Why Our Model Gives Brazil Only 7.3% to Win the World Cup

Let's be honest up front: our model got the Brazil–Morocco match wrong. We had Brazil at 45% to win the opener, and they only managed a 1-1 draw — a goal from Vinícius Júnior rescuing a nervy, second-best performance after Morocco bossed the first half-hour. It was one of just three misses in our public record so far (we are 5/8), and all three were draws. We will come back to that pattern, because it matters. But the draw is not the story. The story is a number we published before a ball was kicked: our model gives Brazil only 7.3% to win the 2026 World Cup — fifth, behind Spain, France, England and Argentina.

That is a genuinely contrarian number, and we want to defend it with respect for a five-time champion, not a troll's smirk. Brazil is a magnificent football nation. But "magnificent" and "the team most likely to lift this specific trophy" are different claims, and the data leans cooler on the second one than the Seleção's support does.

The draw didn't change Brazil's math — it confirmed our framing

Here is the part that gets lost in the day-three panic: our model still gives Brazil 98% to escape Group C and reach the round of 32. The opener barely dented that. Brazil sits in a group with Morocco at 95%, Scotland at 74%, and Haiti at just 5%. A draw against the second-strongest team in the group costs Brazil almost nothing in qualification terms. You can read the full group picture on our groups page, and the remaining fixtures — Brazil vs Haiti on June 19 and Scotland vs Brazil on June 24 — are very likely to send Brazil through.

So if Brazil is 98% to advance, why only 7.3% to win it all? Because reaching the round of 32 and surviving seven knockout games against the world's best are wildly different propositions. The title number isn't a verdict on the group stage. It's the product of running the entire bracket 50,000 times.

The flattest favorite field in modern World Cup history

To understand 7.3%, you have to understand the field. No team clears 16% in our model. Spain leads at 15.9%, France at 15.4%, England at 12.7%, Argentina at 10.8%, and then Brazil at 7.3%, just ahead of Portugal at 7.2%. This is the flattest title race we have measured in modern World Cup history. There is no juggernaut, no clear king — just a cluster of very good teams separated by single percentage points.

In a flat field, every favorite's number gets compressed, and "favorite" stops meaning what it used to. Bookmakers imply Spain around 18%; we say 15.9%. The market has carried Brazil as a top-four belief for years on reputation and history. Our model — which does not copy bookmaker odds, runs an Elo rating into a Dixon-Coles Poisson scoring model, and simulates the bracket 50,000 times — keeps landing Brazil a notch below that consensus. You can see the whole ranking and how it moves on our predictions hub and the model calls board.

Knockout variance is the great equalizer — and it hurts the almost-favorite

The deepest reason Brazil's number is modest is structural, and it is the same reason every nation should temper its dreams. A World Cup title requires winning a chain of one-off knockout matches. Even a team that is a 65% favorite in each of four knockout games has only about an 18% chance of winning all four. Variance compounds. One bad 90 minutes — a deflected goal, a red card, a penalty shootout — ends everything, regardless of pedigree.

For a true juggernaut, that variance is survivable because their per-match edge is large. For a team in Brazil's tier — excellent but not dominant, fifth in a field where fifth and sixth are nearly tied — the math is unforgiving. Brazil has to beat variance four or five times, and in a flat field they enter several of those games closer to a coin-flip than the romance suggests. That is the whole ballgame. The 7.3% isn't pessimism about Brazil; it's realism about knockout football.

Morocco's quality is real, and we said so first

The Morocco result deserves more than "Brazil had an off day." Morocco is genuinely very good — World Cup semi-finalists in 2022, Achraf Hakimi at right-back, a defensive structure that smothered Brazil's first half-hour. We flagged Morocco as a dark horse before the tournament, and our model gives them 2.4% to win the whole thing and 95% to advance from Group C — nearly level with Brazil's own escape odds. Read our case on the Morocco team page. That a team this good held Brazil isn't an upset that breaks the model. It's evidence the model already priced: Group C was never a stroll, and Morocco belongs.

Owning the miss — and what it actually tells you

We publish our misses, so here is the uncomfortable bit. All three of our wrong calls — Canada–Bosnia, Qatar–Switzerland, and Brazil–Morocco — were draws. That is a known structural tendency: scoring models slightly under-weight the probability of the deadlock, and a tournament this tight has produced a lot of them. We own that. But notice what the Brazil miss does and doesn't imply. It means we were too confident Brazil would win that one match. It does not contradict the 7.3% — if anything, a Brazil side that needed a Vinícius special to draw with Morocco is behaving exactly like a team our model rates fifth, not first.

Our backtest record is public: 62% accuracy, an RPS of 0.175, and 2.3% calibration error, all walk-forward and open-source. We are not claiming to beat the market. We are claiming to be honest and well-calibrated. You can interrogate the whole approach on our methodology page and run Brazil's bracket yourself in the simulator.

The bottom line

Brazil will almost certainly reach the knockouts, and they have the talent to win any single game against anyone. But "can beat anyone once" and "most likely to win seven in a row" are different things, and in the flattest field we have ever measured, the data puts Brazil fifth at 7.3%. The Morocco draw didn't reveal a crisis. It revealed a contender — a very good one — in a tournament with no king. We may be early on this. We don't think we're wrong.