Football match result prediction: A business analytics approach

• Autor(a) principal: Romão, Diogo Miguel Filipe
• Data de Publicação: 2023
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10071/30835
• topic: Football predictions; Machine learning; Investments; Previsões de futebol; Investimentos

Currently, there is a gap in the market for predicting football match outcomes before a match takes place. To address this challenge, a predictive model based on pre-match variables was developed, focusing on analytical techniques, including machine learning (ML). This study focuses on football match outcome predictions and their relevance in the fields of sports, investments, and entertainment. The goal is to identify the most effective algorithms for predicting outcomes. This study focus on five European leagues: the English Premier League (EPL), Spanish La Liga, Italian Serie A, German Bundesliga, and Portuguese Liga Bwin, for the seasons 2019-2020, 2020-2021, and 2021-2022. The predictive model proved promising in anticipating football match results. The C5 algorithm showed the best results, with an 85.40% accuracy on test data. Additionally, the model was validated in a new sports season, demonstrating its effectiveness in predicting future outcomes. This is essential to ensure the model reliability in future contexts and reinforces its utility for investors, football clubs, and fans. Through financial analysis, profitable strategies were identified, and a monitoring dashboard was built to track the financial performance of the predictive model and to evaluate the effectiveness of the developed investment strategies. In summary, this study contributed to the understanding of football predictions, benefiting investors, football clubs, betting houses, and fans by providing a robust approach to anticipate match results.