When the Source Returns Nothing: The Craft Discipline of Data Sports Writing

2:47 a.m., Brisbane time.

I open the file on my second monitor. Nine sections, nine tables, and almost every cell in those nine tables carries the same phrase: insufficient information. Tournament name blank. Team name blank. Performance record blank. Revenue structure blank. Overall risk rating: cannot be assessed. I read it twice, then a third time more slowly — an old habit, because a clean-looking table is usually a table that has been misread.

Outside, Coronation Drive is empty. The fridge hums once and goes quiet. I sit with that file for nearly forty minutes without opening a text editor.

An editor in Melbourne once told me: the hardest part of this job is writing when there is nothing to write about. I laughed. Tonight I did not.

What I have in front of me is something the Southeast Asian esports scene encounters nearly every week, though it rarely calls it by name: a source that returned nothing. The decision I have to make, before typing a single word, is what kind of nothing this is — and whether I am allowed to fill it in.

Two kinds of silence

An empty result has two different origins, and they demand completely different responses. The first: the source article was never retrieved. Dead link, login wall, a page rendered only by JavaScript, or a server blocking the crawler. The first-stage analyst had no text to work with, so it returned nothing. The second: the article was retrieved, but the extraction layer returned an empty object because of a schema mismatch, a truncated output, or a field-mapping error.

These two demand opposite actions. The first requires re-fetching the source. The second requires re-running the extraction layer with a corrected schema. Re-running blindly when the cause is the first kind will reproduce exactly the same empty result, once more, at the cost of another night.

The diagnostic sits in the fields that require no interpretation. Article title, outlet name, content type — any extraction layer captures these if it has text in hand. When all three are blank, and the content type is marked unclassified, the failure happened before classification even began. Which means it happened upstream, in the collection layer. This is inference from silence, the weakest form of evidence I know, but it still beats guessing.

Why esports loses data more easily than football

Football has a common denominator. Opta, StatsBomb, Wyscout — many providers, but the same basic definitions of a shot, a pass, a duel. When two journalists in two countries discuss the same match, they are on the same scale. Esports has no such thing. Each title has its own statistics system, some opened by the publisher, some entirely closed, some opened only to league partners. A superb metric in one title can be meaningless in another, and the reader is never warned.

I once told a colleague in Sydney that our job is translation between incompatible numerical systems. He said I was pessimistic. Three months later he sent me a piece comparing two teams' win rates across two different titles and asked whether it looked fine. I said it did not. He said I was pessimistic.

In the A-League, I get called a rebel simply because I bring a laptop. But bringing a laptop here is not an act of rebellion. It is an act of self-defence. Because when a league runs on feeling and a league table, it is easy to write a sentence like "this team is reviving" without knowing who they beat, by how much, and in how many minutes.

The information chain in esports is also longer than in football, and every link loses something. The publisher announces a patch. The tournament organiser interprets the patch into rules. The club interprets the rules into tactics. The coach interprets tactics into a press conference. The reporter interprets the press conference into news. The aggregator interprets the news into a headline. The reader receives the headline. Four interpretations, four compressions, and not one of them marked as compressed.

A Vietnamese article citing a Chinese forum citing a Korean livestream is such a familiar structure that nobody finds it strange anymore. When that chain breaks, the result is not the absence of news. The result is news without a root — still circulating, still commented on, still argued over, missing only the one part that made it news.

What a metric can say, and when it must stay quiet

When the numbers speak, the stadium must learn to be silent. But I learned the second half of that much later: the numbers must also know when to stay quiet, when they are not yet thick enough to speak.

A team that is 7-1 in a single-map format is not the same as a team that is 7-1 in a three-map format. In a single-map format, one pistol round can decide the whole match, and a sequence of eight maps has a confidence interval so wide the number is decorative. Add uneven opponent strength, add a mid-season patch, add a team that has already secured its seed and is resting its core players. Those four variables are enough to turn an 87.5 percent win rate into an answer to a question nobody asked.

In 2026, I learned this by getting my copy struck out. After round 23 of the A-League season, I found that Jamie Maclaren had scored 8 goals with an expected-goals figure of 14.2. That gap of 6.2 goals, to me, was a discovery. I wrote a direct critique, cited the data, and concluded the striker finished poorly. The editor struck out nearly all of the numbers, kept the prose, and the prose had nothing left to stand on.

I fumed in silence. Then I did something that, in hindsight, was better than the article: I sat and rewatched 19 Melbourne City matches from that season. Not to find more numbers. To find positions.

It turned out most of Maclaren's chances came from half-blocked crosses, from balls he had to chest down with two defenders closing, from passes arriving half a beat late. Expected goals added all of that into a single figure, and the figure was methodologically correct. But my framing turned a striker who had to create his own chances into a striker who wasted them. The metric was not wrong. My interpretive frame was wrong.

The first lesson everyone knows: do not throw numbers at the reader. The second is less discussed: when a metric is accurate but contextless, it can still produce a false portrait, and a false portrait outlives the article. Every number has a story; my job is not to ruin it.

The following year I retold that story differently. The piece on the danger zone opened with an image of a run — a defender turning his back for half a second, a striker changing direction — and only then introduced the metric. Same data. Different reader. Since then I have held a rule: never write a number without a person behind it.

The part of the data that cannot be measured

In 2026 I was invited to write tactical analysis on France against Argentina in the World Cup round of sixteen in Russia. I became absorbed by Kylian Mbappe, who reached a top speed of 37.6 km/h in the move that produced the decisive goal. I stayed up two nights breaking down frames, measuring every touch, reconstructing the running line.

Then something uncomfortable surfaced: all my tables together could not explain why that move was beautiful. Speed explains why the defender could not get there. It does not explain why I had to watch it seventeen times. Mbappe's feet always tell the truth, but I still need the numbers to translate — and translation always loses something in the middle.

Since then my writing has carried sentences that serve no spreadsheet. The spin of the ball. The tilt of a shoulder before acceleration. A goal is a moment, xG is a fate, and I choose to record both.

This connects directly to the empty file on my screen. If I accept that data cannot measure the most important part of a move, then I must also accept that data cannot measure the most important part of an article. The absence of numbers does not mean the absence of a story. But it does not give me licence to invent a story to fill the gap either. Those two sentences have to stand together, and I see very few writers holding both.

The summer with no new data

In 2026 the pandemic froze every competition. I was 33, and I lost freelance contracts with two broadcasters in the same week. No matches, no new metrics, nothing to analyse. My spreadsheets opened onto old data only.

One night I reopened Liverpool's 4-0 win over Barcelona and built a table myself on Andrew Robertson's running distance: 12.4 km, of which 2.1 km was sprinting. Nothing new. No discovery. Just a player who ran further than anyone on the pitch, and a man in Brisbane counting the metres again.

I wrote a long piece about missing the noise of Anfield. By morning it had been shared more than 4,000 times. Not because I found a new number. Because I dared to write about things that seemed unquantifiable, while keeping the quantified part in place as the floor.

The empty summer taught me this: with no matches, memory still shoots from distance. And in my trade, an empty result is not a holiday. It is a maintenance window. When new data does not arrive, the work is to recheck the method, recalibrate the scale, reread the archive. Tonight's empty file is exactly such a window, except I did not open it myself.

Borrowing language, and the cost of borrowing carelessly

In 2026 I wrote a book on a European Championship. Italy at the time were on a 34-match unbeaten run with an average PPDA of 9.8 — a number saying they pressed hard. I rewatched every match, and happened to watch sport climbing at the Tokyo Olympics at the same time.

I became obsessed with Janja Garnbret, with the way she would settle her body on a wall that appeared to offer no hold. That feeling was identical to the way Jorginho receives the ball under pressure from three. I began using a concept borrowed from climbing — spatial hold — to analyse central midfielders. Not as decoration. To describe more precisely something football vocabulary did not have a word for.

I still believe borrowing language across disciplines is intellectually legitimate, on one condition: a borrowed term must be defined in the paragraph that uses it, or it does not go in. Piling up unexplained terminology is the fastest way to make an analysis read like an advertisement for its author.

And here is the link to the empty file. When data is absent, language is the only thing left to fill with. That is the most dangerous moment. An article without numbers automatically borrows credibility from adjectives. Alarming. Groundbreaking. Historic. Those words need no evidence, only tone. Readers cannot verify an adjective. They can only verify a number, and the number is what is missing.

A gap is not a clean bill of health

In any esports analysis, one group of checks goes first for me: competitive integrity. Match-fixing, account boosting, equipment cheating, joint liability of coaching staff. This is the group with the heaviest consequences and the one most easily skipped when there is no data, because with no data there is nothing to check.

An empty report confirms nothing. It also denies nothing. But in operational practice, a blank cell is usually read as a tick. No financial problems found, so the club is fine. No sanctions found, so there is no violation. This is the most common and most dangerous reading error in the entire analytical chain, because it converts ignorance into a conclusion.

The governance structure of esports makes that error worse. The publisher is both rule-maker and a commercial stakeholder in the very competition it writes rules for. There is no independent arbitration mechanism equivalent to a court of sport. When a dispute arises, the party entitled to publish the finding is also the party affected by it. For a journalist, that means fewer independent verification sources, not more.

Money, and the numbers nobody wants published

The financial structure of an esports organisation has a feature European football has already lived through and is still paying for: salary-to-revenue ratios that routinely exceed sustainable levels. Add two specific items — dependence on publisher distributions, and the amortisation of franchise slots. Together they form a system where one season of competitive failure can trigger a restructuring of ownership.

When the financial section of an analysis is blank, readers do not see a gap. They see a page with no bad news. That is why I never leave this section empty in my own work without stating clearly: no data, not verifiable. Phrasing matters as much as fact.

On transfers, I hold a professional bias I do not hide: agents are the largest hidden cost in the system, and the noise they generate distorts market prices. But that bias is only worth something if I can prove it with two numbers: transfer value and actual competitive value. An analysis without both makes the agent section storytelling. I have written too many such stories not to fool myself again.

This is also where I separate two kinds of value when assessing a signing. Commercial value is measured in followers, shirt sales, reach. Competitive value is measured in rounds played under pressure, situations where a player must create alone, minutes in a losing game state. The two frequently move in opposite directions, and transfer reporting almost always reports only the first.

Heat cycles and expectation gaps

Every sports story has a heat cycle: ignition, spread, backlash. The data writer has a rare advantage in the second phase — the ability to measure the gap between market expectation and actual capability before that gap closes itself.

But to do that, one must know which phase the article being read belongs to, where the author stands, and whether the source is first-party reporting or aggregation. When all three are missing, the analyst has one honest option: state that it is undetermined, and explain why the fact of being undetermined matters.

This is the point I want to stress, and also the part most likely to be cut in editing. A piece saying "insufficient data to conclude" does not sound compelling. But it is entirely different from a piece with no conclusion. The first is a result. The second is an abandonment.

The counterintuitive angle: the industry's problem is not a shortage of data

Here I go against the grain. When an analysis fails, the reflex is to demand more data. More sources, more metrics, more tables. I think most failures in esports do not come from too little data. They come from too much unverified data, presented in exactly the same format as verified data.

A table with numbers looks like a table with numbers. Readers cannot distinguish which cell came from a publisher's official API and which came from a forum post copied four times. The interface does not distinguish. Only the writer can, and writers are rarely paid to do it.

The consequence is a paradox: the more sources there are, the less trust there is, because the cost of verification rises faster than the volume of information. In that environment, an empty result honestly published is worth more than a full table of unknown origin. It is a rare signal: here is someone who checked, and stopped at the right moment.

I know this sounds like a defence of slowness. But I have watched this industry for 23 years, and I have never seen a club go bankrupt for lack of analysis. I have seen several go bankrupt in silence, followed by a wave of surprised articles asking why nobody saw it coming.

The trap of "no evidence"

There is a second risk few data writers discuss. An empty result can be used as a tool. A communications department need only quote the accurate sentence "no evidence of financial problems" to convert it into "the organisation has no financial problems." The first sentence is true. The second is false. The distance between them is my entire job.

The writer is responsible for closing that distance before publication, not after the piece has been misquoted. The way to close it is simple: state what was checked, what was not, and why it could not be checked. Three lines do not weaken an article. They make it impossible to distort.

At 39, I have learned that data also feels pain when it is distorted. And the only person who can prevent that is the one holding the table before it goes to print.

What I did with the empty file that night

I did not write about the tournament. I do not know which tournament. I do not know the team, the player, the date. Any sentence I could write on that subject would be invention, and an invented sentence with correct grammar is still an invented sentence.

Instead I did three things.

I rechecked the source path. If the original text could still be retrieved, I would re-run the extraction layer with a corrected schema. If not, I would look for an archived copy. If no archive existed, I would close the item and record the reason, rather than approximate.

I wrote out the minimum five inputs any analysis needs before it can be considered grounded: the game title, the named entities, at least one quantitative fact if the original contains one, the publication date, and the outlet. Missing any one of them, every downstream conclusion sits at the lowest confidence level.

And I wrote this piece. Because a blocked pipe is also a story. It may be the least-read story of the week, but it is a true one, and it teaches something no analysis of a specific match can teach: our trade is not measured by how many numbers we hold, but by knowing what we are missing and being able to say so.

The long shot in memory always finds the top corner; in the spreadsheet it flies straight at the keeper. A decent writer is one who dares to record both — even when the goal is completely empty and nobody is shooting.

Signals for the next cycle

What I am waiting for is not a more complete report. I am waiting to see whether, over the next six months, the Southeast Asian esports scene develops one small habit: whenever an analysis appears fully loaded with numbers, someone asks where those numbers came from, which date they were taken, and who published them first.

If that habit forms, empty files like tonight's will become rarer — not because the pipeline gets better, but because people begin treating verification as part of writing. If it does not, we will keep getting seasons described vividly with metrics nobody knows who counted.