Atalanta, a PPDA of 9.2 and the Repricing of a Mid-Table Club

In August 2026, Atalanta sold Rasmus Højlund to Manchester United. The base fee was reported at around 72 million euros, with add-ons that could push the total near 85 million. Højlund was 20 years old and had just completed his first full Serie A season: 32 appearances, nine goals.

I reopened my spreadsheet the night the deal was announced. Nine Serie A goals cannot justify that fee on any goal-based measure. But the market does not buy goals. It buys a profile: a 1.91m striker willing to attack the first pressing zone, to run behind a defensive line, and fast enough that a Serie A centre-back cannot turn in time. The goals are a by-product.

The transfer market does not price players. It prices the system that produced them. In Bergamo, that system has a name and a data trail.

A provincial club and a metric nobody read

In 2026 I was 18, a sports management student in Beijing. Serie A was told through two stories back then: Juventus dominating, and Maurizio Sarri's Napoli playing the most admired football in Europe. Atalanta appeared as a margin note — a club from the Bergamo area, modest revenue, a transfer budget big enough only to buy what the giants had passed over.

Gian Piero Gasperini took charge in the summer of 2026. In his first season he took the team to fourth. The press called it a one-season wonder.

I spent three months processing data from all 38 matchdays, calculating PPDA — the number of passes a team allows its opponent per defensive action. The lower the PPDA, the earlier and more aggressively a team presses.

Atalanta averaged 9.2 in 2026-17, the lowest in the league. They forced opponents into 11.4 turnovers per match, level with Juventus. The striking part was elsewhere: that squad contained no player who had ever been valued above 20 million euros.

I wrote a piece predicting Atalanta would hold a top-four place, with the league-wide PPDA table attached. It drew 200,000 reads. When Atalanta finished fourth, an invitation to write deep analysis for the 2026 World Cup followed.

Atalanta was the baptism, pressing was the scripture, and I was the monk under the xG dome.

The machine sits in midfield

Gasperini built a back three with wing-backs pushed very high. In possession, one centre-back steps into midfield as a second pivot; the two forwards split the opponent's escape routes; the midfield squeezes inside. On losing the ball, the whole block surges forward.

That system manufactures a rare asset class: repeatable, measurable actions. Recoveries in the opponent's third. Progressive carries. Third-man runs. These are the lines a scout can write into a report and a sporting director can pay for.

Based on my experience tracking these matches, one detail rarely gets mentioned: most of Atalanta's defensive value does not live in tackles. It lives in a block so high that opponents are forced into long passes. A long pass is a pass with a higher turnover probability. PPDA does not measure tackles. It measures the opponent's patience, and Atalanta was the fastest team in Serie A at taking that patience away.

The sales ledger

In the summer of 2026, three names left Bergamo almost simultaneously: Andrea Conti to Milan, Franck Kessié to Milan on loan with an obligation to buy, and Alessandro Bastoni to Inter for a reported 31 million euros before being loaned back for another season.

In January 2026, Dejan Kulusevski joined Juventus for around 35 million euros. That same year Amad Diallo moved to Manchester United in a deal worth up to 40 million. In 2026, Cristian Romero went to Tottenham for around 50 million. In 2026, Højlund went to Manchester United. In August 2026, Teun Koopmeiners joined Juventus for a base fee of 54.7 million euros plus add-ons.

Read that list alone and the easy conclusion is that Atalanta is a transit hub. The other direction gets far less attention.

In 2026, Atalanta bought Ademola Lookman from RB Leipzig for around nine million euros. Lookman was 24 and had drifted through Everton, Fulham and Leicester without settling. On 22 May 2026, in Dublin, Lookman scored a hat-trick in the Europa League final as Atalanta beat Bayer Leverkusen 3-0, ending the German side's 51-match unbeaten run.

That is the entire model in one name: buy a profile the market has undervalued, place it in a system that amplifies exactly its qualities, and let the value surface.

Alongside the selling, Atalanta kept buying: Mateo Retegui from Genoa in 2026 for around 22 million euros plus add-ons, Ederson from Salernitana, Marten de Roon retained for years as the pivot. The wage bill stayed well below Juventus, Inter and Milan, while net transfer income was positive in almost every window.

Results confirm the model. Between 2026-17 and 2026-24, Atalanta finished in the Serie A top four five times. They won the Europa League in 2026-24. A club that once had nothing to sell now sells almost everything except its coach.

The contrarian angle: PPDA is a map, not the territory

The metric has one large blind spot, and that blind spot is why I never treat PPDA as a conclusion.

PPDA measures a collective. It does not separate the centre-back who steps up to set the tempo from the wing-back who runs 12 kilometres a match, nor the forward who presses the ball carrier from the one who cuts the passing lane. When the team wins, the metric belongs to everyone. When the team loses, it stays silent about everyone.

In too many analyses I read, the heat map has replaced watching the match. A red patch in the right channel gets read as a wing-back pushing high, when most of the touches there actually belong to an opposition winger being funnelled towards the touchline. A heat map answers where a player was. It does not answer what he was doing, for whom, and under what pressure.

There is a structural risk too. A fee of 72 million euros for Højlund shows the market has learned to price the Atalanta label. Once every scout knows how well this club develops and optimises players, the arbitrage narrows. A trade only lasts as long as few people know about it.

And some products do not repeat. Kulusevski left Bergamo for 35 million euros but needed several seasons to rediscover form that matched the fee. Amad Diallo needed a few loans before earning regular minutes at Manchester United. The system that made them did not travel with them.

The biggest risk is human, not analytical. Gasperini has led Atalanta since 2026. Every metric I can calculate is the metric of a collective coached by one man for nearly a decade. When he leaves, the spreadsheet does not transfer with him.

Tactics are the winner's account of the match; data is the loser's first draft.

Signals for the next cycle

What matters in this transfer window is Atalanta's capacity to keep players, not the list of names they could still sell.

Three signals. The wage bill: every top-four finish raises the pressure to pay key players more, and this is the one variable the model cannot optimise with data. The next cohort: Giorgio Scalvini, Marco Carnesecchi and younger names will show whether the system still produces talent or is simply selling accumulated stock. And Champions League revenue under the new format, money that could extend the cycle for a few more seasons or turn Atalanta into just another ordinary club.

Data does not lie, but it still keeps a corner of the truth to itself. Atalanta's corner is this: a system can create value, but only a club can keep that value around long enough to win with it. The question for next season is not how much they sell, but who they dare to keep.