FanPost

NHL Advanced Stats: Applying the Primer to the Second Line Centre

Beastmode: On Hold. - Rich Lam

The other day, I made my first Fanpost here at Nucks Misconduct, laying out a primer for advanced stats that we can use this year to guide our analysis of the Canucks.

I thought it would be a good idea to examine those stats via an "application" shortly after so that the point and their usefulness can be driven home early. I should caution again that advanced stats, especially in hockey, can’t be taken for gospel and are not a replacement for watching the games and forming objective qualitative opinion. Instead, they can help verify what we see with our eyes and possibly point us in the direction of some things we may have missed.

With one of the big mysteries of this truncated training camp being the role of the second line centre, I thought analyzing potential options for the spot would be a good way to "apply" the primer from the other day. What follows are a few tables showing the advanced stats the primer laid out for the following players who may or may not be in the running.

Chris Higgins (ruled out – likely 3rd line winger)

Maxim Lapierre (penciled in as 3rd line centre)

Jordan Schroeder (rookie, no NHL data available to evaluate)

Manny Malhotra (likely 4th line centre)

Andrew Ebbett (competing with Schroeder)

Tyler Bozak (Leafs’ Luongo bait?)

Ryan Kesler (The irreplaceable)

Obviously, right away our analysis has a big wrench thrown into it since we can’t rely on these stats to evaluate Schroeder, who has 19 points in 30 games with the AHL’s Chicago Wolves so far this year. But we can evaluate the other candidates and see if the rumoured "Ebbett v. Schroeder" battle for the role is even the right one. I’ll show the tables, briefly re-explain the stats, and provide some commentary.

Even Strength Per-60 Statistics - 2LC Candidates
G/60 A/60 Pts/60 Pen/60 Draw/60
Jordan Schroeder N/A - No NHL Time
Andrew Ebbett 0.82 0.41 1.24 1.2 0.8
Maxim Lapierre 0.66 0.74 1.4 0.8 1.1
Manny Malhotra 0.55 0.87 1.42 0.6 0.7
Chris Higgins 0.91 1.1 2.01 0.5 0.8
Tyler Bozak 0.67 1 1.66 0.5 0.5
Ryan Kesler 0.68 0.79 1.46 1 0.6
Pen/60 & Draw/60 = penalties taken and drawn

Here we have just some basic stats, translated to per-60 minute rates. The point of this, for refresher, is just to standardize scoring and penalties across even playing time, so time on ice doesn’t have an impact.

Nothing really stands out too much here. Higgins looks good, except that he spent a lot of time on the wing and, as we’ll see later, played with pretty good teammates usually. We see that Lapierre is the best at drawing penalties, while Andrew Ebbett goes to the box a little too often. Were special teams to become a particular strength or weakness of the club, this would be helpful information.

Even Strength CORSI Statistics - 2LC Candidates
CORSI C-Rel C-QOC C-Rel-QOC C-Rel-QOT
Jordan Schroeder N/A - No NHL Time
Andrew Ebbett 7.83 2.6 -0.054 0.372 -5.571
Maxim Lapierre -10.79 -20.7 -0.985 -0.018 -6.205
Manny Malhotra -20.72 -32.6 -0.57 -4.587 1.289
Chris Higgins 8.5 2.2 -0.173 0.956 1.534
Tyler Bozak -4.99 -4.9 0.029 0.34 0.361
Ryan Kesler 13.24 11.2 -0.31 0.581 2.955
Corsi = Team shot diff w player on
C-Rel = Team shot diff w player on - Team shot diff w player off
C-QOC = Corsi of opponents
C-Rel-QOC = C-Rel of opponents
C-Rel-QOT = C-Rel of linemates

I included brief explanations in the table, but allow me to further elaborate: CORSI is basically shots (shots on goal, shots blocked and shots wide) minus shots against. It is designed to be a sort of plus-minus for shots, which have shown to be very strongly correlated with future goal differential as well as scoring chance differential. The higher the CORSI, the more opportunities the player is a part of compared to opportunities given up.

However, as we learned in the primer, there are some mitigating factors to make note of. Corsi-Relative takes into account the team context, so you’ll notice all of these Canucks take a hit in Corsi-Rel, since the Canucks has a whole were very good and thus performed well with or without these individuals. Corsi-QOC and Corsi-Rel-QOC are metrics to measure the quality of competition players faced. So players with a high Corsi-Rel-QOC played against a team’s best players, while a low Corsi-Rel-QOC played against an opponent’s lesser players. Finally, Corsi-Rel-QOT measures the quality of a player’s teammates, where a low grade indicates playing with one’s lesser teammates and a high mark indicates, you know, centering the Sedins.

From this table we learn a little bit more about the players in question. Primarily, we learn that Ryan Kesler is very good and will be tough to replace. We also see that, despite playing on a "checking line," Manny Malhotra didn’t play against particularly strong players or with particularly weak players. He was also awful in terms of generating chances (but we’ll see something more on that in the next table). We also see that Ebbett performed pretty well, with difficult competition and without the benefit of strong teammates (ding ding ding).

Even Strength Other Advanced Statistics - 2LC Candidates
Z-Start% Z-Fin% PDO GVT +/-per60
Jordan Schroeder N/A - No NHL Time
Andrew Ebbett 37.9 49.5 994 1.3 0
Maxim Lapierre 22.2 45.8 1009 2.8 -0.22
Manny Malhotra 13.2 40.6 1003 1.6 -0.94
Chris Higgins 46.6 49.8 1019 8.4 0.71
Tyler Bozak 52.5 54 996 6.3 -0.5
Ryan Kesler 48 49.1 1008 8.4 0.51
Z-Start% = % of shifts start in offensive zone
Z-Fin% = % of shifts ended in offensive zone
PDO = >1000="lucky" with Sv% and S%, <1000="unlucky"
GVT = Worth __ goals over a fringe NHL replacement

Our final table shows zone start and finish rates, a metric I don’t particularly like called PDO, and Hockey Prospectus’ Goals Versus Threshold. The biggest thing that should jump out at you is how strong Lapierre and Malhotra were at moving the play from their own end to the offensive end, which is a very valuable skill. We should also note that Malhotra’s poor CORSI scores likely come, at least in part, due to the fact that he’s almost exclusively starting his shifts in his own end. Ebbett also wasn’t bad here. Finally, we see that Bozak has a strong GVT, but didn’t jump out in any of the other stats we examined which is a) likely due to the formula being based on points and b) a signal that, even though it’s a nice tidy stat to describe what HAS happened, it might not be our best bet when trying to gauge future performance.

ANALYSIS

Based on these tables and stats, you’d probably lean towards Andrew Ebbett as the best option, given the choices. Considering Ebbett has just 58 points in 163 career NHL games and has 21 points in 29 AHL games this year, he might not stand out as an option based on "normal" stats. It’s important to note, however, that Ebbett played just 18 games in the NHL last year so his stats come from an especially small sample size and may not be terribly reliable. He might be worth a shot as a temporary fill in, though, given his success in his audition last year.

You might also suggest that Max Lapierre is worthy of an audition – while his CORSI scores were poor, he played against neutral opposition with poor teammates and did a great job moving the play out of his own end. While Malhotra did some of those things, too, he did so with better linemates against lesser competition.

Finally, while Chris Higgins stacks up well in some categories, it’s pretty difficult to judge how a winger would move to the pivot based on anything other than seeing it happen.

APPLICATION

All of this is to say that it’s not an exact science. As I mentioned ad nausem in the primer, these are meant as a supplement to watching and making your own judgments. Based on this data, it looks like Ebbett is a solid choice. However, hedidn't play a lot last year, we don’t know how Higgins would transition and we can’t evaluate Schroeder. I still thought it was a timely example to use as an illustration.

The best use may be in a situation, say, 10 games into the year. While we’d still be working with small samples, we could at least compare the second, third and fourth line centres based on their performance. We’ll also have some observation to rely on at that point and can "test" to see if the stats back up what we’re seeing.

I hope all of this makes sense and was explained in a clear way. As I finished, I kind of regret not cherry picking an example that showed a more "clear winner," but such is life. If you have any further questions or want to discuss any of these elements more, leave a comment or show me them tweets.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join Nucks Misconduct

You must be a member of Nucks Misconduct to participate.

We have our own Community Guidelines at Nucks Misconduct. You should read them.

Join Nucks Misconduct

You must be a member of Nucks Misconduct to participate.

We have our own Community Guidelines at Nucks Misconduct. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9355_tracker