We continue our analysis of the performance of the Washington Capitals forward group for the 2022-23 season by taking a closer look at the performance of each and every line combination deployed last season for each individual forward.
The refined look attempts to glean additional insight into what worked and what didn’t work on a line-by-line case, and assist in identifying specific “needs” the team might have for the upcoming season.
Previous Line Assessments:
Today’s focus is on Dylan Strome. Strome recorded 23 goals and 42 assists in 81 games played for the Capitals last season. He finished with a personal expected goals for percentage of 52.51%, fourth-best on the Capitals.
ALL LINES DEPLOYED
The following graph plots all of the line combinations deployed with Strome for the 2022-23 season at five-on-five (sans lines that included Marcus Johansson, Lars Eller and Garnet Hathaway). The graph includes the total time each line was on the ice (TOI), the percentage of offensive zone faceoffs each line was on the ice for (OZFO%), the expected goals differential (xGF – xGA) and the expected goals for percentage (xGF%) deployed at five-on-five. [Click to enlarge].
[The statistics used in this post are courtesy of Natural Stat Trick and the NoVa Caps Advanced Analytics Model (NCAAM). If you’d like to learn more about the statistical terms used in this post, please check out our NHL Analytics Glossary]
Note: The thin horizontal red line above bifurcates the positive and negative expected goals for percentages for all of the line combinations.
WHAT WORKED
The Ovechkin-Strome-Sheary line became a mainstay for the Capitals last season, mostly because it was Ovechkin’s most efficient line combinati0n. The trio logged 268 minutes of time on ice at five-on-five and generated an expected goals for percentage of 52.4%. That’s pretty good. However, It will be interesting to see what the Capitals do should Connor Sheary not re-sign with the team this off-season.
They could go with the Ovechkin-Strome-Aube-Kubel top line, which recorded an even better efficiency (68.10%) but spent less than a quarter of the time on ice together at five-on-five (43.8 minutes). Aube-Kubel was also excellent on the fourth line, but could theoretically be replaced by Beck Malenstyn.
The Ovechkin-Strome-Oshie line also worked well, but in very limited minutes. The Milano-Strome-Oshie was given more time on ice but with expected goals for percentage of 51.2%.
WHAT DIDN’T WORK
Former Capitals head coach Peter Laviolette kept giving the Ovechkin-Strome-Wilson line time on ice together, but to no avail, as the trio logged an underwhelming expected goals for percentage of 41.6% in 74 minutes of time on ice at five-on-five. I would expect to see this line combination be given another shot under the next coaching regime. However, it should be reiterated that Wilson worked best with Nicklas Backstrom at center last season.
The Sheary-Strome-Wilson was also given significant ice time last season, with somewhat disappointing results.
CONTEXT AND CAVEATS
We will most likely see the return of Ovechkin and Strome on the top line next season, but who plays right wing is still very much an unknown. If Sheary is not re-signed and no other right wingers are brought in during the off-season, look for Aube-Kubel to step into the role, at least initially. Also look for Oshie to get time on the top line and it’s very likely the new coaches will revisit Ovechkin-Strome-Wilson in hopes of finding better performance than last season.
Next up is Anthony Mantha. At the end of this exercise we will compile the optimal line combinations for the Capitals from the 2022-23 season.
By Jon Sorensen
Yes, yes, Yes! It’s the Whale!
Ned Lamont is Governor of Connecticut, and he made his pitch to Bottmann for the NHL to relocate the Coyotes franchise to Hartford. Mayor Luke Bronin wants it too. Bunch of articles in NHL-space about this
The Brendan Shanahan Whalers gave the Caps FITS in 1994 before the defection to Carolina
Thanks for the stats on Strome and his lines. What a GIFT to the Caps and GMBM
I don’t see it happening. They need a central division location, and the XL Centre is no NHL arena.
Houston is 6th largest US tv market, Atlanta is 7th. I’d give them the edge in markets. Can their basketball arenas handle NHL hockey? If Houston, no division changes. If Atlanta, could go Atlantic and move Detroit to Central or Met and move Columbus.
I’m with ya on this, should be regional move. Also heard Salt Lake tossed around.
Also hearing XL Center (Hartford) is in bad shape.
Also like the idea of moving Detroit back, so who knows.
Well, SLC does have winter sports while Houston and Atlanta don’t, and they are also NBA. Could see a lot of things going for it. Don’t know anything about the existing facilities for any location. I agree, not building something new or heavily renovating an arena could be important. Travel to Houston or Atlanta would be easier for most of the league and Central time is probably an easier schedule than Mountain.
We could really go wild card. The ECHL Solar Bears can seat 17,353 with 1,428 club seats and too many suites to count in the Amway Center.
Wow, Solar Bears play where Orlando Magic play?
yep. both teams have same owners.
Interesting. Sounds like an owner who would love an NHL team
I’ve often wondered about an AHL team. But travel expenses would be a lot lot lot higher for AHL than ECHL with ECHL in Ft Myers, Jacksonville, Savannah, Atlanta, Charleston, and Greenville and only Charlotte in AHL.
Detroit prefers to be in the East so that fans can watch more of the Away Games. That being said, Detroit is not with any of its traditional rivals, so they seem like a bit of a misfit. In some ways, a slight realignment could help give Detroit more natural geo geographical rivals. Maybe more Pitt and Columbus to the Atlantic and move TB and Florida to the Metro. The downside of that is removing Pittsburgh from its traditional rivals in Washington and Philly. But Detroit and Columbus would be a natural geographic rivalry even though the teams have not been good at the same time. They’re just over 3 hours away from one another. Buffalo is only 3 hours from Pitt. Detroit is actually only a little over 4 hours from Pitt. (I’m thinking a little bit out of the box.)
I actually liked the Bubble Games when we really focused on divisional games. I’d like to have more division games than we have now. I would like all first month and all last month games to be divisional or, at worst, same conference. Cut back travel those two months, make those games count more.
I do read articles about players who would like to see reduced travel. AHL and ECHL do back to backs against the same time. Wouldn’t mind seeing that in the NHL.
So I’m open to rethinking alignments.
Concur zero chance for Hartford. Handicapping:
1) Houston 2) Salt Lake City 3) Atlanta (not again!?)
Quebec City and Milwaukee have NHL-ready rinks and would be a fanatic franchise. But the Name of the Game is Favoritism, and Gory B will satisfy his $$$ friends first
Wilson was absolutely not better with Backstrom than he was Strome, Wilson produced absolutely nothing with Backstrom last year, these advanced stats are about as useful as a leaking fart. No game as ever been won by advanced stats. Scoring goals and not allowing them wins you game. Strome and Wilson were the caps most productive duo for the last month or so.
You couldn’t be more wrong. Sorry, but it’s fact. Games are t won by any stat, but stats explain what happened in past games. Please read other posts.
Wilson line performances: https://novacapsfans.com/2023/05/16/assessing-tom-wilsons-line-deployments-for-the-2022-23-season-what-worked-and-what-didnt-work/
Additionally, Strome and Wilson were rarely linemates in the last month. I understand dismissing the value of something you don’t understand, it’s human nature. I would be happy to explain the stats and how they are beneficial.
I’d also be happy to consider your stats that say these stats are wrong.
How many pts did Wilson and Backstrom actually produced together at 5v5? How many pts did Wilson and Strome produced together 5v5? Backstrom had 10 pts at 5v5 in 39 gms and was – 21, he also went from the end of january to early March without a single pts 5v5. Expected goal for or against don’t necessarily translate in goal or even mean good defense. It always depends on the game situation so advanced stats have to be taken into context. Team leading by 2 goals or more will likely defend more than attack so their advanced stats will likely take a hit and vice versa. Does’nt mean they were’nt the better team when they needed to take the lead. I don’t need to be schooled about how to play, watch ,coach or evaluate player thank you. I called that if Backstrom came back he would sink the team, look what happened..
You’re changing the focus of the debate, which was, which line worked best for Wilson. You are now focusing on attacking Backstrom, which is fine, but has no weight on which line was Wilson’s most effective line. You are using hypotheticals regarding game situations, score, etc, but no hard data, which does ‘t have any direct relation to the debate. It’s clear you are triggered by Backstrom, which is fine, but not relevant to what is being presented here.
But you love advanced stats so you should have no problem pulling out those stats, I watched every games Wilson played with Backstrom so I know they did’nt produce together they may have produce a goal together at 5v5 maybe 2 that’s it. Strome and Wilson were the Caps best duo over the last month. So go on check the data who was Wilson more productive with at 5v5 Strome or Backstrom?
Backstrom. I provided the link before: Here is Wilson’s line data:
https://novacapsfans.com/2023/05/16/assessing-tom-wilsons-line-deployments-for-the-2022-23-season-what-worked-and-what-didnt-work/
We all watched every game, and we all probably saw it differently, that’s why hard data has more value. You are t producing anything concrete other than what you think you saw. You want me to hold more value in what you saw versus real data?
It takes three players to make a line, not two. Who were their linemates, respectively, for these GF-GA?
Throwing my two cents into a heated discussion) What I think explains the seeming discrepancy between the eye test and these advanced stats, is taking into account the defensive side of their play. Whatever Ovi-Strome-Wilson produced offensively was often negated by what they gave up defensively. They couldn’t hold on to the puck for too long, and pretty much, as soon as they’d lost it, the other team was gone with it. Yes, this line can backcheck, sometimes forecheck effectively to get the puck back or at least hold up the opponent’s move up the ice. But they can rarely strip the other team of the puck, especially in their own end. They are just pathetic in front of their own net or when chasing someone back into their zone! So they have to be changed at that moment (if the opposing team didn’t score by then already), and someone more capable defensively goes on the ice. Strome is much better defensively than either Ovi or Wilson, but he is just a tad bit too skinny and is shrugged off the puck very easily all too often. Backy is a lot better defensively, and even though he couldn’t really produce offensively (at least not this latter part of the season), he helped compensate for Wilson’s defensive struggles.
Overall, these metrics, together with shots for/against, and corresponding percentages, are rather predictive metrics that help see who controls play more (possession), and which team generates more chances while certain players are on the ice. They are more forward-looking, and there’s a ton of statistics out there explaining the Corsis, Fenwicks and other indices that have sprung up on their basis. Eventually, possession and a higher chances for ratio (xGF%) translates into goals. Even if they are scored after a certain line or players go off the ice, and someone else has to pick up their slack. I hope this helps in the discussion.