What’s been driving the Leafs’ improvements on the PK?

Back in September I wrote about the one area I was concerned about in regards to the Leafs heading into the season: the penalty kill.

For those of you who, understandably, refuse to consume two of my posts in one sitting, the Leafs gave up a tonne of shots at 4v5 and the only reason they didn’t give up a tonne of goals is because Frederik Andersen stood on his head. They were bottom five in every shot metric and 4v5 sv% is extremely volatile, so if the Leafs didn’t fix their process they were destined to give up way more goals on the penalty kill this year.

Good news! We’re nineteen games into the season.. and they’ve improved! They’ve been pretty much league average across the board, which is a significant improvement over last season.

They’re giving up less total shot attempts and less unblocked shot attempts, resulting in a much better expected goals against number. They’re still conceding as many actual shots on goal against, but I’d bet on that number decreasing if they continue to give up fewer total shot attempts. Unfortunately the Leafs are still giving up goals at a higher rate this year than last due to the 4v5 sv% predictably slipping a little bit. It’s still been about league average, but it was just so good last year that a bit of decline was more or less inevitable.

Looking at fenwick save percentage (FSv%) vs. expected fenwick save percentage (xFSv%), you’ll notice that the ’16/17 Leafs goaltenders performed significantly above what should be expected based on the location of unblocked shot attempts given up. Their 90.62 xFSv% ranked 24th in the league, implying that on top of giving up a high rate of shots, they were giving up high quality shots.

Early on this year, though, things are looking a little better. Their expected save percentage ranks tenth in the league at the moment, which implies that they’ve been successful at cutting down on quality shots conceded on top of merely giving up less shots. They aren’t getting the ridiculous 4v5 goaltending that bailed them out last year, but they’re still performing slightly above expected.

So, what’s changed? The coaching staff hasn’t, but the personnel has. Most notably, free agent acquisition Ron Hainsey has played pretty much every second of every penalty kill and Nikita Zaitsev isn’t too far behind him. Hainsey has literally been on the ice for 85% of the time the Leafs have spent on the penalty kill, the highest percentage since the lockout by almost 12%. Insanity. The visuals below (plucked from the amazing hockeyviz.com) should help put into perspective just how much the Leafs are leaning on the two defenders this year relative to how much they leaned on anyone in ’16/17.

It seems doubtful that Mike Babcock plans on leaning on these two to this extent for the entirety of the season and as the season goes on you’ll see some different names rotated in with injuries and such, but they’re clearly going to use them a lot. I’d like to see Jake Gardiner used a little bit more on the penalty kill when they reign back Hainsey’s minutes a little bit due to his ability to defend zone entries, which I’d argue is the most important asset of penalty killing. Or maybe I’m wrong and the Leafs’ world class sports science department thinks that Hainsey can continue to play this much on the penalty kill without the wheels falling off, but it seems rather unlikely. It seems even less likely when taking into consideration that he’s also playing against top competition at 5v5.

The penalty kill has also once again seen a heavy dosage of Zach Hyman, Leo Komarov and Connor Brown up front, with a few other players rotating through the final spot, which is more or less what they did last year.

Whatever route they decide to go in the personnel department, it’s going to be interesting to see if the early improvement is real or not and whether Hainsey and Zaitsev don’t have their legs fall off their body at some point.

*data via corsica.hockey