With the tournament now down to its final four, scorecasting analysis suggests that Argentina’s likelihood of retaining the trophy is up to 42%. Spain are on 22% to win, France 20% and England outsiders at 15%. There’s a 10% chance of England beating Argentina 1-0 – and a 21% chance of a 1-0 defeat.
With this World Cup’s quarterfinal ties completed over the weekend, the tournament reached a total of one hundred matches played and just four remaining: the semifinals over the next couple of evenings, the final on Sunday and, the day before that, the game no final four team wants to be in – the playoff for third place.
How often do the four best teams make it to the semifinals?
This World Cup has delivered its seeded four teams as the last four. Those four are both the top four teams in FIFA’s ranking system as well as in our Scorecasting Elo rating system. It’s the first time this has happened with respect to FIFA rankings since their classification began in 1992.
Our Scorecasting Elo ratings find that the 2014 semifinals (Germany versus Brazil and Netherlands against Argentina) were also contested between the four best teams in the world at the time. At that time, FIFA ranked the Netherlands at 15 and Argentina at 5.
But before that tournament, only the 1934 World Cup comes close, where the teams ranked first (Austria), third (Italy), fourth (Czechoslovakia) and fifth (Germany) competed in the semifinals. On average, the lowest ranked team has been tenth, with South Korea (twenty-second) in 2002 by some way the lowest ranked team to reach a semifinal, followed by Chile in 1962.
That the top four have made it this year is a purposeful feature of FIFA’s tournament design, structuring the knockout stages such that the top four teams, provided they won their groups, would only meet at this stage.
Knockout tournaments, which the World Cup became after the first 72 matches were complete, do not tend to deliver the best four teams as the final four since over 90 minutes unexpected outcomes can occur – and frequently do in football. League seasons, which take place over 38 or more matches, are more likely to deliver the top four as the best four teams since they are based on a larger set of matches – a bigger sample, so to speak.
Have there been fewer shocks in this World Cup?
The best four being the final four also hints at something else: by and large, this has been a World Cup that has not delivered shocks. There have been near misses – perhaps most notably Egypt coming close to beating holders Argentina.
The shocks that have happened – perhaps most notably Paraguay eliminated Germany alongside Cape Verde’s draw with Spain – feel like they have been relatively minor. Is this true?
We can answer this question by applying our scorecasting method back to all previous World Cup editions, and generating the kinds of probabilities that we have used for all three match outcomes in this tournament.
A standard measure of surprise is the mean squared forecast error – often called the Brier score in sporting settings. It adds up the distance between what we predicted for all three possible match outcomes, and yields a larger number for a more surprising outcome. If we then average that over all matches in every World Cup, we have a measure of how surprising each tournament was.
Our Brier score measure finds that Spain’s 0-0 draw with Cape Verde was the biggest surprise of this World Cup, achieving a Brier score of 1.6, followed by Curacao holding Ecuador (1.5), and Uruguay drawing 2-2 with Cape Verde (1.4).
Table 1. The biggest surprises of this World Cup
| Team | Team | Brier | |||
|---|---|---|---|---|---|
Spain  | 0 | – | 0 |  Cape Verde | 1.6 |
Curacao  | 0 | – | 0 |  Ecuador | 1.5 |
Uruguay  | 2 | – | 2 | Cape Verde | 1.4 |
Qatar  | 1 | – | 1 |  Switzerland | 1.4 |
England  | 0 | – | 0 |  Ghana | 1.3 |
Sweden  | 1 | – | 1 |  Japan | 1.2 |
Overperformers shaded
Below this was Qatar’s 1-1 draw with Switzerland (1.4), England’s 0-0 draw with Ghana (1.3) and Sweden holding Japan (1.2). Germany’s draw (before the defeat on penalties) was the next biggest shock, followed by South Africa defeating South Korea and an Erling Braut Haaland-inspired Norway toppling Brazil 2-1 in the last 16.
With this measure, we find that the 2026 World Cup was the least surprising World Cup since the 1970 edition in Mexico. Despite those values above 1, the average Brier score across the tournament was just 0.47, below Qatar 2022’s 0.62, Russia 2018’s 0.58 and Brazil 2014’s 0.61.
Chart 1. How surprising was each World Cup?
By this score, the most surprising World Cup in history was 1958 in Sweden (0.72), followed by Spain 1982 (0.65) and Mexico 1986 (0.62). But this has been the best attended World Cup in history – both in total attendance and average attendance – which indicates that there is more than the possibilities of shocks that drives people to go to sporting events.
Ranking all World Cup shocks in this way delivers a list of the greatest giant-killings. Cape Verde holding Spain to a 0-0 draw is the eleventh largest shock in World Cup history. The biggest shock was England losing 1-0 to the United States in 1950 (1.9998) – a match in which England had expected goals of 13.2!
France beating Mexico 4-1 in 1930 (1.98) was the second biggest ever shock, followed by Algeria beating Germany 2-1 in 1982 (1.96). North Korea’s 1966 1-0 win over Italy is the sixth largest shock ever, and Senegal’s 1-0 win over holders France in 2002 is the tenth largest ever (1.71).
Is football coming home?
Throughout the tournament, England’s chances have been edging up, from 9% before a ball was kicked to 15% before the quarterfinals. Despite making progress to the final four, England’s probability of winning the World Cup remains at 15%. This reflects the fact that in our model, they are facing the toughest possible semifinal – against our tournament favourites Argentina.
Chart 2. Every team’s path to the trophy
Argentina’s chance of retaining the World Cup is up to 42% having edged past Switzerland in the quarterfinals. Spain are now 22% to win the competition, and France 20%.
Hence England remain the outsiders still despite reaching this advanced stage. There’s a 29% chance that they will compete in the World Cup final on Sunday, and hence a 71% chance that they will instead participate in the third-place playoff. Spain are 55% to beat France and make the final, and hence France are at 45%.
Table 2. The final four
| Team | Elo | Elo rank | FIFA rank | Reach final (%) | Win cup (%) |
|---|---|---|---|---|---|
 Argentina | 1486 | 1 | 2 | 71.0 | 42.3 |
| Spain | 1433 | 2 | 3 | 55.2 | 22.1 |
 France | 1419 | 3 | 1 | 44.8 | 20.4 |
| England | 1367 | 4 | 4 | 29.1 | 15.2 |
Due to rounding, percentages may not sum to exactly 100%.
A Spain-Argentina final is thus the most likely showpiece match at the New Jersey-New York Stadium on Sunday at 39%, followed by France versus Argentina (32%), then Spain versus England (16%) and France versus England (13%).
Chart 3. The most likely final
SpainvArgentinaFrancevArgentinaSpainvEnglandFrancevEnglandWhat will happen in each semifinal match?
On Tuesday evening, France have expected goals of 0.86 and Spain 1.04 – a reflection of the Spaniards’ remarkable defensive strength, having conceded just one goal in the entire tournament up to this stage.
As such, a goalless 90 minutes is far from inconceivable, with a 15% probability. A 1-0 Spanish victory is slightly more likely at 15.4%, and France are 13% to win 1-0. Spain are about 39% to win in 90 minutes, France at 29% to do so, and hence there’s a 32% chance that the match goes to extra time.
On Wednesday evening, Argentina have expected goals of 1.16 and England 0.55, reflecting what is likely to be another tight encounter. A goalless 90 minutes is at 18%, while an Argentine 1-0 win is at 21%, and England winning 1-0 is at 10%. England are only at 18% to win in 90 minutes, with Argentina at 51%, and again extra time is 32% likely to happen.
