Sports Analytics Club: The Peak stat of the week

Looking into “Corsi” and what it means for hockey

Dani Chu and Matthew Ryers explain some of the major sport analytics statistics. Arshika Chandranath / The Peak

By: Dani Chu and Matthew Ryers

With the advancement of technology in the sports world, new stats pop up almost daily to try and better explain individual and team performance. In this column, we plan to discuss the history, usefulness, and flaws of advanced statistics in sport. We will be covering a wide range of sports including everything from baseball and basketball, to hockey and soccer. This week we will break down the first heavily adopted advanced statistic in hockey — Corsi.

The focus of Corsi is to measure the difference in offensive opportunities between the two teams on the ice. It does so by counting the difference in shot attempts. The reason to consider shot attempts instead of counting only shots on goal is that not all shot attempts generate a shot on goal despite still being an offensive opportunity. A Corsi value for each team is calculated based on the difference between their shot attempts and their opponents. Positive Corsi scores indicate the team spends more time in the offensive zone while negative numbers reflect the opposite. For example, the Vancouver Canucks have a Corsi score of -190 on the season, averaging out to be -3.96 per game. In comparison the Buffalo Sabres, a team with fewer goals for and more goals against than the Canucks, are currently averaging a -4.75 Corsi rating per game. Because of the success of Corsi, recent expansion have been made to measure player Corsi. This modified calculation is then based on shot attempts only when the given player is on the ice. Some of the exceptional individual Corsi totals amongst players who have played more than 25 games this season include Artemi Panarin, Dougie Hamilton Brandon Saad, Jonathan Toews, and Mark Giordano.

As for the history, the statistic was first modelled by Vic Ferrari, the creator of the blog “Irreverent Oiler Fans” — a hotbed for hockey analytics ten years ago. The statistic is named after Jim Corsi, a goalie coach of the Buffalo Sabres, simply because Ferrari liked a picture of Corsi’s mustache on the Buffalo Sabres website. If not for that brilliant moustache, this statistic could have easily been named after the former head coach of the Buffalo Sabres and Dallas Stars, Lindy Ruff. The obsession with naming this newly formalized statistics after Buffalo Sabres’ management is due to Ferrari hearing the outline of this idea on a radio show featuring the general manager of the Sabres.

Finally, the stat is not without its faults. While Corsi does a good job measuring puck possession, and team strength, individual player Corsi is flawed in that it does not isolate a single player. A less skilled player who plays with more skilled players will often have a better Corsi than a skilled player playing with less skilled players. This happens because the less skilled players have their Corsi inflated by their more skilled teammates. 

Thanks for exploring Corsi with us, we hope that you join us as we travel across sports looking at the newest and greatest advancements in player evaluation and strategy.