Nine Analytical Dimensions, One Pitch: The Limits of Modern Football Data

Kazan, June 30, 2026. Minute 64. Kylian Mbappé receives the ball near the halfway line, some forty metres from the Argentina goal. He is nineteen years old. He tosses his messy hair, drops his centre of gravity, and the pitch looks as if it has been torn in two. I am standing behind the camera rig for the documentary The Strange Winds, abandoning a script approved weeks earlier, following the boy with the lens. Four minutes later, he does it again.

Only months later, in the edit suite, did I measure the average speed of those two surges: 37.8 km/h. But that is a number from the following day, from software, from the desk. What I saw from the press row was only a child standing still inside a knot of team-mates, breathing, like a wind that does not know it has passed. I never wrote the 4–3 scoreline into any caption of that film. Real football, for a documentary screenwriter, lives in the shape of a run.

I have told that story many times in production meetings. Every single time, a younger colleague asks me how to put a running stride into a data table. I have never been able to answer.

A framework too beautiful to doubt

Modern football has built an enormous analytical apparatus, and it runs so smoothly that almost nobody notices we are living inside it. Any club above mid-table level has its own analysis department. Every broadcast carries probability graphics. A contract worth tens of millions of euros is entered into the books through a division the audience never sees. Regulators hand down verdicts in the language of accountancy. Agents talk to each other in sell-on percentages.

I have worked with the nine-dimension analysis dossier that many professional departments now use. Tactics. Finance and the transfer market. Results and the public-opinion cycle. League landscape and club positioning. Rules and governance. Coaching and the dressing room. Risk profile. Media and expectation. Whole-industry transmission. Each dimension has its own vocabulary, its own scale, and its own empty cells waiting for data.

That framework is beautiful. It is logical, it is comprehensive, and it gives the impression that if you fill in enough numbers you will understand the match. But once I received a forty-page dossier — all nine dimensions, all headings, all tables — and in every cell the same line repeated: insufficient information. A steel-framed building with nobody inside.

I kept that file. I did not delete it. It taught me more about our trade than any report stuffed with numbers ever has.

The probability cell and the running stride

xG — expected goals — is perhaps the finest invention of data football in twenty years. The model assigns every shot a probability of becoming a goal, based on angle, distance, body part, defensive pressure and goalkeeper position. A shot from the edge of the box may be worth 0.08 goals. A close-range finish, one-on-one, may be worth 0.35. The greatest value of xG is that it separates process from result: a team that loses while generating 2.4 xG did not play badly, it simply did not score.

Alongside xG sits PPDA — passes allowed per defensive action. The lower the figure, the higher and more aggressive the pressing. A team with a PPDA of 6.5 means that for roughly every six opposition passes, they engage in a duel once. Combine PPDA with the xG of transition moments and an analyst can reconstruct almost an entire tactical system without watching the full ninety minutes.

Almost. Because Mbappé's two surges in Kazan were not inside any model before they happened. The model only prices the shot afterwards, once he has gone past the Argentina defence and the ball is in the net. That shot is valued at 0.09 or 0.12, depending on the model. What produced it — 37.8 km/h in the 64th minute of a World Cup quarter-final — the model can only describe, not explain.

I once sat beside a young analyst in a meeting in Shenzhen. He opened a match dashboard, pointed at the xG column and said: this team deserved to win. I asked him: the left winger that day, do you remember him? He went quiet. The dashboard has no cell for the names of the men who run without the ball.

Data counts what happened; it cannot measure what wants to happen.

The books the stands never see

A supporter reads that a player has moved for eighty million euros and thinks of one number. The club accountant thinks of another. The transfer fee is amortised, spread evenly across the years of the contract. A five-year deal worth eighty million euros costs only sixteen million euros a year in the accounts. That is why clubs love long contracts: the same money, a lighter annual accounting burden.

Alongside amortisation sits the sell-on clause. A club that sells a young player for a modest fee usually retains ten to twenty per cent of the value of the next transfer. For long-established academies this is no small sum: a graduate sold for sixty million euros can bring the old training ground a figure close to its entire annual operating budget.

Then there is financial fair play. UEFA applied FFP from the 2026/12 season, obliging clubs in European competition to balance income and spending across a multi-year cycle. England has its own version, the Profit and Sustainability Rules, whose strongest sanction is a points deduction. In November 2026 Everton were docked ten points; in February 2026, on appeal, the penalty was reduced to six. In March 2026 Nottingham Forest received a four-point deduction. Those numbers never appear on a scoreboard, yet they move an entire season's positions.

In much governance literature, the wage-to-revenue ratio is treated as the simplest measure of health, with a reference threshold usually quoted around seventy per cent. Above it, a club can still win the title, but it is spending its own future.

And at the bottom of the financial system sits fear. January is when struggling clubs pay above a player's true value, because they are not buying to build but to survive. I call it the panic premium. Every winter window I witness at least one such deal: the fee pushed up by half again, a four-and-a-half-year contract, and the man signing fully aware he is overpaying.

A club can win on the pitch and lose in the books in the same week.

Goals that arrive earlier than the model

No metric is more abused than xG when results are discussed. A team scoring three from 1.1 xG is lucky. A team scoring one from 2.8 xG is unlucky. In the short run, the gap between actual goals and xG is just noise. In the long run it is fate: almost every side that over-converts xG across half a season returns to the mean.

But there is a paradox the models handle clumsily. If a team keeps beating xG across two or three seasons, people begin to speak of identity, of the quality of individuals who score at decisive moments. At that point xG stops being a judge and becomes a mirror held backwards: it is the model that has not been recalibrated. I have watched enough matches to believe that some teams live on what data calls luck but which is really the skill of applying pressure at exactly the right instant.

Public opinion is less patient. Pressure on a manager is measured by the density of articles in the ten days after a defeat, by the odds on the sack race, by the number of banners in the stands. In China, where I live and work, a coach can be judged over three matches: win and he is a genius, draw and he is a caretaker, lose and he is a relic. The opinion cycle is compressed so tightly that no tactical process has time to form.

The opinion cycle always runs faster than the tactical cycle, and that is why most football projects die before they grow up.

Who stands on which tier

Analysing a club without placing it inside its league context is meaningless. The same squad, the same manager — in one league they are title contenders, in another they are scrappers. The measure is not raw points but the gap between expectation and reality.

Three resources must be compared. First, squad value on the market. Second, financial power, expressed through the wage bill and net spending capacity. Third, academy productivity: how many graduates reach the first team and stay there. These three resources usually rank clubs in a very similar order, and that overlap is exactly what makes modern football more predictable than people assume.

From my own observation across many seasons, at most academies only about one graduate in ten remains in the first team beyond three seasons. That figure says nothing about the quality of the academy; it says that developing players and using them are two different jobs, and very few clubs do both well.

The most worrying signal in that picture is not a low position but a talent flow running backwards. When two or three pillars at their peak attract stronger clubs, the club must choose: sell early and reinvest, or hold on and trade for one more season. Neither choice is free.

The grey zones of the law

Football's legal framework runs on clauses supporters rarely hear named. Third-party ownership — a structure in which investment funds hold a player's economic rights — was banned by FIFA from 1 May 2026, after years of argument about conflicts of interest. Approaching a contracted player illegally, known as tapping-up, is an offence that costs clubs money and reputation alike. Article 19 of FIFA's transfer regulations protects players who are not yet adults, permitting only a narrow set of exceptions.

Above all sits multi-club ownership. One group can hold stakes in several clubs across several countries, and every time two of them meet in European competition the organisers must convene a hearing on eligibility. For people in my trade, this is ground on which every analysis must carry a footnote: the law has not kept up with the money.

And since refereeing technology entered the stadium, controversy has not disappeared. It has moved. Once, eleven men on the pitch argued with one referee. Now millions argue with a line drawn by software, and with the very definition of phrases like clear impact or the position of the arm. I have sat inside the review room at a domestic league. Three minutes for one incident. Those three minutes did not calm the stands; they merely taught them that football's law contains grey cells, and machinery does not make grey cells vanish.

The generational staircase

The dressing room is the hardest dimension to analyse and the most decisive. No metric captures a thirty-four-year-old captain slowly losing his voice, or a twenty-one-year-old who has just signed a new contract and begun to ask for more. In this corner, what I learned from documentary film-making is worth more than any spreadsheet: people tell you the truth once the camera is off.

There is a classic phenomenon analysts call the new-manager bounce. After a change at the top, results often improve for six to eight matches, sometimes sharply. But that effect comes mostly from things that never appear on a tactics board: tension released, roles redefined, a few frozen players suddenly allowed to play again. Once the emotional baseline returns to normal, the team reverts to its true level — and that is the moment public opinion turns on the new man.

Generational transition is a staircase, not a doorway. Nobody steps down all at once. Men stand on different steps for two or three seasons together, and during that stretch the club must both collect points and raise people. Very few manage both at the same time.

The virus named after international weeks

The risk profile of a club is probably its most underrated section. People treat injuries as accidents. But for teams that supply many internationals, injury is a recurring, forecastable cost. The press half-jokingly calls it the FIFA virus. A player flies twelve thousand kilometres, plays twice in five days, and returns with a tight hamstring and a starting berth in a crucial weekend fixture.

Then there is the calendar. A side competing in the domestic cup, the league and a continental tournament can play more than fifty matches in a season, before commercial pre-season friendlies are counted. No model predicts precisely who will break down in match forty-two, but everyone knows someone will. For a manager, load management is an art balanced between mathematics and instinct.

A third risk is rarely discussed: tactical staleness. A system that once confused opponents loses its potency after roughly eighteen months, once rival analysis departments have enough data to neutralise it. The irony is that success itself causes the staleness: winners change little, and those who do not change get decoded.

Rumours, and the men who sell them

No market manipulates information as heavily as the transfer market. A rumour may be released by an agent to apply pressure on a parent club, or to raise a price, or simply to remind the world that his client still exists. Practitioners must grade sources by reliability, from journalists with direct club relationships to outlets that merely aggregate other people's work.

The expectation cycle in football runs almost mechanically. A young player who shines for three matches is called a phenomenon. By his tenth match, if he is still ordinary, doubt sets in. By his twentieth, the very people who praised him are the first to say he was overhyped. I have watched enough of these cycles to draw one conclusion: the content of a story matters less than its position in the cycle.

The flow running up to the national team

In the end, everything in football flows in one direction: from academy to first team, from first team to national team, from national team to the rights and commercial markets. A good academy produces players, and also produces value for the whole system. A national team that performs well across a summer can lift the television value of its domestic league substantially for years afterwards.

That is why major tournaments are always moments of compressed emotion. The fervour of supporters, the pressure of federations, the ambitions of sponsors — all of it funnels into one month. And in that month every forecasting model turns fragile, because the data sample is too small and the emotional variable too large. A missed penalty in the 88th minute has less to do with technique than people think; it has to do with a player knowing that forty thousand people are in the ground and several million more are at home.

Drawing on my experience of watching matches across eight World Cups and eight Olympic Games, I can say that most of the moments remembered longest in the history of major tournaments were never forecast. They came from a player the model had not yet ranked, at a moment nobody had prepared for.

Where every framework stays empty

Here I must state what I believe is the blind spot of the collective memory about data football. The wind of Kazan taught me that youth never rolls according to schedule. And that forty-page file of empty cells taught me something else: a framework is never the explanation. It is a waiting room. The match is the one who opens the door.

Modern football, pressured by data, has become alarmingly uniform. The inverted winger is now the default in every academy, because the models show that reverse-footed players touch the ball more often in dangerous zones and generate higher xG. Arjen Robben is the most copied template of this generation. Meanwhile the natural-footed man hugging the touchline, running straight into the channel and crossing — a player like Jesús Navas — has been filed away into history. The number of crosses from wide areas falls season after season, and few treat that as a loss. I treat it as a loss: we have optimised one half of the pitch's width and forgotten the other.

That is why I do not believe statistics can replace observation. They are not opposed; they sit on different floors. Data answers the question of what happened. The match, filmed the way I have filmed it for forty years, answers a different question: why that exact instant made it impossible to look away.

An analysis piece, in my method, should keep only one number and hold it under a new light. My number today is 37.8 km/h, the average speed of two surges in the 64th and 68th minutes in Kazan. It does not say Mbappé scored twice. It says only that somebody, at nineteen, ran faster than everyone around him on the one day his team needed it most.

Freezing the memory

In the summer of 2026, the stadium in Shenzhen held forty-five thousand seats, and across the two hundred and forty minutes I filmed there, not one spectator was present. I had grass, empty chairs, and the guard's whistle sounding in waves. In the script I wrote one scene: a ball rolling slowly through a goalmouth with nobody in it. I called it the ritual without spectators. The management asked me to add canned cheering to the film. I nodded, and quietly did not.

An analyst could reconstruct that season in thousands of lines of data and still never touch the only thing worth saying: those empty rows of seats sat there like one long rest. No model can score a rest.

Every model learns from the past; only a child's legs know how to describe the future. So if anyone asks me what the nine-dimension framework is for, I will say: to put a moment in the right place. And the moment itself must be watched with the eyes, heard with the ears, and sometimes it demands abandoning an approved script to follow a nineteen-year-old boy running. In a world with no spectators, football is left with nothing but breath and truth.