Three Data Layers Beneath a Goal: Notes from Vietnam's Youth Academies

An A4 Sheet and Two Goals

In March 2026, from the stands at Lach Tray Stadium, I held an A4 sheet printed from a tracking application. A 17-year-old home player had just scored twice in 34 minutes. The numbers were tidy: two goals, three shots, a 66.7 percent conversion rate, 8.4 kilometres covered, four sprints. The scout sitting beside me had already pulled out his phone to message his head office. I held his hand back.

The two goals were real. The A4 sheet was only the surface layer.

I have spent 24 years in this profession learning one thing: a data table never tells its own story. It only speaks when we know which winter this boy has just come through, where he grew up, what coaching curriculum shaped him, and what standard of defence he faced. That day, the 17-year-old scored twice against a back line containing two centre-backs born in 2026 who had already played more than 60 V-League matches. That detail makes the number heavier, or lighter, depending on how you read it.

Data is the topsoil; I always dig three layers deeper.

Four Development Models, One Wrong Ruler

Vietnam's annual season is entering a compressed phase. The V-League, the First Division, national youth competitions and continental qualifiers run in parallel, splitting the minutes available to young players and distorting every dataset gathered from above. A U19 player can play 90 minutes for the youth team on Tuesday, sit on a V-League bench on Saturday, and be called up to the U23 side the following week. Aggregated, a common denominator appears: few minutes, but a dense travel schedule.

That is the terrain. Read the map without the terrain and the data will point you the wrong way.

Vietnamese youth development currently has four model types. The first is the academy of a well-resourced club: residential players, a gym, a doctor, a GPS system. The second is a training centre attached to a company or a social organisation, where the curriculum is good but recovery conditions vary. The third is the local school-linked setup, where players still attend academic classes each morning. The fourth is the late-blooming player found in grassroots football who never passed through an academy at all.

These four produce four different types of data. The most common analytical error is applying the same ruler to all four.

I have made that error.

In 2026, working as a senior expert at a northern youth academy, I underrated a 16-year-old midfielder named Nguyen Duc Nam. My data table had three columns: body mass index, 30-metre sprint speed, and sprints per match. Nam fell below the national U17 benchmark in all three. I wrote in my report that he lacked the physical foundation for professional football. Three months later, Nam made his V-League first-team debut and recorded four assists in his first five matches.

What I missed sat in a column that did not exist in my table. Nam had just returned from a cruciate ligament injury and was in a growth-spurt phase, his body developing faster than his ability to control it. I read the numbers without reading the person. Since then, every data table of mine carries an extra column called biomedical context, and one rule: never pass judgement on a player under 18 without growth data.

Three Layers of a Number

This is how I work, and how I recommend domestic academies restructure their player files.

The first layer is the surface number. This is what reaches the news: goals, assists, minutes, distance covered, sprint counts. In Vietnam, data collection has improved considerably thanks to cameras and tracking systems, but the gap between competitions remains wide. Youth leagues mostly record only goals and cards. When layer one is all you have, every conclusion is an empty conclusion.

The second layer is the quality of the number. A goal does not say where the player received the ball. Of the 17-year-old's two goals at Lach Tray, the first came from a misplaced long pass by the opponent, the second from a counterattack after the opposing side had pushed up chasing an equaliser. Both were real goals. But their predictive value differs sharply from a goal built through four passes inside the box.

So I do not use total goals. I use three supporting metrics: the quality of the preceding pass, the receiving position inside the finishing zone, and the defensive intensity of the opponent in the ten minutes before the goal. All three can be recorded by hand by a disciplined observer, even where no tracking system exists.

The third layer is the condition of success. This is the layer I care about most. A young player only has a basis for development when four conditions coexist: a suitable curriculum, minutes at a suitable standard, a physical foundation on an upward curve, and an environment that does not burn him with expectation.

None of these four appear in any statistical table. They must be excavated.

In 2026, analysing Kylian Mbappe at the World Cup in Russia, I did not start with his four goals. I counted 11 successful dribbles against Argentina, then checked the starting position of each one. Most began on the left flank, in a zone where Argentina positioned their marker late and lacked midfield cover. Mbappe's condition of success then was space, not raw speed. That is why my report predicted France would win based on midfield structure rather than on a star. The report was later used as teaching material by a domestic academy.

When I present this reading method at scouting workshops in Vietnam, the reaction is usually the same: people nod, then return to total goals. That habit is hard to break, because it is fast and because it delivers a tidy answer.

The Third Condition, and the Price of Ignoring It

Of the four conditions, the one most ignored in Vietnamese football is accumulated load.

Young Vietnamese players are often pushed into the first team too early because of a club's need for results, then pushed out again for lack of minutes, then called up to youth national teams mid-season. The result is a familiar pattern: debut at 18, a burst of brilliance over six matches, injury at 20, and by 23 nobody remembers he was ever there.

Injury does not erase a talent's name; it only drops that talent into a lower sediment layer.

In 2026, when global football paused, I accepted an invitation to review an academy in central Vietnam. Historical data showed an 18-year-old striker named Tran Van Cong with a scoring rate of 0.8 goals per 90 minutes, the highest in the academy, but with very few minutes played and frequent cramping from the 70th minute onward. The two metrics contradicted each other. Training grounds were closed, so I interviewed his family online, examining diet, sleep and the daily commute. That administrative data explained most of the problem: he was short of water and short of sleep, not short of talent.

I recommended a professional contract before the league resumed. When the 2026 V-League kicked off, Cong scored six goals in his first season.

I retell this not to boast about one correct call. In my profession, I am wrong more often than right. I retell it to show that goals per 90 minutes and load tolerance must be read together. Goals per 90 alone is an unverified promise.

It took me three years to understand that data also needs a growth spurt.

Data can also be used to say no.

In 2026, during the winter transfer window, I reviewed the loan file of a defender moving from Ho Chi Minh City to Hai Phong. The surface looked excellent: 12 successful tackles across three matches. But when I isolated the three away fixtures, I counted three direct errors leading to goals. All three occurred in the first 15 minutes of the second half, and all three followed a loss of possession by a teammate higher up the pitch. The pattern pointed to one specific problem: the player coped poorly when attacked immediately after his team lost the ball.

I advised the club against a long-term deal. Two weeks later the player suffered an injury and the contract was cancelled. That outcome did not please me. It simply confirmed that match-by-match, action-by-action verified data carries higher warning value than an aggregate figure.

Local Calibration: Do Not Grade Vietnam With a European Ruler

Born and trained in France, I carry a set of standards wired into my reflexes. Years of working here taught me those standards must be calibrated before use.

A European academy measures a young player's recovery capacity against assumptions about nutrition, sleep, pitch quality and travel distance. In Vietnam, a 17-year-old may wake at five in the morning, ride a motorbike 20 kilometres to training, train twice in hot humid conditions, then go home to study. Placing him on the same load chart as a French player of the same age is a physiologically meaningless comparison.

That is why I always add a step called local calibration before writing any judgement. It has three parts: record actual training conditions, record travel distance and time, and record pitch quality during the sampling week. Those three lines of notes often explain fluctuations that tracking data cannot.

Another example sits in pressing metrics. PPDA, passes allowed per defensive action, is a good tool for measuring pressing intensity. But in Vietnamese youth competitions, poor pitches and a broken rhythm make the metric swing sharply between halves. I usually calculate PPDA in 15-minute blocks rather than per match, otherwise I risk misjudging a midfield simply because the pitch flooded in the second half.

The Effort-Metric Trap

The football data industry is selling us a belief: more numbers means more understanding. In my experience, the opposite is more often true.

Distance covered and sprint counts are usually presented as effort indicators. But running without purpose also produces pretty numbers. A midfielder who covers 12 kilometres in a match, most of it directionless, appears as a warrior, while in reality he is being pulled out of position and opening space for the opponent to exploit.

The only way to distinguish is to place effort metrics beside positional metrics. How far did he run, and what was he running for.

Another blind spot is minutes played. Academies often take pride when a young player gets heavy first-team minutes. But if most of those minutes come in decided matches, or in a position that is not his strength, then the minutes measure the club's convenience, not the player's development.

In 2026, working with a group of young journalists at a major international tournament, I analysed Pedri's data and found his distance covered dropped 18 percent after the 75th minute. I issued a warning that if he was pushed into extra time, the risk of decline was clear. The coaching staff did not rotate. Pedri left the tournament with an injury. I was right and changed nothing, a limit anyone working in data must accept. Since then I have begun studying machine learning methods to supplement the human eye, because I understand my old model is slower than the pace of modern football.

Caution is not weakness. It is the condition for surviving this profession.

Empty Reports and Honesty About Gaps

The type of file I hate most, and the type I check first when scouting documents arrive, is the formally complete but substantively empty report. It has a title, a table, a conclusion, and not one line of data to cross-check.

In my system, such a report is flagged as an empty report and returned to the sender. I do not speculate on behalf of missing data. If a column is empty, I mark it empty and state that no conclusion is yet possible.

This may sound like a dry administrative rule. But in youth football, filling gaps with guesswork is often how a wrong player is sold to a right club. A file missing data but carrying a firm conclusion is a dangerous file, because it manufactures a sense of certainty exactly where none exists.

So I keep one simple professional rule: a report that is honest about its gaps is worth more than a report that fills its gaps with speculation.

A Goal Only Means Something When We Know What He Has Just Been Through.

Back to the A4 sheet at Lach Tray. After the match I sought out the 17-year-old, not to congratulate him but to ask about the previous week. He told me he had slept four hours because of school exams, trained an extra morning session, and played a youth-team match on Friday afternoon. His two Sunday goals came from a sleep-deprived body. That is real data, and it exists in no software.

Three days later, at the next training session, he suffered a hamstring injury. Not serious. But exactly where I had feared.

What I Am Waiting For

The boy at Lach Tray is still nothing yet. He may become a cornerstone, or he may become a name nobody mentions in three years, and both possibilities carry real probability. What I can do is keep recording, keep digging another layer, and keep writing down the times I read him wrong.

If the four conditions, a suitable curriculum, minutes at the right standard, a rising physical foundation, and an environment that does not burn expectation, hold for two more seasons, I will have grounds to write a serious report on him. If one of the four breaks, the numbers will still look good while the player disappears from the map.

A growth spurt is the most beautiful thing the league table cannot measure.

A player is not a number, but the number is where my excavation begins.

I do not excavate stars; I excavate context.

And context in Vietnam is changing faster than the data tables we are using to measure it.