Their season both disappointing and over, the Washington Capitals will be the topic of conversation a lot this summer. They’ve already fired a coach, and they’re bound for more drama around some star and core players like Kuznetsov, Backstrom, and Mantha.
But before we get into that, I want to get us on the same page about what the Capitals were. I recently saw a person characterize the 2022-23 Capitals in a way that was in conflict with what I saw, so I thought maybe all of us could benefit from a check-in with the cold hard facts.
These were your Capitals.
During five-on-five play, the Capitals ranked 17th in goal rate, just below the Islanders and above the Kings. That goal performance ranked slightly below the team’s ability to take shots – they ranked 14th in shot-attempt rate. Factoring in the quality of those shots, using expected goals or xG, the Capitals were 16th. The team similarly ranked 17th in shooting percentage.
Placing between 14th and 17th out of 32 suggests the Capitals were a middling offensive team, a notion backed up by HockeyViz’s heatmap of the team’s offense. The red blobs on this heatmap indicate that the team took more shots from that location, so a red blob near the opponents net would be good.
Micah McCurdy’s model describes the Caps’ offense as being one percentage point below league average. Unremarkable.
Opponents outscored the Capitals 182 to 173 during five-on-five play, so there must be a deficit somewhere on defense. Washington’s goals-against rate ranked 21st in the league, the same as their attempts-against (meaning the rate of shot attempts taken by opponents against the Caps). Taking quality into account, things were fractionally worse – 22nd place. Completing this pattern is the goaltenders’ saving percentage, which ranks 21st.
I should add that the Capitals did notably better at keeping opponents away from dangerous shot locations; their rate of opponent high-danger chances ranks 16th. That goes a long way to explaining HockeyViz’s neutral reckoning of the team’s shot suppression.
Special teams were not dissimilar. The Caps shot-generation during the five-on-four power play placed them 16th in the league. That rank drops to 22nd when factoring in quality using xG, likely a penalty for relying on Ovechkin’s one-timer from the faceoff dot, which, due to its pre-shot passing, is hard to appreciate by models that depend on publicly available data. Ultimately, the team’s actual goal rate ranked 13th, and their conversion rate ranked 16th.
Again, consulting HockeyViz, we see a seemingly below-average Washington power play. In their special teams graphs, brown blobs mean more shots come from a location, so the big brown blob by the left faceoff dot suggests the Ovi Spot. In our minds we could house-rule that brown blob to be more dangerous than indicated.
The penalty kill could be seen as a relative bright spot. The Capitals did not fare well at limiting overall attempts, ranking 25th in opponent shot-attempt rate, but they improved to 11th when taking into account the quality of the opponent’s chances. That helped the team keep opponent goal-rate at 13th in the league, a number helped by strong goaltending, which ranked fourth by save percentage.
Here is HockeyViz’s idea of the Capitals penalty kill.
Putting all this information together, we can see how different compartments of Washington’s game compare to the rest of the league. Below are the ranks for the four states discussed above, plus my subjective grades.
The biggest gap between this rate-and-rank-based understanding and HockeyViz’s visualizations comes in the five-on-five defense, which they measure as better: almost identical to league average. Otherwise we get a summary that isn’t hard to characterize:
The 2022-23 Capitals were a mediocre team, not noteworthy during even-strength play and only slightly above average at special teams.
That is a simple summary, and under scrutiny it completely falls apart. Because the Washington Capitals were never static. I do not recall any sustained stretch when the Caps played like a league-average team. Consistency was a quality this team did not have (which is not always a bad thing).
One obvious example is the team’s injuries, among the worst in the league and visualized well by NHL Injury Viz. This visualization is longitudinal, so the x-axis is time.
Longitudinal is the key to understanding the Capitals. If we look at the team’s season as slices of time demarcated by injuries, we see that this team was unstable. I had to simplify a great deal in the visualization below, which is still pretty busy, but I hope the effect will come across. The line in this graph shows the percentage of shot attempts taken by the Capitals over time – so above fifty is good. The line is split into sections based on player absences, which I’ve had to simplify for legibility.
We see a modest team that takes a month to get going, but who finds their groove in late November and December, before the Carlson injury starts a long decline that gets progressively worse even after his return, when the team evidently gives up.
Splitting the Caps’ offense and opponent offense into separate lines, we see this more clearly. The blue line is Washington’s expected-goal rate (offense) compared to the league median rate. The red line is the same for opponents against Washington.
The offense is spiky to an uncommon degree, starting poor, improving (seemingly independent of injuries) through Christmas, and then crashing around the same time Carlson exits the lineup, and Backstrom and Wilson return. After that there’s fits and starts as Washington generates good offense for brief spurts before falling quickly back to league average. I think that instability is extraordinary.
On defense, the team starts slow, then becomes fantastic in midseason, then becomes terrible at the end – by this metric they were fourth worst in the league over the final 25 games of the season.
In totality this was a stupefying dull team, but that’s only in aggregate of the thousand different manic iterations that existed of them. For a moment they were genuinely one of the best teams in the league, driven foremost by excellent defense, but then lineup changes happened, and they got much worse. Then they sold at the deadline and were miserable for a month and a half.
I don’t mean any of this to criticize the coaching staff or to pin the season’s failure on individual players. I’m just finding facts. This was a complicated season, a maddeningly frustrating one for this writer, and we’ll all struggle to characterize it succinctly, except to say it was bad.
(It was bad.)
This story would not be possible without Natural Stat Trick and HockeyViz. Please consider joining in supporting them.
Headline photo: Alan Dobbins/RMNB
