World Cup 2022 update II: Ramenez la coupe à la maison?

Shock defeat for Brazil and another heart-breaking exit for England have left France as World Cup favourites. Their chances of lifting the trophy are 40%, just ahead of Messi’s Argentina (39%). Croatia (12%) and Morocco (10%) remain outsiders, but in a tournament full of upsets, nothing is certain.

Here at the Economics Observatory, we do have something of a bias, in case that has not been obvious from our World Cup forecasts so far. We hoped that England would ‘bring football home’, to paraphrase. But as most who use the phrase are keen to point out, its use has always been ironic: to symbolise that biennial hope that never quite fulfils.

So, we wrote in hope more than confidence, since even our quarter-finals update only gave England a 12% chance of winning the competition. World Cups are absorbing contests precisely because the best team does not always win. Brazil were considered by some margin the favourites in Qatar, but have fallen by the wayside. If even Brazil, with all its incredible talents, cannot win a World Cup for over a quarter of a century, should we be surprised that it’s taken England so long to become genuine contenders again?

A World Cup is gruelling, with four rounds of life-or-death knockouts. Even if Brazil were (as indeed they were) about 80% to win their knock-out ties, four in a row at 80% only gives a probability of 41% that Brazil wins all four. So, disappointment is natural, but England got as far as was most likely for them – arguably a little bit further.

How does this work? Our odds tables have given the probability of a team reaching a stage of the World Cup based on how many teams can reach that stage. So, if you count up, our probabilities for teams making the quarter-finals was 800%, since eight teams make it. But equally, that means that for each team individually, the probability they make that stage is too high. Because it’s the probability they reach the quarter-finals conditional on them having made the Last 16.

If we look back at England’s chances (which were submitted to the RSS, rec.sport.soccer, John De Lange Sophisticated Prediction Contest), England had a 23% chance of going out in the group stages, a 26% chance of exiting in the last 16 (England could have faced Netherlands had results gone differently) and a 23% chance of making it as far as the quarter-finals.

Even Brazil, who were at the outset favourites with a 27% chance of winning, had a 23% chance of exiting at the quarter-final stage (as they did). It is this uncertainty of outcome that attracts people to watching the World Cup in their millions.

Equivalently, England’s 6% chance at the outset should mean that, if they are at 6% for every tournament, they should expect to win one every 16 competitions (or 32 years). It means that going 56 years without one is a stark statistical aberration – that implies England have had just a 2% chance of winning the World Cup each edition since 1966, which seems rather low. So, to provide some statistical hope to the hope that Gareth Southgate’s team provided on the field in Qatar, we can reasonably hope to see a World Cup or European Championship win in the not-too-distant future. But as Ted Lasso has famously said, it’s the hope that kills you…

But what about the highly unexpected final four, with two semi-finalists that had low expectations of making such progress. Despite Croatia reaching this stage four years ago in Russia, they had just an 8% chance of getting this far at the outset of the tournament. Morocco’s chances were just 4%.

Figure 1: Chances of reaching the final, by combination

Figure 2: Chance of progression and elo score, by round, by team

Source: Author's calculations

Unsurprisingly, the most likely final is France against Argentina, at 50%. A final between Morocco and Croatia has just an 8% chance of occurring. There’s also a 20% chance of the first World Cup final since 1950 to not involve a European team, if Argentina and Morocco progress.

Finally, there’s a 23% chance of a repeat of the 2018 final between France and Croatia. It would be the second time a repeat final has happened in successive World Cups, after West Germany and Argentina contested both the 1986 and 1990 finals.

France have a 72% chance of reaching the final, and Morocco have the corresponding 28% chance of causing a huge upset. Argentina are 69% likely to reach the final, and Croatia 31%. But Croatia prevailed 3-0 when the two met in 2018 in Russia, and had just a 20% chance of beating Brazil in the quarter-finals. Anything can happen, but the most likely thing is that ‘le football rentre à la maison‘, as they say in France (probably).

Where can I find out more?

Evaluating strange forecasts: The curious case of football match scorelines – study by James Reade and colleagues.
Going with your gut: The (in)accuracy of forecast revisions in a football score prediction game – another research paper by James Reade and colleagues.
Futbolmetrix’s blog: Discussion of football (futbol, calcio, soccer) and numbers.