Superficially, the Oilers were okay on the penalty kill, finishing in a tie for 17th overall in the NHL with the Vancouver Canucks, 0.1 points ahead of the Columbus Blue Jackets. Dig a little deeper, though, and the numbers get ugly.

What went wrong, what went right, and how can the team get better?

Edmonton ranked 27th in shot attempts against per hour while shorthanded; clearly they spent far too much time in their own end of the rink. They did a decent job of defensive zone coverage once in their own end, climbing up to 23rd overall in terms of shots against/hour, but that still isn’t a good number.

Cam Talbot saved the team’s bacon, as a high save percentage made a bottom-third NHL team look respectable when down a man.

It’s worth taking a moment to look at which players were helping and which players were hindering the Oilers on the penalty kill.

The chart above divides into two sections: the 2015-16 season, and then a four-year span from 2011-15 for context. Each section tracks three numbers: total ice-time, shots/missed shots (Fenwick) against per hour, and relative Fenwick against per hour. All numbers are at 4-on-5 only and are either courtesy of or generated from totals at stats.hockeyanalysis.com.

The last number is key. It shows, in an average hour, the difference in the number of shots and missed shots a team face with and without a player on the ice (as calculated by me). So, for example, using Matt Hendricks this year, Edmonton’s penalty kill faced a little under one less shot per hour when he was on the ice than it did when he was off it, suggesting that the unit he was a part of slightly outperformed the penalty kill as a whole.

Now, there are all kinds of caveats here. Linemate effects are probably a thing; PK forwards tend to play in pairs and stick together for a long time, and I can’t split these players up by performance with and without certain partners. We’re also looking at relatively small chunks of time. That’s why the four-year numbers on the right side of the chart are so important.

But enough explanation. Let’s look at the players.

Connor McDavid only played a half hour, but looks like superman through this lens.

Ryan Nugent-Hopkins and Iiro Pakarinen didn’t play a lot, but both were part of effective forward pairs when they were on the ice. Nugent-Hopkins has some pretty decent history here over his first seasons in the league and this is usually an area where young players struggle; I’d say it’s worth continuing to use him on the PK even with injury risk.

Matt Hendricks was a mainstay this year on the PK and did okay, but history suggests that he isn’t an elite penalty-killer. He’s going to be back for another year, but this is something worth keeping in mind when his contract expires next July 1.

Mark Letestu and Anton Lander sort of fall into the same category in that their numbers aren’t good this year, but both have been quite good in previous seasons. Letestu is almost certainly coming back and given his history I’d be comfortable using him as a regular. I feel the same way about Lander, but of course there’s a decent chance that he doesn’t return.  

Finally, we come to Lauri Korpikoski. Korpikoski has been an unmitigated disaster on NHL penalty kills for a half-decade now. The Coyotes were a whopping 8.5 events worse per hour over a four-year span when he was on the ice; Edmonton this year was 12 events worse per hour, which works out to an extra shot or missed shot against every five seconds when he steps on the ice. The man is a menace.

By the numbers, then, Letestu looks like a mainstay on a pretty decent unit, as does Nugent-Hopkins. McDavid and Pakarinen have been impressive over short runs and Hendricks is serviceable. If Lander’s back, he can contribute here. Korpikoski should never take another shift as an NHL penalty-killer.

After a long explanation above, here we can jump right to the players.

First, the kids. Griffin Reinhart looked really good in limited minutes, but as with McDavid and Pakarinen it probably makes sense not to read too much into that. Darnell Nurse, Oscar Klefbom and Brandon Davidson all struggled by the shot metrics, which isn’t at all surprising given their age. Klefbom looks better if we toss in last year’s results, but again we’re talking about a small run of minutes and it doesn’t pay to read too much into them.

Now we turn to the veterans.

Adam Pardy is the kind of player who looks like a natural fit on the penalty kill, and he’s been good whenever used. It’s something Edmonton may want to keep in mind when it makes a decision on him this summer.

Andrej Sekera’s numbers are surprisingly good, at least to me; despite being a touch undersized he’s found a way to be effective on the penalty kill in the NHL.

Mark Fayne had a good year, mostly playing with regular partner Sekera on the PK. His historic numbers don’t look too great until one remembers that he spent the bulk of his time with New Jersey and that the Devils had the best shot-preventing penalty kill in the league when he was there. Being a touch below average on the best PK team in the league doesn’t represent any kind of problem in my view.

Eric Gryba had a tough year by the numbers, but I don’t really believe it. He would have been out a ton with either Davidson or Nurse (both regular even-strength partners) and Gryba’s previous work was excellent. If he’s re-signed, he’ll rightly be a mainstay again.

Defence is likely to see a fair bit of changeover, but of the players on this list the veteran quartet of Sekera, Fayne, Gryba and Pardy are all good choices for penalty kill work. Davidson, Nurse, Klefbom and Reinhart meanwhile are all likely to go through growing pains as they put in NHL time.

As a general rule: veterans good, kids bad, especially on defence.

Letestu and Nugent-Hopkins are the two key holdovers on the penalty kill up front. McDavid looked great and Pakarinen looked good in limited duty. Lander is another good option if he returns, while Hendricks fared much worse than I expected and would seem to be only serviceable. Korpikoski should never play another NHL penalty kill shift.

Sekera, Fayne, Pardy and Gryba are a good veteran core for a penalty kill defensive unit. All the young guys other than Reinhart struggled, and Reinhart didn’t get enough minutes to really be confident in his results. Nurse, Klefbom and Davidson appear to have been liabilities when Edmonton was down a man this year and will need to improve.