UEFA EURO 2020: a "pure game of chance"?

TL;DR

This study analyzes UEFA EURO 2020 match data, finding that goal scoring and waiting times fit Poisson and exponential models, highlighting the role of independence despite the game's complexity.

Contribution

It demonstrates that goal scoring in UEFA EURO 2020 can be modeled using simple probabilistic distributions, challenging the notion of football as purely chance-based.

Findings

Poisson distribution fits goal counts well

Exponential distribution models waiting times effectively

Homogeneous team abilities support model assumptions

Abstract

We analysed the distribution of the number of goals scored in each of the 51 football matches played in the UEFA EURO 2020 final phase as well as the waiting times between scores (also considering censored times). We found that the Poisson model fits the score data and the exponential distribution fits waiting times quite well. Such a good fit could be considered somewhat counterintuitive and unrealistic given the memoryless property of the exponential model. However, some peculiar features of this study have to be considered: the abilities of the teams were relatively homogeneous; the time span was short; there was no distinction between home and away games; only the total number of goals scored in each game was considered. Although UEFA EURO 2020 can certainly not be considered a "pure game of chance", this competition can be seen as an intriguing example of the pervasive real-world ubiquity of the concept of independence and its ability to encompass complex situations in a single, parsimonious model.