PPDA 9.2 and the Price of Pressure: What the Transfer Window Misprices

January in Bergamo, a night cold enough to cut skin. I sat in the temporary stand of the Gewiss Stadium, a battered notebook in my hand, my eyes fixed on the data column that only I cared about: the number of passes the opponent completed before losing the ball. When the match ended, the number on the screen was 8.7. I wrote it down and underlined it twice. Eight years later, when Gasperini's protege was sold to a big English club for a fee the press called 'a hundred-million-euro windfall,' the outlet I contributed to cited only his goals. Nobody mentioned the 8.7. Nobody mentioned how the system had produced him. The name was sold, but the machine stayed behind, and it kept turning.

That is why I am writing this. The January transfer window is at its hottest, and people are pricing players with the numbers easiest to see. Goals, assists, betting odds, social-media fame. Meanwhile the thing that actually creates modern football value — the ability to press high, to force opponents into mistakes, to win the ball in the final third — gets pushed to the bottom of the balance sheet. I am not saying the people in the industry are stupid. I am saying the market pays for the outcome, not the process, and in a decade when the process has become the main battlefield, that is a systemic mispricing.

Gasperini arrived at Atalanta in 2026. I was eighteen, a sports-management student in Beijing, and I spent three months of that season tearing apart the full data of all 38 Serie A rounds. I remember the first afternoon I saw it. The Serie A table ranked by PPDA — the number of passes a team allows its opponent before winning the ball back. Atalanta's average that season was 9.2, the lowest in the league, level with Juventus, the table leaders. The press called Atalanta a 'mid-table phenomenon,' a provincial club with no ambition. I looked at the data table and saw the opposite. A team that does not let its opponent string together more than ten passes before losing the ball is not a phenomenon. It is a machine. I wrote a prediction that Atalanta would qualify for the Champions League. The piece drew two hundred thousand reads. That season they finished fourth.

PPDA is not a pretty number to print on a shirt. It is a record of a collective's decision-making speed. When a team presses, they do not merely run. They run according to a structure, a signalling system, a script rehearsed to the point of reflex. Gasperini did not train his players to run more. He trained them to run at the right time, in the right direction, against the right man. That is why Atalanta's PPDA sat strangely low while their running volume was not even top of the league. They did not run for the sake of running. They ran to lock.

I once spent an entire evening rewatching footage of three Atalanta matches against the event-data table. Something strange appeared. Their number of ball recoveries in the opponent's final third was not much higher than mid-table teams. But the number of times they forced the opponent to pass backward or out to the touchline was half again as high. They did not need to steal the ball. They made the opponent hand it over. This is the difference between pressing as an action and pressing as a system. Anyone can see the action. You have to read the system to see it.

And the transfer market always buys the action and sells the system.

When Atalanta sells a striker, the big papers count his goals at Bergamo, multiply by a season coefficient, and arrive at a fee. But where did those goals come from? They came from opponents losing the ball in midfield because they were squeezed, and the ball being played into the box in three passes. When that player moves to a new club, without that structure, the goals vanish. Not because he got worse. Because the machine stayed in Bergamo.

I call this the transfer paradox of the pressing system. You buy a player. You cannot buy the system that produced him. Within eighteen months, his value will be repriced by fans, and someone will say: 'Turns out he was just a product of the system.' They say it as criticism. But it is praise for the system and criticism for the buyer.

Data does not lie, but it still keeps a corner of truth to itself. Goals are truth. A ball recovery in the opponent's third is also truth. It is just that the second truth does not get broadcast that night.

---

In mid-January, I had a phone call with a friend who works as a scout at an Eastern European club. He asked me how to read a midfielder his team was chasing. I asked him back: does your team need someone to win the ball or someone to keep it? He went silent for a few seconds. It turned out the whole board had never asked that question. They were only looking at the player list praised on ESPN forums.

That moment made me realise something that had haunted me for years: people are buying players by heat map, and the heat map has become football's new divination. A red stain spreading across a pitch diagram, a hot streak on the right wing, a hotspot in midfield, and everyone nods knowingly. But a heat map does not tell you whether a player ran at the right moment or too late. It does not tell you whether his run opened space for a teammate or merely created the illusion of effort. A heat map only draws where a man has been, not why he was there.

In the scouting world, we distinguish two kinds of people: those with energy and those with impact. Energy is total distance, number of duels, number of accelerations. Impact is the degree to which those actions change the state of the match. What use is a player who runs twelve kilometres and opens no useful space? What use is a player who runs nine but in that nine makes six movements that force defenders out of position? The answer lies in the zone event data has not yet covered. And that is precisely the zone the transfer market underprices most.

I once ran a small comparison during an internship at a television station in Belgrade back in 2026. I picked out twenty midfielders most rumoured for transfers in Europe that summer. For each, I recorded two numbers: their rumoured valuation on news sites, and their inverse PPDA — the pressure that player contributes when his team is without the ball. Not one player with an outstanding inverse PPDA had a high rumoured valuation. All the names ranked highest on the news sites had pressing numbers at average or below. The conclusion was simple: the market pays for scorers and creators, not for the men who make opponents err.

A year later, at the 2026 World Cup in Russia, I met that lesson again in a different form.

When Croatia reached the quarter-finals, I was contributing to an online football magazine. I analysed their data and found something that made me sit down. Croatia's average xG was just 1.1 per match — suspiciously low for a side reaching the knockout stage. On paper, they should have gone out in the round of sixteen, against Denmark. But they won. They won because Subasic saved five of twelve penalties he faced, a rate of 41.7%, more than twice the average of goalkeepers at that tournament. They beat Russia in a shootout. They beat England in extra time. They reached the final.

I wrote a controversial piece: Croatia did not need to control the ball. They needed to drag the match into the penalty shootout — their kingdom. The piece angered many. But when they reached the final, I understood something I had never fully grasped: xG measures chances, not the ability to convert chances under extreme psychological conditions. That is the limit of xG in knockout matches, when psychology, experience and set pieces outweigh any model.

Tactics are the winners' account of events; data is the losers' original manuscript. Croatia lost the final to France. People wrote that Croatia 'lacked class.' But look at the data: they shot more, controlled more, and lost only to moments the model could not predict — a disputed penalty, an own goal, a shot from outside the box. The winners' account tells of France's strength. The losers' original manuscript tells of a team that went all the way by a road no one believed traversable.

The Croatia story taught me two things about the transfer window, and both matter more than people think.

First: a player's value is not entirely in the standard metrics. Subasic was an average goalkeeper if you looked only at goals conceded. But in a shootout, he was the third-best in the world. If you are a club preparing to buy a keeper for knockout ties, you should pay for that edge. In reality, nobody does. Because no column on the standard news sites measures shootout psychology.

Second: a player's value can be hidden by the very system that produced him. Croatia played an aggressive defensive block, and Modric shone. But there was another midfielder — Brozovic — running himself to exhaustion in every extra time to cover for Modric. In the transfer windows that followed, Modric moved to big clubs for enormous fees, while Brozovic took a humbler path. But if you have a system that needs a cover man, Brozovic's price to you should be above his market price. He is the classic undervalued modern player: the man doing the work nobody sees.

---

In 2026, when the whole world shut, I wrote my master's thesis on football without crowds. I compared 142 Bundesliga matches with spectators against 106 after the lockdown of the 2026-20 season. One number kept me up for nights: home-win rate fell from 43% to 32%. That is the macro truth. But when I isolated Dortmund — a side with a PPDA of 8.1, one of the lowest pressing numbers in Europe at the time — their home-win rate dropped from 67% to 38%. A fall more than three times the league average.

What does this mean for the transfer market?

It means a large part of a team's value comes from outside its players. It comes from the presence of the crowd, from the roar, from psychological pressure on referees, from the rhythm disruption of away teams forced to play in a cauldron. When you buy a Dortmund player at a high fee because he played well in the Bundesliga, you are buying part of a value that came from the stands. When that player moves to a team without those stands, that part evaporates.

This truth led me to write my first principle, one I still apply in every transfer analysis: every data table is a scripture, but once you have read it you must know how to let it go. If you cannot let go, you will buy a player merely because he happens to be in the right system at the right moment.

I learned this after a personal wound. I wrote a forty-page draft on the impact of empty stadiums and delayed it for weeks because I wanted to check the referee variable further. While the draft was waiting for me to finish, a German analyst published matching results. I realised that absolute perfection is the enemy of timeliness. In data analysis, as in transfers, you have to decide inside a short window, with incomplete information, and live with the consequences. The analyst and the buyer share the same pain: both are driven by data, and neither ever has enough.

---

Back to the current January window. I sit in Beijing, tracking hundreds of lines a day. On the forums, dozens of names are mentioned. I keep my own tracker with four criteria: original source, age, importance within the system, and off-ball pressure metric. The names with a high fourth metric but a low valuation are the ones I circle in red.

I sell players by minutes run, not by fame on TV. I said this to a colleague after finishing an analysis of a deal between Liverpool and a Serie A side. I do not deny that fame has value. But fame is built on moments, and a team's value is built on minutes. In a ninety-minute match, goals take a few seconds. Everything left — running, pressing, covering, opening space — is durable value.

What worries me in this window is that the glow of a few players is eclipsing a bigger picture. People talk of a young striker valued at three figures of millions after one breakout season. But that season, how many times did he win the ball in the opponent's final third? If that number is low, his scoring at the old club does not guarantee scoring at the new one. Goals inside a non-pressing system are goals in a match where the opponent was squeezed into defending. Those goals are beautiful, but they were born in a condition the new system may not reproduce.

This is the counter-intuitive point I want to raise.

People believe a goal is the final unit of value in football. A striker with twenty goals is a good striker. One with ten is average. People believe goal count is proportional to the fee one should pay.

But transfer data over a decade shows a suspicious correlation. Goal count correlates strongly with instant valuation, but does not correlate strongly with value delivered to the buying club over the following thirty months. This is a correlation, not a causal relation, and I must say that plainly so I do not fall into the trap I always warn others about: correlation is not causation. But a correlation that is strong and repeats across seasons, leagues and positions is enough to raise a serious question about how the market misprices.

If a goal has value, one must say it was born in a specific tactical context. And if that context is not recreated at the new club, the value of the goal must be discounted. That discount does not happen. The buyer pays the price of the old context but receives a new, unfamiliar one.

---

I once told a friend who works as a player agent a personal story I rarely bring up.

When I was twenty, I worked at a small television station in Belgrade, running a sports-data segment. My job was preparing data tables for commentators before each match. One time, I prepared a detailed breakdown of a young striker and asked the main commentator to talk about his pressing ability. He refused. He said: 'The audience wants to hear about goals.' I replied: 'If the audience does not hear about pressure, how will they know where the goals come from?' He went quiet. In that broadcast I got thirty seconds. In those thirty seconds I talked about the team's PPDA. That evening I got two messages from viewers asking what PPDA was.

That was the start of the work I do today: translating data into human language, not to teach people, but to give them a different view of the match they just watched.

In this transfer window, I think of that Belgrade evening every time I read a rumour. There is a type of rumour I call the 'self-made rumour.' It starts from an anonymous social-media account, is cited by a mid-tier site, then quoted by a major paper with the phrase 'according to a source close to the situation.' After three steps like that, the rumour is no longer a rumour. It becomes a social fact. But it is still a rumour.

When an agent inflates a price, the player's real value does not rise. Only the buyer's expectation is lifted. And when expectation is lifted while value stays flat, a gap begins to accumulate. That gap is a debt a club will pay in eighteen months, when fans start asking why the player is not worth the fee.

I hold no grudge against agents. They do their job. But scouts and buyers need a filter. That filter has three layers.

Layer one: the source of the rumour. If the source is not a reputable local journalist where the player plays, reliability falls sharply. Transfer rumours in England usually come from reporters with networks in Europe. Rumours in Spain usually come from club-friendly press. Distinguishing the two matters more than people think.

Layer two: motive. If the rumour appears exactly when a player needs a new contract, the motive is usually a raise. If it appears exactly when a club needs to sell to balance its books, the motive is usually to create an assumption. The buyer needs to understand who needs what behind a rumour.

Layer three: data. This is the layer I care about most. A player who appears in a rumour but has a high off-ball pressure metric, a superior progressive-passing metric, and a good metric for involvement in space-opening situations makes that rumour more valuable. If the opposite, the rumour is just trying to sell you a name.

---

Let me return to Atalanta once more, because it is the most complete lesson I have.

This club built a machine over nearly a decade without needing a single number-one star. They sold their best names, every season, and still stayed in the Champions League group. They do not change blood. They change parts. When I look at that machine, I see a model people call the 'Atalanta model,' but I do not like calling it a model. I call it an architecture. Architecture differs from model in that architecture has walls, a foundation, rooms. A model is only a blueprint.

Atalanta is a baptism, pressing is scripture, and I am a monk under the vault of xG. I wrote that line years ago, when people still called me 'the madman defending a provincial club.' They do not call me that now. But I keep the line in my notebook, because it reminds me that data has no function of worshipping strong teams. Data only describes the truth.

And the truth of this window, when I look at it through data eyes, does not lie in the expensive names. It lies in young players with high pressing metrics in smaller leagues. It lies in covering midfielders nobody mentions. It lies in goalkeepers with good shootout save metrics priced low. It lies in players whose off-ball running is recorded in the right system but undervalued because they do not score.

I am not saying the expensive names do not deserve their fame. Some of them are genuinely excellent. But the market prices them at a level not matched by the data they generate, but matched by the social story built around them. And every time price and value separate, some club pays the bill at the end of the cycle.

---

There is one thing I always tell my students when they ask how to analyse transfers.

I say: do not be the first to believe, and do not be the last. Be the third. The third has time to apply the filter, time to cross-check data, and time to distinguish the name from the value. The third is not swept along by day-one noise, nor does he still have the chance to buy at the best price. But the third is the only one who buys at the right price.

In this window, I set myself one rule: for every big rumour, I spend thirty minutes reading the player's off-ball pressure metric and comparing it with the equivalent position in similar leagues. If the gap exceeds two standard deviations, I note it down. And when I note it down, I always write the same line at the end: this is data, not prophecy.

The map is not the territory. That is the line I remind myself of every day. A data map only draws what we have measured. The territory is where the match happens, with variables nobody can measure: psychology, luck, the moment. No model captures the moment. That is why I still go to stadiums, sit in cold stands, hold a notebook, and record the numbers the screen does not record.

---

Tonight in Bergamo, the air turns cold again. I look at the data table of a match just finished. The home team's PPDA is 9.4. The score line is three. Of those three goals, two came from ball recoveries in the opponent's half. No bulletin tonight will mention it. But I know some scout somewhere is quietly noting that 9.4, and in a few months a name will appear in the news with a fee people cannot understand.

The question for the next transfer window is not who will arrive, but who is making their team better in the seconds without the ball. If you can answer that with data, you will buy a player at his true value. If you answer it with feeling, you will buy a name, and then relearn an old lesson: a player's value is not where people look, but where people glance.