In what has long been considered a “lost season” for the Buffalo Sabres, contributions from their younger assets have been a breath of fresh air. None have stood out more than rookie centerman, Dylan Cozens.

Despite posting modest base statistical scoring numbers (and missing time due to injury), the 20-year-old has been one of the best forwards on the team. This has especially been the case since Don Granato was promoted to interim head coach (which we’ll expand on later).

Analytics enthusiasts have frequently commented about Cozens’ underlying metrics in his first professional season. There is a lot to like about his game, both using the “eyeball test” and in perusing the data. In knowing this, I decided to see how his advanced metrics stacked up against past rookie pivots in the NHL. One player whose first-year production compares very closely to Cozens’ is Sabres’ captain, Jack Eichel.

Before we go any further, let’s make one thing particularly clear – this article is not designed to make any claims that Dylan Cozens will be the next Eichel. It’s not designed to say that he’s on a comparable trajectory towards stardom. It’s merely a contextual observation of the statistical similarities between Eichel’s 2015-16 stats, and what we’ve seen so far this year from Cozens.

With that disclaimer out of the way, let’s dive into the numbers.

Establishing Situational Context

In a lot of statistical comparison articles I’ve written, I tend to provide stats followed by context. I decided to do it the other way this time because the environments in which Eichel and Cozens produced their respective rookie impacts are particularly interesting (and frankly, important).

In this section, we’ll compare situational factors like quality of teammate, quality of competition, and deployment settings. We’ll start with Cozens who’s RAPM metrics paint a picture of a player who has done more with less-than-inspiring linemates, and he’s done it against stiff competition.

According to Evolving Hockey’s RAPM data, of all the NHL players currently under the age of 21 (of which there are 32), Cozens has the fourth-lowest QoT (quality of teammate) rating. His teammates’ per-60 RAPM xG rate currently stands at -.042 (which translates to one standard deviation below average).

This may come as a surprise. After all, Cozens is currently playing alongside Arttu Ruotsalainen and Anders Bjork (a trio that has experienced some success as of late). It’s important to remember that, before the trade deadline, Cozens has spent time with a wide array of linemates, particularly since he started the year on the wing and then transitioned back to his natural center position.

Equally noteworthy is the quality of competition Cozens has faced this season. Under the same criteria, Cozens has faced the third-most difficult slate of competition at a per-60 RAPM xG rate of .032 (1.05 standard deviations above average). Compounding on that number is the fact that Cozens has been deployed in a defensively-tilted OZS rate at 48.15% on the year.

Comparatively, Eichel’s QoT and QoC landscape wasn’t quite as treacherous. Applying the same under-21 criteria to the leaguewide metrics in 2015-16 (Eichel’s rookie campaign), the then 18-year-old fell pretty close to the middle of the pack in both categories.

Of the 59 under-21 players who dressed in 2015-16, Eichel’s linemates’ RAPM xG numbers (per-60 rate of .008) ranked 28^th, and his competition quality registered similarly at 25^th overall (per-60 rate of .021). On top of that, Eichel had a slightly more favorable deployment ratio, holding an OZS rate of 53.42%.

Another key difference here that must be acknowledged is their respective ages. Eichel was a year younger in his rookie NHL season. Though he did spend a season in the NCAA before his draft year, that age difference is significant. That said, it’s pretty clear that comparatively, Cozens had a slightly more difficult environment to work in from an analytics standpoint.

On-Ice Impacts

Now that their respective background contexts have been established, let’s determine how Eichel and Cozens’ rookie on-ice impacts stack up to one another. From a flyover view of the above RAPM side-by-side graph, the similarities are immediately evident.

In terms of standard five-on-five on-ice shot quality differential, Cozens current overall xGF percentage of 44.21 compares closely to Eichel’s rookie season mark of 44.75%. The same can be said for their first-year Corsi rankings (47.59% and 45.86%, respectively)

So, we’ve established some pretty direct similarities – both contextual, and statistical. If these two players are so similar, why aren’t Cozens’ base scoring numbers as impressive as Eichel’s were? In 81 games as a rookie in 2015-16, Eichel posted 56 points. Cozens’ current output of 11 points in 31 games (10 registered at even-strength) would only place him on a 29-point trajectory over 82 games.

There are a couple of explanations. For one, Eichel spent a lot more time on the power-play as a rookie than Cozens has. He actually served on the Sabres’ top man-advantage unit in 2015-16, registering 246 minutes (average of three minutes per game) during the year. Conversely, Cozens has only spent a total of 36 minutes on the power-play this season on the team’s secondary unit (average of just over a minute of PP time per game).

But what about their scoring rates at even strength? If we remove Eichel’s power-play production from the equation, he paced out at 35 points over 82 games. Obviously, that narrows the gap between them by quite a bit.

We could even go a step further and examine their production rates (both base and expected) on a per-60 basis. From that standpoint, Eichel and Cozens held nearly identical xGF/60 marks of 2.19 and 2.18, respectively. In terms of actual on-ice GF/60 rates, Cozens’ rookie mark of 2.93 thus far is substantially higher than Eichel’s inaugural on-ice rate of 1.99.

Other Comparable Rookies

Eichel isn’t the only player in recent memory whose rookie impacts draw close parallels to Cozens’. In researching the data for this piece, I stumbled upon several others. Strictly in terms of underlying statistical production, Clayton Keller’s 2017-18 campaign is probably the closest comparable I found.

In Keller’s first NHL season, he experienced very similar QoT and QoC scenarios to Cozens in terms of environmental difficulty. Also, Keller registered strong xGF marks (2.26 per-60) while posting negative defensive impacts. Obviously, Keller and Cozens aren’t really comparable from a physical standpoint, but both use their above-average speed to help create offense.

In terms of style tendency, Kirby Dach is probably a good comparable for Cozens. Both of them are larger-framed players who do their best work orchestrating an offensive counter-rush. Like Cozens, Dach’s base stats as a rookie weren’t anything to write home about (23 points in 64 games last season), but his underlying numbers were relatively similar.

An interesting trend here is that there are several high-end centers in the NHL whose RAPM metrics were poor defensively and better offensively in their first NHL seasons. Even Nathan MacKinnon’s initial relative data marks followed this trend.

Telling you that players in their late-teens and early-twenties tend to improve with time isn’t exactly groundbreaking. I’ll also reiterate that I am in no way saying that Cozens will be the next MacKinnon, Eichel, or Keller. That said, high-end offensive talents show early and often give up a lot more on-ice quality than they produce. Cozens is no exception to that trend.

Granato’s Impact, Closing Thoughts

All of the aforementioned advanced metrics take into account Cozens’ 2020-21 campaign as a whole. That said, I’d be remiss if I didn’t note the underlying data improvements he’s experienced since Granato took over for Krueger (primarily on offense). A lot of that improvement has to do with philosophical changes implemented behind the bench.

Under Krueger, Cozens was posting negative RAPM impacts at both ends of the ice. As a young offensive centerman, he was being utilized in a defensive deployment ratio. That could work if the head coach allows said centerman to use his speed and transition ability to create chances off the counter-rush, but that simply wasn’t the case in Krueger’s system.

Now that the Sabres are running a more “high-event” style, Cozens can utilize his transitional tools and it has resulted in a steady increase in adjusted xG production. The clip below from the Sabres’ April 20th matchup against the Boston Bruins is a good example of what I’m referring to.

Dylan Cozens, fastest man alive pic.twitter.com/VuXnmKCaId

— The Charging Buffalo (@TheChargingBUF) April 21, 2021

The way he was able to leverage his superior skating ability to control end-to-end possession and single-handedly flip the ice is something we saw far less frequently under Krueger. This all probably ties into the general confusion Granato alluded to when asked about what the Sabres have done differently under his leadership.

Cozens’ Corsi-for marks have also experienced a substantial improvement as of late. This again is likely a direct result of the Sabres employing a more aggressive forecheck, that has helped them maintain offensive zone-time. The Bjork-Cozens-Ruotsalainen trio has been particularly proficient in this regard.

Whether or not Cozens turns into one of the top-end talents we’ve compared his rookie numbers to remains to be seen. Still, the similarities he’s drawing to some of the top young talents in the NHL are inspiring. It isn’t too bold to say that if he had received more power-play minutes (and didn’t play on the worst team in the NHL), we might have been talking about a Calder Trophy candidate.

Either way, it looks like the Sabres finally nailed a high-end first-round draft selection, something they’ve failed to do consistently for well over a decade.

Photo Credit: Mike Stobe/NHLI via Getty Images