1. Dell might also be just average, but that’s alright
Aaron Dell has been getting as close as he’s ever been to an NHL starter’s workload the past few months. Partly because of Jones’ injuries and perceived down year, Dell has already started as many games halfway through this season as he did all of last season. Sharks fans (and one would imagine, the Sharks organization) has been patiently waiting to see what the team has in Dell. Up until recently, Dell had been showing well.
The Rangers game before the All-Star break really showed regression in action, however. Per corsica.hockey, the backup netmider’s save percentage was 10 percentage points below what one would have expected it to be, given the shots he faced. On average, an average goalie in his position that night would have saved three more goals than Dell did. That’s a pretty horrendous performance, even when you consider the Swiss cheese imitation the team did defensively in front of him. The thing about goalie stats is they’re cumulative, so a really bad night or a really good night can drop or boost their numbers quite a bit.
Even with the rough late-January stretch of play, Dell is still performing like an average goalie on the whole this season. If you look at his career to date, you can see that it’s been mostly good. Below is a chart of his goals saved above average in a 5-game rolling average (chart and data from corsica.hockey.)
This chart also makes it clear that one really bad performance has a bigger impact on a goalie’s oeuvre than a really good performance.
Even after Dell’s worst stretch of his career, he’s still showing out to be about an average goalie this season. For $625,000, an average goalie is a great value. It won’t remain that way if he keeps slipping toward mediocrity, though.
2. Fourth line offering some consistency in wake of Jumbo’s injury
Even before Joe Thornton’s injury, the team’s top three lines started to show some signs of wear. All three of Joe Thornton/Joe Pavelski/Timo Meier, Logan Couture/Tomas Hertl/Kevin Labanc, and Joonas Donskoi/Chris Tierney/Mikkel Boedker saw a reduction in the proportion of shots or expected goals they controlled. The game against the Rangers was a get-well moment for the Hertl, Labanc, Couture line; but they remain the only group of the three with top-line numbers.
Thornton’s injury changed the team composition, which was to be expected. Pete DeBoer wasn’t afraid to send his lines into the food processor against the Blueshirts, suggesting that right now there aren’t many players exempt from an in-game shake up.
The fourth line of Barclay Goodrow, Joel Ward, and Marcus Sorensen has remained intact during the final three-game stretch of January. As of this writing at the second intermission of the game in Detroit, with that trio on the ice at 5v5, San Jose is taking about 55 percent of all shots and generating just a bit less than 57 percent of all expected goals. That’s awesome play from the bottom of the lineup, and if that quality continues San Jose might have an advantage over shallower teams.
3. DeMelo does some things well, but Heed is a better partner for Dillon
Now that January is behind us, Dylan DeMelo and Tim Heed have played a nearly equal amount of 5v5 hockey this season, and it appears the two might have more in common than meets the eye (data pulled before the game in Pittsburgh.)
Individually, the two players are nearly indistinguishable as far as the percentage of expected goals they generate, as well as the rate of expected goals for and against they help the team produce. That’s just one component of evaluating a player, but we’ll allow it to make a point: there may not be as much as we think that separates DeMelo and Heed, at least in the eye of the coaches.
However, it’s plain to see that Heed is the only one for Dillon. The pair does have the advantage of starting substantially more shifts on offensive zone faceoffs than the Brenden Dillon/Dylan DeMelo pairing, but their numbers are hard to ignore, zone starts notwithstanding.
The Dillon/Heed pairing helps the Sharks take the highest percentage of score- and venue-adjusted shots — the best predictor of future goals.
FF% = Fenwick for percentage. Fenwick = unblocked shots. The Dillon/Heed pair also helps the sharks take the highest percentage of unblocked shots compared to the rest of Dillon’s defense pairs. While their offense can’t live up to that of Dillon and Burns, their defense allows 10 fewer unblocked shots against per hour of 5v5 ice time than the Dillon/DeMelo pairing.
Finally, they turn the highest percentage of shots into the highest percentage of unblocked shots into the highest percentage of expected goals of any Dillon pairing. I’ll grant you, the difference between the individual components (expected goals for and expected goals against/60) of the Dillon/DeMelo pair and the Dillon/Heed pair here is pretty slim, but let me ask you this: do you prefer a defense pair that should outscore the opposition by three-plus goals every hour or the pair that might outscore them by one goal every few games?
4. Unfortunately, this team is pretty average right now and not getting better
A regressing backup goaltender, an injury to the team’s most important player, and defense pairs that aren’t quite living up to the team’s early-season billing as a defensive juggernaut are combining to create a bit of a monster. Even before Thornton’s injury, however, the team was struggling through January.
The red rectangle highlights the month of January. Though the team started off on a decent foot, the performance proved to be an apex. San Jose has been outshot the entire second half of January, despite managing to outscore opponents at 5v5 during that time. If you squint, you can sort of see how the team’s rate of shots against has gradually risen as the season has progressed. Now that they’ve lost one of their best (defensive) players for the foreseeable future, the outlook is less rosy even than the red shaded part of that graph representing the team being outshot.
With this graph, you don’t even have to squint. You can see the team slowly allowing a higher rate of 5v5 shots against. January marks the worst it’s been all season. That trend is similar when we look at expected goals.
The team picked it back up after Christmas somewhat, but January seems to mostly be the continuation of an alarming trend.
Repeat Dell regressing, the team only having one viable line, and the continued allowance of shots and expected goals against into February and you end up with a team that just isn’t very good.
5. The coaching staff should do a good job keeping together what’s left of the Sharks
There might be hope. DeBoer and his staff have been pretty astute observers of their team’s performance this season. They made changes to the power play after a rough start. They put together three lines that, until recently, were humming along quite nicely. They seem to understand the value of being able to roll four, effective lines. They are able to look past scoring slumps and spurts to identify which players are performing well and which aren’t. Even their more controversial decisions — insisting Hertl play wing and icing DeMelo over Heed — have two sides that could be argued fairly using numbers.
I have no doubt the coaching staff is aware of the team’s current shortcomings. They recalled Danny O’Regan again, which suggests they’re looking at some different line combinations to be a catalyst on both sides of the puck. After a few extra days of rest for the All-Star break, the coaches iced three new lines at the top of their lineup. In Pittsburgh, Tomas Hertl returned to center, where he gives the team its best chance to win. Against Detroit after two listless periods, DeBoer returned to his trusted Logan Couture/Tomas Hertl/Kevin Labanc line, suggesting he’s aware of which lines have worked in the past. If anyone is going to keep this team afloat until Jumbo’s return, the coaches are going to have to optimize things as well as they can. So far this season, they’ve mostly done that.