Swiss Football League Predictions

On this page you can find our predictions of Swiss Football League. The winning probabilities are calculated with an elaborated model based on Swiss football stats from the last ten years. The predictions are constantly updated.

About our model

The predictions are made with a random forest approach based on the following indicators:

  • Current ranking of the teams
  • Market value of the teams, average per player (source: Transfermarkt.ch)
  • Elo rating of the teams (source: http://clubelo.com/)
  • Average points scored of the teams in the last three months
  • Average points scored of the teams in the last year
  • Average points scored of the team since season 11/12

The model was developed and tested based on over 1.500 Super League matches since season 11/12. The model reaches an accuracy of approximately 0.56. This means it can determine the outcome of a football match (win/draw/loss) correctly in 56 percent of the cases.

To predict the final standings of the season, 100 simulations were performed. The average outcome of the final score of each team results in the final ranking. 95 % of simulations predict a final standing within the values CI low an CI high (model uncertainty).

You want to challenge our model?

Here you can play our new game “Beat the Robot” and win cool prizes worth over 150 Swiss Francs!

3 Comments

Leave a Reply