Drafting Defensemen's Effect on National Hockey League Outcomes

Abstract

Any draft in the league’s “salary cap era” (2005– Present) would go differently with hindsight & with knowledge of the ultimate realized value of players. This study aimed to explore whether National Hockey League (NHL) teams that choose to draft defensemen in the first and second round of the draft more often will perform better as a team. The study investigated data from 12 seasons 2007-08—2018-19 in the National Hockey League with 30 teams per year. The study incorporated several independent control variables and conducted an OLS regression on defensive draft investments’ ability to predict regular season winning percentage. The OLS model was a generally good fit with an adjusted R 2 of .632 (F(15, 344) = 42.04; p<.001). The results indicate that defensemen selected in the draft’s first two rounds are responsible for a tangible effect on NHL team outcomes. It was found that the number of Draft Selections had a statistically significant relationship with regular-season winning percentage (p=.006) and each additional defenseman drafted in those rounds corresponded to a 0.7% improvement in a team’s season winning percentage (p<.001). Additionally, for every standard deviation increase in the use of non-drafted defensemen (free agents or transfer minutes), you would expect to see a reduction in team winning percentage of 4.6% (p=.003). A logit regression was also used to analyze the same variables’ ability to predict a playoff berth (Model fit: χ 2 (15,360) = 205.22, p<.001). While traditional performance factors of Offense Quality and Goalie Quality were significant predictors of playoff qualification (p<.001), the defensive draft variables were not significant. Offensively-minded defensemen also did not present a statistically significant effect on winning percentage or playoff berths for NHL clubs in either model