England’s footballers have a 3.8 per cent chance of winning the 2018 Fifa World Cup, according to a Monte Carlo simulation applied to all the teams in the tournament.
Whilst the England national football team has realistically long odds, 2014 winners Germany are ranked as having a 13.3 per cent chance of retaining the trophy.
These are outcomes from an uncertainty model devised by the University of Adelaide’s Steve Begg, who rates Australia as having a 0.1 per cent chance of winning the tournament.
Begg, a Professor of Decision-making and Risk Analysis in the University’s Australian School of Petroleum, has developed a ‘Monte Carlo simulation’ of the competition based on team rankings with other input including recent form.
The key idea behind the modern Monte Carlo technique is that rather than trying to work out every possible outcome of a complex system, enough possibilities are modelled to be able to estimate the chance of any particular outcome occurring.
“The outcomes of many decisions we make are uncertain because of things outside of our control,” said Prof Begg. “Uncertainty is crucial in predicting the chance of an oil or gas field being economic. In the World Cup, it determines the many ways the whole tournament might play out. What makes it so hard to predict is not just uncertainty in how a team will perform in general, but random factors that can occur in each match.”
In the simulation, Begg is said to have generated 100,000 possible ways the whole tournament of 63 matches could play out. Although there are many possible options – almost 430 million outcomes in just the Group stage – this is “more than enough” for an assessment of the probability of how far each team will progress. He can run 800 different simulations a second.
In his model, two key uncertainties are a team’s “tournament form” (their general level of performance entering the finals) – and a team’s “match form” (the extent to which the team plays better or worse than its tournament form in a given match). The possible scores for each match are derived from the likely number of goals, based on scores from all matches in the last three World Cups, allocated to the two teams based on their relative match form.
Inputs to the model are based on FIFA rankings over the past four years, modified by Begg’s knowledge of the game and team; for example Russia will have a higher “tournament form” because of host advantage, and poorer teams have a relatively greater “giant killing” upside than better teams.
On current inputs, Begg has calculated the Australian national team has a 14 per cent chance of advancing through the groups stage, 3.8 per cent of making the quarter-finals, 1.2 per cent of the semi-finals, 0.3 per cent of being in the final, and 0.1 per cent chance of being the 2018 World Cup Champions.
“This may be disappointing,” he said, “But to make good decisions, it is really important to base beliefs on evidence and reason, not what you would like to be true.”
“Probability is subjective, it depends on what you know,” said Begg. “It doesn’t need data; you use what information you have to assign a degree of belief in what might happen, and thus make decisions or, in this case, a judgement on who wins. The crucial thing is that your information and reasoning is not biased.”
The 2018 FIFA World Cup Russia kicks off today, June 14, 2018 with its first game between hosts Russia and Saudi Arabia.
This makes Alan Turing’s achievement look easy!
The bookies (and even insurers) put their money where their mouth is: is he offering 30:1 against England? Does this forecast have any use to anyone and why produce it?
The bookies are offering 16 : 1 against England winning today. Would be very happy to put a tenner on at 30 : 1 against. The bookies are also only offering 40:1 against Russia, his 393:1 would be a good long-risk punt. I suspect that this guy would not survive long in the gambling field, and might have to pawn his computer.
For fun 🙂 (Bookies set their odds where their punters money goes, plus an amount to ensure a profit, not on the actual assessment of the probabilities)
I think that you will find that bookies and actuaries start with their own assessment of probability then modify as experience / better information comes in. I can see no use nor reason for this probability assessment, was hoping that someone could tell me why he bothered publishing this. If that is his statistical-fun …. how sad!