Jason Mackey at the Trib recently posted these two tweets about the Penguins recent penalty kill efforts:
“Wow. Penguins have allowed 12 PPG over their past 36 games. They’re 120 for 132 during that stretch (90.9%).”
“Even better, they’ve been shorthanded 147 times – second most in the NHL. Yet their 87.8 percent success rate ranks second.”
I’m pretty impressed by these numbers. I think we all remember how bad the penalty kill unit looked at the beginning of the season. Looked like the Pens couldn’t catch a break, couldn’t stop taking penalties, and like we were in for a long year. From Mackey’s tweets, it seems as though they still haven’t quite figured out how to stop taking penalties (second most penalized team in the NHL), but it certainly appears as though patience has paid off, and the penalty-killing unit has figured itself out.
As always, I wanted to dig into these stats a little further. I did some reading on how best to evaluate penalty kill units, and I was surprised at what I found.
Generally, the first thing I look at when evaluating play is shot attempts, since we have so many more shot attempts than goals. Consequently, when I’m looking at a player’s defensive ability, I like to look at shots against. It stands to reason, then, that shot attempts against would be the best way to measure penalty killing ability.
Not so, say Eric Tulsky and JLikens.
As Tulsky writes:
“Penalty kill save percentage turns out to be a lot more reproducible than power play shooting percentage -- teams have a lot more control over how many shots they stop than over how many of their shots go in. As a result, JLikens was able to show recently that this changes how we should look at penalty kill units. He found that if you take a random 41 games from a team's season and try to predict their PK% in the other 41 games, you do much better if you use PK% in the first set to make your guesses than if you use SOG/60. In other words, teams don't have all that much control over their shooting percentage, so a good power play is one that generates a lot of shots. But teams do have control over save percentage, and just using shot rate to evaluate penalty kills leaves out the skill of the goalie, so you actually do better to use PK%.”
To date, the Pens have played 40 games, just one game shy of the 41-game, halfway mark. The Pens (87.9%) currently sit only behind the Chicago Blackhawks (89.7%) in PK%.
Last season, the Penguins finished fourth on the penalty kill, with a PK% of 85.0%. So it should be said that the Pens have had very good penalty killing units for the last two seasons. Still, I’d like to take a look at what’s kept them dominant this year (and last).
With Tulsky and JLikens advocating that we look at both shot rate and the skill of the goalie, that’s what we’ll do.
2014-2015 Shorthanded stats, sorted by shots against per 60 (SA60)
For the sake of comparison, here are the 2013-2014 shorthanded stats, once again sorted by shots against per 60 (SA60)
|(Tables courtesy of hockeyanalysis.com)|
As you see in the bolded column, the Penguins are giving up just about the same number of shots on the penalty kill this year as last year. We have a smaller sample for this year, but overall, they gave up 1.2 more shots per 60 last year (58.3) when compared to this year (57.1).
(Additionally, I looked at shorthanded scoring chances against per 60, courtesy of War on Ice, and the Pens are giving up slightly more shorthanded scoring chances this year than last—54.7 per 60 this year compared to 54.5 per 60 last year, 23rd and 25th in the league respectively—and the closeness of these two numbers to each other, and their position in the league, leads me to believe it hasn’t been a huge factor in explaining performance.)
Leaving the analysis here, however, would exclude goalie talent, so we’ll also look at Fleury’s performance.
Among goalies with at least 20 starts, Marc-Andre Fleury’s .930 SH Sv% is second in the league, behind only Colorado’s Semyon Varlamov (.963).
For comparison, Fleury finished 2nd in the league last year, with a .911. First in the league was New Jersey’s Corey Schneider, who posted a .918 shorthanded Sv% on the year (minimum 40 starts).
Fleury was excellent last year, when compared to his peers, and he continues to be quite good this season. What’s interesting to me is that although his SH Sv% is much higher this year, Fleury’s position in the league hasn’t changed, which I believe can be attributed to the small sample size. Fleury is still playing dominantly, but the numbers are higher now than they might be through 82 games because we are only halfway through the season.
Overall, I don’t think it’s likely that Fleury maintains his .930 SH Sv% (Varlamov even less so), but I think it’s clear that Fleury has been excellent for the Penguins, even as they have defensively been a below-league-average unit in front of him.
Note: Martin, Agnew, and others have noted that the Penguins are using a similar penalty killing system to last year, which may explain similarity of shot rates against.