Objectively projecting Darnell Nurse

Updated: January 11, 2018 at 2:06 am by Jonathan Willis


Darnell Nurse is a fascinating player, and his long-term
career arc is of extreme importance to the Edmonton Oilers.

The most common method of projecting that sort of career arc
is the old-fashioned one: Assessing the player’s physical skills and then imagining
what he’ll look like in his prime. The one we’re going to use today is a newer method, which involves looking at what happened to players with
comparable underlying statistics.

Neither method is definitive, as last I checked it’s still
impossible for humans to foretell the future. Both, however,  are helpful in this sort of conversation.

The first thing I had to do was assemble a list of similar
players. Hockey-Reference is indispensable for this sort of thing, and I used
its player index function to put together a
list of NHL defencemen
since the dawn of the analytics era (2007-08) at roughly
the same stage of their career as Nurse and with similar points/game totals.
Then I narrowed down that list using the following criteria:

  • I selected only players in their draft+3 season
  • Each player had to have played in at least 60 percent of his
    team’s games and have averaged at least 18 minutes per night
  • Each player needed to have a negative Corsi rating relative
    to his team

I was left with a dozen players, including Nurse himself.

The Master List

Player Team Size Season GP PTS/GP ATOI Raw Corsi Rel. Corsi
Marc Staal NYR 6’4″, 207 2007-08 80 0.13 18.8 56.1 -0.2
Cody Ceci OTT 6’3″, 205 2014-15 81 0.26 19.3 49.4 -1.3
Luke Schenn TOR 6’2″, 229 2010-11 82 0.27 22.4 48.1 -1.6
Cam Fowler ANA 6’1″, 207 2012-13 37 0.30 20.4 47.5 -1.7
Brett Pesce CAR 6’3″, 200 2015-16 69 0.23 18.8 50.0 -2.1
Jared Cowen OTT 6’5″, 238 2011-12 82 0.21 18.9 51.5 -3.1
Marc-Edouard Vlasic SJS 6’1″, 205 2007-08 82 0.17 21.6 52.0 -4.6
Jonas Brodin MIN 6’1″, 195 2013-14 79 0.24 23.9 46.2 -4.9
Darnell Nurse EDM 6’4″, 213 2015-16 69 0.14 20.2 45.5 -5.8
Erik Gudbranson FLA 6’5″, 216 2012-13 32 0.13 18.8 46.9 -6.4
Zach Bogosian ATL 6’3″, 219 2010-11 71 0.24 22.4 47.2 -6.7
Jack Johnson LAK 6’1″, 230 2007-08 74 0.15 21.7 40.9 -10.4

That’s an interesting group of players, and I would divide
them into some broad categories:

  • Home runs: Marc-Edouard Vlasic
  • Mileage may vary: Marc Staal, Cam Fowler, Erik
    Gudbranson, Zach Bogosian, Jack Johnson
  • Disappointing players: Luke Schenn, Jared Cowen
  • Too soon to tell: Cody Ceci, Brett Pesce, Jonas Brodin,
    Darnell Nurse

Vlasic should be uncontroversial; he’s a brilliant shutdown
defenceman, one of the finest in the game of hockey. Schenn has evolved into a
serviceable player but nothing like what he was originally projected to be,
while Cowen is currently locked in a buyout dispute with Toronto.

The five players in the middle will gather differing reviews
depending on the source; all I can offer is my own take.

Staal and Gudbranson both play tough minutes but probably
aren’t suited to top-pair work; the former in particular hasn’t been the same
player since losing Anton Stralman as a partner. Fowler’s the rare puck-moving
defenceman who doesn’t drive shot metrics and also probably belongs on the
second pairing. Bogosian has never gone a full season without getting out-shot,
nor has he ever outperformed his team’s average over a full year and he’s
played for some pretty bad teams. I’ve written
about Jack Johnson previously
and not much has changed since then.

A Closer Look

Player Season QualComp ZS Diff. PTS/60 Pen. Diff. Raw Corsi Rel. Corsi
Jack Johnson 2007-08 0.085 85 0.51 -0.6 40.9 -10.4
Marc-Edouard Vlasic 2007-08 0.034 8 0.31 0.0 52.0 -4.6
Zach Bogosian 2010-11 0.024 28 0.53 -0.3 47.2 -6.7
Marc Staal 2007-08 0.011 -92 0.43 -0.3 56.1 -0.2
Erik Gudbranson 2012-13 0.006 -31 0.35 -0.3 46.9 -6.4
Luke Schenn 2010-11 -0.011 12 0.85 -0.1 48.1 -1.6
Cam Fowler 2012-13 -0.016 -11 0.48 0.2 47.5 -1.7
Jared Cowen 2011-12 -0.025 -85 0.64 -0.3 51.5 -3.1
Darnell Nurse 2015-16 -0.028 51 0.31 -0.5 45.5 -5.8

Numbers above courtesy of Behind
the Net

  • Quality of competition is an interesting metric here.
    Johnson faced brutally tough opposition, while Vlasic and Bogosian also played
    tough minutes early on. It’s interesting to note that our disappointments –
    Schenn and Cowen – are down at the lower end of the scale, as is Nurse.
  • Zone start differential is the difference between total
    number of offensive zone and defensive zone starts (a positive number means a
    tougher assignment). Johnson, Nurse and Bogosian are at the tough end of things
    here, while we see some explanation for Staal’s relatively strong Corsi number.
  • I’m honestly not sure how useful 5-on-5 points/hour over a single
    season is for evaluating defencemen. Schenn and Cowen fared best, while Vlasic
    was at the bottom of the list here. That disparity is good news for Nurse, who
    along with Vlasic sits at the bottom of the chart.
  • Penalty differential is the gap between penalties taken and
    penalties drawn per hour. Again, there’s little clear order here, save that the
    guys who struggled on the Corsi side of things tended to struggle here as well.

Of the players on this list, the ones I see as having the
greatest similarity to Nurse statistically are Bogosian and Gudbranson, but it’s
impossible to rule out any of the outcomes on this list. At the low end, that
means Cowen and Schenn; at the high end it means Vlasic. It’s tempting to look
at his physical tools and project at the high end of the spectrum, but that
sort of reasoning would have applied equally well to Johnson and been faulty.

What we do know is that for the most part, defencemen who
played a lot at the same age as Nurse while a) not scoring much and b) getting
out-shot don’t end up having long careers as top-pairing options. Vlasic is the
exception, and did it in tough minutes on an excellent team, which naturally
tends to depress things like relative Corsi.

My guess would be that Nurse eventually settles in as a solid
second-pair defenceman, but at this stage there’s a wide variety of possible outcomes.
We’ll know a lot more a year from now.

For those interested,
I used different criteria but got similar results when I performed a
comparable experiment
for the
Edmonton Journal back in March