To predict the outcome of the world cup 2018, we first have to understand what the predictors are for a successful world cup. To do so, we gathered data since the first tournament in 1930. 19 world cups were played since. For all teams that competed in these tournaments we gathered the historical data.
We now know, for each team/country, the final rankings of each world cup. We used a combination of econometrical techniques to estimate the impact of the different variables. Having quantified the impact of the different variables by making this model, we can make a prediction for the world cup in Russia! We do this by something we call 'team strength'. This strength depends on how “favorable” the variables are for each team. For instance, playing at home is a good indicator for a successful world cup. Also, older teams have a tendecy to perform better than younger teams. For example, the average age difference between countries that make it to the semi-final or go all the way to the final is one year.
What has to be understood is that we are only modeling a probability that one team wins or loses. The bigger the difference is in team strength, the more likely one team will beat another. This website allows you to play the world cup 2018 as often as you like, and gives a good indication of how much chance each country has of winning. However, as the dutch like to indicate the randomness in football by the saying; "the ball is round" ;)
List of variables used in the model:
- Average goals scored/conceded in the WC qualifications
- Average number of goals scored/conceded on previous WC’s
- Age of the players
- Continental cup titles (champions league, … )
- Difference in temperature to Russia
- Previous world cup titles