Statistics to change the game

For those wanting to predict their favourite player’s success, or make sense of team selection in sports, understanding the statistics off the field is crucial.

James Nash on soccer field

James Nash, who was in the first group of students to complete Victoria University of Wellington’s Master of Applied Statistics with Associate Professor Ivy Liu, has become an expert in exploring the applicability of various machine-learning algorithms for reject inference, and is now applying this knowledge to his favourite pastime: sport.

“I’ve always loved sports and grew up constantly having conversations with my friends about who the best players were, or which team we thought would win the game. With statistics, I have learnt how to create a ranking and justify my selections.”

As part of the Master of Applied Statistics, James was given the opportunity to undertake a placement at the storytelling data company, Dot Loves Data, as an analytics intern.

“My placement involved working 200 hours over a five-week period on a large project on the applicability of reject inference methods to a cricketing setting.

“Reject inference is used to estimate the likelihood of a bowler taking a wicket within an innings based on ball-by-ball data, and our model found that bowlers who deliver a high number of dot balls and a low economy rate have a higher probability of taking wickets.”

This research, rating the attacking performance of a nonwicket-taking bowler in limited overs cricket that James completed during his placement, was published in the popular Mathsport 13 proceedings, in collaboration with his supervisor, Dot Loves Data partner, Dr Paul Bracewell.

For now, James aims to keep working in analysing big data and says, “The dream is to be the lead statistical analyst for Liverpool Football Club.”