World Cup 2026 after the opening group games: is football coming home?

The 48-team World Cup is 24 games old – each team has played once, with Colombia and Uzbekistan finishing off the first round of fixtures in the early hours of this morning. What have we learned so far?

Perhaps most notably, if a team changes its style of play, its attacking and defensive strengths may no longer be accurate. England were expected to score 1.3 goals last night to Croatia’s 1.1, reflecting a rather dour, resolute style for both teams over recent years.

Rather than a 1-1 draw, or a 1-0 win either way (which had between them a 36% chance), England scored four to Croatia’s two (the first England 4-2 World Cup win since a certain match in 1966). This scoreline had a 1% chance of occurring – based on estimates of attacking and defensive strengths over the last three years. How much should a system update – and how quickly?

How our Elo model learns

Our Elo rating system builds that in. Its formula is essentially an ‘adaptive expectations’ formation for macroeconomists of a vintage generation, as it learns from each mistake it makes, but perhaps most pertinently, an Elo rating system is entirely backward looking. Many of the other forecast models employed, including those that rely on bookmaker prices, or Transfermarkt transfer values, have an element of ‘forward-lookingness’ about them – current and future player ability is worked into the valuation of teams and players.

As an example of the Elo updating, the prediction for Spain against Cape Verde was 0.88, indicating a very likely Spanish win. With a draw occurring, the outcome of that match was not a 1 (Spain win) but instead 0.5 (draw). The Elo error correction was thus -0.38 for Spain and +0.88 for Cape Verde. A K-factor was then multiplied by -0.38 to give us Spain’s new (lower) rating (1425, down from 1432), and by 0.88 to give us Cape Verde’s new (higher) rating (1090, up from 1080).

Note that teams who have played more international matches have a lower K-factor in our method, hence Spain (around 900 matches) move less than Cape Verde (266 matches). We’re still learning more about Cape Verde – arguably a beautiful feature of this expanded tournament format.

Who gained and who lost in round one?

If the change in Elo since the tournament started on 11 June is our metric, then Ivory Coast have gained most through its last minute 1-0 win over Ecuador, gaining 21 Elo points. This lifts them from the 39th best country in our Elo ranks to 34th.

The next biggest winners are Australia and Austria at 12 points, then England at 11 points, Scotland moving 10 and Cape Verde with their 10 point move. But Cape Verde are the biggest positive movers in terms of rank, becoming the 82nd best national team from 93rd before the tournament.

What does the first round of group matches mean for the overall tournament?

Argentina’s great start has naturally strengthened their position – they are now 27% to win overall. They are also 82% to win their group, up from 65% before their 3-0 win, and so their path becomes clearer. It will have been affected by the possibility that Uruguay (Argentina’s likely last 32 opponent) are weaker than previously thought, and the likelihood that Lionel Messi and his team face Colombia rather than Portugal in the quarterfinals after the latter dropped points in the first set of fixtures.

Chart 1. Win probabilities after the opening group games

England have jumped to fourth favourites at 10%, with France and Spain both on 11%. The French drop back a touch, which is likely to reflect changes in their path to the final – Brazil may drop into France’s quarter of the draw if they finish second behind Morocco in their group. England are now cast-iron favourites to win their group (98%), having beaten the only other team likely to top the group last night.

Things haven’t changed much for Scotland: Brazil are 57% to win the group (down from 58%), while Morocco are 46% to finish second (up from 43%). But gaining that crucial 1-0 win over Haiti has increased Scotland’s chances of making the last 32 to 81% from 70%.

Scotland are most likely (39%) to remain on three points given tough matches to come. But there’s a 26% chance that they get a draw from one of the matches, and a 21% chance that they manage to win one of them. With three points on the board, they’ll have a 33% chance of being one of the eight best third-placed teams, and with four points, that likelihood rises to 61%, while also increasing the chance that they could nick second place in the group.

Updated group finishing positions

Who could be England’s opponents in the first knockout round and beyond?

Congo at 26% are the most likely opponent in the last 32, way ahead of Portugal/Colombia/Senegal/Uzbekistan at 12-14%. This is because they are most likely to finish third in Group K behind Portugal and Colombia, and England are now almost certain to win their group. Only if Group K’s third-placed team isn’t one of the best eight will England face another team in the last 32 – Senegal being the most likely in the absence of a Group K team.

By finishing top of the group, England’s pathway is established; if all goes according to expectations, they would then face Mexico in Mexico City on 6 July in the last 16, then most likely one of Brazil, Morocco, Japan, the Netherlands and Norway in a quarterfinal. Holders Argentina would likely then stand in their way in the semifinal, with France or Spain being final opposition from the other side of the draw.

Author: James Reade