Fitness, chance, and myths: an objective view on soccer results

Abstract

We analyze the time series of soccer matches in a model-free way using data for the German soccer league (Bundesliga). We argue that the goal difference is a better measure for the overall

Dynamics of tournaments: the soccer case

“…where a and b are essentially variances related with the teams' fitness and statistical fluctuations [24]. Here we have a ≈ 0.075 and b ≈ 1.55.…”

“…Recently, by considering the Bundesliga, a careful investigation pointed that the quality of a team is better described by goal difference than the number of points and that it is constant over each season [24]. However, due to minimalist approach desired in this work, our model does not take some particular ingredients into account, such as, temporal variations over the seasons, home advantage, team-specific characterization, and specific details about the quality factor (e.g., goals difference instead of number of points).…”

“…(6) is an attempt, motivated by the data, to promote a good adjustment of the model by using a minimal number of parameters. Of course, other forms for Q(m) may be employed, for instance, we could also consider it based on a normal distribution, log-normal distribution [32] or linear combination of two normal distributions [24]. When considering our data, these possibilities do not give a significant improvement in the results.…”

“…Our present study attempt to fill this hiatus by using a random walk-like model which reproduces some of these important features. We basically employ the statistical framework based on random walk interpretation of soccer leagues used by Heuer and Rubner [24] (see also Ref. [25]).…”

Die Drei-Punkte-Regel in der deutschen 1. Fußballbundesliga und der Anteil unentschiedener Spiele

“…; w=0,08. 7 Insgesamt gibt es -bei der gewählten Variante der Adjustierung wegen der Saison 91/92 -vier weitere Intervalle (Intervalle 7, 8,9,11) mit einem Unterschied auf dem 5%-Niveau (vgl. .…”