The Night the Data Didn't Arrive: The Discipline of the Sports Analyst
At 2:47 in the morning, it was still raining in Brisbane. I was sitting in...
At 2:47 in the morning, it was still raining in Brisbane. I was sitting in front of a screen with a cup of coffee that had gone cold hours earlier, and on that screen was a data frame with exactly one meaningful line: an empty list of information points. No title. No source. Not a single number. Nine hours before, I had blocked out the schedule, opened three spreadsheets, and built the skeleton of an analysis for a match I was certain had something worth saying. The system returned a blank page.
People assume a sports data analyst spends his life swimming in numbers. In practice, most of my working hours are spent waiting for the numbers to arrive, and they do not always arrive. Some nights the connection drops. Some nights the source page blocks the scraper. Some nights the data arrives but the extractor returns an empty object — the text is right there, only the machine reading it is broken. These two failures are entirely different, and the fix for each is entirely different. To the reader, they look identical: an article with no numbers in it.
This is where the job gets hard. When there are no numbers, there is always a comfortable exit: write without them. Replace data with adjectives. Use "appears to", "seems to", "notably". Nobody can fact-check a sentence that contains no facts. The piece still publishes, still gets read, and I still sleep.
That night I did not take the exit. I typed a sentence no editor wants to receive: "Insufficient information to conclude." Then I sent it, shut the machine down, and sat in the dark for another twenty minutes. Every number carries a story, and my job is not to break it. But when there is no number at all, the job is harder: do not invent one.
What follows is about that gap — a gap far more common in the sports industry than anyone likes to admit. It does not only happen to one man in Brisbane at three in the morning. It happens in every newsroom, every analytics department, every transfer report, every league table, every balance patch in competitive gaming. How a sport handles that gap determines the quality of everything it says for the rest of the season.
Context: the anatomy of a sports data pipeline
A modern sports analysis runs through three layers. The first is ingestion: pulling in source text, video, stat sheets, press releases. The second is extraction: pulling out discrete facts — player names, minutes, scores, transfer fees, publication dates, starting lineups. The third is interpretation: placing those facts into a tactical model or a causal chain to produce a judgement. The third layer can never generate anything the second layer failed to extract.
In footballing terms these layers have familiar names. Layer one is the match. Layer two is the stat sheet. Layer three is the story in the morning paper. When someone says "that was a great game and nobody described it properly", what usually failed was layer two — or layer three simply made things up.
The two failure modes must be separated, because the remedies are opposite. Mode A: the source never arrived. The page was blocked, the video taken down, the match not broadcast, the press release deleted after forty minutes. The fix is to re-fetch — find an archive, find a replay, find a syndicated version. Re-running the analysis model without re-fetching will reproduce the empty result exactly, and the analyst will conclude that the match simply had nothing in it.
Mode B: the source arrived but the reader broke. The text is there, the footage is there, only the stat export came out with the wrong columns, or the extractor mapped fields incorrectly. The fix is to re-read by hand, line by line. For a football analyst, this is the difference between rewatching the tape and reopening a corrupted summary file.
That night in Brisbane was Mode B. I knew because even the mechanical fields — title, source, category — were blank. An article with genuinely no information still has a headline. When the headline disappears too, the fault lies in the machine, not in the match.
The distinction sounds dry and technical. It is not. Across a season, hundreds of editorial decisions get made on the confusion between these two failures. A player has no metrics because the league does not collect those metrics — and the conclusion drawn is that he played badly. A coach is not asked anything at a press conference because the feed cut out — and the conclusion drawn is that he was evading questions. A team's footage is pulled for rights reasons — and the conclusion drawn is that they had nothing worth watching.
One empty object, three entirely different stories, and all three wrong in the same way: treating absence of evidence as evidence of absence.
The economics of having to say something
If twenty-three years in this trade have taught me one thing, it is this: the strongest pressure in this industry is not the pressure to be right. It is the pressure to have something to say.
Sports news cycles do not permit silence. A match ends at nine in the evening Brisbane time; the analysis
