35,000 Students and the GPS That Never Learned to Feel Sad

Hook

Last weekend a number passed through my feed that almost no one in football bothered to pick up. The National Autonomous University of Mexico — UNAM — published the results of a health screening of roughly 35,000 new-entry high-school students, spread across 14 campuses of its preparatory school system. Three in ten female students reported drinking alcohol. Close to half admitted using at least one substance. A university health official, Gustavo Olaiz Fernández, stated that the figure should have been "practically nonexistent."

I read that item three times. Once as a data man. Once as someone living in Europe, where every youth academy has a gym, a GPS unit, and a list of metrics as long as your arm. And once as someone hunting for the blind spot inside his own profession.

Because here is what I must say upfront: UNAM is not a football club. It is a university. The story is about school health, not football. But UNAM is the parent institution of Pumas UNAM — Club Universidad Nacional — the Liga MX side. And that bridge, however thin, is enough to force me to sit down and write this piece.

Because if a university can screen 35,000 teenagers for alcohol, substances, anxiety and depression, the question I cannot avoid is this: why can thousands of football academies across Europe — with more money, more devices, more analysts — not do the same for the very children they are developing?

Context

UNAM is one of the largest universities in Latin America. Its high-school system — preparatoria and CCH — feeds directly into the university. Every year, incoming students take a tool called the Examen Médico Automatizado, EMA — an automated medical exam designed to assess both physical and mental health, and used to build prevention strategies.

What matters: this is not a voluntary online survey. It is an institutional tool applied to an entire intake cohort, run across 14 campuses, at a scale of roughly 35,000 students. In my language, that is a near-census sample. No small-sample problem. No selection bias. No "one match proves nothing."

And the result: around three in ten new female students drink alcohol. Nearly half report using some substance. What pushed the health office to speak was not just the level but the trend: the female rate runs higher than historical expectation. Olaiz Fernández made clear the issue is "not only internal" — the university is positioning itself within a national trend rather than owning the problem alone.

Now the football part. UNAM owns Pumas UNAM. That club runs its own youth system, its own academy, its own intake. So under one institutional roof there are two systems receiving teenagers: one developing footballers, one running general education. The second just screened 35,000 of them for alcohol and substances. The first shows no public sign of an equivalent screening.

I am not saying they do not do it. I am saying we do not see the data. And in my trade, "we do not see the data" is always a finding.

Core

1. The data infrastructure of youth football

Take an elite European academy. Every U15–U18 player carries a GPS device in every session and match, producing total distance, high-speed running, sprints, decelerations, metabolic and mechanical load, and recovery windows. Alongside it: passes, completion rate, box entries, shots, expected goals. At team level: PPDA, recoveries in the final third, transition counts, possession.

That is a mountain of data. I have built those models myself. But I also know what they miss.

They measure the body. They do not measure the person.

No metric tells me whether that boy slept. None flags whether he drank on Saturday night. None detects that a U17 player's 15% drop in dangerous-ball involvement is not a hamstring issue but a state no one has named.

I once watched a case. An academy player's GPS numbers were so clean the fitness staff used him as an example. Distance flat. Sprints flat. But a technical index I compute myself — involvement in dangerous possession — fell steadily for six weeks. No one looked, because the GPS board stayed green. Only when he asked to stop training did anyone ask why. And the answer was not on any of our spreadsheets.

That is the gap. And it is systemic, not accidental.

2. Thirty-five thousand versus zero

| Criterion | UNAM (school system) | Typical European academy | |---|---|---| | Mental-health/substance screening scale | ~35,000 per intake | No equivalent published | | Tools | EMA — automated medical exam incl. mental health | GPS, xG, PPDA, training load | | Subjects | All new-entry students | Academy players | | Purpose | Prevention, intervention design | Performance optimisation, injury reduction | | Publication | Yes | Almost never |

The right column is blank on the first line. Academies broadcast physical metrics as a badge of professionalism while mental-health and substance data on the same children barely exists in public.

3. Three indices every academy should collect

Subjective recovery index: daily sleep quality, stress, physical sensation, mood, on a simple scale. Cheapest data to gather, most ignored.

Adverse-event frequency: not for discipline, but for pattern detection. Three unusual behaviours in one age group in one month is a system signal, not three bad kids.

The ask threshold: how many players in an academy dare to say "I need help" without fearing for their place. That number would be one of the most important metrics an academy owns — more than xG.

4. Reading UNAM's numbers

Three in ten female students drinking runs against historical norms. When a large sample reverses a trend, that is structure, not noise. Nearly half reporting substance use, at n≈35,000, sits outside the noise band. And the reference standard — "should be practically nonexistent" — is a mature data process: definition, standardised tool, benchmark, feedback loop from data to policy.

I must admit as an analyst: this process is tighter than the data process of most youth academies I have worked with.

5. Why the gap exists

Performance culture (high confidence): what wins soon gets measured; mental health does not win in six weeks. Reputational risk (medium): publishing weakness creates exposure, so some avoid collecting. And the old belief that young players are raw material, not people to be asked.

6. Pumas in the bigger picture

I have no internal Pumas data and will not invent any. But under one roof sit a university that just screened 35,000 teenagers and a club with a youth system. If I were their head of data, I would ask: why not run an EMA variant for academy players? The infrastructure, the tool, the intervention design experience — all already exist. Only the decision is missing.

The issue is not missing data. It is the missing choice to collect it.

Contrarian

Our industry is proud of its data revolution. We speak of xG as a new language, PPDA as a measure of identity, of training load and sports science.

But let us be honest. Football's data revolution is a revolution of the body, not of the person.

We measure how far a player runs, sprints, passes, shoots. We even measured breathing via GPS in a bubble season. But we do not measure a young talent's biggest enemy: himself.

I was once rigid. In 2026, redesigning Lyon's programme, I ordered players to hit 120% of a GPS threshold to count as sufficient. Muscle injuries fell from 12 to 5. I was proud. I thought I understood football. But I never asked a player on that rehab list whether he had slept the night before. Whether he had drunk. Whether he was okay.

Here is the most counter-intuitive part. Academies do not avoid psychological data because they disbelieve numbers. They avoid it because they believe in numbers. They know that collecting it will say something they do not want to hear. And in football, the damage is not done by the truth but by the truth made public.

Data never lies, but it knows how to hide. Our job is to make it talk. And football's data revolution has, for twenty years, made it talk about legs while forgetting the rest.

Takeaway

My prediction, a blueprint rather than a prophecy: within 24 to 36 months, at least one major league in Europe or the Americas will introduce mandatory mental-health screening for youth academies, mirroring how sports-medicine standards already mandate cardiac screening. Not because the industry grew kind, but because data like UNAM's is creating a benchmark the industry can no longer dodge.

If that happens, the gap will be filled from outside, not from within. And that is always the most expensive way for an industry to learn a lesson it should have learned itself.

As for the boy who lost 15% of his dangerous-ball involvement for six weeks while the GPS stayed green — he is still waiting for a metric that knows how to feel sad. The question is not whether we have the equipment to measure it. The question is whether we dare to write that number into the sheet.