Photo by Ben Green/NHLI via Getty Images
One of the most challenging tasks for a head coach is constructing the four forward lines and three defensive pairings. Coaches have to balance the lines for offensive output and defensive capabilities, which can often be difficult when battling through roster turnover as a result of injuries, suspensions, and the like. When coaches get the line combinations right, dividends are quickly realized.
Through four games this season, we’re starting to see some trends in terms of line performance. In this post, we’ll take a look at some of the forward lines that Head Coach Peter Laviolette has deployed thus far into the young 2022-23 season and propose new lines that match wingers to their optimal center match.
Statistics used in this post are courtesy of Natural Stat Trick. If you’d like to learn more about the statistical terms used in this post, please check out our analytics glossary.
Recent line deployment performance
Here’s how some of the more regular lines we’ve seen so far this season have panned out in terms of shot attempt generation and chance generation:
There’s some good here, mixed with some paltry results. The Alex Ovechkin, Evgeny Kuznetsov, and Connor Brown line have actually done really well in terms of shot attempt generation, which isn’t something necessarily very new considering Ovechkin’s propensity to attempt shots at a high rate. The concern here is the vast negative differential in high danger chances (HDCF%), where that line only generates one high danger chance for every four chances they allow.
The Anthony Mantha, Dylan Strome, and TJ Oshie line has struggled, and there’s no real way to sugar coat that fact. That’s likely a line that doesn’t stay together long term, especially with a 0 HDCF% and a scoring chance for percentage (SCF%) of 20. They’ve struggled in terms of shot attempt generation as well.
A line that has really excelled during five-on-five play is the Marcus Johansson, Lars Eller, and Aliaksei Protas line. They’ve utterly dominated their competition in shot attempt generation, overall shots on goal percentage (SF%), and scoring and high danger scoring chances.
The fourth line of Conor Sheary, Nic Dowd, and Garnet Hathaway has really struggled through the first four games. Sheary has been able to find the scoresheet but has done so without Dowd or Hathaway on the ice with him. Part of this perceived underperformance is that this line is typically matched up with the opposition’s best offensive lines. On the other hand, the best way to defend against teams’ best offensive players is to keep them hemmed into their own defensive zone. This line hasn’t been able to do that so far this season.
Now, let’s take a look at how these lines fare when we compare goals for percentage (GF%) and expected goals for percentage (xGF%):
Although the Ovechkin, Kuznetsov, and Brown line performed rather well in terms of possession numbers, that hasn’t carried over to overall GF% and xGF%. Part of the sub-50% xGF% figure here is the fact that this line gives up a bevy of high danger chances against.
The offensive output of this trio hasn’t made enough of an impact to justify keeping these three together at this point, and since Connor Brown is now out long term with a lower body injury, we’re likely not going to see this line for a while.
The Mantha, Strome, Oshie line’s struggles continue here. They haven’t been on the ice for a goal for (or against) but see a minority of xGF while on the ice. This usually means that they’re allowing high quality shots and chances against while also not generating those at an equal level of their opponents.
Again, the Johansson, Eller, and Protas line has been dominant. They haven’t been on the ice together for a goal against and own nearly 60% of expected goals while on the ice. There’s not much more to say here other than this is a line that Laviolette should turn to when trying to keep a lead late in a third period.
We see the struggles of the Sheary, Dowd, and Hathaway line continuing here. If you’re hemmed in your zone getting barraged by shot attempts and scoring chances against, you’re not going to have a great opportunity to generate goals or chances.
Matching wingers to centers
Part of how coaches construct lineups is matching up wingers with the center they have the most chemistry with. I don’t have an insight into how close these players are in the locker room or during practice, but we do have the power of analytics to help us out here.
I went through each winger that has played so far this season (so no McMichael or Snively yet) and matched them with the center that they’ve played a reasonable amount with and also performed the best. Here’s how that panned out:
Zooming out for a second, it’s clear just how effective Kuznetsov has been analytically this season. He has yet to pot a goal but has four points in four games (including all four in the last two games). The interesting piece is that four wingers have Kuznetsov has their best match, where Eller only has two, and Strome and Dowd only have one best match.
Using these pairings (and sometimes having to use the “second best” option), we can start to construct a lineup. Here’s what I’d propose as a lineup to roll out for a game when Kuznetsov returns from his one game suspension:
The Ovechkin, Kuznetsov, and Sheary line was just too good in Monday’s comeback win over the Vancouver Canucks. They were absurdly brilliant and were a huge reason why the Capitals were able to complete the third period comeback.
Although Strome and Mantha only had a 33.33 GF% when on the ice together, they had a 53.39 xGF%. This means that they’re pretty unlucky when it comes to scoring goals and if they’re able to keep up the xGF pace, they should be relatively successful soon. Adding the possession monster in Aliaksei Protas into the mix on that line should help invigorate the Mantha and Strome duo.
The bottom six is where it gets interesting. I’d be totally on board with McMichael and Eller swapping positions here, but we’ve seen in the past that Eller and Oshie have rather good chemistry together. McMichael gets some veteran support on his line, which should put him in a good position to succeed.
On paper, Johansson gets a demotion to the fourth line. In actuality, he can help stabilize the fourth line and add some offensive prowess to the mix with Sheary moving up to the first line with Ovechkin and Kuznetsov. Johansson fits here because one, you don’t want to have one of your most NHL ready prospects wither away on the fourth line, and two, Johansson will still get power play time.
Conclusion
Four games are a rather small sample size of games in an 82-game season, but we can see the trends that start to arise. With a decent amount of roster turnover and new additions to the lineup, it can be difficult for a coach to find the right line combinations through a few games and the preseason.
We will most certainly see a bevy of different line combinations through the 82-game journey that is the NHL regular season. The best we can hope for is some consistent performances.
By Justin Trudel
We’ll have real data for McM and Snively after the Ottawa game, so my question is a bit premature, but why McM as the winger instead of Snively?
It’s a good point, Fyre. I think C-Mac and Eller could be flipped. Snively was strong on the wing last season. But let’s see how it plays out.
nice analysis and break down….this really gives us a good down-in-the-weeds view of how the lines are doing
Thank you!
I don’t think I’ve read a better article on our line combinations. Well done. At another blog, many folks are clamoring for us to go ahead and extend Strome. For me, he’s a serviceable pickup and useful to us for this year but that’s it. This article confirms what the eyes see but, in fairness to Strome, he’s been saddled with Mantha and Oshie and those two are just not performing up to spec yet.
I think the problem with Mantha, Strome and Oshie is actually Oshie. He has had one good game, but he’s clearly hurting. If it’s something that will take just a short time to heal I’d sit him for a while and play Malenstyn on the 4th line, putting Mantha on the right side and Snively on the left. Snively is not a 4th line player… he needs to be with linemates who have better offensive capabilities than Dowd and Hathaway, and Malenstyn is a better fit on the 4th line than Johansson is. If that means sitting Johansson… I’m okay with that…
??? Which of the two games Oshie has points in was his “one good game”?
Really appreciate it! I think Strome could be in DC for a decent term. It’s pretty early to say whether or not he’ll pan out, but it’s certainly early to want him to be extended. He’ll be a restricted free agent after the season, which gives the Caps a bit more flexibility with the roster, especially if Backstrom can find a way back on the ice.
Thanks for the insight, Justin.
Thank you for reading!
Protas isn’t just a fancy stats hero… He also passes the eye test. He is effective on the forecheck and the in the neutral zone with his size and newfound speed. He still gets lost in the D zone a bit, but again often makes up for it with his reach. I think he has been the Caps most effective all-round forward so far.
Sheary has also been very effective in spots, and deserves the shot at 1RW. I think the problem on the first line was Brown, who didn’t show anything before getting hurt. Sheary could make a big contribution there, since he is a relentless puck retriever on the forecheck and has the hands to be a legitimate threat that other teams have to account for.
The rest of the forward crew has been “meh” at best. As I said elsewhere, I think Oshie has been playing hurt. The rest just haven’t been playing very well.
Agreed on Protas not just being an underlying statistical monster. He looks like he belongs at the NHL level, I was more saying that because his tremendous performance hasn’t translated to the box score statistics quite yet. He will get there and could be a real asset in DC for years to come.
I think Brown just had some growing pains with a new team, linemates, and a new system to learn. I think he would have come into his own with a bit more time.
I don’t like breaking up the Johansson-Eller-Protas line. Strome has shown versatility in being able to work with multiple forwards, so I’d rather see him make changes than Eller. My lineup would be:
Ovi-Kuzy-Sheary
Mantha-Strome-Oshie
Johansson-Eller-Protas
McM-Dowd-Hathaway
I can’t see McM making Dowd and Hath any worse than they’ve been so far this season. I’d actually consider double shifting Protas with Dowd and Hathaway. The only reason I’d play McM is because he’s technically the number one off the bench, but my real preference given who’s out, would be Beck M, so I’d have McM on short leash depending on how well he played with those two. Mantha has scored with Strome on ice, so I’d like to give those two and Oshie a little more room to see how well they play together. If poor numbers result, than I’d reconsider, first off the bench, like Snively or McM.
So far, no matter who’s with Dowd and Hath, that line is currently a problem, so in essence, it’s a decision between the other three lines how the lineup with look.
Mantha-Strome-Protas was a dominant line, but the problem was that the rest of the lines were too weak. Ovi-Kuzy-Sheary make a strong set, but I don’t like the Eller line without Protas. It can dominate defensively, and with the Dowd line not doing that, we have to have a strong shut down line. It can also contribute offensively. Although the Mantha-Strome-Protas is an offensive force, they aren’t as strong defensively as the Johansson-Eller-Portas line, and you are still getting offensive output too.
Anyway, that’s the way I see it.
Thanks Jon, once again, for a marvelous breakdown.
Thanks for your thoughts, as always, Jonathan. Credit for this one goes to Justin, our analytics writer.
I’m on board with a lot of what you’ve said here. I would just be hesitant to have McMichael on the fourth line, mainly because we want him to develop offensively. If you’re going to have him teetering between a fourth line deployment and a healthy scratch on any given night, I’d rather him get first line minutes in Hershey.
Thank you Justin for the great analytics work, and thanks for taking the time to read my two cents. I get where you’re coming from on your lineup too. It’s fun playing lineup theory!
I agree, having McMichael on the fourth line isn’t ideal. I’d rather he was in Hershey too. But if he’s going to play, that’s where I’d put him with the lineup we have currently because it was the only spot I had left that was open. I’d rather play others more suited there, but I’m sure it wouldn’t harm McM to get that experience too.
Some advocate for him to play center for the Caps, but I don’t think that’s the best lineup we can have. I don’t think we’re good enough to ‘test’ guys out. Last year was an exception.
As an Oriole fan, I was excited to see Henderson come up towards the end of the season, but it changed the dynamic of the team I think, in a way that wasn’t helpful for the team. He was rotated around, playing different positions, sitting veterans who had earned their spot on the field, and created an environment more like a team who already were out of the playoffs, not one who was still trying to make it. Like training camp. Like end of the season, test your minor league players in the bigs.
Are we trying to test players, or trying to win games? You really can’t have both mentalities on a team that just isn’t good enough to do that. If that was the mentality, than trade everyone you can, and start afresh. Then play McM to your hearts desire, and make sure he gets on PK and PP. Washington can’t really afford to have a 2C/3C not play PK or PP. That’s a luxury a team doesn’t really have typically.
I am worried about injuries and especially worn out vets towards the end of the season and playoffs. That is why I think we need to carefully integrate the potential subs in before they are actually needed. They need the practice and game experience with the vets and the coaches need the video to help them improve. I think that waiting until the end of the year will be too late.
Interesting to see who is working well with who.
And now for something completely different: