The San Jose Sharks have struggled during their first two games this season. They have attempted just 46 percent of score- and venue-adjusted shots at 5v5 so far. That ranks 22nd out of 31 teams. The team has only shot 4.37 percent, so there is a bit of bad luck to blame for the score lines. Yet, considering their inability to control the majority of shot attempts, the 0-2 start seems warranted.
Logan Couture went so far as to call the team’s effort “garbage,” and Pete DeBoer said during practice this week he’s “looking for any kind of chemistry right now with anybody.” So far, DeBoer has practiced what he’s preached. The lines we know about have looked like this during practice:
Labanc - Thornton - Pavelski
Boedker - Couture - Hertl
Meier - Tierney - Donskoi
Karlsson - Carpenter - Hansen
Perhaps more than anything due to Paul Martin still recovering from or possibly aggravating an injury from last season, the defense pairs received a little shake-up during the game against the Kings. Paul Martin left practice early on Tuesday morning, leaving open the possibility Tim Heed suits up for the first time this season in the near future. Martin’s absence — assuming the Dillon/Burns pairing that ended the team’s second game remains — leaves the defense corps arranged as follows:
Vlasic - Braun
Dillon - Burns
Heed - DeMelo
Some fans are starting to ask, as they have asked all summer, if the team might be making a move for a big-name player given their cap space. Before Doug Wilson opens his rolodex, there are some changes DeBoer can make that should help improve the team’s chemistry and, hopefully, results.
Call up Marcus Sorensen
The Swedish winger’s most common linemate during the 2016-17 season was Micheal Haley. Despite that pairing, Sorensen put up the team’s 4th-best (among forwards) expected goals for percentage relative to the rest of the team (according to Corsica.Hockey). That is, with Sorensen on the ice at 5v5, the Sharks could expect to score 3.45 percent more goals than when he was off the ice — the 4th-best mark among all Sharks forwards last year. He only played 19 games, and time alongside Tomas Hertl and Timo Meier likely helped boost his expected goal numbers, so there will probably be some regression this season. Sorensen has proven to be a more-than-capable 4th-liner who could complement other skaters further up the lineup.
Take Vlasic off of special teams
Between 2014 and 2017, about 120 defensemen (four per team) each season played 100 or more minutes on the 4v5 penalty kill. Per Corsica, While Vlasic was on the ice at 4v5, the Sharks allowed a significant number of unblocked shots per hour. Here are Vlasic’s ranks in that department for each of those years:
2014-15: 90th out of 124 defensemen
2015-16: 83rd out of 125 defensemen
2016-17: 109th out of 118 defensemen
During the 2016-17 season, 70 defensemen played 100 or more minutes on the 5v4 power play. With Vlasic on the ice, the Sharks put up 65.1 unblocked shots per hour, a mark good for 51st out of 70 total defensemen. With Vlasic on the ice on the 5v4 power play, the Sharks had an expected goals rate of 5.39 per hour, which ranks 55th of those same 70 defensemen. This is a fairly rudimentary way to look at a player’s contribution to special teams. There are 3-4 other players on the ice and a lot of other factors. These numbers still suggest someone who is not benefiting the team either when they are a man down or a man up. Reducing Vlasic’s special teams time would allow the team to find players better suited for those roles while limiting Vlasic’s minutes to the crucial 5v5 matchups he’s tasked with on a nightly basis.
Shake up the D corps
While it is clear Braun is no longer producing the top-pairing numbers Sharks fans are used to seeing, the blueline’s real weak link last season was Paul Martin. Sean Tierney (no relation to everyone’s favorite 4th-line center) turns hockey stats folks’ work into visualizations. In a recent visualization, he combined Dawson Sprigings and Emmanuel Perry’s WAR models into one ranking for each player. Here is how the Sharks’ defenders stacked up last season.
Clearly Perry’s model is dragging Martin’s rating down. However, Martin’s shot impact also shows a third-pairing defender, rather than someone demanding the second-pairing minutes he has. Image courtesy of Dominic Galamini. (Look at the shot generation and shot suppression columns on the far right of either pane. The counting stats are unimportant).
By the end of the game against the Los Angeles Kings, Pete DeBoer had Brenden Dillon playing up with Brent Burns. In 104, 5v5 minutes last season, that pair helped the Sharks attempt just above 61 percent of all shots. With the two on the ice the Sharks were expected to score nearly 66 percent of the goals. Those numbers ranked 3rd (shot attempts) and 2nd (expected goals for) of all defense pairings who played at least 100 minutes together last season. While those figures are destined for some regression, DeBoer would do well to play those two together until it stops working.
Pairing Dillon with Burns in second-pair minutes would allow the team to play Vlasic and someone else (Heed? Braun?) in less-extreme minutes. A lineup that looks something like:
Vlasic - Braun/Heed
Dillon - Burns
Martin/Braun - DeMelo/Heed
Offers a lot more potential than the defense corps as currently organized does, if only because a Dillon/Burns pairing looks like it could handle tougher assignments than Martin/Burns if needed.
Play Dell more
We cannot unequivocally say Aaron Dell is a better goalie than Martin Jones. Goalies’ save percentages don’t start to stabilize until around 3,000 shots against. Aaron Dell has only faced 542 shots, so he has quite a ways to go before we have an idea about how good he really is. Martin Jones has faced more than 3,500 NHL shots and his career save percentage is .914 — almost exactly average.
Save percentage isn’t the only thing that goes into evaluating goalies, but it is a good barometer. We can get into things like expected save percentage and goals saved above average and even-strength save percentage, too. In all of those, Jones is average. What we do know is Martin Jones’ numbers are the definition of average. We don’t know what Dell is, but we know there’s a chance he could be better based on his small sample size of results thus far. There is only one way to find out, and playing Dell more will give the Sharks a better sense of what they have in their backup and it will let the coaching staff rest Jones. Playing 65 games a season is no way to last an entire 6-year contract extension (sigh).
Use past results to build forward lines
One of the problems with this team is that they have about four or five top-6 forwards and about a dozen bottom-6 forwards. Ideally, a team is able to play top-6 forwards in middle-6 roles and middle-6 guys in 4th-line roles. Those types of team compositions help produce prolonged success. The Sharks’ current arrangement isn’t awful — it could be a lot worse — but it has the makings of a fairly average team rather than a really good team. In the absence of excess talent, the coaching staff should turn to last year’s results to build their forward lines and hope that the players repeat some of the magic.
As exciting as watching Timo Meier receive Jumbo’s feeds all night would be, the top line has found success with Labanc on its wing. The coaches seem to feel the same way. That’s the only line that has so far survived untouched this season. Somehow, all nine of the other forwards have caught the coaching staff’s ire. Here is a suggestion for how they might line up:
Labanc - Thornton - Pavelski: More or less set in stone for the time being.
The Couture line is really the issue. Lost in all the commotion is the fact he had a down year by his standards, even before his horrendous injury last year. Very few lines with Couture on them posted positive shot attempt and/or expected goal differentials. He and Donskoi have formed a solid pairing the last two seasons. With more time removed from both of their injuries, perhaps the duo could be in for a bounce-back affair. Sliding Meier onto Couture’s other wing brings puck possession and shooting onto what could become an offensive force.
Boedker - Hertl - Hansen was absolutely monstrous with favorable zone starts in 58 brief minutes together last year. They helped the Sharks attempt 56 percent of all shots en route to expecting 69 percent (nice) of all goals. The sample size is small, but it’s worth riding out until it regresses.
That leaves a handful of players for the 4th line. As mentioned before, the team must add Sorensen to their nightly roster in order to improve their chances of winning games. Perhaps a Sorensen - Tierney - Karlsson line can spin some Swedish magic. Both the Tierney - Sorensen - Haley and Tierney - Karlsson - Haley lines kept their heads above water last season. Remove the weak link and voila!
New and improved (mostly) numbers-based lineup:
Labanc - Thornton - Pavelski
Meier - Couture - Donskoi
Boedker - Hertl - Hansen
Sorensen - Tierney - Karlsson
Vlasic - Heed
Dillon - Burns
Martin - Braun
Dell (at least more than 25 games or until his sv% resembles something closer to Jones’)