As a young man growing up in Portugal, Luis Amaral loves to play, watch and talk about football. Amaral and his friends vigorously argue about which players are the "best". But, it's just a matter of opinion.
Unlike baseball and basketball, there is not much statistical information to detail how each soccer player contributes to the match.
Amaral is now a professor at Northwestern University. His love for soccer then combined with his research team's computing skills to measure and rank soccer players based on an objective measure of performance, not opinion.
The results of his research have been published in June 2010 in the Public Library of Science (PLoS ONE). Amaral's analysis and his team were able to objectively release the performance ratings of all players in the 2008 European Cup tournament. Their research results were used to cover the match and became the basis for the team of experts, coaches, and managers who previously usually subjectively chose the "best" players in the tournament.
Amaral, a professor of chemistry and biological engineering at McCormick School of Engineering and Applied Science, says that in football there is relatively little to count and you can count how many players score, but if a player scores twice in a game, that's great. You really can only score two or three goals and two or three assists among potential eleven players.
To find a quantitative way to rank players, he and his students, Josh Waitzman, first created software to pull statistical information directly from the Euro 2008 website.
This type of broad statistical information usually only congregates for important matches, Amaral said. Amaral and Jordi Duch, assistant professors of applied mathematics and computer science at Universitat Rovira I Virgili in Spain, used data to measure players' performance by generalizing methods of social network analysis.
You can specify the relationships in which the network elements are football players. So you can have a relationship between each player they pass the ball. Moreover their goals are both to score, you can include other elements in this network, which is the goal.
The Amaral team mapped the flow of the ball between the players in the network as well as the firing information and analyzed the results.
They see the way in which the ball can travel and end up in shots. The more teams pass (run the ball) and end with a shot, the better the team. Then, whenever the ball is passed to then completed in a shot, the player will get a better score.
It will never happen by chance, that we will get striking results with data from so many experts if our size is not good.
He said this type of analysis can be used outside of the football world as well. Companies can use this method to rank and evaluate the performance of employees working in a team project, for example.
Given that this research was conducted four years ago, and with the growing science of statistical data analysis in football, Amaral's research is certainly a milestone in the growth and development of performance analysis in football.
Now we may just enjoy this as a free app or service on the website. However, the struggle of Amaral and comrades is what we should appreciate in which science and football are able to walk together and influence each other.
Journal sources: Jordi Duch, Joshua S Waitzman, LuÃs A Nunes Amaral. Quantifying the Performance of Individual Players in a Team Activity.