Wednesday, 25 November 2015

Mike Babcock's use (or misuse) of the Maple Leafs roster: Part 1 - Forwards

Toronto Maple Leafs: Players Who Should Flourish Under Mike Babcock
Image courtesy of
Who in Leaf Nation wasn't ecstatic when Mike Babcock was brought on to be the Maple Leafs head coach? Rumoured to have been out of the running completely to Detroit and Buffalo, the Maple Leafs were, from the fans' perspective, a surprise choice as Babcock's professional home.

In his introduction presser, Babcock said, as part of his introduction, "It's the Maple Leafs. And I'm proud to be here today, and I look forward to the process, the battle, the pain, the fun, the journey."

And what a journey it's looking like it will be.

The Leafs have posted a 7-10-5 record in their first 22 games under Babcock's guide. While there certainly have been sources of optimism, the overall outlook on this particular season remains bleak.

Mike Babcock brought with him, to Toronto, a fantastic personality, a hard working attitude, strong systems, and a plan for the future. But what is that plan?

Babcock, again from the introduction presser, after a question from Michael Grange of Sportsnet, "We want to build a team, off the ice and on the ice, that the fans of the Toronto Maple Leafs can be proud of. We have to create an environment that's safe for a player. What I mean by that is, when you win every day, it becomes pretty safe for the players."

So let's take a look at the environment he's created for the players, and how he's using them.

I'll start with forwards today, and some time soon I'll do the defense and goaltenders in a Part 2 post.

The simplest way is to use Stephen Burtch's Expected Corsi (explained herebroken down into a statistical writeup here, and data shown here) to represent coaching usage. At the same time, we can present Burtch's dCorsi to show each player's results in this usage. I'll additionally break down some of the factors that contribute to a player's Expected Corsi to show what Babcock is using these players for. All of the factors that go into the model can be found in the statistical write up linked above.

I calculated a rough league average for Expected Corsi For (per 60 minutes of ice time) (ECF60) Expected Corsi Against (per 60 minutes of ice time) (ECA60), which were 51.59 and 52.08, respectively. We'll use these numbers to establish whether each player is getting harder or softer minutes. I calculated them by averaging the top 20 values of each stat, the bottom 20 values of each stat, and finding the average of those two averages. Very messy, and probably not close to the actual number, but gives us a number halfway between the low/high extremes that we can use.


Player TOI/G ECF60 ECA60 ZSO% rel CorT% CorC% CF% rel GF% rel dCorsi60 dCorsiImpact
Nazem Kadri 15.42 62.59 54.98 20.38% 49.51% 50.3% 3.96% -5.3% -3.71 -20.99
Tyler Bozak 13.29 55.85 56.01 9.76% 49.05% 50.64% 2.5% 11.67% 0.16 0.64
Nick Spaling 12.86 48.68 58.82 -24.65% 48.86% 51% -10.96% -7.5% -9.97 -34.21
Peter Holland 9.72 48.4 57.63 -12.66% 49.42% 49.61% 1.63% -13.48% 9.96 27.43
Byron Froese 10.16 44.29 60.96 -33.04% 49.12% 49.85% -5.91% -15.24% 2.65 7.17

The results stats (CF%, GF%, dCorsi) are only there for the curious. This post is discussing player utilization only, so that's all I'll talk about. Scope is important.

And boy is there a lot to talk about!

With this graph you can take away many points about usage for each line. The centers have stayed basically put in terms of their position in the depth chart, so we can analyze lines by their center. Namely:

Babcock is playing typical top 6 bottom 6 zone start structure

You can see from the top 2 centers (Kadri and Bozak) versus the bottom 2/3 centers that the top lines are getting more offensive zone starts. Their ECF60 numbers are well above average.

There is a distinct checking line

The 3rd line (Spaling) is getting the toughest competition while the top 2 lines and the 4th line are being sheltered. This line also has the lowest CorsiT% for Spaling, which is indicative of a line that the coach is okay with being out-possessed due to their superior defensive zone coverage. Its ECF60 is well above average.

The second line isn't very strong

Bozak's CorT% is the lowest of the team, even lower than the 4th line! That's not ideal. Ideally CorT% would decrease with line order. Seems crazy to say this, but it seems we need to support Tyler Bozak better.

The top line is being put in a position to score

Kadri's ECF60 of 62.59 is very high, significantly that the average I calculated above of 51.59 (actually, good for 7th in the league for players with more than 1 GP). This is a very good situation to be putting your top line in.


Player TOI/G ECF60 ECA60 ZSO% rel CorT% CorC% CF% rel GF% rel dCorsi60 dCorsiImpact
James Van Riemsdyk 14.45 63.32 54.4 24.21% 49.69% 50.20% 10.45% -2.16% 5.43 28.75
Leo Komarov 13.84 62.11 55.21 49.46% 50.27% 50.64% 8.91% 1.11% 5.12 26.00
Joffrey Lupul 12.57 48.71 58.29 -17.98% 49.22% 50.48% -7.67% 4.00% -5.18 -23.89
Shawn Matthias 12.32 56.12 56.4 8.60% 49.32% 50.34% -1.35% 3.47% -5.28 -22.76
Michael Grabner 8.53 47.73 57.62 -13.42% 49.10% 50.16% -1.41% 8.09% 4.97 14.12
Daniel Winnik 11.59 48.07 59.35 -22.80% 49.10% 50.16% -7.76% 25.76% -5.52 -21.31
P.A. Parenteau 10.86 54.27 56.1 3.21% 49.02% 50.20% -2.51% -12.50% -4.7 -18.71
Brad Boyes 9.28 52.81 56.2 1.35% 48.64% 50.06% 4.8% -12.97% 11.37 29.90

Since the utilization of wingers is much more spread out, we can only use their numbers to see their individual usages. This is still interesting, but not as definitive as being able to extrapolate for a whole line like above.

Still, though, there's so much to talk about!

Like I said above, I'm only discussing utilization, so I'm definitely going to leave out how amazing Brad Boyes has been. I definitely wouldn't need to mention how underutilized he and Grabner are.

Also, it would be completely breaking my scope to even begin to discuss the disappointing performances from Winnik, Parenteau and Matthias.

And lastly, it would be inconceivable to mention how much better Van Riemsdyk and Komarov have been at possession than Kadri has.

So I won't. I'll talk about utilization. For instance:

Notice how close everyone is in ECA60

Competition and other factors that strongly contribute to allowing scoring chances against seem to have been distributed reasonably evenly around the Maple Leafs. Namely, see that all the teams' CorC% is within 1% of each other. This is why people talk about quality of competition stats being near useless in large samples.

Notice how close everyone is in CorT%

This one isn't quite as expected as the similarity in CorC%. This shows the Leafs team is balanced from top to bottom in possession ability, which would be great if they were a really good possession team.

Michael Grabner hasn't been given a chance to succeed offensively

With an ECF60 of only 47.73, for the offensive talent he has shown he can be in the past, he hasn't been given a real chance to produce offensively. It would be nice for him to get a chance on the second line with Bozak.

Joffrey Lupul has been succeeding offensively even though maybe he shouldn't be

Lupul finds himself in a much similar situation as Grabner. And yet, he has a GF% of 4.00% and 5 goals. Good for him, but likely an unsustainable production level if his usage remains the same.

Well, that's about all I can draw from this. If you find anything interesting here or in data somewhere else about utilization that I missed, please share in the comments!

No comments:

Post a Comment