Free Report: 10 Powerful Technical Chart Formations

Advanced Hockey Stats - An Introduction

One of the most interesting, and sometimes polarizing, developments in hockey is the growing usage of advanced statistics, also known as “fancy stats”. Before your eyes start to gloss over, please don’t let the monikers fool you. There is nothing particularly “advanced” or “fancy” about these stats.  Such metrics require nothing more than middle-school math to calculate, but provide a layer of insight that was previously unavailable.

Take the “+/-“, or plus-minus, as an example. In essence, a player gets a plus every time he is on the ice for an even-strength or short-handed goal. Conversely, a player for the team that got scored against, whether it is even-strength or short-handed, gets a minus.  This stat is often cited as a way to distinguish good defensive players but it is clear that it is lacking in descriptive value. It tells us nothing about a player’s quality of competition, quality of teammates, usage, and ability to drive possession, all important contextual information that is lost in the traditional plus-minus stat.

This brings us to advanced statistics. In this article, we’ll be covering some of the most popular possession, usage, and quality of competition statistics.

(Some of the best advanced stats resources out there:,, and  

Possession Statistics – Corsi and Fenwick

Two of the most popular possession stats are Corsi and Fenwick, and they serve as proxies for possession. Corsi is very similar to plus-minus but has one big advantage, a much larger sample size (more shots than goals). The math is rather simple:

Corsi = (SF + MSF + BSF) – (SA + MSA + BSA)

Where: SF=on-ice shots for, MSF=on-ice missed shots for, BSF=on-ice blocked shots for, SA=on-ice shots against, MSA=on-ice missed shots against, BSA=on-ice blocked shots against.

The thinking is that the more shots on net a player can direct, or is a part of generating, while he is on the ice, the better the player is at possession. Fenwick is essentially the same thing as Corsi, minus the blocked shots.

As with most statistics, looking at them in a vacuum does not provide any context and renders the statistic less useful. That is why there are stats like Corsi Relative, which measures the difference between a player’s Corsi and the team’s Corsi when he’s on the bench. Simply put, it uses the team as a point of reference. It is a useful stat, and a good starting point, when trying to figure out who might be the strongest possession players on your team.

Usage Statistics – Zone Starts and Time%

Some other useful statistics for analyzing a player is his zone starts and what percentage of his TOI (time-on-ice) is divided between even strength and special teams. Specifically, Offensive-Zone Starts (OZS), Neutral-Zone Starts (NZS), and Defensive-Zone Starts (DZS) are metrics that tell you where a coach prefers to use his players. Additionally, you might also want to know how a coach uses a player and a good way to look at this is through Even-Strength Time% (EVTm%), Power-Play Time% (PPTm%), and Short-Handed Time% (SHTm%). Combining these two statistics, you can acquire a more complete understanding of a player’s zone and time usage. 

Again, these stats are good starting-points and can be relied upon for a general understanding of a player’s usage. However, one can imagine that a player with greater DZS would suffer from a lower Corsi rating due to the play’s proximity to his own net. Some work has been done on adjusting possession metrics for zone starts (by removing the first 10 seconds following a face-off) and if you’re interested in finding out more, start here with this article by David Johnson.

Quality of Competition – Corsi Rel QoC and Corsi Rel QoT

One of the great things about these statistics is that they build on each other.  Earlier, we talked about Corsi, Corsi Rel, and what they meant for a player as far as possession. Zone starts and time percentages answer the questions “where” and “how” a player is being deployed. 

Still, these stats don’t give us an indication of the quality of competition a player faces, or the quality of teammates a player plays with. This is where statistics like Corsi Relative Quality of Competition (Corsi Rel QoT) and Corsi Relative Quality of Teammates (Corsi Rel QoT) can be useful.

Corsi Rel QoC is the average Corsi Rel of opposing players, weighted by head-to-head ice-time, while Corsi Rel QoT is the average Corsi Rel of teammates, weighted by the time they’re on-ice together. If better players typically have the puck more often, and consequently take more shots, a player who sees an opposing team’s top lines each night should have a high Corsi Rel QoC rating. The same is also true for a player’s linemates. Better linemates will likely have the puck more often, and therefore take more shots, leading to higher Corsi Rel ratings. Therefore, a player who’s consistently playing with strong possession teammates will have higher Corsi Rel QoT ratings.

Putting It Together

Here at Sporting Charts, we’re all about finding interesting ways to visualize sports statistics and as a finance geek of sorts, I couldn’t resist putting a chart together while using some of the stats we just talked about.

Let me start off by posing a question: do you know who plays the toughest minutes on your team, or in the league? Well, given what we know about usage and quality of competition stats, we can set certain parameters (i.e. minimum number of games played, etc.) and plot the top-30 players in Corsi Rel QoT against their OZS. Here is the resulting chart, courtesy of

Qo T Scatter Chart


Qo T Top -30

Players in the upper-left hand corner face the toughest competition and usually do so in the defensive or neutral zone. Here, you’ll find players like Oliver Ekman-Larsson, Dion Phaneuf, David Backes, and Jay Bouwmeester. On the lower-right hand corner, you’ll find players who also face the toughest competition but do so while usually starting in the offensive zone. It’s no surprise that you’ll find players like Pavel Datsyuk, Henrik Zetterberg, Marleau and Mikko Koivu in this quadrant. All are gifted and offensively-minded players, and coaches will usually line-up their best players against them.

This chart is also useful for identifying some unsung heroes, guys who face the toughest competition and in the most difficult situations, on a nightly basis, but do not get a ton of recognition. Players like Matt Cooke, Jay McClement, and Boyd Gordon face some of the toughest competition in the league but don’t get a lot of recognition because they aren’t the most physical or dynamic players.

A Note of Caution

As I stated earlier in the article, stats without context renders them less useful. As we continue to learn more about these stats and implement them in our analysis and vocabulary, it is important to reference the context of these numbers. Nothing ever happens in a vacuum, especially not in a dynamic sport like hockey (save for a penalty shot), and these stats should not serve as the centerpiece of your analysis. Instead, it is another tool in your toolkit when analyzing a player, or team, and is most useful when applied in conjunction with visual observations.

Questions and comments are always welcome.

Follow me on Twitter: @MLHS_Aaron


Latest Articles