienced the writer, the easier the fall. My head always carries a wardrobe of story moulds: the new-manager bounce, the player entering the final year of his contract, the club sliding because of a packed calendar, the promoted side collapsing after half-time for lack of squad depth. As soon as a club name lands in my hand, those moulds assemble themselves into a finished article. The prose flows. The numbers do not exist.

This is why I do not write until I have reopened the match video, even when the data sheet is already on my screen. Data does not lie, but it is very good at hiding surprises. A data file can tell you how many pressing actions Team A made per minute. It cannot tell you that Team A's midfield stood nine metres apart in the 67th minute, and that the gap itself opened the decisive pass.

Based on my experience of watching matches, an empty payload always needs three steps, and every step costs time.

Check whether the original document arrived. Log the status code of the fetch request, log the length of the document body before parsing. If the body is empty or the status code reports an error, the culprit sits in retrieval. If the body is full and the parsed output is still empty, the culprit sits in processing. Those two faults look identical from the outside, and the fixes run in opposite directions.

Pin down the identity of the source. Outlet, byline, publication date, publication time. This is the most perishable data and the one that cannot be rebuilt. A headline can often be recovered from the body. A source identity cannot. For a transfer story, the credibility of the outlet is the decisive variable, well ahead of the wording itself. A line from a reporter with a direct relationship to the agent carries different weight from a line rewritten off social media, even when both use the same sentence.

If it cannot be verified, stop. Do not write filler.

I learned that last step from a mistake. In December 2026, Morocco beat Portugal 1-0 to reach the World Cup semi-finals. In my analysis of Morocco's 4-1-4-1 block, I wrote that they had won 14 of 14 aerial duels. The correct figure was 13 of 14. An account specialising in auditing numbers came after me. I did not argue, I deleted the post, reopened the video, and republished a correction within two hours, opening with a plain statement of where the error sat. The corrected post drew twice the readership of the original.

Since then, every number in my work is cross-checked against at least two independent sources, and the date the data was captured is printed at the bottom of the post. The method makes me slower than outlets that simply copy numbers from one place. It also means I never have to publish twice.

On 19 June 2026, Japan beat Colombia 2-1 at the World Cup. Carlos Sanchez was sent off in the third minute after a handball in the box, Shinji Kagawa opened the scoring from the penalty spot, and Yuya Osako headed the winner. I drew six diagrams by hand for that piece, measuring the average distance between the midfield and forward lines at 22 metres for Japan and 35 metres for Colombia. A former Vietnam international shared the post. Two days later it had 4,200 reads.

Those six diagrams were drawn on paper, not in an app, because I need to see the gap through my own hand before I trust it.

There is one skill that appears in no data course, and I consider it more important than reading expected goals fluently: recognising when you are holding an empty template before you write a single word.

When the stadium is empty, the sound of the ball becomes data. I listen and I write it down. When there is no sound of a ball to hear, the correct move is to close the notebook, not to imagine the noise.

In my early years I thought the hardest skill in this trade was reading a match. I have changed my mind. The hardest skill is refusing to write about a match you only hold a frame for.

The majority believe the enemy of football analytics is wrong data. My experience says otherwise. Wrong data can be argued over, corrected, reissued. An analysis that looks complete but is hollow leaves no trace to argue with. It drifts through the system, gets shared, gets cited, and becomes the foundation for the next post. Three months later nobody remembers where the original number came from, including the person who wrote it.

The experienced writer is the most dangerous person in that situation. I carry enough material to construct a very convincing piece about a match I have not watched a single minute of. Audiences carry enough inertia to believe it, because the layout is standard and the prose flows.

Part of the responsibility belongs to the reading habit. We are taught to trust complete form very early: clean tables, a full contents list, correct terminology. In football, that trust is reinforced by speed. Everyone wants the analysis within an hour of the final whistle.

The majority wait for the star; I look at the space behind him. There is another space worth looking at first: the space inside the data file in my own hands.

Next time a data sheet is put in front of you alongside a confident claim about a match, try two questions: when was this document fetched, and from where. If the person offering the number cannot answer the second, the rest of the table does not need reading.

Tactics are not magic. It is just that some people look a little longer. And the person who looks longest is the one who knows when to say he has nothing to say yet.

A hand-drawn diagram from the 2026 World Cup still reads tonight's match. I do not draw in an app, I draw by hand as I did in 2026, but only after checking that the match genuinely exists on my hard drive.