Randy Carlyle Coaching Myths Part 4: The Penalty Kill

After debunking the myths that Carlyle's systems reduces shot quality, that he's been better than Wilson, and that increased team toughness is worth a damn, the last bastion of hope by those who will not concede that Randy Carlyle is a bad coach has been the Leafs' penalty kill.

After spending every year since the first lockout among the bottom five teams in the league, except 2005-06 when they were mercifully 24th, the Leafs currently find themselves with the 3rd best penalty kill in the league. Many look at that vast improvement and immediately attribute it solely to Carlyle's coaching system. In comparison to the inept system that Ron Wilson ran, which used ridiculous fronting and an insistence to keep two hands on one's stick at all times, 4 monkeys trying to fuck a football would be an improvement.

Draglikepull wrote a couple of great posts last year comparing the Leafs' penalty kill to the Minnesota Wild's kill. In the first post he found that the goalies were facing a fairly equivalent workload but seemed to be giving up higher quality chances partly as a result of the team's insistence on fronting. But was it really all the fault of the coach's implemented system, or did the terrible goaltenders at Wilson's disposal also play a part, and more importantly how does that translate to judging the Leafs' newfound PK success under Carlyle?

What I did was take the Leafs' PK numbers for this season as well as the five previous seasons; times shorthanded (TS), powerplay shots against (PPSA), powerplay goals against (PPGA), shots against per time shorthanded (SA/TS), the PK% Rank, and the PK save percentage (PKSV%); and then found the average for the previous five seasons and looked at what this year's PK would be like with the goaltending from the previous five seasons as well as the SA/TS.

Here are my findings: (links to spreadsheet; data as of April 6th, 2013)

First, there is a slight improvement in SA/TS for this season. The Leafs are 7th in the league for SA/TS behind only STL, BOS, NJD, PHI, SJS, and MTL; five good defensive teams and then Philadelphia. Though with players like Jay McClement and Nikolai Kulemin playing so well, their PK acumen definitely plays a role in the improvement. Just think about how less frequently you're hearing the phrase "to the line but not out" with Jay McClement clearing it with relative ease. Secondly, even if this year's team gave up as many shots as the Leafs of seasons past, it would only drop the PK% from 86.7% to 84.4% (3 ranks) while only adding 3 goals against.

However, when you look at the PKSV%s of this year and the previous five there is a huge difference. A 3rd ranked PK down to a 22nd ranked PK? Oof? A drop from .890 to .835 in PKSV% would result in a full 50% more goals against on the PK (8) which would be worth nearly 3 wins in the standings, and currently have the Leafs sitting 9th and out of a playoff spot.

Also when comparing the Leafs' PKSV% to league average, we can see how horrid it's been, and how amazing it currently is.

As shown here, here, and here predicting a goalie's (or team's) PKSV% from year to year is a fool's errand. There is simply far too much variance from year to year to predict it, and most likely we should just assume the PKSV% to regress to the league average of ~0.867. Thus looking at how the Leafs PK was constantly far below league average for the previous 5 years should tell you all you need to know about whether or not the Leafs had actual NHL quality goaltending (they did not). It should also tell you that don't expect this run of .890 PKSV% to continue next season. And as it will fall, so too will the PK% itself. If you do the math for this year's PK with a PKSV% of .867 instead of .890....

...you'd see that we'd drop to a slightly above average (81.7%) PK. Something to definitely celebrate compared to year's past, but not something with which to heap outsized praise upon Randy Carlyle especially since he doesn't run it anyway, Scott Gordon does. That's a guy that deserves some praise. Mirtle recently spoke with an analyst that noted that goaltending played a larger part in the improvement for what it's worth but that the team is also giving up lower quality chances:

According to a recent analysis done of Toronto's play by the Sports Analytics Institute - an advanced statistics group which currently works with two other NHL teams - the answer is a little of both, with the biggest differences being better goaltending and better shorthanded play.

SAI analyst Mike Boyle, for example, calculates that after Saturday's game, the quality of shots the Leafs goalies have faced is down roughly 3 per cent at even strength and 13 per cent on the penalty kill this season over last year, making their jobs easier.

But "goaltender effectiveness" is also up 17 per cent, which is consistent with how many more pucks they have been stopping over last season when Reimer was hurt and Jonas Gustavsson appeared in 42 games.

So in conclusion, just as it's the reason for the Leafs overall success, the vastly improved goaltending provided by Scrivens and Reimer is a major reason for the team's improved penalty kill. Other factors include improved personnel (thanks Burke!), and Scott Gordon taking over the penalty killing duties after being stowed away on the powerplay last year. It's not just some new PK system but a variety of interconnected factors.

P.S. - I didn't include shot location in my study because there are only 98 shots with coordinates available on theninjagreg's shot tracker and I didn't feel that was enough to draw any sort of conclusion about shot location as it's only 25% of the shot total with coordinates we have for last season. So while the average distance is 3.5 feet farther out, the shots are too scattered to say for certain that there is any appreciable difference here, and frankly I don't really see any less shots coming from the "home plate" area.