Brazil Have Not Played Like Brazil in 20 Years. The 2026 World Cup Was Their Last Chance

Brazil’s expected goals against Norway in the round of 16 were 2.30. Norway’s were 0.76. Brazil had twelve shots to Norway’s nine. Every number that fits on a broadcast graphic said Brazil were the better side at MetLife Stadium on 5 July, and they lost 2-1.

Start any argument about Brazil’s identity crisis there. The data was right that night. So was the eye test. They were describing different matches.

I watched it expecting the usual post-mortem. Instead I spent the next week working through the numbers from all five of Brazil’s games, and most of what I found contradicts what was written about this team in July.

Why did Brazil lose to Norway when the xG said they won?

Of Brazil’s 2.30 xG, 1.55 came from two penalties. Take those out and their open-play expected goals were roughly 0.75. Norway’s were 0.76, from 34% of the ball.

That possession figure should have led every report. Thirty-four per cent. Against Norway, a country in its first World Cup since 1998, a country that had to be talked into believing it belonged there. Norway had 66% of the ball and 92% pass accuracy to Brazil’s 88%, and Norway set the rhythm of a knockout tie against the five-time champions.

The penalties tell the same story from the other end. Bruno Guimarães, 28, of Newcastle United, had his 14th-minute spot kick saved by Ørjan Nyland, after Kristoffer Ajer fouled Matheus Cunha and Ismail Elfath went to the monitor. Neymar, 34, of Santos, scored the second in the 100th minute, after Leo Østigård fouled Casemiro. Both were given for contact in the box. Neither finished a passage of play Brazil had built. Expected goals cannot tell a penalty won by a stumble from a penalty won by twelve passes, and on 5 July that was the whole match.

Erling Haaland, 26, of Manchester City, scored in the 79th and the 90th: a header from Andreas Schjelderup’s cross, then a finish into the bottom corner. That made it goals in 14 consecutive internationals. Norway have played Brazil five times now and have never lost.

Has Brazil lost its footballing identity?

Yes. Just not the one everyone keeps naming.

The consensus that formed within about forty minutes of full time was that Ancelotti’s European pragmatism had smothered Brazilian flair. The pressing data says otherwise, and I want to argue with this one properly, because it is everywhere and it is lazy.

Brazil under Carlo Ancelotti were not cautious. Of the 33 goals they scored across his tenure, eight came from winning the ball near the opponent’s box. One in four. That is a high-press conversion rate that would sit comfortably among the most aggressive sides in Europe. Vinícius Júnior and Matheus Cunha pressed from the front rather than dropping off. The build-up used a 2+2 shape with the centre-backs splitting wide, and the instruction on regaining the ball was immediate verticality, not circulation. Nobody parked a bus.

The flair did not fail either. Vinícius Júnior, 26, Real Madrid’s left winger, scored in all three group matches, joining Jairzinho in 1970, Romário in 1994 and Ronaldo and Rivaldo in 2002 as the only Brazilians to manage it. He arrived off a club season of 22 goals and 10 assists in 53 appearances, 16 of those goals in La Liga from 36 matches. Neymar became the second Brazilian after Pelé to score at four different World Cups. Gabriel Martinelli, 25, of Arsenal, won the round of 32 against Japan at 95:00. Ten goals in five matches, two clean sheets. The forwards did their job.

What Brazil have lost is the ability to make a knockout match happen at their tempo instead of the opponent’s. That is a midfield problem, and it has nothing to do with flair.

Was the 2002 Brazil team really a flair team?

No, and this is where the received identity turns out to be a forgery.

The last Brazil side to win a World Cup was built on Gilberto Silva and Kléberson screening a back three. It beat four European teams in a row by controlling games rather than improvising through them: Belgium 2-0, England 2-1, Turkey 1-0, Germany 2-0 in the final. Three of the four were one-goal margins from a team that defended its structure and attacked in transition. Ronaldo, Rivaldo and Ronaldinho were the finishing mechanism. The engine was a midfield that could take the ball with a European No. 8 on its back and still play forward.

The joga bonito everyone says Brazil abandoned belongs to 1970. So when people ask Brazil to “play like Brazil again”, they are usually describing a team that has not existed for 56 years and prescribing it as a cure for the loss of a team that existed 24 years ago. Different problems. Opposite solutions.

Brazil in 2026 had more raw attacking talent than the 2002 side. What they did not have was a midfield. Look at who Ancelotti took: Casemiro, 34, of Manchester United, off a Premier League season of 9 goals in 34 appearances with his contract expiring; Fabinho, 32, of Al-Ittihad; Lucas Paquetá, 28, of Flamengo; Bruno Guimarães, 28; and Danilo, 25, of Botafogo. Two of the five central midfielders were 32 or older. Ancelotti confirmed after the exit that the international careers of Neymar, Casemiro and Danilo were finished. He knew before the rest of us did.

Why does Brazil keep losing to European teams at the World Cup?

Since the 2002 final, Brazil have played six World Cup knockout ties against European opposition and lost all six. Over the same period they have played seven against non-European opposition and won all seven.

The European list, in sequence: France 0-1 in 2006, the Netherlands 1-2 in 2010, Germany 1-7 in 2014, Belgium 1-2 in 2018, Croatia on penalties in 2022, Norway 1-2 in 2026. Against Ghana, Chile twice, Colombia, Mexico, South Korea and Japan, seven from seven.

The counterargument I take most seriously is the one Brazilian media leaned on all July: bad luck. Guimarães missed a penalty in the 14th minute. Ancelotti admitted a tactical error. Over a longer run of matches, the superior xG wins out. It is a reasonable position, and it does not survive the arithmetic. Give Brazil a 60% chance in each of those six ties, roughly where the betting markets had them for most of them, and the probability of losing all six is 0.4 to the sixth power. That is 0.41%, about one in 244. Six tournaments, five managers, three generations of players, same outcome. At some point the word for that stops being luck.

Norway knocked Brazil out on 5 July 2026, twenty years and four days after Thierry Henry’s volley in Frankfurt ended the 2006 campaign. Twenty years, one pattern.

Did the 48-team World Cup format flatter Brazil?

Badly. Expansion is the only reason Brazil qualified automatically, and the group it handed them inflated every tournament-level number the team produced.

Brazil finished fifth in CONMEBOL qualifying, their worst campaign on record: 28 points from 18 matches, 8 wins, 4 draws, 6 defeats, 24 scored and 17 conceded, ten points behind Argentina’s 38. They scored 1.3 goals per game, their lowest rate since the current qualifying format began 30 years ago. They lost 4-1 to Argentina, 2-0 to Uruguay, and 1-0 away in Bolivia on the final matchday. Under the old allocation of four and a half places, fifth meant an intercontinental playoff. The 48-team format gave them a direct berth instead, then drew them with Haiti, the first team eliminated from the tournament, and a Scotland side that won once in the group.

Now split the five matches into the two nobody-games and the three real ones. Against Haiti, roughly 1.56 xG to 0.23. Against Scotland, 4.46 to 1.13. Those two carry the aggregate on their own.

In the competitive matches the picture inverts. Morocco drew 1-1 and out-created them, 1.53 xG to Brazil’s 1.23, with Brazil on 51% of the ball. Japan gave Brazil 69% possession and around 2.07 xG, and Brazil still needed Martinelli at 95:00. Norway gave them 34% and 0.75 open-play xG. Brazil’s tournament profile looked healthy because two of five opponents could not defend a corner.

Then there are the host conditions. Summer kick-offs in North America mean mandatory cooling breaks, and Ancelotti’s own account of the exit was blunt: “I think we lost the World Cup during the hydration break against Norway,” the mistake being “thinking the team had lost control of the match when they hadn’t.” A structural feature of a USA-hosted tournament handed him a three-minute window to make a decision he did not need to make.

How far can this Brazil team realistically go in 2030?

Opta’s supercomputer gave Brazil a 6.4% pre-tournament chance of winning in 2026, sixth favourite, and a 96.9% chance of escaping Group C. The group model was almost exactly right. The trophy model was generous.

My honest read for 2030, with Neymar, Casemiro and Danilo gone and the rebuild starting from Endrick, 20, Wesley, 22, and Rayan, 20: a 55-65% bet to reach a quarter-final, something like 8-10% to win it. Better than 2026, but not because the forwards improve. Because the midfield cannot get older.

The ceiling is entirely a function of whether Brazil produce one ball-progressing No. 8 who can play under a European press. They have not produced one since Kléberson.

What metric would actually measure Brazil’s decline?

Something that scores whether a possession survives contact. The share of possessions starting in your own half that face at least one opponent pressing action and still reach the final third with the receiver facing goal.

Nobody has built it, and the obstacle is data, not analysis.

Expected goals measures the end of a possession. Possession percentage measures its volume. PPDA measures how hard you press without the ball. None of them measure the middle of your own possessions, which is exactly where Brazil have been failing for twenty years. Event data records the pass that was played. It cannot record the option that was not there, and the absence of a forward-facing receiver is precisely the failure. You cannot tag a non-event.

Tracking data could do it. But tracking data is licensed league by league and owned by clubs, and international football is the one competition with no continuous longitudinal dataset. FIFA collects it tournament by tournament and does not release it. There is no four-year tracking record of the Brazil national team anywhere, which is why every argument about this team, mine included, ends up being made with box metrics and memory.

Brazil went out with a better xG than Norway and 34% of the ball. Until someone builds the metric that can hold both of those facts at once, we will keep arguing about flair.

Sources: Opta Analyst — Group C preview and probabilities, Opta Analyst — 64 best stats of the 2026 World Cup, FIFA — Brazil 1-2 Norway match report, FIFA — Ancelotti reaction, ESPN — Ancelotti on the hydration break, ESPN — Brazil 1-2 Norway report, FOX Sports — Brazil v Norway box score, 2026 World Cup qualification (CONMEBOL)