Mr Vincent Lee

Profile

I am a second-year PhD researcher in the School of Mathematical Sciences at Queen Mary University of London, supervised by Prof. Vito Latora. My research focuses on football analytics, specialising in the mathematical and data-driven modelling of football. Prior to starting my PhD, I completed an MSc in Data Analytics with Distinction at Queen Mary University of London.

One strand of my research focuses on on-field decision-making. While most existing football performance metrics evaluate the action that occurred, I develop counterfactual models that assess players’ decisions relative to the alternative actions available within the situation. My current work uses football event data to model shooting decisions by constructing a decision space of feasible shot placements and quantifying the observed decision relative to this counterfactual set. This work has been submitted to the MIT Sloan Sports Analytics Conference. I aim to extend this approach to other on-field decisions and develop a more general framework for quantifying decision-making in football.

I also study competitive relationships between football teams using network science and statistical modelling. This includes using graph-based centrality measures to rank teams, signed network representations to characterise uncertainty and instability in competitive relationships, and probabilistic models of match outcomes to estimate underlying team strength.

More recently, my research has expanded to longitudinal player development within professional football academies. I investigate how biological, physical, football-related and psychological characteristics interact and co-develop over time as a complex system. In particular, the study focus on distinguishing temporary developmental advantages from indicators of longer-term potential, and on understanding how different developmental pathways relate to subsequent player progression.

Supervisor: Prof Vito Latora

Research

Research Interests:

• Ranking and Rating Systems
  Bradley–Terry models, Davidson models, Elo-type rating systems, ranking algorithms

• Network Science
  Centrality measures, signed networks, directed networks, temporal networks, multilayer networks

• Probability and Statistics
  Statistical inference, probabilistic modelling, regression, Bayesian methods, stochastic processes

• Mathematical Modelling
  Dynamical systems, optimisation, latent-variable models, longitudinal models

• Statistical Learning and Machine Learning
  Supervised learning, high-dimensional statistics, predictive modelling