Great Lakes Analytics in Sports Conference
Texas Tech University
Outshooting Opponents in Scoreless First Periods
and Game Outcomes in NHL Playoff Hockey
Much of hockey analytics focuses on shots rather than goals. Although shots don’t contribute to winning quite as directly as do goals, shot differential (i.e., outshooting one’s opponent) appears to reflect predominance of puck possession, and is correlated with winning. Further, shots have some statistical properties (i.e., greater frequency and variability) that make them preferable to goals. However, the validity of shot differential can be obscured by “score effects,” as when a trailing team fires many shots at the net in a desperate attempt to rally (i.e., high shot volume in this instance correlates with losing). One can laboriously examine play-by-play sheets to restrict one’s analysis to shot attempts taken with a tie score. However, a potentially more manageable way of ruling out score effects is to study games in which no scoring takes place in the first period (or first and second period), to see if shot advantages of various magnitudes while a game is scoreless presage victory in the game. The present study offers a small-scale analysis of NHL playoff games over three seasons (2015, 2016, and 2017), to see if outshooting one’s opponent in a scoreless first period (and by what magnitude) predicts increased likelihood of winning the game. Sixty-eight games were identified in which no first-period scoring occurred. Games were divided into three categories to simplify analysis: those in which one team outshot the other by one or two shots in the first period (n = 21 games), outshot the other by three or four shots (n = 21), or outshot the other by five or more shots (n = 26). Games were also classified as to whether the team with more first-period shots won or lost the game in regulation time or went to overtime (a regulation tie). It was hypothesized that the larger the advantage in first-period shot differential, the greater the likelihood that the heavier-shooting team would win. Teams that outshot their opponent by one or two shots during a scoreless first period won in regulation only 24% of the time (losing 52% and tying 24% of the time). In contrast, teams that outshot their opponents by three or four shots won 57% of the time (losing 33.3% and tying 9.5% of the time), and teams that outshot their opponents by five or more shots won 50% of the time (losing 31% and tying 19% of the time). In conclusion, outshooting one’s opponent during a scoreless first period -- and doing so by larger margins -- appeared to show which team was better that night, in terms of ultimate game outcome.
Back to Great Lakes Analytics in Sports Conference