Loading...
Repository DOI
Publication Date
2024-10-08
Type
doctoral thesis
Downloads
Citation
Abstract
In modern soccer (football), extensive motion tracking data are collected, capturing players’ movements at a rate of 25 times per second. Traditionally, these data have been utilised for tactical analysis, focusing on aspects such as team formation and identifying motion patterns. However, my research aims to leverage these data in a novel manner, prioritising player performance, injury management, rehabilitation, and player welfare. This thesis seeks to develop new statistical methods to identify personalised patterns of movement, enabling the creation of tailored training sessions that address the physiological demands specific to each player’s position. Additionally, by analysing the types of movements and associated physical forces, sports scientists can design rehabilitation programs for injured players more effectively. To this end, advanced modelling techniques are incorporated to enhance the analysis of motion tracking data. The bivariate generalised linear model (GLM) offers a sophisticated approach to jointly modeling angular change and speed change in player trajectories. By characterising trajectories into interpretable parameters, such as angular change and speed change, this model provides valuable insights into the underlying patterns of player movement. Furthermore, the bivariate GLM facilitates the clustering of trajectories based on the estimated parameters, allowing for the identification of similar movement patterns among players.
Publisher
University of Galway
Publisher DOI
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International