clock menu more-arrow no yes

Filed under:

A new identity: Finding the optimal defence pairings for the Montreal Canadiens

New, comments

Using microstats to project the best lineup on defence the Canadiens can ice in 2017-18.

Washington Capitals v New Jersey Devils Photo by Elsa/Getty Images

Recently, I used Ryan Stimson’s player categorization system to identify how the Montreal Canadiens can maximize their offensive output. Stimson’s system also applies to defencemen, so it’s certainly appropriate to focus on the blue line as we continue our look at the optimal deployment of the Habs’ roster.

Like the forwards, Stimson’s clustering has identified four distinct archetypes: “puck-movers,” “all-around,” “volume shooters,” and “defensive-oriented.” Not surprisingly, all-around defencemen, consisting of some of the league’s elite, contribute the most to their team’s offence (expected goals for per 60 minutes, or xGF60), while stay-at-home defensive-oriented players contribute the least.

Full methodology can be found here: https://hockey-graphs.com/2017/04/04/identifying-player-types-with-clustering/

Interestingly, volume shooters come out ahead of puck-movers. Given all of the attention given to puck-moving defencemen (PMD) following the Nashville Predators’ run to the Stanley Cup Final, this may come as a surprise.

It is important to note that Stimson’s label does not correspond to how PMD is conventionally used by media scribes and fans. Stimson’s criteria for a puck-mover describes a player who skates well with the puck and drives transition play, but who rarely shoots. When Stimson says “puck-mover,” he’s not talking about players like Erik Karlsson, but more about players like Nathan Beaulieu.

As he did with the forwards, Stimson examined which archetype pairings generate the most expected offence in the form of expected-goals-for percentage (xG%).

Interestingly, although all-arounders are generally the elite of the NHL, the best pairing from an xG% perspective is an all-arounder with a volume shooter. In fact, a pairing of two volume shooters is the second-strongest combination, ahead of anything featuring a puck-mover.

This is good news for the Canadiens, who can no longer assemble an all arounder-volume shooter pairing with the departure of their only all-around defenceman to Ak Bars Kazan. At the moment, the Canadiens feature only players categorized as volume shooters or defensive oriented, with no NHL-level data available for Jakub Jerabek.

Detailed metrics for all players can be found here: https://public.tableau.com/profile/publish/Player_Types/Dashboard1#!/publish-confirm

Here, it must be noted that Mark Streit, despite his reputation, is classified as a defensive-oriented blue-liner by Stimson’s clustering. Looking at the stats, the Swiss veteran has not accumulated particularly admirable puck-moving or transition-play statistics over the last two seasons, but is decent at shot generation.

With only two player archetypes available, the best statistical set-up for the Habs is five volume-shooters and one defensive-oriented player, which would translate to:

But let’s be realistic: there is very little chance that the Habs will voluntarily ice a lineup not featuring both Karl Alzner and Jordie Benn. With a lineup of four volume shooters and two defensive-oriented players, the permutations become:

Regardless of configuration, the entire right side is locked in. Shea Weber reprises his role on the first pairing, Jeff Petry plays on the second pairing, and Jordie Benn slots in on the third.

The question thus becomes who will play with those three players.

Going by the model

Using Stimson’s model, it is optimal to slot a defensive-oriented player next to the likewise defensive-oriented Benn. In a competition between Streit and Alzner, the inevitable victor is the high-profile signing Marc Bergevin made on the opening day of free agency.

That leaves two volume shooters to occupy the positions across from Shea Weber and Jeff Petry. Based on available data, the remaining volume shooters, Schlemko, Davidson, and Redmond, line up as follows:

None of these players are first-pairing quality, with all three falling into the second-pairing-level tier (33rd to 66th percentile). It is Davidson who has the best offensive metrics of the group. However, the disparity in sample size between the former Oiler (420 minutes of even-strength ice time) and the former Shark (1089 minutes) makes for a tough definitive relative evaluation. Based on this evaluation, the two players are not far apart, indicating that perhaps Davidson could earn an early-season trial run as Weber’s partner, but the more experienced Schlemko likely has a leg up.

The configuration featuring a double defensive-oriented player pairing thus translates as follows:

Security in balance

The second configuration, featuring two defensive-oriented—volume shooter pairings, allows for Alzner as an option next to Weber. But the ex-Washington rearguard has not displayed the offensive abilities, and especially the passing prowess, necessary to properly augment a special weapon such as Weber’s shot.

As it stands, no matter what, Weber will have to compensate for the weaknesses of his defensive partner. That individual should at least be someone who has the ability to boost Weber’s game on some level, and Schlemko or Davidson, with their passing and transition prowess, are better suited for that endeavour.

Alzner slots next to Petry, and Davidson next to Benn. This creates two more balanced pairings that are less efficient in generating offence based on Stimson’s model, but may be more palatable to traditionalist coaching sensibilities.

The wild card

Bergevin landed a European defenceman on a low-risk/high-reward contract this off-season, but it wasn’t Mark Streit. Jakub Jerabek may be an NHL neophyte, but the native of Plzen is fresh off an outstanding KHL campaign and has donned the coat of arms of the Czech Republic at U18, U20, and senior levels.

It’s hard to place Jerabek’s KHL numbers in the context of Stimson’s model, but given that the Czech was fifth among defencemen in KHL scoring last season, he doesn’t fit the defensive-oriented archetype. However, despite this offensive output, Jerabek is not a prolific shooter, having amassed 79 SOG in 59 GP (67th among KHL defencemen). This, combined with the fact that only a single one of his five goals came at even-strength, indicates that Jerabek would be most likely classified as a puck-mover.

This gives the Habs some additional flexibility, but the problem is that volume shooters are generally superior to puck-movers in Stimson’s model. For Jerabek to be optimally utilized, he needs to force one of the defensive-oriented players — Alzner or Benn — out of the lineup. More realistically, Jerabek can also supersede those in the second-pairing tier, currently occupied by Schlemko and Davidson. Of course, if that were to occur, then Bergevin has signed a first-pairing NHL defenceman for $700,000.

Depth up front, questions at the back

In contrast to the known quantities present within the Canadiens’ forwards, the defence corps offers more questions than answers. Realistically, a legitimate argument can be made for any of the new acquisitions to play on any pairing.

The main issue facing the Canadiens is that while Schlemko, Davidson, Benn, Alzner, and Streit are all very serviceable NHLers, none of them rise above their peers, and none are, on the surface, capable of taking Markov’s vacant spot in the lineup. With very few known and established quantities on the blue line aside from Weber and Petry, come October, J.J. Daigneault might actually be the busiest man on the Habs bench.

(Again, a special stick-tap to Prashanth Iyer and Dom Luszczyszyn for the inspiration, and Ryan Stimson for all of the data and graphics.)