Maple Leafs player usage chart: February 4

Hockey fans familiar with the work of Hockey Prospectus‘ Rob Vollman may recognize the “Player Usage Chart”, which is an excellent way to visualize the way players in a system are being used, and how successful they are.

Below the jump, I have an early season look at the Toronto Maple Leafs’ chart. Even though the season is but eight games old, a couple of things popped out at me. The horizontal axis of the chart represents the percentage of times a player started in the offensive zone relative to the defensive zone. The vertical axis of the chart represents the quality of competition a player has faced by Behind The Net’s Corsi Rel QoC measure. Finally, a players’ circle and its respective colour show how successful a player is at his role.

A more detailed explanation on the graph from Puck Prospectus:

What are Offensive Zone Starts (Horizontal Axis)?

Offensive zone starts is the percentage of all non-neutral shifts started in the offensive zone.  A common misconception is that it’s the percentage of all shifts started in the offensive zone, but it ignores those in the neutral zone and is therefore perhaps poorly named (like most hockey statistics).  Think of it more as a representation of whether a player is used primarily for his offensive talents, or defensive.

What is Quality of Competition (Vertical Axis)?

Quality of Competition is the average plus/minus of one’s opponents over 60 minutes, except that it is based on attempted-shots (Corsi) instead of goals.  In this particular variation we are using Relative Corsi (explained below).  Players who face top lines will have high QoC’s while those with the easier task of facing mostly depth lines will have negative QoCs.

What is Relative Corsi (The Bubbles)?

Corsi, another poorly named statistic, is simply a player’s plus/minus, except that it’s measured in attempted shots instead of goals.  In this case it’s calculated over 60 minutes, and Relative Corsi is calculated relative to how the team did without him.  As explained by Corsi bubble innovator Eric Tulsky, a big blue bubble represents someone whose team attempts a lot more shots than their opponents while he’s on the ice, and a big white bubble is someone whose team is usually getting outshot badly.  In Tom Awad’s variation the bubble is sized according to a player’s ice-time and shaded dark green or dark red based on their Relative Corsi.

What our early season chart shows is how difficult a task has been given to Nik Kulemin and Mikhail Grabovski early this season. While both players are facing the same difficult QualComp minutes compared to last year (that chart can be found here) they’re starting far more shifts in the defensive end of the ice as the first and third lines get more offensive starts.

Despite this, their bubbles are still blue (albeit very small blue bubbles) so they’re doing an excellent job relative to their tough situations this season.

The other thing I can gather from this is that David Steckel is being mis-cast as a fourth line centreman. Last season, he was stretched much further to the left, leading the Leafs in defensive zone face-offs taken, and wasn’t beaten too badly despite the situation. Playing primarily between Colton Orr and Mike Brown or Frazer McLaren this season, Steckel’s face-off skills are wasted while playing with players who can’t do anything with the puck. Tyler Bozak is still seeing far more PK minutes than Steckel as well, even as Toronto’s second PK unit continues to be lit up for scoring chances against.

Cody Franson and Mark Fraser are having moderately successful seasons, but they’re minutes are limited, and when they play, they’re out against easy competition. Korbinian Holzer has only played once this season, but he had a poor game, so hopefully he can make amends during this call-up to replace Carl Gunnarsson, but given Gunnarsson’s position on the graph, that’s a pretty tall order.

Mike Kostka is right there with James van Riemsdyk. That’s quite surprising to me since van Riemsdyk has seemed like he’s helped the Phil Kessel line push play. He’s brought some good things to the powerplay, but evidently he’s having some trouble at even strength. The team misses Clarke MacArthur more than Joffrey Lupul, and I think he’s a more versatile player who can score from a checking role in a pinch and is one of the Leafs’ better pickup in recent years.

This is quite early in the season, but it’s a depiction as to how the Leafs players are being deployed. Note the graph is fairly linear: Players with high Quality of Competition rates are more likely to have lower offensive zone opportunities. Last season, other than Steckel, the Leafs were deployed seemingly at random. I prefer Carlyle’s line matching of the first three lines but it’s clear he’s mis-cast the fourth.

Again, thanks to Rob Vollman. His Twitter feed is @RobVollmanNHL and he posts excellent stuff on HockeyAbstract.

  • Komisarek on the top pairing with Phaneuf – does that look funny to you?

    And Steckel is playing under five minutes per game – I assume he’s hurt. If healthy it would be interesting to move either him or McClement to the wing and make a premier shutdown line.

    Grabovski’s great.

    • Ha it looks like it. Phaneuf’s been playing with Kostka.

      Don’t think Steckel is hurt, but when the Leafs brought McClement in in the summer, they did the following:

      A – Put McClement between Orr/Brown and Steckel in the press box

      When MacArthur and Lupul got hurt, they

      B – Put McClement on the second line with Grabovski and Kulemin

      C – Put Steckel on the fourth line

      D – Stopped playing the fourth line

      McClement saw a lot of PK time and had the occasional defensive specialty shift with Grabovski and Kulemin, so he was hovering around the 11:00 mark most nights. If Steckel is hurt, then their fourth line deployment makes even less sense. I don’t know why they wouldn’t try to move Steckel up with say, Komarov and van Riemsdyk, for a second “specialist” shutdown line to boost his minutes.

      It’s a weird situation.

  • KesselSnipes

    Thanks, this was interesting. Would love to see an update in a month or so. Another idea, not sure if it’s possible, would be to do do a ‘diff’ between the current graph and the updated graph, so you can visualize how the usage of players changed, as well as their success.