When the Football Feed Learns to Read Blink Rates

When the Football Feed Learns to Read Blink Rates

On Tuesday morning, in the transfer feed I open every day, one headline carried the “football” label. I clicked it. There was no team, no player, no scoreline. The story was about an 80-second trailer: the woman in the frame blinked six times per minute, while the usual reference range is cited as 15 to 20. A body-language analyst, a clinical psychologist and a behaviour expert each concluded the face was insincere. The subject is Elizabeth Holmes, the founder of Theranos convicted of fraud. The report ran in The Express Tribune; the film is titled You Can See Everything, premiered at Telluride and reaches cinemas on October 16.

I sat with that line longer than it deserved. Years ago, an off-topic piece landing in the football section annoyed me for a minute and then faded. Now it matters, because the feed is my workplace, and one wrong label takes down the whole filtering system readers build around it. The “football” label is a promise about content; when the promise breaks, every filter built on it collapses at once.

Reading people is something football has always done, it just rarely says so out loud. Before every transfer window, recruitment departments watch not only match footage but interviews, dressing-room reactions, the way a player greets the staff. In negotiations, sporting directors and agents read each other through the pauses between sentences. For penalty shootouts, sports psychologists are brought in to dissect pre-kick routines. Football was never a stranger to behaviour. It simply keeps behaviour at the reference level, while the final call still rests on footage and match data.

The second layer of context belongs to information infrastructure. A general-interest paper that carries football coverage usually gets tagged at section or feed level, so every article in that stream inherits the same label. I have seen a school basketball tournament filed next to Premier League news, a golf course opening filed beside a centre-back's injury. In those cases readers spot it instantly, because the article still contains a ball, a player, a scoreline to check against. This time there was nothing. The piece left no trace of football, so readers had no way to detect the mismatch. They just saw an unfamiliar name wedged between two familiar transfer stories.

Transfer season is a season of noise, and noise has to be filtered by what can be verified. A loan with an obligation to buy can decide the financial fate of a small club for three seasons: the clause structure, the activation trigger, the wage split, the accounting date. In modest-budget leagues across the region, one badly timed purchase clause can lock a wage bill for the next two seasons. That is the information fans need and deserve, even when it is drier than a face avoiding the camera.

The first question I asked myself was not whether the article was right or wrong, but what a reader in a football feed would learn from it. The answer: a method of reading people with no baseline. A rate of six blinks per minute was set against a 15-to-20 reference range, but blink rate shifts with lighting, dry eyes, contact lenses, medication, the concentration of reading, and whether a person knows the camera is running. Someone speaking on camera is not in the same conditions as someone speaking in a meeting room. An 80-second edited clip cannot be called a sample at all. Moscow taught me that preparation begins with calling people by their right names — and calling a name correctly, in the fullest sense, means knowing exactly what you are measuring on that person.

In football, that discipline is stricter. When a club weighs a contract, the name that needs calling is not the one on the shirt but the whole circumstance: hometown, language, previous league, family environment, how the player reacts when substituted in the 70th minute. Without those, a blank face at a press conference says nothing. With a baseline, the same face says a great deal: it tells you the player is tired, angry, or trying to look calm about bad news.

In 2026, in empty stadiums, I could hear clearly that every match still had its own breath. Players called to each other, slapped the post, talked to the referee, and their voices carried without a crowd to cover them. That period taught me behaviour can only be read against a reference point. I knew what a certain player's sigh meant, because I had sat beside him through two hundred sessions before.

Football's data side went another way. Metrics such as expected goals or passes allowed per defensive action share three properties that expression analysis lacks: they measure the same thing on every subject, repeat over time, and compare across matches. They are dry, but they can be checked. Expression analysis is the opposite: vivid, memorable, and nearly impossible to verify unless the analyst publishes the method.

So why does the feed pick the vivid one? Based on my experience watching matches and watching the feeds themselves over many years, I think it comes down to the economics of attention. A line about how someone blinks draws a quicker click than a line about the structure of a release clause. Aggregation systems need traffic, so they have little incentive to fix a wrong label as long as the label delivers reads. A technical error is being fed by a real demand.

In football, body language is trustworthy only when tied to a baseline observed repeatedly — the same person, the same context, across many moments — while a conclusion drawn from one isolated frame is a guess dressed as data.

What worries me is not one article. It is that the same logic is walking into recruitment rooms. More clubs use automated tools to score candidates' attitudes from interview footage. If the input is isolated clips with no reference point, the output will be scouting myth: a player rated unprofessional simply because he blinked little in front of a camera on a morning after no sleep.

The easiest explanation is to blame the tagging algorithm. I do not think that is the whole story. If the football feed keeps importing behavioural content from outside, part of the blame belongs to football writing itself: we have not produced enough methodical, verifiable behavioural work to fill that space. When sports journalism revolves only around transfers and scorelines, the door stands open for vivid content with thin grounding.

In Chengdu, I once followed a story running the other way. A 19-year-old goalkeeper, number 12, cried while describing cartilage damage during quarantine. I helped him turn his account into a public post; it spread, and the club decided to hire a psychologist. That was also reading behaviour, but with someone accountable, with context, and with a concrete consequence attached. The difference between the two examples is not the subject matter. It is who stands behind the conclusion.

The signal to track in the coming months is not a film trailer but how clubs and feeds publish their methods: what they measure, across how many observations, and who audits it. If the answer is still isolated clips with no baseline, fans should build their own filter rather than wait for the system to fix itself. At 68, I place my trust in the next generation of storytellers.