When Machines Read Football: Lessons from an Empty Analysis and the Future of Sports Journalism

The Groupama Stadium had been empty of fans throughout the 2026-20 season, but nothing is as empty as an analysis with nothing to analyze. That is what I realized while reading a technical report recently — the analysis clearly stated "insufficient information" in every section, from tactics to finance, from the dressing room to league structure. A nine-process analytical framework meticulously designed, but the input contained only two letters: N/A — no team name, no player, no match, no date.

I have been following Olympique Lyonnais for over two decades, from the early days when Jean-Michel Aulas built the project to climb back up, through the seven consecutive Ligue 1 titles, and the seasons of struggling in mid-table. I have witnessed coaches change, players come and go, and a club disappear from the Champions League map after just one wrong transfer decision. But never have I seen an analytical system so "empty" — not because it lacks data, but because from the very beginning, the source simply did not exist.

Context: The race between speed and depth

Throughout 41 years in the profession, I have watched football change in ways we now call "the speed of information." In the old days, a journalist could spend an entire week writing a feature about a substitute in the youth team — and the newspaper would still publish it, because readers needed those stories. But since digital platforms rose to dominance, what the sports media industry demands is no longer depth, but speed. Transfer news must be available within 30 minutes of a tweet. Match analysis must be published the moment the final whistle sounds. And to meet that demand, automation systems emerged — tools that can read hundreds of sources simultaneously, extract information, and output analysis in minutes.

The nine-dimensional analytical framework I am referring to is a typical example. It was designed to evaluate every aspect of a football story — from on-field tactics and club finances, to match results and league standings, compliance, dressing room status, risk profiles, media narratives, and industry-wide transmission effects. Nine processes, each divided into dozens of sub-criteria. A theoretically perfect system — but theory does not read football.

Core: Three layers of problems hidden behind an empty analysis

The first thing I realized when reading the report carefully is that it is not just about missing information. It is about a system designed to process information, but lacking a mechanism to detect when that information does not exist. Stage-1 — the information extraction step from the source article — returns a complete schema with empty fields. Instead of reporting an error, it neatly fills in "N/A." And Stage-2 — the deep analysis step — receives a complete framework, not missing a single field, but with every value reading "insufficient information." This is not a technical failure. It is a failure in philosophical design.

The second layer of the problem lies in "misleading-completeness risk" — the danger of artificial completeness. An analysis with full structure, adequate headings, adequate sub-sections, but no content. It looks official, looks professional, looks as if someone actually did the work. But in reality, it is just a shell without a soul. In sports media, where deadlines weigh heavily and viewership pressure drives every decision, such an analysis could easily be deemed "clean" — no issues — and pushed downstream as if it were trustworthy.

The third layer, and perhaps the most important, is the lesson about football's nature as a human phenomenon. The report listed dozens of metrics it could not evaluate — xG, PPDA, possession rate, wage expenditure, net debt, FFP, PSR — all useful tools when data exists. But football is not just numbers. Football is the gasping breath of a player in the 89th minute when he knows he is racing against time to preserve a draw. Football is the gaze of a coach looking down at his watch when he realizes he has made the wrong substitution. Football is the weeping of an elderly supporter in the stands at season's end, when the club he has followed for 40 years drops to a lower division. No automated system can capture those moments — and when the source contains nothing but an empty URL, the system cannot distinguish between "no information" and "information does not exist."

Contrarian: Why this emptiness matters

There is an interesting paradox in this report. It contains no football analysis whatsoever — no match, no player, no coach — but it provides highly valuable information: it shows that the content extraction pipeline is malfunctioning from the start. "Article Source: N/A" and "Article Title: N/A" appearing together suggests the fetcher retrieved nothing from the source — possibly a network error, possibly a paywall, possibly a page rendered in JavaScript that the parser could not read.

In 41 years of writing about football, I have learned that the most important thing is not what happens on the pitch, but what happens on the margins — in the dressing room, in the press conference, in conversations never broadcast. An empty analysis is the same. Its value lies not in what it contains, but in what it exposes about the system that created it.

Many will say this is proof that AI cannot replace sports journalists. But I think that is a superficial view. The issue is not that machines cannot analyze football — it is that machines cannot read the source. And to read the source, someone needs to know where it is, how to approach it, and when it is trustworthy. That is the job of a journalist — not an algorithm.

Next signals: When systems need human intervention

The report proposes a "minimum-content gate" — a minimum check mechanism before allowing data to continue down the pipeline. The idea is: if the source article has no title or at least one information point, the system reports an error instead of returning an empty schema. This is a correct technical improvement. But it only solves half the problem.

The other half lies in this: when a sports journalist stands before a match, he is not just collecting information. He is reading football. He watches how a player moves when off the ball and knows that player is having fitness problems. He hears a coach's voice in the press conference and can guess what is happening in the dressing room. He knows that an "empty" article can be as valuable as a "full" one — because sometimes, the most important thing is to acknowledge that nothing happened, and explain why.

Olympique Lyonnais once had a season where every statistic was good — high possession, many shots, high xG — but the results were terrible. If an AI system only looked at numbers, it would conclude the team was playing well and had no problems. But a journalist present at the stadium every week would see something different: the players had lost faith in each other, the coach had lost control of the dressing room, and the club was in a painful transition phase that no one wanted to acknowledge. That is why, no matter how advanced technology becomes, sports journalism still needs humans — not to collect data faster, but to understand what that data means in the context of real people.

Returning to that empty analysis. It contains no football stories, but it reminds us: before building complex analytical systems, we need to ensure the source actually exists. And to ensure that, we need people who know how to find the source — journalists like me, who have stood at the edge of the pitch for decades and know that the real story never lies in the statistics, but in the eyes of a player just stepping out of the tunnel, in the clenched fist of a coach looking at the clock, and in the sigh of a supporter when the match ends with no goals scored.

When will I trust an analysis? When it is written by someone who has sat next to a coach in the dressing room after a defeat, who has seen tears in a goalkeeper's eyes after a corner kick, and who has understood that every number in the standings represents hundreds of shattered dreams and dozens of lives changed. That is when an analysis has value — not because it is full of information, but because it is written by someone who knows how to ask the right questions.