England have a 3.41% chance of winning the World Cup, according to Oxford mathematical model

Brazil are most likely to win this winter's World Cup in Qatar, while 'outsiders' England have a 3.41% chance of lifting the trophy.

That's according to a study from the University of Oxford, who have analysed team rating data that focuses on every first-team international game since 2018.

In terms of the favourites, Tite's Brazil are most likely with a 14.72% chance of winning their sixth World Cup title, while Argentina [14.36%], Netherlands [7.84%], Spain [7.03%] and France [6.37] make up the top five.

Belgium, Portugal and Denmark feature in the top ten alongside Germany and Uruguay, who are more likely to win the tournament than Gareth Southgate's England, according to the study.

Rob Page's Wales side has a 1.07% chance of winning the competition.

The ten countries with the best chance of winning the @FIFAWorldCup - according to @JoshuaABull's Oxford model:

1. 🇧🇷 14.72%
2. 🇦🇷 14.36%
3. 🇳🇱 7.84%
4. 🇪🇸 7.03%
5. 🇫🇷 6.37%
6. 🇧🇪 6.31%
7. 🇵🇹 5.60%
8. 🇩🇰 4.94%
9. 🇩🇪 3.84%
10. 🇺🇾 3.55%
#WorldCup #WorldCup2022 pic.twitter.com/kZQoUVWoC0

— University of Oxford (@UniofOxford) November 17, 2022

Joshua Bull, who was crowned the winner of a Fantasy Football competition that had over eight million entrants in 2020, created the model based on an observation that the difference in the expected number of goals that two teams will score is 'very well' correlated with the difference in their ratings.

In the short video below, Bull has explained the study in further detail.

"If I look at how different the ratings of two teams are, I should be able to predict what the score line will be or what the difference in the scoreline should be, which is good enough to tell me who who is likely to win," he said.

"So for every simulated match, we just have to look at who the teams are, look at what their ratings are, according to eloratings.net and use that to predict what the xG difference will be.

"I can then take that xG difference, convert that into two different Poisson distributions, and use that to predict what the number of goals in that game might be for each side.

"That lets us simulate a real match between those two teams based on the two predicted xG that we get from looking at that difference. So in order to predict the winner, all we have to do is repeat this a lot of times every match in the tournament, and see who comes out on top."

In fact, they repeated it 100,000 times for every match in the tournament to determine the winner.

The model has been created by analysing team rating data from https://t.co/6Rutrl4ORq, focusing on every first-team international game since 2018.

In this @OxUniMaths video, @JoshuaABull explains the maths behind the model: pic.twitter.com/dokvlen1CN

— University of Oxford (@UniofOxford) November 17, 2022