The Star Trajectories of Kent Johnson and Kirill Marchenko

Evaluating Franchise Wings' Early Career Trajectories via EvolvingHockey Metrics

The Beginning

This little project started and spiraled out of control like many of my ideas do. I was curious about comparing players, their trajectories and their values, and wanted to put some information together so that I could start fulfilling that curiosity.

Initially, the question was understanding Kent Johnson on his trajectory many years ago. There was a problem, is he going to be good? is he core?, and we can use data to move toward a solution. Now we have more data, more of an idea of where Kent Johnson is going and how he compares, but also potentially more questions at the same time.

This season, though, that question changed and morphed into something else which is also why I’m writing about it here. Jason Robertson has been allegedly explored as a trade option for the Dallas Stars (which, as you might see in the data is a bit mind boggling). I wanted to look into what sort of assets made sense and realized that I was nesting an entire breakdown and line of inquiry into a sub-heading of another article. It’s a problem for me. The whole Armchair GM series was nearly all written as a single document until I realized that each part needed more exploring and perhaps it’s own space to breathe.

Here we are now, which is a data driven look at the careers and progression of a select group of players that I have identified as “franchise wings”. I think we all know the type, the Patrick Kanes and Nikita Kucherovs, second to the goal of a franchise center but, in some cases, equally if not more important to actual winning.

There are plenty of other players included in the cohort to provide the context of what a player who might never had the winning to become a “franchise winger” and perhaps plenty of “franchise wings” not included because I didn’t really think about it too hard and there’s only so much space on a visualization.

Worth mentioning as a “data warning” is that I am starting from a subjective and/or results distinction when creating the cohort. As I’m writing this, Dom Luszcyszyn is publishing his work on some similar cohort analysis that comes from data first. I think this is the direction that analysis should move in, providing bear and bull cases for a player or situation rather than trying to spit out a binary decision. Data should be used to accentuate our decision making primarily because human brains, biased though they can be, are also powerful processors and data is only good as it was designed to be.

What I’d also like to mention is that because this isn’t a genesis-in-numbers analysis there’s opportunity for bias. By only looking at players who became good or who are reputationally “franchise”, I am perhaps filtering out a ton of players who may have performed just as well but never found the team or results. Thereby signaling that this data is important and “guarantees” success because I never thought to include the ones who didn’t.

This is called survivorship bias and it’s not something I intend to fully deal with today. That’s why, most of the time, I cop out of generating full conclusions. Instead of digging deeper and doing some hard proofs, I just say “use this at your own peril”. I think we are going to use data to make decisions like that anyway, inform some of our probably already held opinions and rationalize them with points that we like, but it’s worth mentioning that this bias is and will happen and we can at the very least that we’re human and there’s little we can do about it.

The Performance and The Winning

Something I wanted to do with this graph was associate these seasons with years in which their team won Stanley Cups or at the very least made deep runs in the playoffs. I couldn’t do it primarily because DataWrapper had restrictions on which line symbols I could edit. That will perhaps have to be saved for a different type of study.

I know we love to talk about Cups or Bust but repeated Conference Finals are still very good, highly frustrating due to expectations, especially for ownership who gets to make a ton of money. We should optimize for Stanley Cup wins but this is a vicious and volatile sport, perhaps a “shoot for moon, land in stars” type beat.

In any case, wisdom about player peaks has been oscilliating age windows. Old hockey man wisdom was peaking age 27-30. Early analytics suggested an audacious 23 year old peak. Dom now finds that 25-27 is probably the peak window.

In the coming data, I think we’ll see that these sorts of peaks mostly hold true. I don’t have a strong desire to create a better age curve or to get into the finer details of player value that aren’t captured in these metrics. As far as I can remember, penalty differential is by far the most susceptible to age which isn’t captured in any of this data.

What I observed, at least by casually marking seasons that resulted in playoff runs and Stanley Cups, is that winning tends to happen later and oftentimes even later than D+10 which held true for Kucherov, Stone and Panarin. Patrick Kane and Nikolaj Ehlers are some of the only players to win big very early in their careers. Technically, Jordan Kyrou too but he played so little in that first season it doesn’t really register.

I didn’t include the full career bracket primarily because that wasn’t the objective. Teams only control so many years of a player and therefore should really only count on having 7 or so years of that player. While the best players on the best teams often stay for nearly their entire careers, it’s not a guarantee and perhaps fanciful to project for entire careers.

The Objective

The primary goal is to evaluate underlying performance and examine what still needs to happen for Kent Johnson or Kirill Marchenko to become true Cup Winning Caliber Wings or ascertain whether small peaks should be capitalized on to find true Cup Winning Caliber Wings for our Presumed Cup Winning Caliber Center in Adam Fantilli.

This all started years ago with Kent Johnson and because there was limited data and I had a limited skillset, I did what I could. Mostly, this is looking at aging curves but I think there’s still plenty of utility in comparing disparate aspects of the aging curve, comparing an individual with other players and thinking about how they stack up. That first season, Kent Johnson compared most neatly with Mikko Rantanen and Matt Boldy. He scored, but he was behind as a playdriver. That said, his A3Z metrics were uncommonly good , specifically his ability to create transition.

From a player makeup perspective it’s now interesting to think about a few questions. How is he different from those players, how is he alike? What do they do that he doesn’t? What did they do prior to breaking out or what factors contributed to sustained development? Fun questions to think about and building blocks toward the film work that could highlight specific skills that are critical.

I digress, let’s get into the work.

The Cohort and The Charts

This isn’t a perfect cohort but a general grouping of wings who were all-stars or contributed in the top-six of successful teams. This doesn’t include players who were maybe centers at some point, like Steven Stamkos or Joe Pavelski, and it doesn’t include players who don’t have their whole careers captured by the numbers, like Alex Ovechkin or Rick Nash or plenty of others. I am assuredly missing some very good wings, and including some young players who maybe don’t belong, but I think it still illustrates the purpose well. The cutoff for season inclusion was 20 games, I think RAPM data needs more than that but it was pretty good for continuity.

There’s a reason most of these lines are grey and part of that is the inherent interactivity of DataWrapper. When I first graphed Kent Johnson’s cohort it was an unintelligable barrage of color. Now, specifically chosen players are highlighted and their trends amplified.

What I like about this approach is that it really shows the inherent volatility in performance from year to year. Maybe it should be less volatile if it’s accurately reflecting player quality and it’s instead the numbers method that is volatile, but either way the point should be taken well. Ups and downs are natural within a players’ career, even the best wings have peaks and valleys.

Player quality and skills are likely the same year to year but their expression, which we like to call performance, isn’t. Even worse the effect of performance on winning, or results, is even more volatile.

Where creating a pure aging curve, or even generalizing it to a range of outcomes, would give us the same information, homogenizing it into a single average, or line or range, would rob us of some of the complicated nature of the evaluation. Players, and evaluating them, is inherently probabalistic but it’s always the specifics, single outcomes or outliers that ultimately bring a Stanley Cup and so it’s fun to chase those dragons. I think these charts do a great job and the work of color and grey helps bring that to life.

If you’d like to examine the trajectory of specific players, please use the interactivity of DataWrapper. You can highlight specific points but it might be easiest to use by mousing over a name on the legend.

There are some special quirks with certain players and the cohort lines. First, I’ll explain the micro-cohorts.

Franchise Driving Wings are players who were on teams that did win, either the Stanley Cup or made Conference Finals, and played on a line where they were the primary driver. That doesn’t necessarily mean they are better, though it is something to explore with respect to this data, but that they had their own lines, on the high end think Patrick Kane, Matthew Tkachuk, Nikolaj Ehlers. The “complementary” wings played alongside Elite Centers, think Mitch Marner, Mikko Rantanen.

There is a host of more subjective additions either way. Did Jason Robertson drive his own line or do Roope Hintz and Joe Pavelski get some credit? What about Nikita Kucherov with Stamkos and Point? Did Johnny Gaudreau drive a line or was that really Tkachuk? It felt like William Nylander drove a line but does John Tavares really not count? Brandon Hagel really drives it lately, but what is he without Anthony Cirelli?

In the end, I did concede some of these points to the players who really drove lines, using the data as justification. It’s a bit messy but this was mostly exploratory anyhow.

For specific players, I think I need to mention a few.

Nikita Kucherov was a special frustration. He missed the entirety of his D+9 season, where he won a cup, and then only resumed in his D+10 which would have been invisible on the chart as a single point. So he gets a special star.

Valeri Nichuskin also has a time gap, returned and played a season with the Stars and then you can tell when he joined Colorado. Technically his last year in Dallas shouldn’t count either because it was only 8 games but I kept it in anyway.

Mark Stone, Artemi Panarin and Johnny Gaudreau all had some of their best seasons just after D+10. None of Stone’s time with Vegas (his last season was partial with Ottawa prior to the trade and his data from Ottawa is used), and as such winning, is captured on this chart. Only a single season of Panarin in New York is captured and very little of the winning.

Johnny Gaudreau’s super season, the 3rd best goaldriving season in tracked history (behind Panarin’s last point on the Goal Differential graph below but also behind 12-13 Johnathan Toews) isn’t shown. It coincides with Tkachuk’s D+7 monster season but that’s something we’ll explore later.

Relative to trades, I tried to use the data that made the most sense. There’s some subjectivity there with respect to which partial season I used. Mikko Rantanen’s last season was split, some of Martin Necas, Kevin Fiala and the last season of Mark Stone. It’s not perfect but it’s there.

The RAPM Data

I’ll be moving through these full graphs with the horizontal axis as years post-draft. Perhaps this would be better served as a full age curve but I find that you’re drawing a line somewhere anyway so this is what I went with.

I’ve created a couple of micro cohorts in players that I have subjectively identified. If a player played on their own line, or was otherwise the driver of the line moreso than the center, they are called the “driving” wing. If they played next to a presumably elite center, they’re classified as a complementary wing. As an example, Mitch Marner is a complementary wing and William Nylander is a driving wing.

These aren’t necessarily important on their own, though the data tells its own story, but could be used for greater roster construction purposes.

I’ve tried to rein in the vertical axis but some of the extreme outlier players are squishing the differences on the grander scale making year to year often look the same when some small differences are indeed performance.

This first section will be primarily devoted to RAPM data which is a model created by Evolving Hockey and developed from work in Basketball. Essentially, this is a regression model that pretends players start on an equal playing ground and are all “average” and determines their performance relative to what would have been expected.

Essentially, it accounts for teammates, deployment, zone starts, rest, score state. Most things that we know have effects on scoring and shots being taken. It then spits out three separate stats (really six separate offensive and defensive impacts) that represent rate impacts per 60.

If a player has a 10 Corsi Differential, adjusted for all that we can, it means they’re responsible (mathematically) for 10 more shot attempts in his teams’ direction over 60 minutes.

Given the way we can measure impacts, I think measuring offense is safer, more accurate and more important, generally. NHL coaches often create offensive structures that are either laissez-faire or promote individual creativity within whereas their defensive structures seem primarily effort or position based. Less skill expression, not none, and more diffused responsibility.

Perhaps I’m talking myself into re-running this exercise for all six of the broken out model components but for now we’ll include differential because I think that’s an important way to think about hockey from a holistic perspective.

We can always break down specifics when we get to individuals as part of the zoom-in and zoom-out methods I’ve come to enjoy. The whole picture is best understood through a time sequence of top-down gestalt, then zoom-in toward specific details and then zooming out to the gestalt and seeing if it changes the understanding.

Defense is good, offense is great (perhaps most under control of the individual), winning your matchup in whatever way is the best. Exaggerating the difference between what your matchup creates and what you create is the pinnacle. More on this later.

Corsi Differential. Shot attempts relative to situation. The best marker we have of possession as demarcated by shots. After tracking the entirety of the last season, I think it’s pretty good. If you’re not good enough to turn possession into a shot attempt, does it matter what you’re doing with the puck? (this is an oversimplification).

Worth mentioning that the franchise driving players are always averaged out to be greater than the complementary players but ultimately they are all well above NHL average. In D+5, the cohort average is 3.9 RAPM C± which, if we look at this past season, would rank 90th among forwards (Sam Reinhart, in fact) which puts this group of players as averaging out to be “first liners”. That’s a poor way of looking at it but it’s just an example.

Here, we really get to see the evolution and developmental leap of both Kent Johnson and Kirill Marchenko.

Where Johnson started out as a poor play-driver he was let’s say commonly poor. Lots of highly drafted wings had those impacts and they grew as they adjusted to the NHL. In that respect, Kent Johnson’s play-driving grades out well ahead of the curve.

He hasn’t reached peak play driving yet, and perhaps you doubt some of the numbers if you think he siphoned some credit from Fantilli or Sillinger, and still has to get much better if he wants to win in the NHL. That said, he’s ahead of superstars like Matthew Tkachuk, Nikita Kucherov, Johnny Gaudreau. He matches the play-driving growth of Mikko Rantanen.

In this specific metric, Kent Johnson is indeed ahead of Jason Robertson who was a rookie debuting in his D+4 (Dallas development). He landed with a bang whereas Johnson toiled under some poor head coaches and was even demoted to the AHL last year. Those struggles perhaps color our perception of the player where Robertson has only been known as excellent.

That Johnson’s best is ahead, or must be if he’s going to contribute, should be a good thing. I’ve watched his game closely for years now and can see quite easily the areas for opportunity. I’ve been critical at times of certain blindspots, which do date back to his draft year, which might preclude his forward steps or might be just the things he needs to take them. I can’t say I watched any of those other players in the same detail, they may have similar blindspots or they may have been good but subdued for other reasons.

If Johnson takes a step forward like Robertson or Svechnikov or Tkachuk did, we might have a truly special line-driver on our hands (important considering next July 1 he becomes eligible for an extension). If he doesn’t, well then Rantanen and Voracek did the same thing so there’s no reason to panic.

Kirill Marchenko, as dominant as he was, is behind Kent Johnson from a pure possession driving perspective. He has scored goals every season but only this one did he truly tilt the ice. Last season, then, did indeed represent an investment season where his time with the Russians and carrying Sillinger and Texier to matchup line “success” were portents to a true breakout (perhaps something Kent Johnson did with Sillinger and Pyyhtia though I wouldn’t call that a success).

Perhaps Marchenko didn’t have as high of a ceiling in terms of play-driving because he shared ice with Monahan and Werenski much more frequently that Johnson did. While Johnson did win his minutes, Marchenko often dominated them but equal credit was given to linemates.

Marchenko and Johnson, at the very least, earned inclusion on this list with their performance in 2024-25. They’ll need continued performance to truly cement their status among the cohort but their developmental leap has earned them opportunity.

Here’s where I also need to mention the cutoffs at the top of the graph. I had to do it primarily because the absolute heights of Matthew Tkachuk were making it impossible to understand the meat of the cohort. In Tkachuk’s last season in Calgary and his first in Florida, Tkachuk reached truly special heights in possession play-driving earning a +14 followed by a +17. Tkachuk has never been under the Driving Wing cohort average and has dramatically improved the peak potential. He is, without a doubt, the best possesison driving wing that the NHL has seen since it started tracking data in 2007 (he owns two of the top 4 performances) and the best possession driving player since Patrice Bergeron (who owns 2 of the top 5, 4 of the top 20, 6 of the top 40 and 8 of the top 60 C± seasons since 2007).

Which brings me to the next player, Jason Robertson. Apparently on the trade block his peak season of +11 RAPM C is the 14th best season since 2007. It’s a number not beat by a wing other than Justin Williams in 2012-13 and only nearly replicated by Mark Stone’s 2018-19 partial Ottawa season (he was traded to Vegas in the middle).

There have been some other spectacular seasons by players not included, Valeri Nichushkin in three seasons from 2019 to 2022, and Zach Benson’s most recent season and Alex Ovechkin’s 2009-10 all finding double digits.

Robertson, Tkachuk and Stone all share different yet similar play characteristics. Where many teams might be chasing wingers who control play like centers, this group drives play from the wing in a purely wing style. They’re the epitome of Darryl Belfry’s Spiders and should be appreciated for how special they are from a play driving perspective.

Perhaps Robertson’s trend is concerning but his last, injury riddled season still graded out as above average for a driving wing. Tkachuk, similarly, came down from the summit but this past season has recreated another dominant season.

If I’m Don Waddell, I have no problem on betting on Jason Robertson. If I’m Jim Nill, I’m not relinquishing him at any cost. I think we can frame this in a different way. Perhaps the mistake has already been made by Jim Nill who failed to capitalize (though Eastern Conference Final runs are ultimately a good result) on absolutely massive seasons from an excellent play driver.

Shoutout Zach Benson whose playdriving through the first two seasons is already difficult to believe.

As much as driving play is an important factor in winning any hockey game, getting to the interior and attacking the dangerous areas (and preventing the same) is perhaps more important. Whether or not you find it reliable or a better metric could depend on a couple of factors but should be considered in relation to the overall possession corsi measures. Different coaching systems could also have different prioritizations which will also have some effect. Jim Montgomery has famously gotten great results out of a danger over volume approach where Carolina, though it’s changing, and LA have loved point shots and hunting for rebounds.

In this metric, we get a bit of a different flavor for both of the young Blue Jackets. Where Kent Johnson was dominant in shot attempt differential, Kirill Marchenko gets a lot of credit for driving danger. Per-shot, this probably means Marchenko was much better at getting to the slot, or making it easy for his teammates to get there, on a regular basis.

Both forwards, however, grade out exceedingly well on an xG driving basis. Both put up seasons that would be considered good season for any of these franchise defining wings. If we believe their actual goals will come back to earth, their chance foundation is at least still pretty high.

In relation to Jason Robertson, Marchenko and Johnson performed at interesting levels. Kent Johnson was equal at the D+4 season, and Kirill Marchenko was equal at his D+7. In between these two matching seasons, Robertson matched a height that neither has had an opportunity to surpass. Maybe Kent Johnson gets there over the next two seasons and maybe Marchenko continues his growth curve and becomes one of the games best wings. Maybe, though, they both dip and Robertson bounces out of his injury season and maintains his place among the elite drivers from the wing.

I think Marner’s performance over his career is interesting too. While he’s regarded as an excellent transition winger and play-driver by many, he only graded out okay from a pure possession standpoint. Perhaps Matthews who is an outright play-driving monster, siphons away some of the credit. Where Marner does excel, however, was in driving danger. That’s interesting to think about especially as it relates to a certain Adam Fantilli strength (scoring from the slot).

It’s hard to believe what has happened to Andrei Svechnikov. He entered the NHL with a bang and signed what looked to be one of the best value post-ELC contracts possible. Over the last two seasons, however, his performance has dropped. Will the Hurricanes position this as a valley in a career that will be redeemed or will they believe this trend a result of Carolina’s diversification. As a play-and-danger-driver, Svechnikov was most unimpeachable prior to this past season.

All of the aforementioned underlying metrics ultimately matter very little if it doesn’t result in a tangible difference in actual goals scored. Though a team is more likely to win in the future if they control chances, and perhaps an individual player more likely to do the same, if they never materialize as actual goals then we were essentially forgetting about what matters.

I find a lot of the xGAR or WAR or sG discourse to be fairly interesting. Most of these metrics, including I’m sure plenty of the above, are optimized to make a larger model or their own more predictive of future outcomes. That’s all well and good, they’re great tools. But at the end of the day, if we use them to rank players then we’re evaluating the now by the promise of what might come. That’s a little backwards, no?

In any case, RAPM G± doesn’t necessarily only describe what happens in the now but it does describe what matter most at the end of the day. I think about this with respect to Brady Tkachuk specifically. If you looked at some of this on-ice impacts prior to goals you’ll think he’s a preeminent ice-tilting wing. But then you get to the actual goals and he’s just fine. A teammate issue? A sign a breakout is coming? Or is he finding it tough to beat the pad-stuffing allegations.

Now let’s get to the elephant in the room: Kirill Marchenko. From a goal driving perspective, the season we just watched was one of the most prolific in history. Literally the 9th best in history just behind this past season’s Leon Draisaitl and 2012-13 Sidney Crosby.

There are some other “one-hit” wonders on the list around him, Pascal Dupuis in 2012-13, Jared McCann in 20-21, Chris Kunitz in 12-13 (all Penguins), Pavel Buchnevich in 2022-23 with the Blues.

Marchenko’s goal-tilting underlyings have grown steadily and are now backed by bona-fide chance-tilting just the same. Still, from a raw goals-lead-to-wins perspective, this might just have been peak Marchenko. If it isn’t, or he maintains around the same level, the Blue Jackets might be in for a long period of stellar play.

The problem with his metrics, though perhaps I should get into this later, is that too much of it is driven by defense. I said the same last offseason, but it’s unlikely Marchenko continues to solo-suppress goals against. I believe he does indeed have a certain defensive aura but we haven’t necessarily seen the same year over year repeatability of defensive metrics.

Of the other top seasons in total goal differential, the only other players with similar share of the differential driven by suppressing goals against (42% of Marchenko’s .508 were from defense) were Derek Stepan’s 41.8% of his 2012-23, 4th all-time ranked 0.574 for the New York Rangers, Pavel Buchnevich’s 44% of his 2022-23 0.472 for the St. Louis Blues, Tomas Tatar’s 59.9% of his 2022-23, 30th ranked all time 0.434.

Later, those players did not go on to repeat this performance but they have remained really good players for a good chunk of time (Tatar has aged out at this point). The players who often found themselves repeated on this list are Nathan MacKinnon, Leon Draisaitl, Auston Matthews, Sidney Crosby, Matthew Tkachuk, Pavel Dastsyuk. They generate their impact through primarily offensive conributions but are bona-fide superstars and top of the league players.

We might have just seen the ultimate peak of goal differential dominating Kirill Marchenko. If he repeats this performance, we might be witnessing the birth of a Selke-type impactful wing (his all-time GA suppression ranks 68th in 24-25 and 104th in 23-24).

If we combine just the past two seasons, Marchenko ranks 2nd in G±/60, behind Leon Draisaitl, ahead of Nathan MacKinnon and Auston Matthews. If we include all seasons Marchenko has been in the league he ranks 8th, behidn Matthews, MacKinnon, Bergeron, Pastrnak, Tkachuk, Eichel and Draisaitl. He doesn’t necessarily have the shot and hcance profile to back it up but the goal totals are certainly there.

Driving actual goals has always been Kent Johnson’s best skill, at least among these adjusted shot metrics. He’s really early and has maintained performance above the average each season he’s been in the NHL and he’s gotten better each time.

He’s still supremely young and the most important growth is yet to come. In comparison with Jason Robertson, he now looks a little behind with a lot of ground to make up. The good news is that plenty of star wings were in a similar position. Mikko Rantanen, Brady Tkachuk, Sam Reinhart, Jesper Bratt, Jake Guentzel had each never had a goal-driving season as good as Kent Johnson at this point in their careers.

Diving in a bit deeper and filtering some data all-time can help put this into some context. I have filtered RAPM GF/60, and combined all minutes (min. 1500 to hopefully filter out single seasons) at a max of age 22 (moving away from the Draft Year metric because it’s not an option on EH website).

Kent Johnson ranks 45th in GF/60 (50th in total Goal Differential). Ahead of Kyle Connor, Gabriel Lendeskog, Filip Forsberg and just behind Mitch Marner, Andrei Svechnikov, Sebastian Aho, Elias Pettersson. Jason Robertson, though he barely meets the minutes cutoff at 1859 in 128 GP, ranks 1st. Ahead of Sidney Crosby, Connor McDavid, Evgeni Malkin and Alex Ovechkin. He doesn’t have the early seasons of decreased production dragging him down but he is indeed extremely good.

If you’d like a future offensive dynamo, look no further than JJ Peterka who ranks 6th in RAPM GF/60. He’s a defensive liability but not outside of the realm of McDavid, Draisaitl. The warning signs would be Ryan Strome, Jordan Eberle, Kyle Connor, Evander Kane, Matt Duchene, Alex Galchenyuk who all were technically better defensively.

I started leaving him off CBJ target lists primarily because the objective is to get better defensively but he’s put himself in elite territory and he certain passes the test visually.

Nils Hoglander also jumps out as a breakout candidate given his adjusted play-driving at a young age.

Points and Production

So, though I believe play-driving and matchup-winning are more important than individual point production I would be remiss to dismiss the metric as having no importance. If a player can finish possession with goals, they can probably be placed alongside other good players and be expected to do the same.

There are certain player skills that are reliable and repeatable and can always be translated from skill to peformance with regularity. It’s perhaps why transition skill is so valued. If a player can solo carry the puck through the neutral zone, they’re probably going to be able to do that regardless of whatever else is going on.

As the NHL is changing, and neutral zone defenses becoming more sophisticated, I think teamwork is becoming more important in each facet. We’ll have to continue to find the type of players that can continually repeat their impactful performances which is why I’m been using a variety of metrics to paint the full picture.

The next step in that picture is individual point production which I will be separating into two facets: 5v5 primary point rates, filtering out secondary assists which have been seen as statistical noise, and PP points/60, without filtering.

I’m using both of these as proxies for individual contributions within the context of a players’ ability to win matchups. If they constantly set up goals or score them, and their team tends to do that more often, it’s likely that they were indeed the reason for that happening with greater frequency.

If the player can’t score points, but do drive differential, they’re still useful. Potentially extremelly useful but all of the best players find a way to do both. Even the most defensive versions of stars found ways to score points like Aleksander Barkov who posted 96 or Patrice Bergeron who scored 79 in 65 games.

Here, we can perhaps see the reasonf or the reputational differences between Andrei Svechnikov and Mitch Marner. Though Svechnikov has driven play at a high level, where it took Marner a bit, Marner has always been the superior primary point producer. Worth mentioning that Marner has pretty much always been a superior producer to two certainly elite complementary wings in Jake Guentzel and Mikko Rantanen.

Both Svechnikov and Marner have tailed off as of late but their differences are ultimately drastic.

Here, we see the continued progression of Kent Johnson and Kirill Marchenko. Both, in terms of creating points at 5v5, look absolutely like franchise wings on the backs of tremendous jumps. They didn’t get lucky and produce secondary assists.

The question, with respect to those two, is simply that this year was different and there’s not much prior information that they should be at this level. Kent Johnson’s growth has been steady, so perhaps that is less exciting, but they both had obscene finishing this season.

A large portion of their primary points are driven by goalscoring while shooting 19.5% and 15% respectively. I think they’ll be excellent finishers through their career but I’m not sure expecting this high of shooting percentages is the appropriate move.

Still, I’m not looking into point or shooting% distributions for the rest of the players so we’ll have to be content with them being on track while being perhaps tentative of their future scoring performance (they both graded out well in terms of xGF/60 so we can’t be altogether too worried).

One of the most critical aspects of being a star player is performance on the powerplay. If I’ve learned anything from some of the cup winners, it’s that your core should be your powerplay. You should pay them, perhaps because powerplay scoring is a major part of winning especially through the playoffs, but also because points get paid so they better be able to perform in the most offensive situations.

This is most important relative to whether or not you’re giving out a long term contract. Your team will have non-core players critical to winning but they’re usually elite 5v5 producers and paid far under their worth (good defensively helps with that too).

In any case, Svechnikov redeems himself with powerplay scoring where Marner further cements his critical scoring talents. Jason Robertson’s peak seasons mirror his underlying 5v5 play though they should technically have little to do with it. We can explain his most recent season as injury but perhaps losing Joe Pavelski has something to do with it too.

Kent Johnson, aside from his Pascal Vincent unlucky streak on the powerplay, has had a promising start to his career and an especially promising past season as the poweplay QB, more or less. (I know that’s usually meant for the Defenseman at the point, but the play ran through Kent Johnson this season and I think it makes way more sense to consider Kucherov the QB than it does Hedman).

Kirill Marchenko, though his second season is disguised as being poor despite him being the best producer, has been influential but not elite on the powerplay. I’ve spent enough time writing about how I think he could be pushed to a different position to unlock greater success but it’s also possible he isn’t part of a truly elite unit. That has implications with respect to long-term contract positions and not necessarily his inclusion on the team.

Those implications, I think, necessitate some exploration this next season. Try to change around the tactics, I think Florida is doing some interesting things on the PP in the playoffs but also there are plenty of teams with dominant units that CBJ would be well aligned for, to see if you can get more from Marchenko. Otherwise, perhaps try him out as a penalty killer and see if he really is that kind of defensive suppressor.

I think it’s worth mentioning with respect to “repeatability” that the powerplay is pretty much always a controlled situation. I’ll get into it later with respect to Marchenko and his primary method of acquiring dangerous puck touches but if a player is bona-fide powerplay dominant, they’ll be a reliable source of goals.

The CBJ Bull Case

The quick hit is simple, when Don Waddell said that he believes CBJ have the pieces in the room to be contenders, Kent Johnson and Kirill Marchenko look like they are big reasons for that assertion. They’re exemplary 5v5 point producers and bona-fide first line xGF drivers.

The Bull Case for the Blue Jackets is an easy one to write, it’s borne out in these players’ jump this past season. This season, two young forwards under team control broke out and positioned themselves within the cohort of future franchise wings. Those wings tended to win Stanley Cups or make Conference Finals Appearances after their current ages which means all the Blue Jackets need to do is surround them with talent and watch them flourish.

Both players found ways to contribute on the powerplay and helped, or completely, drove separate lines. They play different styles but each has secondary characteristics that mean they should work together as well. There’s a ton of reason to keep Kent Johnson and Kirill Marchenko and use this short window of very cheap contracts

The Robertson Bull Case

There’s really little to doubt about Jason Robertson from a player acquisition perspective, and I think the charts above make it clear. He is one of the few wings who have posted truly elite play-driving sequences, he scored 109 points and he’s been part of late playoff competitive teams.

As a wing, there are few skills he doesn’t have. Perhaps blistering pace but considering his play-driving and the similar lack of footspeed from other similar stylistic wings in Matthew Tkachuk and Mark Stone, you have to believe it’s not an overt problem.

He can manage the puck after the blue-line especially as an anchor, he can create give and go sequences, he can promote the success of his line by being heavy on the puck, not just with his body but with his constant readiness to beat the defense with an easy pass through their layers. He’s pre-pass preparedness is elite and his vision and space management is incredible. He truly unlocks transition by building play even if he’s not directly carrying the pucks across the line and getting credit.

If you want to slide a player next to Adam Fantilli and help him promote success, Jason Robertson is pretty much it (Mitch Marner is good too but there are certainly contract concerns). He has the east-west, play-building game conquered that should perfectly unlock Adam Fantilli or even Cayden Lindstrom in time.

He’s all-brain, he’s good defensively and he contributes on the powerplay. Nothing about this body of work suggests he’s going to age poorly. If you get 10 good years out of Jason Robertson, which would put him at JvR’s current age (who wasn’t ever this good), Adam Fantilli will be 30. That’s about as big of a window as you can ask for.

If you want to put it in terms of Evason’s line construction, Robertson is a sublime small space winger who projects to fit in really any combination of the Blue Jackets’ top six.

I think it’s worth mentioning the differences in his play year over year. Robertson hasn’t returned to his 109 point summit and that might have something to do with roster construction. Robertson’s 5v5 primary point rates have been slipping but the 109 point summit was reached via 40 powerplay points.

One critical difference has been Robertson’s playstyle. Not that he’s a different player but he’s certainly asked to carry more of a burden outside of the offensive zone. Why? Well losing Joe Pavelski might have something to do with it, maybe Roope Hintz is losing some heat too. Either way, I would caution that the reduction in points and offensive zone contributions is inherently poor.

If you allow me a dalliance with HockeyViz, we can certainly see that a reduction in Pavelski’s abilities may have resulted in a decrease in pure netfront volume. The Blue Jackets, if they had anything last season, had net-front players. Perhaps Jenner and JvR aren’t long for the team or don’t have a place next to Robertson, but Monahan would certainly be eligible.

There, Robertson’s demonstrated versatility can perhaps be a weapon and his offensive zone dominance when liberated a lever for a coach to pull. Ask him to drive a line and he’ll give you good possession restults. Ask him to create offense on a stacked line, he can do that too.

The Kent Johnson Future Bear Case (or Why He Should Be Used as a Trade Piece)

For Kent Johnson, the problem is primarily that a lot of his possession driving, his RAPM CA/60 is the biggest source of differential. He’s an otherwise run-of-the-mill shot attempt booster. I believe his intelligence as it relates to forechecking and his neutral zone pressure diffusion could explain some of the difference but perhaps not all of it. He’s a smart player but loses some engagement in some critical situations perhaps because he doesn’t have the outright strength that wins battles outright in the NHL.

Among elite wings, though, there are few that have dramatic shot attempt suppression numbers. I wonder if that’s because Kent Johnson tends to slow the game down and drag the game into a lower event state. He still wins his matchups and minutes that way but his suppression might be overexaggerated. He looks like a good defensive player but really just fewer things happen. A nuanced difference but maybe an important one.

When we look at some superstars we see the profile of players who are dominant offensively but putrid defensively (Draisaitl, McDavid, Crosby, MacKinnon). In reality, they simply play higher event games which means they exaggerate the differential between their offense and defense.

Let’s take a 60% CF% for example. If there are 10 shot attempts while that player is on the ice, 6 CF and 4 CA, they have created a 2 shot attempt differential. If they make the play more high event and maintain their shot attempt share, say 50 shot attempts while they are on the ice, they have 30 CF and 20 CA. They have now won their minutes by 10 shot attempts instead of 2. (This is a big explanation for Marner’s lack of playoff scoring as well, he actually wins his minutes just the same they’re just lower event and played more defensively, whether that’s him or his coach is to be determined).

So now, fictional Sidney Crosby has now allowed 20 CA per game where fictional third liner only allowed 4CA in the same amount of minutes. That must mean Crosby is bad defensively. In reality, it’s just number manipulation.

So the question for Kent Johnson is whether he’s well positioned to adapt to a higher event environment or whether he can enforce his “slow the game down” approach as he plays with and against better players. Personally, I think he can but this is the bear case so we have to ask tough questions.

I’d also caution some concern with his xGF/60 driving but there’s reason for optimism either way. The concern, for me, is that most of his xGF/60, outside of a late stretch with Fantilli and JvR and a small stretch alongside Monahan, came from being paired with beefy netfront wings who created tip plays.

Boone Jenner and JvR both created a lot of danger with these tips where Kent Johnson did a lot of the work to allow them to get into those positions. They’re related, and Kent Johnson deserves credit, but it might not be the same style of danger chance driving, orchestrating passing plays that result in inner slot chances, that we see from true elite wings in Kucherov, Pastrnak, Tkachuk, Marner or Robertson and that might be what the Blue Jackets need.

When looking at Kent Johnson’s profile, I think we can see some of that borne out. His deflection assists/60 are high, chance assists are low but his high danger assists/60 are indeed high.

Consistent with his play last year is his tendency to hold pucks in search of dangerous opportunities. Though that undermines some of his shot production metrics, it might explain why the goal driving is there and might also explain some shot attempt suppression (while searching for a high danger pass, he is killing shift clock, and he’s smart enough not to throw pucks away and create counter opportunities).

He may be more aligned with a coach who prioritizes and builds structures around creating danger, I’m not disparaging Evason here, but it’s clear that he needs to improve or “solve” his approach in the offensive zone. He’s great in motion and in transition but when the game slows down he needs to find better plays (maybe that’s just alongside better teammates too, too much of this time has been with Cole Sillinger and Adam Fantilli wasn’t exactly the picture of teamwork either) or find plays that keep motion.

There are some red/orange flags there but mostly Kent Johnson’s growth has been nothing but glowing. His late season play, especially against Buffalo and Washington and especially when he was playing against elite competition, was very exciting. He’s got room for growth, an appetite for learning and a season to build off of.

The Kirill Marchenko Future Bear Case (Or Why He Should Be Used as a Trade Piece)

Kirill Marchenko, on the other hand, is a fascinating case. What he had in vibes last season, undermined by “poor analytics” while winning matchups alongside Yegor Chinakhov and Dmitri Voronkov, he has now actualized. He came by his domination in a unique way alongside Sean Monahan but continued it alongside Adam Fantilli as well.

While he did have an outright dominant goals driving season, perhaps you think he isn’t that kind of defensive suppressor. These two seasons of back to back dynamic goal suppression paint him akin to Aleksander Barkov. The goals, most likely, are going to start going in sometime.

There’s also reason to believe, if Darryl Belfry’s assertion that superstars play off the pass holds true for this specific context, that Kirill Marchenko’s puck-stripping playstyle, though it is a repeatable skill, might not result in the same ceiling of offense. Perhaps Monahan is critical to dominance in that playstyle, which is why I like that they could lean into a Hagel/Cirelli dynamic, but Marchenko created plenty of passing danger on Fantilli goals.

He’ll need to get better at playing in possession and threatening through passing rather than relying on puck winning or exploiting teams in transition to take those next steps. Specifically, Robertson remains an excellent example for how to diffuse pressure at the blue line. Marchenko’s space passing in transition and slot passing in general often have a volume over precision approach, something that will need to be addressed if he wants to succeed in the playoffs.

I really have nothing bad to say about Marchenko, in total. He addressed nearly every weakpoint of his in a single offseason and has only left improving specific aspects of playmaking efficiency. He’s got plenty of time to add that dimension to his game but he’s already one of the best play-builders and primary puck carriers on the team. If anything, this exercise has suggested that he’s in fact more unique and better positioned to be a star player than I anticipated.

His bear case pretty much only exists insofar as his inclusion in a trade deal. He’s older than Kent Johnson which means he might have less runway, and maybe he’s just a really really good top six wing instead of the top line driving superstar that his might contract might demand he be paid like.

The Bear Case for Acquiring Jason Robertson

If there is a bear case, perhaps it’s Robertson’s acquisition cost with a tiny hint of footspeed and transition ability.

If you think rush offense is critical to winning in the playoffs, or footspeed is the salve that heals all transition wounds in the intense environment of the playoffs, Jason Robertson might come at a cost. You would, at the same time, have to believe that Kent Johnson or Cayden Lindstrom are part of the acquisition and they they represent true puck-carrying or transition aces that should succeed in that environment. Johnson’s NHL Edge footspeed is only slightly higher whereas Marchenko comes out dramatically ahead (Robertson has skating equal to Tkachuk or Stone who have had no problems in the playoffs).

It’s a little foolish but I would once again mention that Kent Johnson was part of one of the best entry units with the man advantage and has been an excellent powerplay producer. The powerplay is huge in the playoffs and stripping a primary puck carrier away might introduce problems. The answer to those problems would be an actualized Adam Fantilli whose skating speed should more than make up for it but he wasn’t in that role last year.

Perhaps it’s why Dallas is moving him, because he can’t necessarily be responsible for all of transition while maintaining offensive zone dominance and might be paid like it. Moving him to PP2 after aqcuiring Rantanen might mean paying him money he earned with powerplay production and no longer putting him in that same spot. If he doesn’t eclipse Kent Johnson’s powerplay performance or role dominance, or Fantilli doesn’t step up to the same degree, the Blue Jackets might be introducing friction in a critical area.

This is really the critical avenue of evaluation. Jason Robertson is a complete Bull Case. The Bear lies entirely within the cost of acquiring him and the potential that you give up in moving on from one of the bright young stars currently on the CBJ roster.

The Verdict

If the original genesis of this research question was: Should the Blue Jackets trade Kent Johnson or Kirill Marchenko for Jason Robertson? My initial gut instinct is torn in both directions. It doesn’t necessarily feel like either direction is a wrong answer. Giving up a good player and maybe some capital to get Jason Robertson makes sense, staying the course and betting on the answer being in the room also makes sense.

This past season, though Robertson was injured, the two young CBJ wings give a lot to bet on and with the certainty that they want to be here. Jason Robertson, however, is a truly unique talent and any team should always be on the hunt for superstar players. Huberdeau was a 115 pt player but not a true driver and Florida won big on the Tkachuk trade.

That’s the difficulty of the thing, ultimately, that we’re not necessarily searching for the most logical and likely to succeed answer, or we could be and take a volume approach, we’re looking for the one that gets a Cup. Dean Evason’s arrival coinciding with the breakout, and the teamwide outrageous finishing, has introduced both optimism and uncertainty for both players. It’s juxtaposed the fear that trading them away means you’ve lost a future “homegrown” superstar with the reality that Robertson already is one.

It’s just difficult to fight our pattern craving brains. Robertson’s trajectory has been downward and his GM wants to move him, but the charts show that that these valleys have happened in the past and been overcome, where Kent Johnson and Kirill Marchenko are on dramatic upward trends and are people we already know, the future for them feels limitless.

If we’re coming from a brutal, cold and calculating angle, (or zooming out for some more durable analytics perspective) I think Kent Johnson is the player to be offered. He looks like a future superstar, Jason Robertson is one, maybe we shouldn’t make it more complicated than that. I presume Nill is chasing the tough angle in moving Robertson but it’s possible he’s chasing his resonant waves of prospect talent or restocking after moving futures for Rantanen. I’m going to be honest, I can’t fathom why he would entertain moving this player in the first place.

Johnson’s trajectory is fantastic and his importance to the higher functions of creating offense critical but in pretty much all metrics we’re hoping that he still takes more steps that either Marchenko or Robertson have already taken. He’s got time, perhaps we should have some faith that a 22 year old player will continue on a tremendous trajectory, but the degree to which he needs to improve is the difference between being a superstar and being an excellent player.

Perhaps Kent Johnson does become the next Patrick Kane or Nikita Kucherov, we can’t rule it out with the data or with the skills that I have, but perhaps he tops out as Mitch Marner or Kevin Fiala or Martin Necas and is a good player who doesn’t quite bring the juice when the lights are brightest (he does have a World Juniors Golden Goal under his belt). He was the lifeblood of the powerplay but, at 22 years old, couldn’t really do it on his own without Sean Monahan’s presence.

If the Blue Jackets can get away with a deal that doesn’t involve either Marchenko or Johnson, it probably involves Cayden Lindstrom. If it doesn’t involve Lindstrom, I would probably run to the table to take the deal and question Jim Nill’s approach.

Perhaps there’s an angle for Voronkov and Chinakhov as depth versatility with big impact potential. Maybe it’s Boone Jenner for low-cost veteran leadership and a powerful netfront game. Both tertiary packages also presume multiple first round picks and more futures. The NHL seems like it’s going loopy with “toughness” and “playoff” discourse and I have seen few GMs immune to those effects.

The Blue Jackets can probably win an all-futures trade, competing most directly with Montreal perhaps but Utah does have the 4th overall pick to dangle. Depending on total cost the Jackets would still have a lot of work to do to remodel the defense without whatever capital Robertson would cost. Heck, they can probably win the auction outright if they’re willing to move on from Johnson, Marchenko or Lindstrom. If Waddell leverages the Jiricek return to swing Robertson, that will more than make up for the loss of star potential.

The TL;DR

Let’s say it this way though, there’s little reason to trade Kent Johnson or Kirill Marchenko right now if you believe in the players and their trajectory (frankly I don’t see why Robertson would be traded right now before his contract expiration). They were part of a stellar locker room and drove play independently of their linemates. They were both part of an absolutely stellar entry system on the powerplay. They both have improved significantly year over year.

If you have questions about whether their play is sustainable, Jason Robertson is one of the best ways to capitalize on breakout seasons possible. It would be a strange bet, to be sure, given Waddell’s infatuation with his roster players and the locker room this past season.

However, I have little doubt that Jason Robertson is a player who will be good for a very long time and I can say that with some certainty. He is a player with every skill domain conquered who plays literally exactly how you’d want a wing to play. He is not reliant on skating speed, and has the brain and skills to age quite gracefully. If he’s a good player for 10 years, you’ve bought as many cracks at a Cup as you can reasonably expect at this juncture.

If the Blue Jackets can find a similar, non puck-touch dominant player for cheaper, perhaps that’s a better move. Maybe they can get Svechnikov for cheap and hope he breaks out as a complement to Fantilli. Maybe they can spend some of the assets and place a big bet on Zach Benson instead. If they can get Mitch Marner for a reasonable contract that might make more sense too considering he’s otherwise free.

Werenski, Monahan, Marchenko and Robertson are about as good of an insulating and mentor cast as you could create for Adam Fantilli (or Kent Johnson or Cayden Lindstrom) and would dramatically increase the likelihood of a cup in the nearer future.