World Cup Prediction Model

Posted 28th November 2017

World Cup Predictor Hero.png

Opta’s World Cup Predictor has been built using Opta’s extensive database of past international and World Cup matches

How is the model built?

Each team has an attacking and defensive strength calculated based on past performances. Given these attacking and defensive strengths – and several other World Cup specific variables – for each game we can assign a likelihood to each potential result (either team to win or a draw). For example, consider if Team A were playing Team B: the probability of a Team A win may be 20%, a draw 30% and a Team B win 50%. From these probabilities or likelihoods we can simulate the entire tournament. Once we simulate the tournament enough times we can imply the likelihood of any team to win the World Cup or reach a particular stage of the tournament.

Which matches are the attacking and defensive strengths based on?

The team attacking and defensive strengths are based on all official results from the four years before the World Cup. This includes all World Cup qualifiers, confederation championships, confederation championship qualifiers, friendlies and the FIFA Confederations Cup. In addition, these strengths are based on results at the previous two World Cups (2010 and 2014).

What additional World Cup specific variables are considered?

Based on historical World Cup results we have calculated the effects of several additional factors. We include two “home advantages”: one for the host nation and another for all the teams from the host confederation (in this case UEFA). We also include an advantage for teams that have won a World Cup since 1970, this serves as a proxy for teams that have World Cup winning pedigree.

What does the model miss?

The model does not include any player specific information because we don’t know the squads for the tournament yet. If Messi were to get injured tomorrow Argentina’s chances of winning the World Cup would decrease significantly but our model wouldn’t internalise this until Argentina start playing more matches and suffer as a result of missing Messi.