If you follow hockey at all you have probably heard the term/phrase “corsi” or “fenwick” or “possession drives production” a few hundred times. These advanced statistics have seemingly created a civil war-like atmosphere among the hockey community and it breeds a childlike or high school-like grudge match. I am here to tell you once and for all that it is not as black and white as both sides make it out to be.
First off, the basic information you need to know is that corsi/fenwick are fancy analytics in the hockey world that track shot attempts while a player or group of player is on the ice. There are specific instances of this (i.e. corsi in all situations, 5v5, etc…) but there is no need to dive into the specifics at this point. The only thing that is relevant is how the one side, the Pro-Fancy stat community, views these stats. Pro-Fancy stats people act just like kids in middle school/high who think they are always right no matter what. They refuse to understand where the other side is coming from and they just think that their ideology is correct because, well, they said so and “LOOK AT OUR CHARTS!!!!” They will sit there until they are blue in the face telling you that the team with the most shot attempts (highest corsi/fenwick) are going to win 100% of the time. Any rational hockey fan knows that this isn’t always the case. Sure, this logic makes sense if you live in a vacuum where only one metric is counted to determine a victor but that is NOT the case. When this is pointed out to the analytics people, they quickly dismiss the argument as either invalid, being a small sample size, or they quickly turn and blame another aspect as to why their models didn’t prove anything.
A big defense the analytics crowd uses is they cherry-pick certain teams as their “poster boy(s).” For example, the Pittsburgh Penguins from 2015-2016. This Penguins team either led or was towards the top of the league in corsi% and they always throw this out there to “prove” that FANCY STATS ARE DA BEST! However, with a team comprised of players like Crosby (best in the world), Malkin, and Kessel I would HOPE their team has great possession/shot total statistics. Saying a team is good because they have good corsi is literally trying to prove something by using itself. As someone who has majored in multiple STEM fields, it pains me to see someone do this because you CANNOT use something to prove itself. That’s just not how it works. Other examples this community likes to use are the LA Kings and Florida Panthers. First of all, the Kings and Panthers are fighting for their playoff lives. Second of all, the last few years when the Kings have won the cup, it wasn’t their “incredible corsi” that stood out…it was the clutch performance of goalie Jonathon Quick and skater Alec Martinez, among others. Third of all, the Panthers fired their head coach this year, brought in a new regime that catered to analytics and now they are still fighting for their playoff lives (and newsflash…it’s not looking good). Of course the pro-stats community writes this off as an anomaly or something else from left field that really just is them refusing to admit they MIGHT BE WRONG. The new Panthers coach has recently gone on record saying Corsi/Fancy stats do not tell the whole story and anyone who thinks they do does not understand how the game is played.
From Florida’s coach.
From Darryl Sutter and The Kings.
**Honorable mention** The Hurricanes have a Top-5 corsi in the league last I checked and they are towards the bottom of the league in the overall standings…but ¯\_(ツ)_/¯
Another aspect people do not understand is that fancy analytics are changing the way players are playing the game…and arguably not in a good way. A player on the Carolina Hurricanes a month or so ago was talking to reporters about corsi and player contracts. They mentioned how these possession metrics were being used as negotiation pieces for both players and GM’s/teams. He said that because of this, players are going to change the way they play the game in order to do what’s best for them and get the biggest possible payday. This means taking shots whenever they can on the ice, even if it does not benefit the team by taking a large number of incredibly low-percentage shots (yes a clear sighted shot into the goalies pads with no screen counts as a positive corsi event).
Corsi and other possession stats are a GREAT tool to use in player management and personnel decisions…but there needs to be a context attached to them. This aspect is something that the fancy stat community tends to ignore completely. Take Adam Clendenning for the NYR for example. He is a journeyman defenseman who has good possession stats/is a puck moving defenseman. The analytics crowd love him but he is awful in his own end. Apparently that doesn’t matter to the analytics community since their go-to excuse is “if you have the puck more you won’t have to defend as much!” This logic just screams stupidity. This isn’t the 2000’s Devils or a defense filled with Bobby Orr’s.
Credit: Sports Illustrated
Steve Valiquette, from MSG, uses new analytics in his own way but has gone on record numerous times saying these new analytics are very misleading and do not paint the whole picture unless they are viewed and utilized properly. There NEEDS TO BE CONTEXT. At the end of the day, I will trust and believe a former NHL player (especially a goalie) who saw thousands of shots in his career over someone on social media who has never played hockey above a high school level or at all who does nothing but throw charts and graphs in my face saying “SEE?!!?”
As a diehard NYR fan, how many times have we seen the Rangers blown out in shot attempts but come away with a victory? If you ask the fancy stat people, they won’t have a solid reasoning for this other than “HANK CARRIES THIS TEAM!” (which he does, he is amazing, but that is not the point here). This instance which happens time and time again just shows that there is more to hockey than just shot attempts for and against.
People think these stats help you predict what will happen in games. If this was the case, why even play the season if it’s already predicted? It’s because there are other things that play into results. These stats don’t help you predict…they help you ASSUME things. Bottom line…corsi is not the end all be all of the NHL. There are a LOT of different aspects that go into building a successful team. Corsi IS ONE OF THESE ASPECTS…but it is NOT the only aspect. This concept is anything but black and white.
Author: @BenR_Experience via Twitter.