Can Goaltenders Have an Effect on Possession Statistics? Part 1: Evaluating Known Puck-Handling Goaltenders

Cover Pic

A majority of analytic research done on possession has focused on a forward or defenseman’s ability to control the play. Possession numbers (both actual and relative to an individual’s team) are often an effective way of evaluating players. That being said, very little research has been done focusing on a goaltender’s impact on the team’s possession as they don’t have a direct impact in the offensive zone (in fact, all goalies start 0% of faceoffs in the offensive or neural zones (that was a joke)).

While the goaltender’s main job is to stop the opponent from scoring, there are multiple ways a puck-moving goaltender can influence possession. This includes stopping an opponent’s dump-in attempt and leaving the puck for a defenseman to pick up, poke checking an opposing player near the net, deflecting a shot to an area away from opposing players, passing the puck to a defenseman to start a breakout, or even catching the opponent in the middle of a line change by firing the puck up to a forward waiting near the far blue line.

Much like forwards and defensemen, some goalies must be better at driving play than others. In an age where possession is crucial, having a puck-moving goaltender that positively affects possession statistics can provide a large advantage over an equally talented goaltender that doesn’t touch the puck.

A perfect example of this includes a game from last March between the Tampa Bay Lightning and Montreal Canadiens where Ben Bishop picked up two assists. On the Lightning’s first goal, Montreal had dumped the puck into the corner and all five skaters went for a change. Bishop recognized this, came out of the net and fired the puck up to Steven Stamkos who was waiting by the red line. Montreal was caught off-guard by this quick breakout pass and broke down in their defensive coverage allowing Stamkos to pass the puck to Vladislav Namestnikov who went in alone on Carey Price. On Tampa’s second goal, the sequence starts with Montreal clearing the puck while shorthanded. Bishop comes out of his net settling the puck down for defenseman Anton Stralman. Stralman takes the puck around the net and catches Jonathan Drouin behind the Montreal defensemen to lead him in on a breakaway. By settling down the clearing attempt and putting it in a position for his defenseman to play it with ease, Bishop is able to jumpstart the offense.

While Bishop’s first pass had more of a direct impact to the team’s goal than the second, both were able to influence the team’s puck possession. It is important to know that not every touch of the puck from a goaltender will lead directly to a goal or even a shot like they did in this example with Bishop.

A goalie must be smart about playing the puck as any potential misfire or turnover could catch him out of position. After turning over the puck to an opposing player, a resulting odd-man rush or scoring chance can occur very quickly. In a recent game between Philadelphia and Boston, Flyers goalie Steve Mason controls a dump-in behind his net and fires it up the far side glass. The Flyers skaters aren’t able to get over to the puck quick enough and Mason’s clearing attempts results in a turnover to Bruins forward Max Talbot. Talbot finds Kevan Miller open in the high slot and he is able to put the puck in the net. It takes just seven short seconds for the Bruins to capitalize on Mason’s miscue.

I’ve identified a small group of goaltenders who are known to play the puck to see whether or not they have an impact on the team’s possession. This includes Bishop, Mike Smith, and Carey Price. I’ve also included the retired Marty Brodeur in the group as he is known as one of the best puck handling goaltenders of all-time.

To determine a goaltender’s impact on possession, I’ve used a series of WOWY (With or Without You) metrics to determine whether or not a goalie’s team performs better with them in net. The goalie’s WOWY metrics compares the team’s possession statistics in games that they are between the pipes and compares it to games where they aren’t in net.

For my research, I used data from the past three full seasons along with this season’s data up through the end of December 2015. As Marty Brodeur is retired, I used data from the ‘07-‘08 season through ‘11-‘12, his last as the prominent goaltender for the New Jersey Devils.

As hockey is a team sport, it is impossible to exactly pinpoint the individual effect a player has to the team, but from this information, we will be able to make conclusions based on the long term trends of several seasons’ worth of data. The data may be skewed from unequal sample sizes, but that is expected as top goaltenders play a majority of their team’s games. Another source of data skewedness can occur from a team having a backup goalie that is bad with handling the puck and having a negative effect on the team’s possession.

To start, I compared these goalies at even strength (5v5), specifically looking at shot attempts. To see if more offense is create with or without a goalie in net, I analyzed the goalie’s combined Corsi For and Against per 60 minutes of ice time (Corsi/60).

Corsi Per 60 WOWY

Half of the goalies (Bishop and Price) had a higher combined Corsi/60 with them in net compared to without them between the pipes. With or without Brodeur in net, New Jersey’s games had much smaller shot attempt numbers than anyone else in the league. This makes sense with their style of play (New Jersey had over 3 less shot attempts per 60 minutes of ice time than any team in the league during this period). While Bishop and Smith were impacted by less than one total shot attempt, Price (+5.29 Rel Corsi/60) and Brodeur (-5.43 Rel Corsi/60) saw larger shot attempt differentials.

Next we’ll look at the goaltender’s CF% WOWY graph:

Corsi Percentage

Bishop and Brodeur’s teams saw a much higher CF% with them in net. Smith’s Coyotes also saw an increase but to a smaller extent. Price’s Canadiens actually saw a dip from a positive Corsi percentage to a negative Corsi percentage with him in net.

It is unfair to analyze a goalie’s Corsi/60 or CF% WOWYs because it is comprised of two variables: Corsi For and Corsi Against. The prior graphs show us an incomplete picture as we don’t know whether a goalie has larger differences between their Corsi For and Corsi Against WOWYs that more so canceled each other out when combined together or whether the goalie really only had a miniscule effect on shot attempts. To get a better look, we have to further break down the graphs:

Corsi For Per 60 WOWY

Analyzing the Corsi For per 60 minutes of ice time (CF/60) graph, all three of our known puck moving goalies playing in the league today have a positive impact on shot attempts for. Bishop had the highest impact as Tampa Bay attempted almost two additional shot per 60 minutes of ice time with him in net. Brodeur’s puck handling abilities didn’t seem to translate to an increased rate of shot attempts for.

Progressing from shot attempts, the next question to ask is whether a positive CF/60 WOWY translates into a higher rate of scoring chances for:

Relative Possession Metrics For

For the most part, an increased CF/60 led to an increase in scoring chances for per 60 minutes of ice time (SCF/60 Rel) and high danger scoring chances for per 60 minutes of ice time (HSCF/60 Rel). Smith and Price saw positive relative statistics for all three relative possession metrics where Bishop’s impact to scoring chances was much lower than his impact on shot attempts and even saw a slight decrease in high danger scoring chances for.

A majority of the relative statistics were less than one attempt so it may seem that they have a very small effect especially when compared to a position player that plays fewer minutes in a game, but over the course of a season where goaltenders play an upwards of 3,000 5v5 minutes, the impact is more apparent.

Looking at the other side of the puck, can puck-handling goalies limit the number of shot attempts that they face? A puck moving goaltender may be able to limit the amount of time an opponent has the puck in the goalie’s defensive zone, thus limiting the amount of shots attempts against. We’ll analyze their shot attempts against (CA/60) WOWYs to get a better picture:

Corsi Against Per 60 WOWY

Bishop, Smith, and Brodeur all faced less shot attempts per 60 minutes than their team saw without them in net. Price, who had a positive impact on shot attempts for, faced over four more shot attempts per 60 minutes of ice time than Montreal’s backup goalies. This increase highly influenced and increased his Corsi/60 WOWY, making him the goaltender with the highest Corsi/60 statistic. Brodeur had the clearest impact of all four goalies, facing over four and a half shot attempts less than the backup goaltender for New Jersey. This impact has helped Brodeur enter into the conversation of the best puck-moving goaltender of all-time.

As we did before, we’ll see whether a positive CA/60 WOWY (meaning the goaltender saw fewer shot attempts than their team’s backup goaltender) translates into a lower rate of scoring chances against:

Relative Possession Metrics Against

As similarly seen on the relative possession metrics (for), a positive shot attempt WOWY translates into positive impact on scoring chances. Following the pattern from the CA/60 graph, Bishop, Smith, and Brodeur all had positive impacts on CA/60 Rel, scoring chances against per 60 minutes of ice time (SCA/60 Rel), and high danger scoring chances against per 60 minutes of ice time (HSCA/60) while Price saw the opposite. With Bishop in net, Tampa Bay is able to consistently keep opponents out of the danger zones shown by his HSCA/60 Rel being his highest relative possession metric. While Price has a +4.33 CA/60 Rel, his teammates in Montreal have done a good job at not allowing opponents to shoot from the danger zones (he has just a +0.22 HSCA/60 Rel).

In conclusion, puck-handling goaltenders have an impact on possession at even strength. Bishop and Smith proved to have positive impacts to their team’s possessions both offensively and defensively where Price only showed a positive impact offensively and Brodeur only showed a positive impact defensively.

Next: Part 2: Evaluating Known Puck-Handling Goaltenders on Special Teams

Next: Part 3: Evaluating Other Top Goalies

Follow Steve Ness on Twitter at: @QuickkNess

All data from


3 thoughts on “Can Goaltenders Have an Effect on Possession Statistics? Part 1: Evaluating Known Puck-Handling Goaltenders

  1. Pingback: Can Goaltenders Have an Effect on Possession Statistics? Part 2: Evaluating Known Puck-Handling Goaltenders on Special Teams | Insight into the Business Side of the Hockey World

  2. Pingback: Can Goaltenders Have an Effect on Possession Statistics? Part 3: Evaluating Other Top Goalies | Insight into the Business Side of the Hockey World

  3. Pingback: Dater’s Avalanche Mailbag: Matt Duchene is still here, but why?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s