Big data reveals how much footballers are really worth
The value of star footballers is a subject intensely debated among football fans, but big data might become the go-to method for determining the price tags of the stars. Researchers from the IT University in Copenhagen and University of Liechtenstein have developed an algorithm that calculates the value of players based on vast amounts of data.
Business IT DepartmentResearchbig dataalgorithms
Written 20 June, 2017 07:19 by Vibeke Arildsen
Football is a popular sport, but also big business. In 2016, English clubs alone spent more than a billion pounds on acquiring new players. The market values of players are discussed among experts, sports journalists and fans on websites like Transfermarkt, and these estimations play an important role in negotiations.
But in the future, big data might become the go-to method for assessing the value of footballers, according to researchers at the IT University in Copenhagen and the University of Liechtenstein. They have developed an algorithm that calculates the value of footballers using objective data on everything from goals to red cards.
"Our idea is to use big data to set an objective price, just as for instance financial analysts use data and statistical methods and not gut feelings to estimate the price development of stocks," says Oliver Müller, an Associate Professor at ITU and co-author of the study.
According to the big data algorithm, Danish Tottenham midfielder Christian Eriksen has been undervalued through the last four seasons. Right now, Eriksen’s estimated worth at Transfermarkt is 35 million euros, while the algorithm puts his current value at above 43 million.
Big data potential in the football world
In the United States, basketball and baseball clubs have long been using data to estimate the value of players, and according to Oliver Müller, the newly developed algorithm shows that a data-driven approach also has great potential in the world of football.
You get immediate feedback after each match and do not have to wait for lengthy discussions among fans and experts. This allows the clubs to know how much to pay for certain players.
Oliver Müller, Associate Professor at ITU
"With just a few taps on the keyboard, you can find the value of any player based on factual information about his age, position and performance. You get immediate feedback after each match and do not have to wait for lengthy discussions among fans and experts. This allows the clubs to know how much to pay for certain players," he says.
The algorithm also allows for hypothetical calculations. On average, a goal will increase a player's value by about 2.5 percent, for example, while an assist can boost the value by around 1.5 percent.
Finally, the algorithm can identify undervalued players on the market, which can be especially useful for clubs with smaller budgets.
How the algorithm works
The football algorithm is based on data about more than 4,000 players in the five biggest European leagues. It calculates values based on more than 25 factors, from age and height to statistics on goals, passes and tackles.
The algorithm also considers popularity, explains Oliver Müller.
"Some players are worth a lot because they are popular. Therefore, the algorithm also checks how frequently people search for the players on the internet and how often people post about players, e.g. on Reddit or YouTube. These things are also valuable since a popular player increases the sale of jerseys and tickets," he says.
Messi should cost even more
Perhaps surprisingly, the algorithm thinks that some superstars are even more valuable than current estimates suggest.
You get immediate feedback after each match and do not have to wait for lengthy discussions among fans and experts. This allows the clubs to know how much to pay for certain players.
Oliver Müller, Associate Professor at ITU
"Many people think that players like Messi and Ronaldo are overvalued, but according to our algorithm, they are actually worth even more because their performances are so exceptional. At the peak of his career in 2011-12, Messi scored 50 goals for Barcelona, and the crowd at Transfermarkt estimated his valued at 100 million euro. Our algorithm thinks that he was worth about 238 million at the time," says Oliver Müller.
The algorithm is not available to the public, but the researchers are currently looking for collaborators who can help move the project forward.
Read the paper 'Beyond Crowd Judgements: Data Driven Estimation of Market Value in Association Football' by Oliver Müller (ITU) and Alexander Simons and Markus Weinmann (University of Liechtenstein).
Oliver Müller, Associate Professor, email olmy@itu.dk
Vibeke Arildsen, Press Officer, phone 2555 0447, email viar@itu.dk