Off the Charts: The Radim Simek experience

As of this writing, we are 11 games into the Radim Simek experience. Maybe more importantly, we are 11 games into the Simek-playing-with-Brent Burns era. The question of who will be the team’s “Wookie Whisperer” has long rested ever so patiently on the tips of San Jose Sharks fans’ tongues ever since Paul Martin hung up his skates.

The Simek experience has so far been a success, not only for the newly anointed Beast Tamer pair’s on-ice contribution, but also for how they’ve potentially enabled coach Pete DeBoer to reimagine the way he deploys Marc-Edouard Vlasic. Though 11 games of 5-on-5 playing time is still a small sample size, it’s enough to give us a decent idea of what DeBoer may be trying to do with the new partnership.

At first glance, the new middle pair’s on-ice counts are impressive. The biggest change from the Burns — Joakim Ryan pair is that Simek — Burns appears to exchange offense for defense. Simek seems more deliberate; he and Burns play at a pace of 92.8 total shot events per 60 minutes, compared to 114 for Ryan and Burns.

Burns and Simek generate less offense, but they seem to make up for it with their stronger defense. The concern is that as we progress from shot to unblocked shot to expected goals, the new pair collects a smaller and smaller percentage of all events. With the pair on the ice, the Sharks create an only slightly higher rate of expected goals than their opponents, especially when compared to the Burns — Ryan pair. Where Burns and Ryan were unlucky to have a poor on-ice save percentage, Burns and Simek have been very fortunate to have a high on-ice shooting percentage. At some point the team will stop scoring as many goals with this pair on the ice at 5-on-5, and the team’s share of goals will be closer to 50 percent.

The pair’s relative-to-team statistics show that there may be some diminishing returns to their improved-compared-to-Ryan defense. As the shots turn into chances, the Burns — Simek pair’s advantage dwindles to the point where they produce a lower share of expected goals relative to the rest of the Sharks. However, the duo helps produce an expected on-ice save percentage (unblocked shots) that is 0.31 percent higher than that of the team. If that number holds, the expected goals share may not have as much of a detrimental impact on the team.

Relative-to-team statistics can also help us understand how the team around the new pair has been playing. Though Burns and Ryan helped the team generate a lower share of shots (57 percent compared to 59.2 percent), the relative metrics might show a team struggling a bit more with the Burns pair off the ice. Without Burns and Ryan on, the rest of the team took 55 percent of all shots. Without Burns and Simek on, the team takes 56.9 percent of all shots. With Burns and Ryan off the ice, the Sharks generated 55.2 percent of all expected goals. Without Burns and Ryan, the team generates 56.1 percent of all expected goals. There isn’t much of a change at all! Despite the differences in relative metrics, the team around the new pair appears to be enjoying the same success.

Contextual Factors at Play in Analyzing On-ice Metrics

There is another major component to this new pair’s on-ice metrics: context.

Teammates have the strongest impact on a player’s on-ice statistics. By looking at the defense pairs’ most common forward teammates, we can extrapolate which forward lines they spend the most time with.

Burns and Ryan played behind the Evander Kane — Antti Suomela — Joonas Donskoi line. It appears the Simek — Burns pair have played with the Kane — Joe Pavelski — Donskoi line, as well as the Marcus Sorensen — Joe Thornton — Kevin Labanc trio.

Kane — Suomela — Donskoi were worse than the rest of the team at taking and allowing shots. That line was also worse than the team on both sides of the expected goals equation. That Ryan and Burns spent most of their time with this trio and still helped the team take a higher proportion of shots and generate a larger proportion of chances relative to the team is impressive. Kane — Pavelski — Donskoi have been better at preventing shots against than Kane — Suomela — Donskoi, but the Pavelski-centered version has been worse at generating shots, for a greater net negative impact on the team’s shot share. The Pavelski trio, however, has been much better at limiting expected goals against.

Simek and Burns have benefitted from playing with a Joe Thornton-centered line, as well. Sorensen — Thornton — Labanc offered positive impacts on both sides of the shots ledger. They have been about team-average when it comes to expected goals. Burns and Simek have had the advantage of playing with stronger teammates and more settled forward lines than did Burns and Ryan. That is a distinct advantage for Simek and Burns and is likely to positively impact the pair’s on-ice metrics. Teammates tell just part of the contextual story.

Zone Starts Impact Shot Share; May Produce Team-wide Ripple Effect

Before we get into the zone starts, it’s worth noting maybe the more important contextual factor: trust. DeBoer plays the Burns — Simek pair, on average, 2.5 more minutes at 5-on-5 than the Burns — Ryan pair each night. Furthermore, DeBoer is a bit more trusting of Simek than he was of Ryan when the game is close (within one goal). Beyond that trust, there is potentially signal in the zone starts DeBoer assigns to the pairs. Everything is pretty comparable except for offensive and defensive zone starts. DeBoer realizes how much better defensively Simek makes the Burns pair. As a result, he gives them fewer offensive zone starts and, so far, a few more defensive zone starts.

Micah Blake McCurdy of HockeyViz has shown that zone starts do impact a player’s ability to generate or prevent unblocked shots. Zone starts generally have a larger impact on the amount of shots a team takes than they do on the shots a team prevents. Still, defensive zone starts lead to more unblocked shots against and fewer shots for. The increase in defensive zone deployments makes it even more impressive that Burns and Simek are better defensively than Burns and Ryan were. The more defensive deployment might also help explain why the pair’s offensive numbers aren’t as exciting.

That DeBoer is able to assign more defensive zone starts to Burns and Simek may have a ripple effect on the rest of the defense corps.

Though it appears DeBoer has been trying to lessen Marc-Edouard Vlasic’s defensive zone starts as the season’s worn on, his most recent defensive zone start peak — with Simek a lineup regular — is the lowest of his season. DeBoer also appears to be assigning Vlasic more on-the-fly shifts, which have the most pronounced effect on limiting opponent shots and on creating more offensive shots.

The rate of on-the-fly shifts Vlasic has enjoyed has increased of late. Given the Vlasic and Justin Braun pair’s poor on-ice numbers, this trend of zone start changes should help, provided it continues. There is more at play here than just Simek enabling DeBoer to lighten Vlasic’s load. We are noting the trend in Vlasic’s zone starts in the event they are correlated with Simek’s deployment, though that conclusion may be incorrect and the trends may be short-lived and a result of plenty of other confounding factors.

For the time being, however, we can enjoy the fact Vlasic and Braun seem to be finding a bit more success in the shot-differential column of late.

Context Makes it Difficult to Isolate Player Impact

At first blush, the Simek — Burns pair seems to be providing a larger positive impact than did the Burns — Ryan pair to the team’s performance. However, by looking at the context in which the two pairs have existed, the differences between the two might not be as great as first appearances would suggest. Though Burns and Simek have contended with a more difficult zone deployment, they’ve also enjoyed skating with stronger, more stable forward lines (Ryan — Burns played with one distinct forward line and multiple Logan Couture-centered lines).

According to relative-to-teammate numbers calculated by Corsica, Simek has had a more pronounced defensive impact relative to everyone with whom he skates. His individual impact on shots shows a stronger positive signal than that of Ryan. Simek’s impact on expected goals is much worse due to his poor offense. These metrics jibe with those of the Brent Burns-led defense pairs of which these two play a respective part. The goals-above-replacement (GAR) metrics from Evolving Hockey also show that Simek’s individual impact on the team’s goal scoring far overshadows that of Ryan.

On-ice goal stats — which the Evolving Hockey model attempts to isolate on a player level — are rife with things like shooting and save percentages, luck-driven metrics from which Joakim Ryan suffered greatly. Eleven games is a small sample size, but it certainly seems as though Radim Simek is so far outplaying his Wookie-whispering counterpart. And, if Simek’s tougher deployment really does help the Vlasic tandem, then Simek’s addition to the lineup will be immense. Our understanding of the impact Simek is providing will evolve as the season wears on and he experiences more teammate turnover and percentage regression. For now, we’ll accept that he has a larger positive impact on the team’s performance.