clock menu more-arrow no yes

Filed under:

By the Numbers: Investigating the Relationship Between Corsi and Goal Scoring

New, comments

The Hurricanes constantly sport some of the best possession metrics in the league but still struggle to reliably produce offense. What’s going on?

Jamie Kellner

The Carolina Hurricanes have been frontrunners in the hunt for the Corsi Cup season in and season out for what feels like a decade. Anyone in this market could tell you that their possession dominance hasn’t yielded favorable results, but rather disappointing stretches of play in which the team is labeled “unlucky”. When a team is this unlucky for a matter of years, it’s natural to begin to question the validity of a statistic that’s advertised as a pretty good metric of a team’s competiveness level.

The Hurricanes lead the league in Corsi For percentage by nearly three percent this season but are an abysmal 30th in the league in goals scored. This narrative isn’t new — since the beginning of the 2014-15 season, Carolina comes in at 28th in the league in regular season goals scored. But in that same time frame they’re second best in the league in CF%, and own a league worst 5 on 5 shooting percentage at 6.7%.

With all of the difficulty that the Hurricanes have experienced over the past few years that contradicts the team’s generally impressive metrics, it oftentimes feels like they play hockey in a parallel universe — a universe with an absence of mathematical logic and reason. I wanted to take a look at the relationship between Corsi and goal scoring in a league wide scope and maybe reveal a greater trend.

The Hurricanes have notoriously outshot their opponents this season and in seasons prior; this is a subject discussed ad nauseam. In looking at the data I wanted to separate the rate at which teams outshoot opponents and instead look at only how much they shoot, ignoring any differentials between their performance and that of their opponent on a given night. To address that I collected only Corsi For, which is defined as any shot attempt for, including goals, shots on net, as well as missed and blocked shots.

Above I have plotted Corsi For against Goals For over the past two full seasons and this season up until December 12th. My immediate observation is that there appears to be a loose trend throughout the league. To put it broadly, teams that shoot more usually score more. That makes sense. But my emphasis in that sentence is on the word usually, because I’m hypothesizing that there is a general point of diminishing returns.

Looking at the data, it seems like teams that score the most goals in the league over the past few years are somewhere in the 60th to 80th percentile in Corsi For. Maybe those teams have uncovered the Goldilocks Zone of shot quality. Maybe their coaches stress a style of play that emphasizes rush shot attempts or odd man situations. Or maybe they just have better finishers. I don’t know the answer.

I’ll do my best to give a ten thousand foot interpretation of this data (because honestly if I try to get more involved I just get confused). To me, the relationship between these data points indicate that more is usually better, but at some point it’s actually worse. When a team is focused on generating an exceptionally high volume of shots on goal, some level of shot quality and randomness is lost.

Unpredictability in this league is valuable, and less of this favorable quality in the offensive zone makes things easier on the opposition. This may explain the “hot goaltender” narrative as well as the team’s inability to consistently turn shots into goals over the past five years.