On Friday, Isabelle Khurshudyan of the Washington Post (subscribe!) ran a great item about the Washington Capitals’ dragging penalty kill. In that piece, the pull quote from Barry Trotz was that the team “can’t kill anything” when a man down. While that certainly feels true, the resulting numbers for special teams after just ten games might not reflect the actual underlying processes.
Trotz was critical of the PK. “Just a lot of extra pucks, a lot of failed clears,” Trotz told Khurshudyan after the disastrous outing in Vancouver. “When you’re spending a lot of time in your zone and they’re putting pucks there, they’re going to find the back of the net. . .we’ve just got to be better back there.”
He’s right; the Caps have to be better. But the early PK results might not be an accurate indication of how the team is playing while a man down, and the same may be true of the power play as well.
I grabbed data from the NHL, Corsica, and Natural Stat Trick to look at an array of stats that describe how the Caps are generating offense on the PP and limiting it on the PK.
Power play
| Stat | Number | Rank |
|---|---|---|
| Conversion % | 22.9 | 9th |
| Shooting % | 7.5 | 17th |
| Caps attempts / 60 | 47.5 | 7th |
| Caps expected goals / 60 | 5.3 | 29th |
| Caps scoring chances / 60 | 49.7 | 10th |
The red flag here is the Caps’ low expected-goals rate. Expected goals are calculated using a model of a team’s offensive profile: how much they shoot, from where, and under what circumstances. For whatever reason, the Caps’ man advantage is not favored by that model.
Micah Black McCurdy has a great visualization for the location and volume of Washington’s first unit.
This looks just fine to me – lots of Ovechkin from below the faceoff dot and lots of Kuznetsov and Oshie from the slot. Personally I don’t think being ranked ninth is overly high or low, though perhaps we should keep an eye on that expected-goals number.
Now onto the so-far dreary PK.
Penalty kill
| Stat | Number | Rank |
|---|---|---|
| Kill % | 73.3 | 29th |
| Save % | 90.5 | 24th |
| Opponent attempts / 60 | 55.8 | 22nd |
| Opponent expected goals / 60 | 7.1 | 12th |
| Opponent scoring chances / 60 | 61.1 | 18th |
The PK is terrible according to its kill rate (T-29th), and they are certainly giving their opponents too many opportunities (T-22nd in attempts), but the share of scoring chances and the resulting expected goals are far more bullish – ranking the Caps as mediocre rather than terrible.
The implication seems to be that that save percentage (or, conversely, opponent shooting percentage) are what’s driving the early results for the PK. Over time, we should see shot volume (and the quality profile of those shots) have a larger impact on the overall kill rate. The problem is that those volume numbers aren’t unambiguously better. While the gap between expected goals and attempts might suggest that the Caps are keeping most opponent attempts to the perimeter, it’s still worrying to see the team in the bottom third at shot suppression.
One final note: The Caps shouldn’t rush to grow their PK numbers too quickly. So far they have committed 54 penalties and drawn just 41. That minus-13 differential is tied for worst with Anaheim for in the league. In addition to underperforming on special teams, discipline is a real problem for this team. They are spending too much time a man down and not nearly enough a man up.
Headline photo: Cara Bahniuk