Nine Blank Pages: When the Football Analytics Room Has Nothing Left to Say

Two in the morning, and I open the report the analytics desk has sent over. Nine pages. Every page has tables, columns, rows, headings set in proper bold. And nearly every content cell says the same thing: N/A. No competition named, no team named, not a single player identified. Nine pages, perfectly formatted to announce that there is nothing to announce.

The first reflex of someone who has written about football for seventeen years is to fill the gap. Swap in a competition, add a few familiar metrics, call a coach, build a conclusion that sounds reasonable. I sat still for about ten minutes before I realised I was about to do the worst thing this job allows: turn silence into an opinion.

Then I remembered a different afternoon, at a training ground with no measuring equipment at all. In 2026 I stood behind the goal and watched a nineteen-year-old forward miss eleven times in a row. The session sheet recorded one line: 0 goals. No column recorded that he stayed forty minutes after the squad had gone into the dressing room. No column recorded that his mother had just been admitted to hospital. That night I posted a short line on the supporters' page, asking who among them had ever thrown a boot after training. The next morning he scored a hat-trick in a friendly. The sheet still said: 3 goals. It never records the rest.

So those nine blank pages left me almost grateful. They remind me that every analytical model begins with a very human question: what did we actually collect? When the answer is nothing, the honest move is to say so.

When football learned to count

Around 2026, expected goals left the statistics forums and entered the club boardroom. The logic was simple: every shot carries a probability of becoming a goal, based on location, angle, and the type of pass before it. Football stopped merely counting shots and started measuring the quality of chances.

Brentford is the example I retell most often. In 2026, then in the Championship, the club closed its academy and moved to a B-team model. The decision was called madness, because an academy is an identity. On 29 May 2026, Brentford beat Swansea 2-0 at Wembley and went up to the Premier League. Owner Matthew Benham built his fortune on probability models, and he carried the same principle into football: find what the market misprices, then wait patiently for the market to correct itself.

The other half of the story sits at Midtjylland, the Danish club Benham bought, where a small side won the league on undervalued players and corners designed like drills. In England, Liverpool set up its own research department in 2026 under Ian Graham, a physicist by training. In June 2026 the club signed Mohamed Salah from Roma for around 37 million pounds. The deal was called a gamble at the time; it now sits among the best value in Premier League history.

Vietnam arrived later, though not much later than most of Asia. GPS vests appeared at training centres from the mid-2010s. V.League clubs began putting someone with a laptop in the stands. From the 2026/24 season, VAR entered a number of matches, bringing a new layer of data on errors and disputes. The national team, under Kim Sang-sik from May 2026, won the ASEAN Championship in January 2026 with a shape rotated match by match, the second leg in Bangkok on 5 January 2026 finishing 3-2 for an aggregate 5-3 over Thailand.

But there is a paradox few people state: the more data you have, the greater the chance of misreading it. Data does not speak. It sits there silently, waiting for someone to assign it meaning, and that someone is usually under pressure to file before noon.

What data sees

Data is good at repeated events. A V.League match can contain four to five hundred passes, dozens of duels, hundreds of accelerations. A large enough sample gives a stable signal, and that is the whole strength of the model: it does not judge once, it counts a thousand times.

That is how expected goals works. A shot from a familiar position is repeated thousands of times in historical data, and the resulting probability is stable enough to judge whether a team is lucky or good. When a side outscores its xG for ten straight rounds, analysts call it a regression signal. Most of the time they are right.

Yet that same repetition limits the field of vision. A model only sees what happens often enough to be counted.

What data misses

In 2026 I followed a transfer involving a right-back moving from Guangzhou to Shanghai. Every outlet was chasing it, each with a source, each with a financial explanation. I knew the real reason, and it lived outside every spreadsheet: his six-year-old son needed foot surgery, and the hospital in the new city was better. I sat on it for nine days. When the deal collapsed over a salary cap, I wrote it as the story of a disappointed family. It was shared twelve thousand times.

Among the numbers of a transfer, I find the heartbeat of a human being. And there is no column for a heartbeat.

The same holds for low-frequency, high-consequence events: a penalty in the 88th minute, an injury in the 30th, a four-hour team meeting after a defeat. Models handle poorly the things that happen once a season. They handle superbly the things that happen two hundred times.

The empty-stadium season of 2026 taught me this the hard way. Players scored, turned to the stands, found nobody there. I proposed a small project: pipe supporters' noise from their homes into the stadium speakers. For one match, more than four hundred people joined a video call. No crowd, and yet I heard applause coming from small screens. No metric captured it, until I brought a decibel meter into the ground myself and wrote the reading next to the scorer's name.

Silence read as safety

Back to the nine blank pages. In a risk matrix, a row marked "undetermined" looks a great deal like a row marked "low risk." A reader skimming will take the flattering interpretation: no risk identified, therefore the club is calm.

In football this misreading appears every week. A club with no reported dressing-room conflict is assumed harmonious. A player with no injury news is assumed fit. A coach without press criticism is assumed secure. All three can be wrong simultaneously, because they are inferred from the absence of information rather than from good information.

That is why the blank report has its own value. It was the most honest document I received that month, precisely because it did not pretend. A report that admits it has no data is safer than a report stuffed with conclusions and backed by no source.

My Dinh, 1 February 2026

I still keep that night's notebook. Vietnam beat China 3-1, the first win over that opponent, on the first day of Tet. Ho Tan Tai opened the scoring in the ninth minute, Nguyen Tien Linh doubled it on 16, Phan Van Duc sealed it on 76. The visitors' consolation came in stoppage time.

What I wrote down was not the scoreline. I wrote that the three goals came from three entirely different situations, and that none of them resulted from holding the ball. The visitors passed more, but most of those passes sat in areas from which nobody scores. Read the stat sheet alone and you could conclude the controlling side lost to bad luck. The rhythm of a match does not live in the scoreline; it lives in the silence between two passes — and the stat sheet does not record silence.

That is the inherent limit: the model measures volume, while matches are decided by timing.

The shape on paper and the shape on grass

A team announces a 3-5-2. On paper, the wing-backs push high, the midfield three control the vertical axis. On grass, when the ball is lost, the wing-backs drop into a back five, the midfield three compress into a lopsided block of four, and the centre-forward becomes the first defender at the centre circle.

An average-position map after the match draws the first shape, because it averages every moment. But teams live in the second shape, and every goal conceded arrives in the instant of transition between the two. From watching V.League matches over the years, most tactical arguments online come from one side reading the announced shape and the other reading the played shape, both convinced they are discussing the same thing.

A good analyst must answer the hardest question: is this team playing this way because the opponent forced it, or because the coach instructed it? The data cannot separate the two. Only someone in the stand, counting how often the manager leaves his seat, can.

The misunderstanding from outside

A popular prejudice holds that data kills football's romance. People picture the analytics room as a cold box where a nineteen-year-old is reduced to a string of digits.

The reality runs the other way. Brentford's promotion is a more romantic story than almost anything English football has produced in the past decade, and the data model freed that story by saving money in places nobody was watching. The danger sits inside the room, in a different person: the analyst who always has a conclusion. The one who has never said "I don't know" is the most worrying person in the building.

In Vietnam the bottleneck is not the volume of data. Clubs already hold more than they use. The bottleneck is the translator — someone who can sit between the spreadsheet and the dressing room, who understands that a 12 percent improvement means nothing if the player sleeps four hours a night because of a newborn. The job needs someone fluent in both languages.

And one more thing I believe after seventeen years: a counter-intuitive finding is only worth something when it comes with evidence from the ground. A conclusion that runs against public opinion, with nobody standing behind it, is just an opinion said louder.

I keep the rhythm for those running on the pitch — that is how I belong to the match. But rhythm can only be kept by someone who knows when to stay quiet, and who says "I have nothing here" instead of building a beautiful report.

The 2026/26 season is entering its most compressed emotional phase. There will be more spreadsheets, more models, more confident people. The metric most worth watching next year may not be a new index at all, but the number of clubs that hire someone to sit between the two worlds. When the next report comes back blank, the answer is not to fill it in. The answer is to fold it, walk down to the training ground, and stand behind the goal until somebody misses for the eleventh time.