Elijah Cavan
Title: NHL Aging Curves Using Functional Principle Component Analysis
Date: December 19th, 2022
Time: 2:00pm
Location: Hybrid, over zoom and in LIB 2020
Abstract
All major league sports teams are interested in projecting the performance of their players into the future. The seemingly most important feature of a model to project future performance is age. On average, players tend to improve from their rookie (earliest) season in the league, until they retire from the league (due to poor performance or injuries, for example). In this project we apply Functional Principle Component Analysis (FPCA) to the careers of NHL players in order to fit individual aging curves for each player. We compare the results of three methods: ImFuncPCA, SOAP and PACE.