Named as one of the best currently active football players in the world, Lionel Messi is also one of the best paid among them all. His Barcelona contract alone is securing him a payment of €40 M per year or €770 000 per week.
Well, recently performed data analytics model confirmed what we all suspected. Even though Messi should be the best paid, he is extensively overpaid. The model and its results are described in a study conducted by a team of Lawrence Technological University in Michigan, which used machine learning and data science to analyze the salaries of 6082 professional football players in Europe. The salary of each was compared to a set of 55 attributes, reflecting each player`s skill set. The model is evaluating scoring and passing accuracy, aggression and vision on the field, speed, acceleration, ball control, physical condition, etc.
The salaries and the above mentioned 55 attributes were combined in a single computational model, allowing researchers to evaluate the salary of each player based on his skills in comparison to the skills of all other players in the same field position.
The computer model evaluated the skills of Lionel Messi and stated that he should be the best-paid football player, followed by Christiano Ronaldo, Luis Suarez, Neymar, David De Gea and Mesut Oezil. Well, the algorithm also states that Leo Messi should receive a salary as of €235 000 a week, which is about a third of his actual weekly paycheck. Such gap rated Messi as the most overpaid football player in Europe, followed by Angel di Maria, Robin Van Persie, and Ivan Rakitic. On the opposite side are the underpaid players led by Bernardo Silva with more than €100 000 difference between his actual weekly salary and the amount produced by the computer model. Following Silva as most underpaid football players in season 2016/2017 are Harry Kane, Granit Xhaka, Timo Horn, and Paco Aclacer.
It is important to underline that the data analytics model is not evaluating the marketing potential of each player in terms of merchandising, PR value, broadcasting rights, etc. It is based solely on their field skillset.
Interestingly, the study shows that the overpaid football players are exceeding the rest just by one attribute and it is the physical strength. Underpaid players are better than overpaid ones in term off agility, acceleration, speed, balance, etc.
The research of Lara Yaldo and Lior Shamir, data scientists from Lawrence Technological University also shows that the skills are in correlation with the players’ salary in all European leagues except in Poland where the prevalent factor for the salary size is the club. Digging deeper within data reveals some more interesting insights like the fact that the different European leagues appreciate a different set of skills. Penalties and vision are more respected in the Bundesliga, while the preferred foot is more important in England and France.
The study also suggests that if there is an objective quantitative method applied to salary determination it may have an impact on players` performance. It could also help for negotiation process simplification as well as determine of a uniform salary scale.
While this data analytics story on football players` salaries insights is more than interesting, we already know that there is a deep relation between sports and data sets. The football team of Milan has a data science team of its own called MilanLabs, ticket pricing is a product of data analytics and lot volleyball statistic is modeled via analytics algorithms for improved understanding of the game.
