Data / entry 11 / May 25, 2026 / 6 min read

Pep Guardiola's last season at Man City, at the underlying-numbers level

The possession-control era ended with City hitting five-year lows in possession and five-year highs in fast breaks. Here's what the pressing data, the attacking model, and the European record actually say about how his final year was built.

The 10-year baseline

• 6 Premier League titles
• 3 FA Cups
• 4 EFL / Carabao Cups
• 1 UEFA Champions League (2023)

The style actually changed in year 10, and the data shows it directly

This is the most counterintuitive data point of his final season. Guardiola built his entire managerial identity on positional possession control, at Barcelona, Bayern, and City. In his last year in charge, City recorded their lowest possession share in five seasons while simultaneously hitting five-year highs in through-balls, fast breaks, and take-ons attempted. The signings that defined the season, Gianluigi Donnarumma, Jeremy Doku, Tijjani Reijnders, were recruited toward directness and transition speed, not control. The manager most associated with slow-building possession spent his final season actively moving away from it.

Pressing intensity across the final two seasons

2024/25 marked the first time under Guardiola that City's pressing intensity visibly collapsed, a PPDA of 12.29, the second-highest among the league's top sides that season, in a year they won zero trophies. The final season partially reversed that: PPDA dropped back to 10.6, closer to Chelsea's league-leading 10.12 than to the passive number from the year before. The pressing data backs up the eye test: this was a more re-engaged, higher-intensity City side than the one that fell apart the year before, even if it wasn't quite the pressing monster of the peak years.

The attacking model: clinical finishing above the expected numbers

This is the mirror image of what Arsenal's defense did this season. Arsenal's actual goals conceded landed almost exactly on their xG model, evidence of a repeatable process. City's attack ran ahead of its own model, scoring more than the shot quality alone predicted, evidence of individual finishing quality, mainly Erling Haaland, doing work the underlying process wasn't fully accounting for. Different clubs, different route to the same conclusion: City's title challenge leaned more on individual brilliance converting chances, Arsenal's leaned more on a system suppressing them.

Domestic mastery, continental disappointment

The final season closes out the exact pattern that's defined City's European record under Guardiola: total domestic control, a single Champions League title in ten years despite being favorites most seasons, and a final-year exit at the Round of 16 with a 3-goal defeat at Real Madrid as the low point. The domestic cup double, going unbeaten through every match in both competitions, was genuinely unprecedented.

Points per game: career vs. final year

Putting the model together

The headline number, second place, undersells what the underlying data actually shows. This wasn't a manager coasting to a quiet finish. The pressing numbers rebounded from the previous season's collapse. The recruitment and the through-ball, fast-break, and take-on data all point to a genuine tactical evolution away from his own signature style, toward more directness, in his very last year. The attacking model shows a team that scored above its underlying shot quality, meaning the goals were real even when the process behind them was closer to normal. And the trophy count, two domestic cups won without dropping a single match in either competition, is a level of control the underlying numbers back up completely. The gap to Arsenal wasn't a story of decline. It was a story of a league that got harder to separate from at the very top.