Corsi: An Overrated Statistic

Before I begin this article I want to clarify that Corsi and Fenwick are not unreliable statistics. They are both good predictors of how players and teams will perform over the course of a season or during a playoff series. The problem is that hockey personalities are beginning to rely solely on these statistics while overlooking some glaring flaws. On top of that they are ignoring other statistical categories entirely leading to the misinterpretation of how players should be perceived.

What is Corsi?

Corsi Statistics
Former NHL goaltender Jim Corsi created the Corsi Rating while coaching in Buffalo.

Since Moneyball was published in 2003, advanced stats have spread into other sports and beyond including criminal justice, music and even government. This effect along with the creation of the Corsi number has everyone from fans, writers, scouts and managers analyzing players much differently than they did a decade ago. Unfortunately, everyone is beginning to overlook statistics that were already in place while focusing solely on Corsi if it were the only way to judge a player.

Corsi Number= (Shots on Net For + Missed Shots For + Blocked Shots Against) – (Shots on Net Against + Missed Shots Against + Blocked Shots For.) Corsi is only calculated for 5v5 play.

Fenwick is the same except it removes the use of blocked shots from the equation. Still, calculating the Corsi number is not enough to analyze a player. From there that number is put into another equation to better determine their value to the team. The Relative Corsi number drawn from this equation essentially tells us if the player plays better or worse than the team average.

Relative Corsi Number = Corsi Number of player – Corsi Number of team when player is not on the ice.

Before we even begin to talk about the numerous problems with these statistics let’s talk about what Corsi truly is. At its very core, Corsi is a plus minus statistic. Albeit, a far better statistic but a plus minus statistic nonetheless. As discussed by David Staples of the Edmonton Journal, Plus Minus has two major problems but Corsi only solves one of these problems, the sample size. Also despite being painted as the face of advanced stats, Corsi is nothing more than a shot counting statistic.

Bias Within Corsi Statistics

Now that we have established what Corsi actually is, a shot counting plus minus statistic, let us focus on the problems. The strengths of Corsi are also its weaknesses.  The fact that it only encompasses 5v5 play is a major drawback that cannot be overstated. Most people think this only excludes special teams, but it also excludes any and all 4v4 play as well. In 2014 the league average was 3.27 Power Play opportunities per game or about 6:17 minutes of special team play. There are no statistics for 4v4 play time but we can assume that per game the average is between 1-2 minutes because of the occurrence of offsetting minor penalties. That is 7-8 minutes of data per game that is blatantly ignored by Corsi. It punishes players who are penalty kill or power play specialists. For instance a fourth line player plays three minutes on the penalty kill and another four on 5v5. Is it truly okay to judge this player based solely on that? What about a power play specialist who gets victimized on 5v5 more but makes teams pay on the power play constantly? In the end games are won by goals not shots.

Quality vs Quantity

(Photo: Andy Martin Jr)
Josh Gorges was 6th in the NHL in blocked shots and had above average Goal Differentials despite lower Corsi statistics since 2012.

The next problem with Corsi is perhaps the most obvious, the shots themselves. Corsi is solely based on counting shots and judges players on the quantity of these shots but not the quality. Wayne Gretzky once said, “You miss 100% of the shots you don’t take.” Still this should not translate to the justification of a boosted Corsi number for missing shots. It is true that a missed or blocked shot can lead to a goal, but that is an indirect path. Hits, takeaways, passes, and player movement off the puck also indirectly lead to goals not just shots and to assume only shots lead to goals is ludicrous.

“What matters is winning, winning requires goals, and a high volume of shots does not, strictly speaking, create goals. Shots are a by-product and not a cause.” via Daniel Wagner of TheScore.

Just as blocked and missed shots should not be judged the same as those put on the goalkeeper, shots on net are not created equally. A shot from outside the zone or a soft wrist-shot from the blue will not have the same quality as a one timer from the slot or a breakaway shot. Typically during a thirty shot performance only about ten of those shots are even considered scoring chances, or quality shots. Also by only assessing shots Corsi ignores goals, or more specifically ignores defense. Perhaps a player did allow more shots against but he forced them to be wide or forced bad angle shots with a low chance of them getting by the goaltender. Corsi does not allow for the analysis of how well a player may be playing defensively in the zone despite allowing these shots.

Half-Truth

The final problem with Corsi is simply due to the nature of the sport itself. Is it truly fair to punish a player because he plays on a bad team or reward a player because he plays on a good team? Good players on bad teams will suffer from this statistic simply because they will never be able to muster up enough shots while bad players on good teams will look far better than they actually are. As stated earlier Corsi is designed to be a comparison of a player to the team average. Because of this design it is one of the reasons why certain players eventually get overvalued and overpaid later. Corsi does well to predict how a team may do in the playoffs after the season is over but it can independently skew data for individual players.

The Whole Picture

(Bruce Fedyck-USA TODAY Sports)
Since 2011, Marian Gaborik only had average Relative Corsi Ratings, but high Goal Differentials. He proved to be a major difference in winning the 2014 Stanley Cup.

This brings us back to the statistic itself and the overzealous act of judging a player. Corsi and Fenwick are not untouchable statistics and they do not tell the whole story. To form a conclusion on a player you have to analyze multiple statistics that are both offensive and defensive. Lately, everyone talks about how the 2014 Los Angeles Kings proved the importance of Corsi. I would counter that argument with the fact that the Kings continued what we already know, teams who finish in the top ten in 5v5 Goals For/ Against Ratio have a much better chance at winning the Stanley Cup. Since 2001, only the Kings in 2012 and Hurricanes in 2006 finished outside the top ten in this category while winning the Stanley Cup. To assume that Corsi is the only indicator of a good team is to ignore every other statistic that teams excel at, including statistics that pre-date Corsi.

Corsi does not have as many flaws as Plus Minus and still helps to show how players can or will perform. It is possibly the best predictor of future events but not the best way to assess a specific player. Goal differential is just one statistic that has begun to be overlooked in recent years. When analyzing players make sure to assess other things such as blocks, time on ice, special team play and other categories. This combined with Corsi can lead to a much better understanding of a free agent or prospect.

9 thoughts on “Corsi: An Overrated Statistic”

  1. Andrew, outstanding, balanced article! Running google searches for the past couple days attempting to find something of this nature and yours is simple and elegant. Well done!

    What I have found in the search (and debating in the wild at Broad Street Hockey) is what you capture perfectly: “The overzealous act of judging a player”.

    There has been a gold rush of data in which so many folks have attempted to stake their claim with a spreadsheet and a click.

    There is no substitute for the experienced eye…but there is nothing wrong with those eyes needing a little help with magnification…and that’s the balance that you and Dubas are pointing to.

    eyetesthockey.com

  2. A healthy blend of eye-testing and analytics is the way to go. It’s difficult to invest a whole lot of merit into real-time statistics because the NHL’s scorers record those stats subjectively; it’s open to human-error, and their scoresheets rarely match at the end of a hockey game.

    There are two tangible statistics in ALL of hockey; goals and assists. The rest are dependent on how one (or all) person interprets them. Statistics and analytics aren’t the problem; human-error and interpretation are the problem.

  3. I’m pretty sure he was hired by Toronto to analyze a wide range of stats beyond Corsi.

    I’d agree with the author in this instance. The eye test still beats a lot of these fancy stats. There are a ton of weak shots that get counted into the Corsi # that have little to no chance of making it past the goaltender. I long for the day when we can figure out a way to measure true possession, the actual amount of time per game a player/team is in control of the puck.

  4. Since I’m sure more fire and brimstone shall rain upon me in the future I wanted to address this here. I defend Corsi within the article, on multiple instances. All I am saying is that like all other statistics it is not infallible, not perfect.

    It is fine to disagree, to be honest I welcome it since it encourages discussion and forward thinking as long as it is constructive. Having the ability to form your own opinion is why I enjoy writing. So please do not take this article too seriously if you disagree. And if you agree, don’t use it as a basis for your own opinion, you have to come to that on your own terms.

    • Oh there are A TON of research articles, pulled from data from Behind The Net that show Corsi to be reliable and it indeed is. That is why I stated that Corsi is not an unreliable statistic. All I am saying is that like all statistics it is not infallible, not perfect.

      • “Corsi” has become more of a buzzword in the hockey community lately. I agree with the points you made. I suppose I could offer you counter-points to each point you did make, but you did well in addressing it’s overuse.

        Individuals should shouldn’t judge a player on Corsi alone.

  5. Welp, Kyle Dubas was just hired as assistant GM for the Leafs. Guess some people don’t think Corsi is overrated…

    • You’re missing the point. Even Dubas himself said he considers a wide variety of advanced statistics, and blends them with old fashioned eye-testing. The author has done nothing but effectively argue that Corsi/Fenwick is not the be-all answer. It’s not overrated unless one subscribes solely to it to judge a player universally.

Comments are closed.