Abstract
Multiple entry drafts exist as part of the pathway between amateur and professional hockey in North America. However, decision making surrounding the selection of goaltenders and their pathway to the National Hockey League (NHL) remains limited. This study explored the associations between playing statistics, selection into the NHL, and participation in the NHL for goaltenders selected in Canadian Hockey League (CHL) drafts between 2005-2015. Data for 795 selections on 792 unique goaltenders in CHL drafts were assessed. Two hundred seventy-six goaltenders played greater than five games in their respective CHL league in their first eligible NHL draft year, or the year in which they were drafted. Logistic regression models were used to identify predictors of NHL selection, revealing that goaltender height (ORs: 1.18-1.31) and goaltender games played (ORs: 1.11-1.14) were associated with increased odds of selection. A logistic generalized additive model exploring the odds of playing at least one game in the NHL showed that career CHL save percentage was strongly associated (OR 8.96, 95% CI 3.59 – 22.4) with NHL participation, alongside height (OR 1.12, 95% CI 1.01-1.24) and games played in a goaltender’s NHL selection year (OR 1.09, 95% CI 1.04 - 1.14). Overall, this model explained 49.1% of the variance seen in playing in the NHL. These findings highlight that some overlap exists between selection and participation in the NHL. We offer some hypotheses for possible decision making in this context, and discuss future directions which may help further optimize evaluation, selection, and participation of NHL goaltenders.
