Modelling time-depended paired comparisons using soccer data
| dc.contributor.advisor | Sjölander, Morne | en_ZA |
| dc.contributor.author | Monareng, Joy | en_ZA |
| dc.date.accessioned | 2025-01-03T11:18:23Z | |
| dc.date.available | 2025-01-03T11:18:23Z | |
| dc.date.issued | 2023 | en_ZA |
| dc.description | Dissertation (M.Soc.Sc.(Psychology))--University of the Free State, 2023 | en_ZA |
| dc.description.abstract | This study presents a comprehensive analysis of soccer ranking models, encompassing the evaluation of existing models, the introduction of a new time-series paired comparison model, and a comparative study of model predictions against official soccer rankings. The investigation into the landscape of professional soccer rankings reveals the limitations of traditional FIFA rankings, prompting the exploration of alternative statistical models. The Linear Thurstone 3 (LT3) model emerges as a standout performer, demonstrating consistent and accurate predictions, particularly in the years 2023 and 2026. However, inherent challenges, including data limitations, tournament dynamics, model assumptions, parameter standardization, and non-round-robin tournament structures, underscore the complexity of predicting soccer outcomes. The study provides recommendations for future research to address these limitations, emphasizing the refinement of models, improvement in data quality, dynamic model adaptation, ensemble modelling, cross-sport comparative analysis, and continuous evaluation. This research contributes to the ongoing pursuit of precision and fairness in professional soccer rankings. The dynamic nature of the field necessitates continuous refinement and adaptation of models to enhance accuracy. By addressing the identified limitations and implementing the recommended strategies, future research endeavours can contribute to a more nuanced and reliable understanding of team rankings in the ever-evolving world of international soccer. | en_ZA |
| dc.identifier.uri | http://hdl.handle.net/11660/12896 | |
| dc.language.iso | en | |
| dc.publisher | University of the Free State | en_ZA |
| dc.rights.holder | University of the Free State | en_ZA |
| dc.subject | Bradley-Terry | en_ZA |
| dc.subject | Paired comparisons | en_ZA |
| dc.subject | Thurstone model | en_ZA |
| dc.subject | Time-depended | en_ZA |
| dc.subject | Correlation | en_ZA |
| dc.title | Modelling time-depended paired comparisons using soccer data | en_ZA |
| dc.type | Dissertation |
Files
Original bundle
1 - 1 of 1