We’re quickly approaching the trade deadline. Now’s a a great time to take stock at how the Washington Capitals are looking — something the team seems to be thinking as well as they called up depth defenseman Martin Fehervary for a big-league audition this week.
With that in mind, here’s an exercise we do from time to time. Let’s see what’s working and what’s not in the Caps lineup.
We’ve done this exercise a bunch of times, and I’ve shared more recent versions on the RMNB patreon, but now we’ve got some better metrics to use and a big sample to draw from. If you’re unfamiliar, here’s a quick guide:
- We order the Caps forwards at left from most ice time to least
- We do the same with Caps defenders along the top from most ice time to least ice
- Where a forward and defender intersect, we show the team’s expected-goals percentage (using NST’s xG reckoning for) when they’re together during five-on-five play
- Then we color-code it so green means good and red means not so much
When it’s all done, you get a pretty grid where the top-left represents a big chunk of the team’s ice time and the bottom-right represents way left. We want a lot of green at top-left.
Unlike previous attempts, I’m going to use expected-goals percentage instead of the team’s shot share. This percentage counts higher quality shots (based on location and context) more than others. Fifty percent would mean the Caps and their opponents break even in expected goals when a given player is on the ice. I’ll also share a version with ice time below.
This exercise is really good for just sorta gently staring at the picture to divine how the team’s doing overall, but also for drilling into problem areas or big successes. We’ll do exactly that below.
Expected Goals Percentage
Time on Ice
Observations
- Wow, that’s a lot of green. Just at a glance, it’s pretty clear this team has faring better than in any season since 2016-17.
- But more importantly than all the green, there are fewer acute problems in red than I’m used to. Even that 2016-17 team struggled when Karl Alzner was on ice. The closest problem like that here would have to be Travis Boyd, the team’s 13th forward, who has more than thirty scratches this season. He’s the purple row in the second table, indicating very low minutes with everyone.
- To have your number one defender-forward pairing, Alex Ovechkin and John Carlson, be so solid (58.3 percent of expected goals) is refreshing. They were 46.1 percent in 2017, back when the team required a ton of kinda inscrutable playmaking talent to outscore expectations. Now, they’ve added controlling the flow of the game to their elite finishing skill, and that’s why Carlson’s a Norris contender and Alex Ovechkin could hit 60 goals.
- Ovechkin’s xG success drops off after the first defensive pairing. He’s underwater with Kempny, Gudas, and Siegenthaler. I feel like I need to point out that — despite Ovechkin having an above-even expected-goals percentage with Nick Jensen (50.4 percent), their actual on-ice goals percentage is just 41.6 percent. I think that’s four parts bad luck and one part Jensen’s play-style clashing with the team.
- The highest numbers anywhere on the table are deep among the forwards. The fourth line of Garnet Hathaway, Brendan Leipsic, and Nic Dowd is probably the best fourth line in the NHL. Even when they’re on the ice with depth defenders, they’re still dominating play– but that 68.9 percent when Dowd and John Carlson play together — nice if you round up.
- John Carlson‘s column is telling. You see those great green spots at the top and bottom of the lineup, but look at the middle. Carlson and Evgeny Kuznetsov see the Caps control just 40.8 percent of expected goals when they’re on the ice together. It’s the worst kind of ice tilt. The team gets 49 attempts per hour — about what Detroit gets — and allows their opponents to get 66 attempts per hour — which is more than any team in the league averages. That said, while Kuznetsov’s performance has been uneven and worrying this season, he seems to have evened out recently as his line gains chemistry:
- Martin Fehervary’s call-up this week invited a conversation about Washington’s stack ranking at defense. While a lot of folks pick on Jensen for poor outcomes or Siegenthaler for inexperience, it seems to me like it’s Michal Kempny pulling up the rear. Kempny ekes out solid numbers with some forwards (Backstrom and the fourth line), but he’s sputtered (43.1 percent of expected goals) with Lars Eller, who is otherwise spotless in a very strong season. I wonder what’s going on. I had worried that he might have lost a step since his hamstring injury, but if that were the case I think he’d be committing more restraining penalties. Instead he’s cut his minor penalty rate in half. Perhaps he’d succeed with a different teammate– he and Carlson both seem to do better when apart anyway.
- I can’t seem to find anyone who plays noticeably worse when they’ve got the good fortune to take a shift with Jonas Siegenthaler. Okay, one, but it’s Evgeny Kuznetsov, and we’ve already stipulated that he’s a weirdo. Siegenthaler is not necessarily an offensive dynamo, but he’s been so solid at suppressing opponent offense that he is undeniably a good player. HockeyViz estimates that Siegenthaler reduces opponent offense by 12 percent— an excellent number. The only problem I can see is his ice time. He could play more, especially late in the season when big-minute players like Carlson need rest or when talent evaluation is happening.
That’s all I’ve got for now. What do you see that I missed, or what do you think I got wrong?
This story would not be possible without Natural Stat Trick and HockeyViz. Please consider joining us in supporting them via Patreon.
Headline photo: Cara Bahniuk
