This project explores whether a goaltender’s height meaningfully influences their performance in the NHL. Using publicly available NHL data, I analyzed over two decades of goalie statistics, prospect rankings, and team defensive metrics. I examined trends in goalie height by season and birth country, and modelled performance using save percentage (SV%) and goals against average (GAA) as outcome variables.
To assess the relationship between height and performance, I used a combination of generalized additive models (GAMs), decision trees, and random forests in R. While NHL teams have increasingly favoured taller goalies, the modelling results show that height alone is not a strong predictor of SV% or GAA. Instead, team context and other factors play a larger role.
The analysis combines data wrangling, visualization, and interpretable machine learning to provide a data-driven perspective on goaltender evaluation and scouting trends.
A PDF version of the report (along with all the files used to produce it) can be read and downloaded from the github repository.
An interactive version of the report can be read on this website by navigating to the Report page.
Data was gathered and validated from the following sources: