Not every forward line or defense pair is created equal. NHL coaches often use a specific pair and/or line as a shutdown unit against opponents’ best players. Others will shelter younger or less capable players in offensive-minded roles, where the risk of mistakes leading to goals against is smaller. With data from Evolving Hockey and HockeyViz, we can see how Bob Boughner’s staff deploys its players at even strength and 5-on-5 and how each player has responded to his role.
Model after model in the hockey analytics world tells us that quality of teammates has the biggest impact on a player’s on-ice results. Other contextual factors, such as quality of competition, zone starts, and game score factor into the larger picture but don’t alter a player’s statistical profile as much as his teammates do. A player who sees a large share of offensive zone starts and enjoys high-quality teammates is more likely to produce positive on-ice results than someone in a shutdown role who plays with weaker teammates.
With Evolving Hockey’s new quality of competition and teammate calculations, we can see Shark skater deployments and and performance at a glance. The following chart examines quality of competition and teammate for every Sharks skater with regular minutes this year. Quality, in this case, is a measure of teammates or opponents’ impacts on shot differential at even strength.
Kevin Labanc, for example, plays with strong teammates against strong opponents. He receives a moderate share of offensive zone starts (average sized circle) relative to the rest of his team. Within that context, he’s done a good job driving shot share at even strength (dark blue circle). Labanc’s impact on even-strength shots, according to Evolving Hockey’s model, ranks ninth among all forwards with at least 100 minutes of even-strength game time this season.
The axes labels here are a mouthful. Let’s start from the beginning.
- “RAPM” refers to Evolving Wild’s model that adjusts for a player’s surroundings (teammates, competition, zone starts, score of the game, and so on). The number it spits out is a player’s adjusted impact on goals, expected goals, or shots. It’s an attempt at isolating a player’s individual contribution to a given facet of the game.
- “Qualcomp” refers to quality of competition. Traditional qualcomp metrics use ice time. For example, Kevin Labanc has played against 165 different individual opponents this season at 5-on-5. His most common opponent is Justin Faulk, against whom he’s played just over 47 minutes. Justin Faulk plays about 30 percent of his teams’ 5-on-5 minutes. Multiply the 47 minutes by 30 percent and so on down Labanc’s entire list of opponents. Add those together and you get 31.86: Labanc plays, on average, against opponents who play 31.86 percent of their teams’ 5-on-5 ice time.
- “Qualcomp RAPM” is the same concept. Except, instead of using ice time, it uses an opponent’s adjusted impact on the game. For this exercise, I chose impact on shot differential. Instead of multiplying Labanc’s 47 minutes against Faulk by Faulk’s ice time, this approach multiplies their time together by Faulk’s shot-differential RAPM and so on down the list of opponents. The result is, in effect, how often Labanc plays against strong play drivers.
- Qualteam metrics are the same, except calculated using a player’s teammates rather than his opponents.
RAPM-based metrics might tell us about the play-driving ability of a player’s teammates or opponents, but it’s unlikely coaches have a clipboard with adjusted metrics. Coaches do understand that Ryan O’Reilly is the Blues top center and match up with his line accordingly. To estimate Boughner’s coaching decisions, we’ll use the more traditional time on ice-based calculation. The following table compares a player’s quality of competition and teammate measures using both ice time (“Coach Deployment”) and RAPM-based measures (“Impact Deployment”).
For every skater, I’ve presented his quality of competition or teammate as a percentile. So, Noah Gregor’s quality of competition at even strength is in the 11th percentile of all NHL players this season. Then, I’ve done basic subtraction. Gregor’s quality of teammate is in the 52nd percentile, so his teammates are 41 percentiles stronger than his opponents. This isn’t exactly how things work, but it’s a quick way of visualizing the disparity between a player’s teammates and opponents.
Also included in the coaching deployment side is how often a player starts a shift in the offensive zone. Measuring zone starts and quality of competition effectively accounts for the same thing, but zone starts are included here because it’s easier to conceptualize a faceoff in the offensive zone than it is a difference in opponent quality. The column to the far right shows a player’s given impact on shot differential, per the RAPM metric.
In many cases, the ice time-based deployment matches up well with the impact deployment. Noah Gregor, John Leonard, Patrick Marleau, and Matt Nieto are sheltered—the quality of their teammates is superior to the quality of their opponents, on aggregate—no matter how you look at it. There are also a few wrinkles. Rudolfs Balcers is sheltered using ice time as a quality measurement. But, he receives some of the toughest draws if you measure player quality using shot-differential impact.
The Sharks’ top line has performed well given its deployment. Evander Kane, Logan Couture, and Kevin Labanc get to play with one another and Brent Burns regularly, which spikes their quality of teammate. That trio also plays against opponents’ best players. Individually, few forwards have had a bigger positive impact than Labanc has on shot differential this season. Couture’s impact is in the top quarter of the league’s forwards, and Kane’s is more subdued, but in the top half of the league. As a unit, those three haven’t been exceptional, but they’ve helped the Sharks maintain a positive shot and expected goal differential whenever they’re on the ice together.
The coaching staff has sheltered Timo Meier and Tomas Hertl somewhat. Those two play with one another, but they see weaker opponents than Labanc’s group. Both players’ impact on even-strength shot differential ranks in the top third of NHL forwards this season. Together, they’ve helped the Sharks outshoot opponents by nearly 20 shots every hour of even-strength ice time.
Nikolai Knyzhov’s circle is hidden and to the left of Tomas Hertl’s. He’s gained more of the coaching staff’s trust as the season has progressed, though he still plays a relatively sheltered, depth role. Despite a low share of offensive zone starts and more than expected ice time with players like Dylan Gambrell, Matt Nieto, and Balcers, he’s produced the 28th-best impact on even-strength shot differential of all 228 defensemen with 100 minutes of playing time this season.
Noah Gregor and Radim Simek have similar depth roles. No wonder, given the two of them have seen each other at 5-on-5 about 30 percent more than we’d expect if deployment were totally random. Though they play with the bottom of the Sharks’ lineup—albeit against weak competition—the two have performed admirably when asked to drive shot differential. Simek’s impact ranks 35 of 228 defensemen with at least 100 minutes of even-strength ice time this season. Gregor’s impact is in the top 42 percent of forwards.
What isn’t working?
Patrick Marleau, Matt Nieto, Dylan Gambrell, Ryan Donato, and Erik Karlsson have all struggled to positively impact shot differential.
Marleau plays late in games when the team needs to score, and has a low share of offensive zone starts relative to his team. Nieto receives a fair share of offensive zone starts, and plays most when the coaches want to protect a lead. Neither player faces tough competition, but both suffer from time with teammates who fail to drive play. Marleau has produced one of the poorest impacts on even-strength shot differential this season and Nieto isn’t far behind.
Gambrell may have the toughest job of any Shark forward. His ice time deployment has been fairly average, but his impact deployment excruciating. His offensive zone start rate is about team average, and his playing time increases as the stakes get higher. It’s no wonder his impact on shot differential has been poor.
Ryan Donato has received sheltered treatment at the hands of the Sharks’ coaches, even if his impact-based deployment belies what the coaching staff thinks given its ice time allocation. He’s spent most of his 5-on-5 ice time with Marleau, and plenty with still-learning John Leonard, which is going to challenge anyone’s ability to contribute positively.
Erik Karlsson hasn’t had the impact many expected or hoped for before this season started. His deployment may have something to do with it. Compared to last season, Karlsson is being tasked with more defensive zone starts and a far-inferior quality of teammate despite facing roughly the same quality of competition. The result is one of biggest negative impacts on even-strength shot differential of all regular NHL defenders this season.
What should change?
There’s only one linemate with whom Dylan Gambrell has posted a positive 5-on-5 shot differential: Noah Gregor. The pair might be helped along by a higher proportion of offensive zone starts than Gambrell typically enjoys. That’s a sign that a more sheltered role suits him better. Giving Gambrell more offensive zone starts and linemates more capable of driving play should improve the center’s on-ice impact.
Matt Nieto is trusted more than John Leonard when the Sharks must defend a late lead. Other than that score-driven difference, the two share a similar deployment profile. Both have been poor play drivers, though the younger forward has been marginally better. When a team is rebuilding, the tie should go to youth. Leonard’s playing time has bumped up a bit; he and Nieto have played nearly identical amounts of late. That’s the right track.
Erik Karlsson has had the best on-ice results, by far, with Radim Simek this season. Perhaps it’s because, of all Karlsson’s partners, Simek has the least defined role. Simek seems to play mostly when there is little to lose and when the coaching staff has an extra zone start to hand out. Maybe that deployment, along with a few more shifts behind the likes of Labanc and Meier, are just what the doctor ordered for the ailing blueliner.
Reuniting Balcers and Meier with Tomas Hertl should help the Latvian winger. Currently, he plays mostly with Dylan Gambrell and Meier against the likes of Brayden Schenn and Jonathan Marchessault—second liners on paper but strong play drivers in practice.
Generally speaking, it may be time to break up the top line. Adding Kevin Labanc to Gambrell’s wing may help the struggling centerman. Logan Couture and Evander Kane should be able to produce even if Donato or Leonard is their third linemate. The Sharks don’t have a wealth of strong play drivers this season, and three of them are sitting on a line together.
Coaches often demote struggling players by playing them less or sending them down the lineup. Unfortunately, this seems like a sure-fire way to erode someone’s confidence further. If a player cannot produce alongside the team’s best playmakers, how are they to find the back of the net with marginal bottom-six guys? Unsurprisingly, many of the Sharks role-playing forwards are struggling to control even-strength play. Pairing some of the strugglers with stronger teammates is the best shot the coaching staff has at rolling out four effective lines. The defense corps is another story. There simply aren’t enough “good-no-matter-what” blueliners on the roster right now. Easing some of Erik Karlsson’s workload may be an effective way of getting him and his compatriots back on track. These deployment changes aren’t guaranteed to improve the team’s performance, but there are too many players with poor on-ice impacts to continue along the same deployment plan for the rest of the season.
Sheng’s Travel Fund
Help fund Sheng's travel! Every dollar goes to the cost of getting to and from Sharks road games.