Guillermo Martinez-Arastey

Project Title: The development of Expected Points in Rugby Union as a standardised metric of performance

Supervisory Team: Dr Matthew Robins and Dr Neal Smith

Project Summary: My PhD research involves the development of a new performance metric in Rugby Union that quantifies the value of a given match situation based on the probability that the team in possession of the ball will score or concede next. Historically, research in the field of Sports Performance Analysis has evaluated performance indicators in Rugby Union through discrete, descriptive and comparative statistics from cause-and-effect-based observations of single performance events. However, analysing such a dynamic interactive team sport is a complex problem that proves challenging to solve through univariate statistics. As a result, literature in Rugby Union has defined over 390 different performance indicators, showing a lack of consensus and consistency when establishing key determinants of success in the sport.

The aim of my research is to leverage modern techniques from the field of Data Science to bring a less binary approach to the view on performance in the sport. Using the learnings from similar established methods in other invasion team sports, such as the NFL and Rugby League, my research explores different machine learning algorithms to develop a metric that contextualises performance, brings a standardised analytical framework and facilitates the benchmarking of team performances under a common language. This approach centres around the idea that Rugby Union success is achieved by scoring more points than the opposition, which will subsequently lead to winning matches and competitions. Therefore, several models are explored to successfully predict the chances of scoring or conceding by a team in possession of the ball given the characteristics of a given match situation. Fluctuations in the probability of scoring or conceding from one play to the next are then used to evaluate performance by establishing whether a team’s actions increased or decreased their chances of success.

About: I completed my MSc in Sports Performance Analysis at the University of Chichester in 2019, after having worked in various analytical roles in the technology industry for 7 years. During my MSc, I was fortunate to be given the chance to get practical experience working in a top-tier Rugby Union club, where I learned both about the sport and the applications of Performance Analysis in the industry. I then decided to start my PhD to further explore the vast opportunities that the introduction of the advances in technology and analytics bring to the field of Sports Performance Analysis and Rugby Union.