International FootballThe Empty-Data Trap: Vietnamese Football Is Confusing “No Data” With “No Problem”

The Empty-Data Trap: Vietnamese Football Is Confusing “No Data” With “No Problem”

Đặng Thế2026-10-17 12:40Tiếng Việt

On a Tuesday morning in Shanghai, a colleague sent me an analysis report. Nin...

On a Tuesday morning in Shanghai, a colleague sent me an analysis report. Nine sections, and every one carried the same line: “N/A — insufficient information”. The tactical table was empty. The financial table was empty. The risk matrix was empty. Right beneath the risk matrix sat a note that kept me at my desk longer than the report itself: “An unidentified risk must not be read as a low risk — it is the absence of data, not the absence of hazard.”

When the whole commentary box says “there is nothing to discuss”, I heard a very quiet gap.

I am 55. I have sat long enough in commentary boxes full of men to know what it feels like to be asked “where are your numbers?” after every judgement. I learned to answer with tables. But one thing I learned later, and far more expensively: telling apart two things that look identical on a screen — no data and no problem.

In the summer of 2026, the stadiums were empty, the leagues stopped, and I had no match to talk about. I went back to Shanghai, opened dozens of old tapes, and counted set-piece goals across the three most recent seasons. The number that came out stopped me cold. The old tapes were sitting there, and I put on my glasses, and I saw the future.

Now I look at Vietnamese football and see a different summer arriving — not a summer without football, but a summer with too many spreadsheets and too few people who know how to read the empty cells inside them.

Over the past decade, football analysis has changed faster than every previous decade combined. GPS vests, semi-automated camera systems, event-data platforms, then xG, PPDA, progressive passes, packing rates. In Vietnam, V.League 1 clubs have started hiring analysts, buying data packages, building video rooms. The PVF academy, the Hoang Anh Gia Lai – JMG academy, Viettel, Hanoi FC — the places with money and intent — have all brought data into their coaching processes.

That is the good news. The bad news sits somewhere else.

When an organisation buys data, it usually buys a report template along with it. A template is an implicit contract: it tells you a match must be described through these sections, in this order, with these metrics. And every template has a blind spot nobody is trained to notice — the empty cell.

In any serious data system, the empty cell has a name: null. Null is where the system failed to retrieve a value. It is entirely different from zero. Zero means “measured, and the result was nothing”. Null means “not measured”. But when the report is printed on paper and placed on the meeting table, both look the same: a silence.

And people, especially people in football, have a very natural reflex toward silence: they read it as calm.

About three years ago I came across an internal template with a beautiful flaw. The “source quality” field read: “assess from the source field of the information points”. But when the list of information points is empty, the source field is empty too, and the “source quality” field has nothing left to stand on. It references itself. The person writing the report was not wrong. The system was not wrong. The template had simply been designed on the assumption that there is always an article in front of it.

That assumption holds 95% of the time. And the remaining 5% is where the worst decisions are born.

The same thing happens on the pitch. A centre-back loses no aerial duel in the first four rounds. The stat sheet reads: zero goals conceded from set pieces. The coach reads it and relaxes. But the right question is: how many real crosses did he face in those four rounds? If the answer is six, then that zero does not measure aerial defending. It measures the fact that nobody has thrown enough balls into his zone yet.

Null dressed as zero. And that zero gets taken into the meeting.

The risk matrix is where this flaw shows its face most clearly. A standard matrix has six rows: sporting, financial, personnel, rules, public opinion, systemic. Each row has a level, a likelihood, an impact, a mitigation. When the input data is empty, all six rows read “insufficient information”.

The problem is the next step. In many of the meetings I have sat in, an empty risk table gets read as “no problems seen yet”. No problems seen yet slides into “there are no problems”. And “there are no problems” slides into “carry on as we are”.

Three slides. Not one of them supported by a number.

I have seen exactly these three slides at a club that started a young player six matches in a row while the medical department had never entered his load data into the system. Nobody decided to play him six times. There was simply an empty data field, and nobody asked why it was empty. In the seventh match his muscle tore. Only then did the report have a number in it.

Football always pays for data in that currency: with injuries, with goals conceded in the 89th minute, with a season drifting by and nobody knowing why.

xG, expected goals, measures the quality of chances. PPDA, passes allowed per defensive action, measures pressing intensity — the lower the number, the more aggressively a team presses.

Both are excellent. Both get abused in the same way: people read them as thermometers when they are photographs.

A team's PPDA drops from 12 to 8 over three matches. The story is told immediately: the coach has cranked up the pressing. It might be right. It might be wrong, for three very human reasons.

First, the opponents changed. Face three below-average passing teams and PPDA falls on its own without anyone pressing better.

Second, the definition changed. The data provider updates its methodology and the whole league's metric jumps a notch in the same week. I have seen this happen. Nobody reads the version notes.

Third, the team is behind. When you are 0-1 down, everyone presses more. PPDA falls not because of tactics but because of the scoreline.

This is why I always demand one thing before believing any pressing metric: the distribution by score state. PPDA while level, PPDA while leading, PPDA while trailing. Those three numbers tell three different stories. Merging them into one average is the fastest way to lie without breaking a rule.

xG is worse, because it sounds more scientific. A team has 2.4 xG but scores once. The conclusion is fired off: poor finishing. Maybe. But maybe the sample is nine shots, and 2.4 xG from nine shots means two extremely clear chances that the opposing keeper saved brilliantly. With nine shots, the gap between xG and goals says nothing yet. A small sample is not evidence. A small sample is noise in a costume.

The thresholds I set for myself: under 30 shots, no conclusion about finishing. Under six matches, no conclusion about tactical trend. Under a full season, no conclusion about a coach.

Back to the summer of 2026.

I counted set-piece goals for mid-table teams across three seasons. The result: 68% of their goals came from dead balls. I wrote a 5,000-word piece titled “Set pieces are the new king of Chinese football”, published on 20 May 2026. A young Shanghai SIPG coach read it and invited me to lecture on tactics at the academy. For the first time, male coaches older than me sat taking notes seriously.

What I remember is not the 68%. It is where that number was sitting

Cầu thủ liên quan