Four XIs built exclusively from the metrics in FIFA's official reports: the defensive team, the attacking team, the balanced team and the breakout team
Every World Cup ends with a "team of the tournament" — and it is almost always a list of top scorers and champions. Here we ran the opposite exercise: we built four XIs using only the process metrics extracted from FIFA's official Post-Match Summary Reports — passes, passing-lane offers, defensive actions, sprints, high-intensity runs, distance and speed — without looking at goals, assists or reputation. The result features Rodri and Pedri, but also three Paraguayans, an Egyptian almost nobody followed, and one resounding absence.
A note on method: we considered the 220 outfield players with at least 300 minutes and the goalkeepers with 3+ full matches in the 104-game database. All volume metrics were normalised per 90 minutes and converted into standardised scores within each position (defenders compared with defenders, and so on). Each XI uses a different weighting of those scores, in a 1-4-3-3 shape. Importantly: FIFA's PDFs do not include individual goals, assists or xG — and the database, on principle, does not invent what was not extracted. This is, deliberately, a team chosen for its work, not for the shine of its finishing. "Offers" follow FIFA's offering methodology: how many times a player made himself available as a passing lane, and how many times he received. Aerial duels and box-specific defending are not separable in the database; the defensive-actions metric aggregates pressures, tackles and interventions. Positions come from the reports' own line-ups.
The balanced XI — the data's team of the tournament
Kobel; Spence, Puerta, Castagne, Douglas Santos; Ounahi, Pedri, Paredes; Doué, Olise, Plata.
In goal, Gregor Kobel (Switzerland) was the only elite goalkeeper across all four indices at once: save efficiency (92 on a 0–100 scale), distribution (87), volume faced (86) and performance under pressure (85). Specialists did more in one department; nobody delivered the full set.
In defence, Djed Spence (England) was the tournament's most statistically dominant full-back: 119 defensive actions per 90 minutes (double the positional median), 59.6 sprints per 90 and a 36.2 km/h top speed. Gustavo Puerta (Colombia, listed as a defender in FIFA's line-ups) combined 88 defensive actions with 58.5 passes per 90 at 91% accuracy. Timothy Castagne (Belgium) was the engine: 64 sprints and 161 high-intensity runs per 90. Douglas Santos (Brazil) closed the left side with 93.1% passing accuracy and 91 defensive actions per 90 — a rare combination of security and volume.
In midfield, the World Cup had an unlikely statistical owner: Azzedine Ounahi (Morocco), the most complete midfielder in the database — 75 passes, 77.5 offers made, 37.9 offers received and 123.5 defensive actions, all per 90 minutes, all among the best in the position. Alongside him, Pedri (Spain), the only one who came close to that balance in the champion side (82 passes and 120 defensive actions per 90), and Leandro Paredes (Argentina), the highest passing volume of the entire tournament among qualified players: 124.9 passes per 90 at 94.2% accuracy.
Up front, Désiré Doué (France) is the tournament's name according to the data: 128.7 defensive actions per 90 (more than any midfielder in this XI), 65.4 sprints and 51.6 offers — the total forward. Michael Olise (France) added creative volume (58 passes and 50.8 offers per 90) to 90 defensive actions across 647 minutes. And Gonzalo Plata (Ecuador) was the workhorse winger: 108.7 defensive actions and 52.3 sprints per 90.
The resounding absence: Kylian Mbappé. With 684 minutes and the highest recorded speed among the eligible (37.6 km/h), the Frenchman misses the balanced XI — and the explanation lies exactly in what this exercise measures: 56.8 defensive actions and 32.9 sprints per 90, roughly half of what Doué and Plata produced. It is the perfect demonstration of the subtitle: a team picked on goals and fame would select him blindfolded; a team picked on process does not.
The defensive XI — the team nobody gets through
Gill; Alonso, Spence, Cáceres, Castagne; Berg, Attia, João Neves; Doué, David, Plata.
This XI has one dominant colour: Paraguay's red and white. Orlando Gill in goal (91 efficiency with the tournament's maximum pressure-faced index: 100), Junior Alonso (136.6 defensive actions per 90, the highest figure among all defenders) and Juan José Cáceres (103.3) in the line. It is the name-by-name translation of what the previous article in this series showed in aggregate: Paraguay was the most reactive team of the World Cup — 28% possession, a transition record — and lived off its wall. In midfield, Patrick Berg (Norway) produced 194.9 high-intensity runs per 90 (no qualified midfielder ran more at intensity), Marawan Attia (Egypt) added 117.5 defensive actions, and João Neves (Portugal) combined 125.9 defensive actions with 94.6% passing accuracy. Up front, the pressing trio: Doué, Jonathan David (Canada, 113.5 defensive actions per 90 — a centre-forward who defends like a holding midfielder) and Plata.
The attacking XI — the team that suffocates with the ball
Maignan; Muñoz, Tah, Danilo, Puerta; Paredes, Rodri, Ounahi; Olise, Luis Díaz, Doué.
Mike Maignan takes the gloves for the best distribution among goalkeepers (index 91). The back line is made of builders: Jonathan Tah (Germany) played 103.8 passes per 90 at 97.1% accuracy — the World Cup's highest precision at high volume —, Danilo (Brazil) and Daniel Muñoz (Colombia), who offered himself as a passing lane 76.8 times per 90, more than any other defender. The midfield gathers the three biggest circulation volumes: Paredes (124.9 passes/90), Rodri (99.2 passes/90 at 93.5% across 725 minutes — the most-used player among the qualified) and Ounahi. Up front, Olise, Luis Díaz (Colombia, 61.7 offers per 90, the most involved winger of the World Cup) and Doué — who becomes the only player present in the first three XIs.
The breakout XI — the team from outside the axis
Shoubir; Freeman, Wan-Bissaka, De Fougerolles, Hakimi; Ounahi, Attia, Eustáquio; Salah, Quiñones, David.
We defined "breakouts" verifiably: the best balanced scores among players from teams outside the Europe–South America axis (CAF, AFC, CONCACAF and OFC) — the bloc that, as we showed in the previous article, played this World Cup at a structural disadvantage. The result mixes genuine discoveries with veterans shining for underdogs. Among the discoveries: Mostafa Shoubir (Egypt), a goalkeeper with elite indices for volume (91), efficiency (86) and pressure faced (95); Alex Freeman (USA, 90.5% passing accuracy with 81.7 defensive actions per 90); Luc de Fougerolles (Canada, 152 high-intensity runs per 90 as a centre-back); Marawan Attia, the Egyptian engine; and Stephen Eustáquio (Canada), owner of the database's most absurd number: 222.4 high-intensity runs per 90 minutes. Among the veterans displaced from European stardom: Aaron Wan-Bissaka defending for DR Congo (99.1 defensive actions per 90), Achraf Hakimi (72.7 passes/90 at 89.5% across 556 minutes) and Mohamed Salah, who at 34 did for Egypt exactly what the balanced XI demands of its wingers: 103.9 defensive actions and 50.8 sprints per 90. Just behind in the queue: Yan Diomandé (Côte d'Ivoire) and Sadio Mané (Senegal).
What this exercise reveals — and what it hides
Three findings run through all four teams. First, the modern forward is measured by what he does without the ball: Doué, Plata, David and Salah get in because they press like midfielders — and Mbappé drops out for the opposite reason. Second, the tournament had an elite midfield outside Europe: Ounahi and Attia appear in three and two XIs respectively, ahead of names from giant clubs. Third, volume is not a synonym for big teams: of the 44 places across the four XIs, 21 went to players from outside the tournament's top-10 FIFA rating.
And what it hides: without individual xG, goals and assists (FIFA's PDFs do not include them per player, and we do not invent data), this method undervalues the pure finisher and the decisive pass — it is a portrait of the process, not the product. Defensive actions aggregate different types of intervention, "offers" measure availability rather than reception quality, and 300 minutes is still a short sample. The official team of the tournament will reward those who decided games; this one rewards those who carried them — and each list tells half the truth.
And when the goals enter the equation?
To close the exercise, we crossed the four XIs with the tournament's final output numbers — complete goals and assists (FBref) and xG and chances created for the leaders (Opta). The two readings illuminate each other.
First, the validations. Michael Olise, whom the process selected for three XIs before we knew his numbers, finished the World Cup as the outright assist leader: 7, breaking Pelé's single-edition record (6, in 1970). And Opta's chances-created list reads like a mirror of our teams: Hakimi (17), Salah (16 in just 428 minutes), Doué (13) and Pedri (13) all appear in the tournament's top 11, alongside the obvious Messi (25, with 8 big chances) and the French trio. Quiñones (4 goals), Jonathan David (3), Muñoz (2, from full-back), Ounahi (2) and Vinícius (4 goals and 1 assist) also confirmed in output what the process indicated.
Then, the contrasts — which quantify exactly what this method cannot see. Haaland scored 7 goals with one of the worst process scores among qualified forwards; Kane scored 6 in the same pattern; and Bellingham scored 7 as a midfielder. Together, the two Englishmen produced 13 goals from a combined xG of just 6.5 — an overperformance of more than 6 goals that no work-rate metric would anticipate. At the top, Mbappé led the tournament in xG (5.65) and converted 10, taking the Golden Boot: elite finishing is a talent of its own, invisible to process.
Finally, we rebuilt the attacking XI giving 40% weight to goals and assists per 90 minutes: Maignan; Kimmich, Muñoz, Hakimi, Freeman; Bellingham, Ødegaard, Ounahi; Mbappé, Olise, Quiñones. Mbappé returns to the team the moment output enters the equation — and the article's conclusion completes itself: process picks those who carry the game; output picks those who decide it. The World Cup's true team of the tournament lives at the intersection.
Update: the World Cup's best players in a single score
After publication, we took the final step: merging process and output into a unified 0–100 score for the 220 qualified players. The ruler: process score plus goals and assists per 90 minutes, with position-specific weights (forwards 50/50, midfielders 65/35, defenders 85/15), a confidence discount for short samples and one important correction — defensive actions measured per minute out of possession, so defenders on dominant teams are not punished.
The top 10: Mbappé (74.4), Bellingham (73.5), Ødegaard (71.1), Puerta (70.0), Bruno Guimarães (67.3), Castagne (66.3), Ounahi (65.9), Olise (65.3), Muñoz (65.0) and Spence (64.3) — with Messi (63.9), at 39, just behind. The podium reads like a summary of the tournament: Mbappé wins on overwhelming output despite weak process; Bellingham is the only member of the elite with both dimensions positive — 7 goals and an above-average process score — and, by this criterion, the most complete player of the World Cup; Ødegaard and Puerta are pure process rewarded. Gustavo Puerta, at 22.9, confirms himself as the number-one breakout, ahead of Alex Freeman (21.8).
The goals-minus-xG delta completes the finishers' portrait: Mbappé +4.35, Bellingham +3.80, Messi +3.23, Kane +2.72 and Haaland +2.71 — the five top scorers all converted far above expectation — while Sarr (+0.02) scored exactly what he created and Hakimi (−1.63) was the biggest underperformer among the leaders.
The Cubarsí problem — what the score cannot see
Pau Cubarsí is the method's stress test. The most-used defender of the World Cup (761 minutes, at 19.4, on the champion side), with truly elite build-up — 92nd percentile for passing volume and 97th for accuracy (96.8%) among defenders — he comes out with a score of 51.8, outside every cut. Half the explanation is context: with 61% possession, Spain barely defends, and the out-of-possession correction lifts him from the 3rd to the 24th percentile in defensive actions. The other half is a structural limitation: counting metrics cannot see the positional centre-back — a defender who works by anticipating and closing passing lanes generates no countable event. By the "coach's trust + build-up" ruler, Cubarsí is perhaps the breakout of the tournament; by the action-count ruler, he is invisible. Both things are true at once — which is why no single score should ever end the conversation.
Data: FIFA's official Post-Match Summary Reports, 104 matches of the 2026 World Cup; 1,252 players processed, 220 qualified outfield players (≥300 min) and goalkeepers with 3+ matches. Metrics per 90 minutes, standardised within each position. The PDFs do not include individual goals, assists or xG; the final sections use official goals and assists consolidated via FBref, xG and chances created via Opta, and ages plus goal cross-checks from an open dataset (CC0, FIFA/Sofascore sources) validated against the official numbers — the gap between the 295 goals with a credited scorer and the official 308 corresponds to own goals. Goalkeeper indices on a 0–100 scale derived from the reports' goal-prevention and distribution tables.