SKM 2023 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
SOE: Fachverband Physik sozio-ökonomischer Systeme
SOE 4: Poster
SOE 4.7: Poster
Monday, March 27, 2023, 17:00–19:00, P2/OG2
How random are team sports leagues? — Maciej Pawlik1, Michał Boruta2, Robert Paluch1, and •Janusz Hołyst1 — 1Faculty of Physics, Warsaw University of Technology, Warsaw, 00-662, Poland — 2Faculty of Mathematics and Information Science, Warsaw University of Technology, Warsaw, 00-662, Poland
To answer the title question, the historical ranking tables of four team sports leagues (Premier League, Bundesliga, La Liga, Svenska Hockeyligan) are analyzed. We investigate the predictability level of a given league by calculating the average absolute rank change ρ between two subsequent seasons of every team. The data are applied to calibrate an agent-based model of team sports leagues where every team possesses a certain strength equal to the sum of strengths of all players that have been selected to the team from a draft list. The sign of the difference between the strengths of two teams indicates the favorite of a given game, and there is a determinism parameter β ∈ (0, ∞) that controls the game predictability. We use the model to simulate many seasons of various leagues and to compute the average absolute rank change ρ of a given team between two consecutive seasons. Plotting ρ as a function of β reveals a continuous transition between phases of luck– and skill–based sports. Comparing the results of our model with the ranking tables of selected professional leagues, we show that they operate close to the transition line but in the deterministic phase. This finding suggests that random factors and abilities of team players are of comparable importance, with an advantage of the latter.