Application of Classification Methods on Fantasy Premier League Data

James Michael Hodge

Diplomová práce

Anotace:

Fantasy Premier League (FPL) allows participants to roleplay as a manager in the English Premier League (EPL) and assemble their own team of real players to compete against other managers. As the popularity of FPL continues to rise, managers are seeking new ways to gain an advantage against other competitors. The following thesis aims to utilize cluster analysis methods on real data from the 2021-2022 season to gain insights into the underlying structure of FPL Goalkeepers and Forwards, as well as to objectively determine which players should be selected. Two modes of FPL, FPL and FPL Draft, will be considered when performing the cluster analysis and discussing the players and strategies that should be considered for team selection. Specifically, agglomerative hierarchical clustering and k-means are used with variables relating to FPL performance, finding distinct clusters and tiers of players that should and should not be considered for team selection. A supplementary goal of this thesis is to compare the performance of agglomerative hierarchical clustering and k-means clustering of FPL data. This thesis provides the framework for how to use cluster analysis in the context of FPL for the additional positions of FPL Defenders and Midfielders.

Abstract:

Fantasy Premier League (FPL) allows participants to roleplay as a manager in the English Premier League (EPL) and assemble their own team of real players to compete against other managers. As the popularity of FPL continues to rise, managers are seeking new ways to gain an advantage against other competitors. The following thesis aims to utilize cluster analysis methods on real data from the 2021-2022 season to gain insights into the underlying structure of FPL Goalkeepers and Forwards, as well as to objectively determine which players should be selected. Two modes of FPL, FPL and FPL Draft, will be considered when performing the cluster analysis and discussing the players and strategies that should be considered for team selection. Specifically, agglomerative hierarchical clustering and k-means are used with variables relating to FPL performance, finding distinct clusters and tiers of players that should and should not be considered for team selection. A supplementary goal of this thesis is to compare the performance of agglomerative hierarchical clustering and k-means clustering of FPL data. This thesis provides the framework for how to use cluster analysis in the context of FPL for the additional positions of FPL Defenders and Midfielders.

Klíčová slova

Fantasy Premier League (FPL), English Premier League (EPL), cluster analysis, agglomerative hierarchical clustering, k-means, Goalkeepers, Forwards

Keywords

Fantasy Premier League (FPL), English Premier League (EPL), cluster analysis, agglomerative hierarchical clustering, k-means, Goalkeepers, Forwards

Jazyk práce: angličtina

Datum vytvoření / odevzdání či podání práce: 26. 6. 2024

Obhajoba závěrečné práce

• Obhajoba proběhla 21. 8. 2024
• Vedoucí: Zdeněk Šulc
• Oponent: Jana Cibulková

Vysoká škola ekonomická v Praze

Magisterský studijní program / obor:
Economic Data Analysis / Data Analysis and Modeling