Oilers: Is the Darnell Nurse criticism fair after his long contract extension?
On Friday, August 6th, the Edmonton Oilers and Darnell Nurse came to terms on a massive 8 year, $74 million contract extension ($9.25mil AAV) that’ll see Nurse locked up until the 2030 offseason.
This is quite a massive increase in pay from his recent 2-year bridge deal that saw him paid just 5.6 million per year and is another in a parade of massive contracts that have been given out to top-pairing defensemen around the league in recent weeks (Seth Jones, Zach Werenski, Dougie Hamilton, Miro Heiskanen, etc).
As with anything Oilers-related, Twitter and the blogosphere have been lit up with a range of opinions in the days since with the general skew of said opinions being that it constitutes a substantial overpayment.
Personally, this isn’t an opinion that I disagree with. $9.25mil seems like an absurd chunk of money to commit to a player that most forms of analysis don’t deem a ‘superstar’. Of course, the market had been set by those earlier signings ranging from 8.45 to 9.58mil, but a contract can be both ‘market value” and ‘overpayment’.
However, I’m not here to discuss whether or not the contract is an overpayment or not. I’m aiming to address some of the more extreme criticisms/assertions made against Darnell as a player. Over the past few days, I’ve seen Nurse tarred and feathered with the broad brush of “bad analytics” and “awful defender”. He often has all his success attributed to his time playing with McDavid and, while there’s some validity to this line of argumentation, these kinds of broad strokes overlook a number of pieces of valuable contextual information. I’ll address each common criticism separately.
“Nurse’s success is due to his time with McDavid”
Perhaps the most common criticism levied against just about any Oilers that has success on the Oilers and it’s one that’s frequently used in the case against Nurse. How much truth is there to it? Well, let’s start with this year:
McDavid w/o Nurse: 54.6% SCF*, 51.9% GF, 9.6% Sh%, 88.2% Sv%, 58.5% Off. FO
Nurse w/o McDavid: 44.8% SCF, 50.0% GF, 8.8% Sh%, 93.3% Sv%, 40.1% Off. FO
Together: 56.0% SCF, 58.8% GF, 11.7% Sh%, 90.1% Sv%, 64.1% Off. FO
* I use scoring chances (SCF) as I believe it’s the best proxy for identifying meaningful possession; Corsi doesn’t distinguish shot quality at all and high-danger chances are more difficult to accurately/consistently quantify (less reliable public numbers)
As you can see, there’s (obviously) a drop-off in Nurse’s possession metrics when he’s away from Connor. This is far from a surprise given how bad the bottom-6 has been and that Draisaitl isn’t a particularly strong player by possession metrics.
The most glaring difference is his deployment when apart from Connor where Nurse sees only 40.1% of his on-ice faceoffs in the offensive zone compared to much more offensive role (64.1%) he plays when with him. This can likely serve to explain some more of the discrepancy between his ‘w/ Connor’ and ‘w/o Connor’ numbers.
Despite this, Nurse managed to keep his goal share even when away from McDavid. This is likely a result of luck on his part but it’s possible that he’s one of those players that helps his team capitalize on their chances at a higher-than-average rate.
Most important to note here is how the two produce much better results when together than either do when apart. McDavid saw his chance share, goal share, on-ice shooting %, and on-ice save % improve when with Nurse than without him. This is the part of the story that’s often omitted in the lower-resolution criticisms of Nurse. He and his pairing(s) had a tangible positive impact on McDavid’s success this past year, as well.
But what about the years prior? What if this was all a result of Nurse’s hot streak this year? Here’s the same set of numbers but with the sample extended to include the past 3 years:
McDavid w/o Nurse: 51.4% SCF, 51.6% GF, 9.9% Sh%, 90.1% Sv%, 57.1% Off. FO
Nurse w/o McDavid: 44.5% SCF, 46.6% GF, 7.2% Sh%, 91.9% Sv%, 41.2% Off. FO
Together: 51.5% SCF, 53.8% GF, 11.8% Sh%, 90.3% Sv%, 60.4% Off. FO
As you can see, this dataset points to many of the same trends. They play a heavily offensive role when together and Nurse and his pairings seem to contribute to a tangible improvement in McDavid’s on-ice results. There’s very clearly a reason why Nurse gets a greater proportion of time with Connor than other Oilers defensemen.
While there’s certainly room for improvement in his ‘w/o McDavid’ metrics, I don’t think it’s all that fair to weigh them too heavily given the sub-replacement level play the Oilers have gotten out of their bottom-6 throughout this sample. With the improvements made to the Oilers’ forward group this offseason, that’ll be an area that I’ll be keeping a keen eye on throughout next season.
(Stats per Natural Stat Trick)
“Nurse is bad defensively”
There’s likely some truth to this position. Nurse has had his share of struggles with over-committing to players already covered by his D partner or chasing players behind the net when a more clear threat is parked in the slot. While this has improved in increments over the past few years, I’m not sure I’m prepared to say this tendency has been fixed. However, I believe this evaluation often overlooks a pretty key component: his playing time with McDavid.
Maybe you’re thinking “but McDavid is the best player in the world; how can playing with him HURT someone”? While he’s absolutely the best offensive player in the world, the same cannot be said of his defensive results.
This isn’t necessarily to say that I don’t think he’s capable of this or that his individual defensive skills (stick checking, positioning, reads, etc) are particularly lacking. It is to say that his rush-dominant style and the fact that his team has generally been hard-pressed to score ANY goals without him on the ice, he’s incentivized- and perhaps even pushed- to play a ‘balls-to-the-walls’ style that often neglects some of the finer details- namely those on the defensive side of the puck.
Don’t believe me? Well, let’s take a look at every D pairing’s goals against rates with McDavid compared to without him:
Larsson w/: 3.22 GA/60 & 2.67 xGA/60
Larsson w/o: 2.38 GA/60 & 2.01 xGA/60
Bear w/: 3.38 GA/60 & 2.43 xGA/60
Bear w/o: 2.71 GA/60 & 2.24 xGA/60
Klefbom w/: 3.11 GA/60 & 2.57 xGA/60
Klefbom w/o: 2.42 GA/60 & 2.27 xGA/60
Russell w/: 2.31 GA/60 & 2.77 xGA/60
Russell w/o: 1.88 GA/60 & 2.02 xGA/60
Benning w/: 3.12 GA/60 & 2.45 xGA/60
Benning w/o: 1.77 GA/60 & 2.1 xGA/60*
* Important to note that 3rd pairing defensemen deployed in the way Benning was have a much easier job and their results should be taken with that grain of salt in mind
Every defensemen seems to see their GA/60 and xGA/60 rates spike when with McDavid. What about Nurse?
Nurse w/: 3.41 GA/60 & 2.79 xGA/60
Nurse w/o: 2.55 GA/60 & 2.32 xGA/60
And so the trend continues. Now take this idea that defensive results dip when with one particular player and map it onto a situation where one defenseman plays a disproportionately greater amount of their time with said player. What can we infer from that? Said defenseman will likely see his overall defensive results take a larger hit as a result.
I believe this explains a fairly big portion of the criticism levied against Nurse’s defensive game. This isn’t to say I think Nurse is an above-average defender, but I don’t believe he’s among the league’s worst as some might have you believe.
NOTE: I feel it’s important to mention that I’m not disparaging McDavid here. I believe his poor defensive results are equally a consequence of the team’s reliance on him to score. As the rest of the forward group improves, I expect to see McDavid focus more on “outscoring” rather than just “scoring”.
(Stats per Natural Stat Trick)
“Nurse isn’t a TRUE top-pairing defenseman”
This is the most absurd of the takes that I see regularly and deserves special attention to refute. Over the past 3 seasons, Nurse ranks 4th in even-strength TOI/game (19:32) behind only Thomas Chabot, Drew Doughty, and Seth Jones and ahead of Kris Letang, Brent Burns, Colton Parayko, and a host of other big names.
As with some of his contemporaries in this ranking, these minutes are at least partly the result of being the best of a rather weak defensive corps, but it absolutely signifies that he has been playing top pairing minutes. Whether he has succeeded in said minutes should dictate whether we say he IS a top-pairing defensemen. So, how has he done?
In his 4085 minutes since 2018-19, Nurse has been a positive influence on his team’s goal share (+4.91 relGF%) and scoring chance share (+2.30 relSCF%) while seeing 50.7% of his faceoffs in the offensive zone. These results rank him 34th and 37th in the league among defensemen that played at least 1000mins in that span. If you shrink the sample size to the past 2 years, Nurse’s positive impacts increases to +6.21 relGF% and +5.01 relSCF% (ranking 32nd and 12th in the league among defensemen w/ over 750mins).
Alongside the above numbers, Nurse has demonstrated his ability to impact the game offensively at a high level with 49 5-on-5 points (3rd in the league), 16 5-on-5 goals (t-3rd in the league), and +0.92 relGF/60 (2nd in the league behind only Makar). Then consider that Nurse has done this with a new primary partner every year (Russell, Bear, and Barrie) and it becomes even more impressive. If these results don’t paint the picture of a true top-pairing defenseman to you, I’m afraid we’ll just have to agree to disagree.
Now, it’s entirely fair to suggest that these offensive results are influenced in part by his playing time with McDavid. This is almost undoubtedly true as players benefit from playing with other great players. However, I’ll only concede this point if people start applying this principle evenly across other defensemen in the league.
Many seem to operate under the assumption that, because McDavid is such a special player, all evaluations of Oilers players should exclude their minutes with McDavid. But what about forward lines that outperform McDavid? McDavid may be the best individual player that the league has seen in a long time but plenty of duos or trios of elite players have better results on the ice than he does.
Lines like Marchand-Bergeron-Pastrnak and Tatar-Danault-Gallagher absolutely dominate their opposition when on the ice but I rarely (if ever) see these mentioned when discussing those teams’ top defensemen. Even when they are, it’s still a rather unfair comparison as there are few teams with as vast a difference between their top-6 and bottom-6 as the Oilers have.
(Stats per Natural Stat Trick)
Conclusion
In summary, despite my belief that it’s an overpayment, the contract is in line with what the market dictates an offensive defenseman like Nurse commands. We all have our right to criticize the contract and even the player (to a reasonable degree), but low-resolution takes like the ones I laid out here should not be accepted.
Other benefits to consider with the contract include a.) The Oilers finally have a durable, top-pairing defenseman that wants to stay here, b.) They locked up one of McDavid’s closest friends for long-term, and c.) Nurse has improved in pretty much every single year since arriving in the NHL. Betting against the contract is fine, but Nurse isn’t the kind of player I’d bet against!
There are going to be people that disagree with my take and other models that value different stats than I do, but that’s fine. That’s part of the beauty of stats and part of what motivated me to write this; there are many ways to look at them and these can tell many different stories. It’s tough to say which one is correct but I think any analysis that doesn’t reflect on some of the context-specific to the Oilers is lacking.
That’s enough from me. What do you think of Nurse as a player? And how about the contract?
(Stats per Natural Stat Trick)