An important element for soccer’s vast popularity is its unpredictability. Since the goal is so rare, surprises due to many different factors can happen. Ajax’s successful run in the last Champions League to reach the semifinals was a welcome change from the previous years. Only the top clubs from the top four European leagues usually reach the quarter finals with Portugal to be the only country from the other leagues to have one of its teams reach the later stages. Before Ajax, there had not been a Dutch representative at quarter finals since 2007.
In a paper that was posted on the Cornell archives on August 23, 2019, Victor Martins Maimone of Oxford University and Taha Yaserri of the Alan Turing Institute in London looked at the predictability of the football during the last 26 years. In particular, they looked at the matches of 11 major European leagues (Belgium, England, France, Germany, Greece, Italy, Netherlands, Portugal, Scotland, Spain, Turkey) which had a winner and they checked if one could establish the winner using the teams’ performances of the previous games. To avoid the obvious limitation arising from the fact that the set of previous teams is different for each team, the two researchers established a directed network of matches and created a fair weighting system to establish the probability of winning the next match. Using their rules, they established that the predictability has increased over the years in seven leagues (England, France, Germany, Netherlands, Portugal, Scotland and Spain) and has remained almost stable in two leagues (Belgium and Italy). In the remaining two leagues (Greece and Turkey), the predictability was high in the past, then dropped and in recent years there is an increasing trend. [Looking at the data, my personal view is that the curve of the Italian league follows the same pattern as that of the Greek league.]
The data also demonstrated that, although the home advantage still exists, it decreases over time for all leagues.
Finally, the trends of predictability are increasing very fast for the richer leagues (England, Germany, Spain). [Why the Italian league (which completes the top four leagues) does not follow the same pattern is a curiosity that deserves to be studied.] This is something we all know through our experience: Richer clubs win games with greater frequency and we are not surprised when they do. To establish this correlation between predictability and influx of money, the researchers used the so-called Gini coefficient which is a way economists use to measure wealth inequality. The result is as expected. To say it in the words of the researchers: “There is a high correlation between predictability in a football league and inequality amongst teams playing in that league, uncovering evidence in favour of the `gentrification of football’ argument.”
Read the full paper “Football is becoming boring; Network analysis of 88 thousands matches in 11 major leagues” at https://arxiv.org/abs/1908.08991