Champions
Contrary to 2010/11, when three out of the five champions were correctly forecast, this ratio was 0% for the 2011/12 season. However, the final appraisal is not completely negative. Indeed, three forecasted champions finished 2nd (Barcelona, Bayern Munich and Paris St Germain). From the expected champions, the true disappointments came from Inter Milan (6th) and Champions League winner Chelsea FC (6th).
Moreover, three league champions out of five belonged to a narrow circle of “three-star” teams, for which our model had detected a concrete possibility of winning the title (Manchester City, Real Madrid and Borussia Dortmund). Another league winner, Juventus FC, was also forecast to finish on the podium (3th place), while the remaining one, Montpellier HSC, was indicated as a possible surprise team.
Top Three
Our model forecast 10 of the 15 teams that finished the season on the podium in their respective championship. Moreover, two out of the five remaining teams are among those whose quality of play during the two preceding seasons convinced us to classify them as possible surprise teams: Montpellier and Arsenal. This confirms the relevance of pitch performance indicators developed by the CIES Football Observatory on the basis of data provided by our partner company Opta.
Ultimately, only Schalke 04, Valencia CF and Udinese Calcio have confounded our predictions by finishing third. With the exception of Chelsea FC and Inter FC, no team that was predicted to finish on the podium finished lower than fifth place (Olympique Lyonnais 4th, Málaga CF 4th, Bayer Leverkusen 5th).
Most positive and negative gaps
The most positive negative gap between rank estimated and finally achieved was for the Spanish side Villarreal CF (-13 ranks compared to our prediction), followed by three other “two-star” teams that could not avoid relegation: AJ Auxerre (-12), Bolton Wanderers (-9) and Wolverhampton Wanderers (-9).
Conversely, the most positive gaps were for “one-star” teams Levante UD (+13 ranks compared to our prediction), Borussia Mönchengladbach (+11), FC Parma (+11, which was considered as a possible surprise team) and Newcastle United (+10).
For more information, click here
Conclusion
From a prediction perspective, this appraisal shows that the statistical analysis of the demographic profile of teams at the start of the season provides valuable projections. However, it also indicates that numerous other factors, some of which are impossible to forecast and measure, should be taken into account to help determine more accurately winning teams.
While aware of the evident limits of the purely statistical procedures followed, we will continue working to improve our models – and continue to enjoy the project as we follow and analyse this fascinating game!