Data-driven evaluation of on-field player performance in football using sensor and video technologies

Abstract

Data has become increasingly relevant and used in football over the years. Technological development has made it possible to gather data from various aspects of the game. However, despite the growing popularity of sports analytics, relatively little research, especially qualitative, has been done on the topic. The purpose of this thesis is to create understanding and practices for taking advance of data for evaluation of on-field player performance in football using sensor and video technologies. This is done by identifying and combining technological possibilities with sports knowledge and suggesting an approach for data-driven evaluation of the on-field player performance. Review of previous literature and semi-structured theme interviews have been used as a method to achieve the purpose of the thesis. The findings of the thesis show that data can be used in the evaluation of on-field player performance in football by assessing players’ physical, technical, tactical, and mental attributes. These attributes have several different metrics, the value of which depends on several factors such as the team's objectives. Furthermore, an approach is presented in the thesis which suggests that the selection of team-specific attributes and metrics guides the user to consider which data is needed to be able to evaluate the desired metrics, which then can be linked to certain technologies and analytical solutions presented in the thesis.