The third period of Tuesday’s game notwithstanding, the Washington Capitals have become a fearsome defensive team this season. In this last entry (for now) of the Twenty Games In series, I’d like to dive deep into Washington’s biggest improvement of 2021-22: the defense.

First, I suppose I have to make the case that Washington’s defense has actually gotten better. Here’s how many shots opponents take per hour of five-on-five play across the last six seasons.

We saw a very good defense in the Jennings year, followed by a crash that stopped just long enough for the Stanley Cup. Since then, the Caps have made improvements in shot suppression each season, finally landing just south of 50 per hour. That’s the sixth best number in the league (as of Tuesday morning at least). This lower shot volume has also been reflected in quality, as Washington’s expected save percentage, .928, is the lowest of this six-year sample, making life relatively easy for Samsonov and Vanecek.

Shot suppression isn’t the sole responsibility of defenders, but I think Washington’s defenders have made exciting changes from last season. Using hockeyviz’s visualization of defender pairings, here’s how Washington’s blue line has changed from last season to this. I’ve added my own annotations. At left is last season; right is this season.

All three pairings have been changed, and all three changes I think have been positive. Let’s dig in.

With Brenden Dillon’s exit, John Carlson now plays most of his minutes with rookie Martin Fehervary. Fehervary’s speed in particular has helped this pairing get very strong, especially on the right side of the ice. Carlson isn’t individually shooting more, but his shifts have been stronger overall when measured in shot-attempt possession (in green below) and expected-goals percentage (in blue below).

Carlson’s 1.11 primary assists per hour is currently second best among all NHL defenders, behind… um… Tony DeAngelo.

Justin Schultz is currently on the mend from an upper-body injury he sustained on Thanksgiving Eve, but before that the 31-year-old defender was having a dramatically improved season. It seems that Trevor van Riemsdyk has better on-ice chemistry with Schultz than last year’s partner, puck-moving dynamo Dmitry Orlov.

Schultz’s on-ice numbers have improved from under 50 percent last season, meaning opponents had more offense than Washington, to solidly above 50 percent this season.

I don’t read this as a condemnation or Orlov at all, who has seen his own improvement elsewhere. Instead, I see this as a vindication of Trevor van Riemsdyk’s value. TvR has already played more games this season than last, and he’s been perhaps Washington’s biggest surprise. We’ll talk more about whom he replaced in a moment.

Schultz-TVR is probably Washington’s most effective shutdown pairing, though maybe not their best overall.

Dmitry Orlov‘s offensive contributions has been increasing for a few seasons, but paired with another surprise-offense stud in Nick Jensen, Orlov is putting up the best point production of his career. With Orlov-Jensen on the ice, the Caps have an outrageous shot-attempt rate: 67 per hour. And Jensen’s goal on Tuesday night was his first of the season not to come directly off a pass from Orlov.

There’s nothing not to love about this pairing.

Here’s that same pairing but from the other side. Nick Jensen’s most common partner last season was septuagenarian Zdeno Chara. Together they had surprisingly strong expected-goals percentage (52 flat), but that’s the only metric in the whole lot that doesn’t point to unambiguous improvements.

There’s an implication to these improvements that the Capitals have undergone addition-by-subtraction: that using two defenders with a combined age of 52 (van Riemsdyk and Fehervary) instead of two defenders with a combined age of 75 (Chara and Dillon) yields better results. Yes, I think so.

Van Riemsdyk and Fehervary were both in relative obscurity last season, which itself could be considered an organizational failing, but at least it is one now rectified. Below are RAPM charts for TVR, Dillon, Fehervary, and Chara. RAPM charts show how far from replacement-level each player could be by various metrics. Higher is better. Each player has two boxes: even strength and power play. You can mostly ignore the power play ones. At left are the added players, at right the removed players and how they are faring in their new destinations.

Poor Zdeno Chara is having a very bad time on the island, on-ice for 67.4 opponent shot attempts per hour– effectively turning every opponent shift into one of those blistering Orlov-Jensen shifts from above. Brenden Dillon is faring better, but not as well as Trevor van Riemsdyk, who I think would have been an upgrade at any time over him last season. I’m not quite sure what to make of Fehervary’s impact on the game yet; I suspect we’ll need to see him in different contexts (i.e. apart from Carlson) before we get a reliable indication of how he drives play.

But for now it’s enough to say that this is the best Washington blue line we’ve seen since 2016-17. And that was a damn good team. My only wish is that Washington could get goaltending that strong again.

This story would not be possible without Natural Stat Trick, Evolving Hockey, and Hockey Viz. Please consider joining us in supporting them.

Headline photo: Elizabeth Kong/RMNB