For die hard NHL fans, the debate over advanced statistics is usually a sore subject. How much weight should we put on these complex stats?When evaluating how your respective NHL team is performing in a given game, what criteria do you use?
Let me give you a scenario.
Your team is battling one of their rivals and the game is tied after one period of play. The intermission report comes on, but you already have your computer out to check in on how your team is really doing.
Those intermission reports are usually fluff, but i digress.
So how is your team doing?
Some people keep things simple and use only use shots on goal and the good ole’ eye test to evaluate their team.
Then there are those who used advanced statistics.
You know, those folks who post a whole bunch of numbers throughout the game to back up their claim of why player #1 is bound to turn it around, while player #2 is doomed because of these complex numbers.
If you have made it this far, then you probably know these numbers well.
Corsi and Fenwick.
Before we discuss the legitimacy of these stats, here is a very simplified and basic rundown of what they measure.
In a very simplified definition, Corsi (developed by former Buffalo Sabre goaltending coach Jim Corsi) measures whether the puck is directed more often towards the opposing or home goaltender while a given player is on the ice.
This stat is measured at even strength and is boiled down to a percentage.
For example, Dan Girardi of the New York Rangers had a Corsi % of 41.3 this past season. This means that the puck was only headed towards the opposing goal 41.3 percent of the time while Mr. Girardi was on the ice.
And then there is the Fenwick stat (named after writer Matt Fenwick).
Fenwick and Corsi are very similar, the only difference being that Fenwick doesn’t count blocked shots into the equation.
Fenwick believes that shot blocking is a skill set in itself, so he emitted it from the statistic.
Back to Girardi. His Fenwick percentage in 2015-16 was a brutal 43%.
So just how reliable are these statistics when measuring a player’s performance? Is it fair to label said player’s performance as “good” or “bad” based on these percentages?
Well sometimes, but you have to look at the situations in which the player usually plays. If the player is a monster defender then you can expect him or her to be inserted into defensive situations (when the faceoff is in his/her own zone).
This has to be taken into consideration.
Dan Girardi known for his defending abilities. He a poster child for what a defensive-defenseman should be. He is certainly no Drew Doughty or Keith Yandle.
By comparison, Patrick Kane tallied a Corsi For of 53.1% this past season. But Kane and Girardi play very different games, and we should expect that Kane will spend more time in the offensive zone that will his defensive counterpart.
This whole argument boils down to a simple chicken or egg comparison. Do you believe that a player causes a loss of possession, or do the situations that the player is inserted to equate to a low statistic?
The main problem which numbers people usually ignore is that there are thousands of variables that affect a player’s advanced statistic.
Variable such as the systems put in place by the coaches, the strength of the opposing team, a players given role, and injuries are just a few factors that contribute to the argument.
Now does it make sense to ignore the numbers altogether? Absolutely not. If a player is no playing up to par, then possession stats can add some insight as to why.
Likewise, saying that a player is performing well solely because of his or her Corsi/Fenwick stats is foolhardy.
In the end, a combination of the eye test and advanced statistics can go a long way to analyzing today’s NHL.
When we remain stubborn and claim that one of these is the only way, well, that’s where the pitfalls begin.