Can you play Connor McDavid too much?

A conversation on Twitter yesterday centered around Connor McDavid’s 27+ minutes of ice time in their 4-3 comeback win over the Washington Capitals piqued my interest. The idea being that McDavid and Leon Draisaitl both are more than capable of handling these outrageous minutes (23:09 for McDavid, and 23:59 for Draisaitl on average) without issue, and that their usage last year (22:50 for McDavid, and 22:35 for Draisaitl on average) can be pointed to as evidence.

But is that true?

I suppose the answer is: It’s complicated, for reasons we’ll get into in a moment.

The Data

Immediately, this conversation brought me back to discussions I was having throughout last year about McDavid’s normally gaudy 5v5 rates (think CF%, FF%, SCF%, HDCF%, GF%, xGF%) being noticeably worse than his previous three seasons — including his rookie year. So, I took a look, choosing (somewhat arbitrarily) to split McDavid’s career into everything prior to the start of last season, and everything since. I say somewhat arbitrarily because it wasn’t scientific by any means, but it wasn’t completely random either — 2017/18 was the last season McDavid averaged under 22 minutes per game. Here’s what I got:

Right away the cause for concern is quite obvious. Despite career high counting numbers, every single one of those metrics has fallen from otherworldly to pedestrian over the last 89 GP. Further, McDavid’s P/60 (points per 60 minutes) both last season and to start this season are below his rate in both 2016/17 and 2017/18. While ~2.8 P/60 is still a gaudy number, it may be a red flag, perhaps, as his rate in 2017/18 was an absurd 3.2 P/60. But is that enough to definitively say that his usage, specifically, is the problem?

Probably not. Here’s why:

Conflating Factors

While we know Connor McDavid’s role and deployment, and that he spent a ton of time with Leon Draisaitl both before and after my ‘cut off’ point, we also know that their right winger went from Patrick Maroon — who is no true first liner by any means — to Alex Chiasson and Zack Kassian, two players whose bodies of work prior to Edmonton (and, in Kassian’s case, in Edmonton) were made up almost exclusively of bottom-half work. We can’t tell from this data alone whether or not McDavid (and Draisaitl) suffers from usage-related fatigue or whether Kassian and/or Chiasson are mostly to blame.

Dom Luszczyszyn from The Athletic suggests that Kassian and Chiasson are mostly to blame. And it makes sense. He writes,

“I don’t believe McDavid is the league’s worst defensive forward, but that’s where he lands based on current measurements (and the fact he still ends up as the league’s most valuable player says a lot about his transcendent offensive ability). It likely stems from overexertion playing so many minutes and low morale from yet another lost season in a finite career. That’s evident in McDavid’s season splits where he put up 58 points in each half of the season, only with drastically different on-ice results. Before Jan. 1, the Oilers had a 51 percent expected goal rate with McDavid on the ice and casually outscored opponents 35-26 at five-on-five. The usual. After that though, the team was outscored 50-41, earning only 45 percent of the expected goals in the process, which is very unlike McDavid.

Those ugly moments coincided with two experiments at right wing: Alex Chiasson and Zack Kassian, two replaceable players who are severe drags when it comes to driving play offering little tangible value at either end of the ice. (OK, fine, Chiasson is OK on the power play). In 531 minutes with Leon Draisaitl on the left and one of those two on the right, McDavid’s expected goals percentage was 49 percent. With literally anyone else on the right side — which isn’t saying much when the team has the league’s worst right winger depth — that got bumped up to 51 percent, all from defensive improvement.

Here’s the real kicker, though. Those numbers are fine albeit unideal considering it’s McDavid we’re talking about, but once you separate McDavid and Draisaitl to maybe find balance things go off the deep end. With McDavid and one of those two on the right, and no help on the left, he was a man on a sinking ship. The team was outscored 12-5 in those 181 minutes and sported an expected goals percentage of 34 in the process. Making McDavid suffer through those kinds of minutes should revoke the “Hockey Man” card for life for anyone responsible.”

While Luszczyszn does touch on the ‘overexertion’ theme I was looking for when I first started on this piece, he makes a strong case that McDavid’s right wingers were the real problem.

Another factor we have to consider is the injuries on defense last year, specifically to Oscar Klefbom, which meant the Oilers were without a single plus puck-mover until Andrej Sekera returned near the end of an already lost season.

It’s also possible that the coaching changes both last season, and over the summer, have played a part, too. I’m not tactically smart enough to be able to point to that definitively either, but it’s something worth considering.

So, Now What?

Well, we’ve identified a concern at least: Connor McDavid’s per minute results have sagged some since the start of last season. The tricky part now is how to solve it. Given that there are likely multiple factors influencing these results, the solution, too, must be multi-faceted.

Obviously, the Oilers have not addressed their depth scoring issues like they’d hoped. The bottom-six group this year is as bereft of offense as the bottom-six group last year, which was historically bad. As such, Connor McDavid must play a ton for the Oilers to have a chance to win. One obvious remedy to this would be to bolster the forward ranks, either via trade or from within the organization, with a handful of Bakersfield Condors enjoying early season success. For me, a combination of trading (Nurse, anyone?) for a proven top-six forward AND the inclusion of one or two forwards from Bakersfield (Yamamoto, Benson, or Marody perhaps?) would be the path forward here, for reasons I’ve explained ad nauseam. Fielding a deeper team might allow for McDavid to be more efficient with his minutes. Dressing a more talented 1RW might provide a platform for his line to be as dominant, in terms of 5v5 rates, as they were two years ago.

Sidney Crosby has never averaged 22 minutes or more per game in his entire career. Granted, he came close (21:57) on two occasions, but still. His contemporary, Evgeni Malkin, has seen one season over 22 minutes per game, during his age-22 campaign. He hasn’t averaged more than 20 minutes per contest in any of the last five seasons. Nikita Kucherov has never seen 20 minutes per game as a seasonal average. In fact, McDavid played 144 more minutes than Kucherov last season — or what amounts to roughly six-and-a-half (!!!) more games — in four less actual games! I think it would behoove the Oilers to try and pare down McDavid’s minutes to keep him fresh and, especially in a ‘lost’ season (in terms of their realistic chances to win the Stanley Cup), to avoid putting unnecessary miles on him when it won’t measurably affect their chances to win it all. Even with this hot start, the team is nowhere close to deep enough to really compete come Spring. I’d rather lose five more games in a ‘punt’ year than see McDavid plateau or peak sooner than he’s supposed to when the team around him is finally set up to challenge for a title.

Ultimately, I don’t think we can say for sure whether or not you can play Connor McDavid too much. But what I think we can say, is the Oilers are getting closer to finding out the answer to that question than a lot of us are comfortable with. Especially so soon after a serious ligament injury.

It will be interesting to watch these indicators as the season progresses and see whether the powers that be — Dave Tippett and Ken Holland — are able to turn this trend around.