polito.it
Politecnico di Torino (logo)

Enhancing Football Scouting: Objective Characterisation of Players and Assessment of Transfer Impact through Machine Learning and Deep Learning.

Matteo Matteotti

Enhancing Football Scouting: Objective Characterisation of Players and Assessment of Transfer Impact through Machine Learning and Deep Learning.

Rel. Paolo Garza. Politecnico di Torino, Corso di laurea magistrale in Data Science And Engineering, 2023

Abstract:

In football, the process of players' identification still strongly relies on the subjective opinions of football experts, such as talent scouts or sporting directors. Especially for medium and small clubs, embracing data analysis represents an unprecedented opportunity to gain a more insightful and objective overview of virtually every player worldwide. This thesis proposes two distinct approaches that aim to enhance the transfer market. In the first section, called "players' embeddings", deep learning techniques are employed to represent football players as vectors in a lower-dimensional space. By leveraging this approach, it is possible to analytically characterise footballers, analyse their similarities and differences, and identify suitable replacements in the event of a player's departure. The second section, referred to as "players' adaptability", places greater emphasis on the qualitative aspects of a football player. Specifically, historical transfer and non-transfer data are utilised to train a model capable of estimating how a player's statistics would potentially change upon moving to a different team. This approach allows estimating the potential impact of acquiring a player and objectively assessing their current team statistics by accounting for the quality of their teammates and the quality of their opponents. The thesis will also show possible applications of the two models in action starting from plausible transfer market scenarios, demonstrating how the two approaches can unleash their full potential when combined together. This thesis was conducted in collaboration with Parma Calcio 1913, where the research was undertaken as part of an academic project with the support and guidance of the Performance and Analytics department. The club's involvement provided valuable insights and real-world context to enhance the applicability of the proposed approaches in the dynamic and competitive landscape of professional football.

Relators: Paolo Garza
Academic year: 2022/23
Publication type: Electronic
Number of Pages: 77
Additional Information: Tesi secretata. Fulltext non presente
Subjects:
Corso di laurea: Corso di laurea magistrale in Data Science And Engineering
Classe di laurea: New organization > Master science > LM-32 - COMPUTER SYSTEMS ENGINEERING
Ente in cotutela: Eidgenössische Technische Hochschule Zürich (SVIZZERA)
Aziende collaboratrici: PARMA CALCIO 1913 S.R.L.
URI: http://webthesis.biblio.polito.it/id/eprint/27736
Modify record (reserved for operators) Modify record (reserved for operators)