The nature of the defence position has undergone more change over the last decade than probably the 100 years before that (well, once we get past the rover, anyway). It really wasn’t that long ago that teams would have one or two bruisers on the bottom pair who would struggle to drive the play, which would often lead to them with worse goal differentials than their teammates, but they would be extremely physical with the opponent. Cross-checks in front of the net to the opponent and glass-and-out with the puck and all that. Think of guys like Douglas Murray, Roman Polak, or Deryk Engelland.
That has changed. Even defencemen that are considered bruisers these days – Radko Gudas, Martin Fehervary, Erik Cernak – are capable of moving the puck to varying degrees. Heck, even Luke Schenn is having a good postseason for Toronto. Those types of d-men are abundant on bottom pairs, but the evolution of the position has brought more skill across all defence pairings. Gudas, for example, finished the 2022-23 season in the 62nd percentile for points/60 minutes at 5-on-5, or easily a second-pair level. He also had over 300 hits. That is the type of guy that is found on bottom pairs in the modern era.
And that brings us to Carson Soucy. At time of writing, his Seattle Kraken team is in the second round of the 2023 playoffs with the series tied 2-2. He has just two points in 11 games so far these playoffs, but Seattle has outscored the opposition 10-4 when he’s been on the ice at 5-on-5, controlling 54.6% of the expected goal share. The numbers are nearly as strong when he’s not been on the ice with Matty Beniers, likely Seattle’s best forward in the postseason, so it’s not a matter of the forwards he’s playing with. His pairing with Justin Schultz has been integral to their success thus far, even playing third-pair minutes.
This playoff success isn’t out of nowhere for Soucy, though. He is turning 29 years old this summer with just 250 regular season games to his credit, and it begs us to ask if he’s been a player that has been hiding on a roster for years and just needs a chance to break out like teammate Vince Dunn. He was a fifth-round pick in 2013, so there’s not a junior/college pedigree. However, he has done a lot of things at the NHL level, prior to the 2023 playoffs, that indicated Soucy is not only capable of playing at the NHL level, but thriving, and helping his team achieve real success.
We are here to go through Soucy’s career, starting with his draft in 2013, and moving all the way to today. He is a free agent at the end of the season, per Cap Friendly, and a deep playoff run could pay off handsomely for Soucy in more ways than one, so let’s dig into what he brings. We will be using Hockey Reference, Frozen Tools, Natural Stat Trick, Evolving Hockey, Hockey Viz, and Corey Sznajder’s tracking data as primary sources unless otherwise indicated.
The Early Years
The 28-year-old rearguard was drafted in the fifth round of the 2013 Entry Draft by Minnesota. That he was playing in the Alberta Junior Hockey League, and not the Western Hockey League, should indicate what level of prospect the Wild were considering at the time. There is a reason Soucy spent four years in college followed by two seasons in the AHL.
I tried to find old scouting reports on Soucy and found nothing. Literally nothing from January 2013 through January 2016. His Elite Prospects profile wasn’t even created until the summer of 2016, over three years after he was drafted.
There was a profile on him by Brandon Sleik of The Athletic in 2018 that gives us a bit of insight into his younger days. The first tidbit being he was small in his teenage years, not hitting his growth spurt until high school, which is what limited him to the AJHL and not the WHL. The second tidbit is that while he was considered a bruiser-type in the AHL, his off-puck skills, his skating, and his defensive-zone decision making all sharply improved. The last note is that Soucy admits he was slow on the uptake in the AHL, coming from college, but started to find his game halfway through his first season, cemented himself in the second AHL season, and then it was off to the NHL.
We will come back to those notes about his AHL improvements later.
Minnesota: 2019-20
Soucy first appeared in the NHL at the end of the 2017-18 season, a time of year when a lot of teams give their college players a brief tryout. He didn’t become a quasi-regular until the 2019-20 season.
That first season saw Soucy perform well even if he was injured in February, ending his season that would have ended less than a month later anyway thanks to the COVID pandemic. Overall, the team carried 51.7% of the expected goal share at 5-on-5 with him on the ice, outscoring the opposition 36-22. He was a bit fortunate with team shooting/saving percentage when he was on the ice, but he was still very good in a variety of areas anyway: his expected goals against impact at even strength was 37th out of 158 regular defencemen (min. 750 minutes), which rates at a top-pair level. His expected goals-for impacts weren’t good, but he was also on a team that had tremendous issues generating offence. Remember that this is the year before Kirill Kaprizov lands in town and the team’s leading scorer was Kevin Fiala with 54 points in 64 games. At 5-on-5, the Wild were 16th out of 31 teams by shot attempts per minute and 25th by expected goals-for per minute. Playing similarly offensively to the Arizona Coyotes doesn’t often do a lot of good.
But it was the defensive side that showed out. A lot of his tracked offensive metrics – assists on teammate shots, zone entries and exits, high-danger assists – were below-average-to-poor. The defensive metrics, though, were all good-to-great. Here are his tracked zone defence metrics for his rookie year:
When you can be a rookie defenceman in the NHL and deny the opposition the blue line and controlled entries at an above-average rate, it’s noticeable. Looking across the league that year, he performed similarly in these respects to players like Niklas Hjalmarsson and Brandon Carlo. Hjalmarsson is now retired but was an elite shutdown defenceman in his heyday, and Carlo is that shutdown guy today. Perhaps there was a reason Minnesota allowed fewer goals per minute with Soucy on the ice than any other regular defenceman that season by 33%.
Of course, everyone knows what happens next: the league is shutdown due to COVID with Soucy on the shelf due to injury. He plays four games in the Bubble Playoffs, the Wild are eliminated by the Canucks, and that’s that.
Minnesota 2020-21
I’m always hesitant to rely on Bubble Season data. The schedule was obviously very weird, players were in and out of the lineup, lots of AHLers were used, unusual offseason etc. Now that we have two full years of post-Bubble data, the actual Bubble season isn’t overly helpful but we need to look at it anyway.
Let’s touch on the defence first, since it was his early strength. Just above, we outlined how he was above-average in a lot of tracked defensive stats, but wasn’t high-end anywhere. That changed in the Bubble season as Soucy was heavily targeted by the opposition – 66% of carries against him with the league average around 56% – but the rate of scoring chances off opponent entries was below average. To illustrate this, here is a graph that shows where Soucy is among carries against (X-axis) and opponent zone entries leading to scoring chances (Y-axis), all on a per minute basis:
The red circle, including Soucy, has a Carolina Hurricanes defenceman inside of it, and that guy is Jaccob Slavin. Yes, one of the closest defensive comparables to Soucy that season, by a few very key defensive metrics, was Jaccob Slavin. Not bad.
That bears out when looking at his public stats, too. That season, Soucy had the lowest 5-on-5 goals against rate among all Minnesota defencemen (again) with an expected goals against rate similar to Jonas Brodin. Across the league, he was 39th out of 133 defencemen by expected goals against impact at even strength, once again a top-pair level. We now had two years of Soucy being really good defensively, and improving, even if the Bubble Season data is not the most reliable.
Offensively, things got better. In fact, a lot better. Whereas the 2019-20 season saw Soucy’s stats slip significantly, at least in comparison to the hopes he or the Wild had after his offensive improvements in the AHL (remember those?), there was improvement in the Bubble Season. Almost all of his tracked metrics were below average as a rookie, or considerably well-below average. That changed in the Bubble season, and for an example we’ll use controlled zone entries. His rookie season saw him well under the league average, and in line with names like Luke Schenn and Deryk Engelland (hey, there they are again!). Conversely, the Bubble Season saw that rise to an above-average rate, and in line with names like Mikhail Sergachev, Victor Hedman, Thomas Chabot, and Seth Jones (inside the red rectangle):
Controlled zone entries are just one piece of the puzzle, but it’s always nice to see improvement in an area that can help offensively. Yes, even in a Bubble Season with unreliable data.
He also showed somewhat consistent shot quality, too. These things can be difficult for defencemen – their shot quality is almost always poor – but a big sample can point us in the right direction; the individual expected goals leaders among defencemen at 5-on-5 over the last three seasons are Dougie Hamilton, Brandon Montour, Roman Josi, Zach Werenski, and Shea Theodore. Soucy wasn’t elite like that, but he was in the top half of the league over his first two seasons, and some fortunate shooting helped lift him to inside the 75th percentile in goals per minute. His tracked scoring chances per minute at 5-on-5 were in line with names like Seth Jones, Adam Fox, and Dmitry Orlov.
At this point, at a minimum, Soucy had shown to be capable of being reliable defensively in a bottom-pair role, which is all most coaches ask from their bottom pair. Soucy had signed a three-year extension before the season, so he had two years left in Minnesota. Of course, the story is far from over at this point.
Seattle 2021-22
The 2021 offseason brought the Seattle expansion draft and Soucy was selected by the Kraken after being exposed by the Wild. We won’t get into the mechanics of the expansion draft, but my hope was that the move for Soucy (and others) to a fresh team would give him a chance at a bigger role. He did, of sorts, earning a career-high 17:40 per game, but that was still bottom-pair ice time in a Seattle context. In fact, he was starting way more often in the offensive zone with Seattle (45.9% of the time) than the year before (30%). He was facing slightly tougher competition, but still not often eating the toughest minutes:
A bit tougher competition, but more ice time per game and way more offensive zone starts. His usage was starting to change.
Soucy’s defence was still good, though. His expected goals against impact at even strength was 59th out of 172 regular defencemen, or the 63rd percentile. That is just outside a top-pair impact, which does represent a slight drop from his first two seasons, but on a bad expansion team against tougher competition with more ice time? It seems more than fine.
On top of that, once again, Soucy had the lowest goals against per minute on his team at 5-on-5. We are now three seasons and 169 games into his career and Soucy’s goals against rate at 5-on-5, relative to his team, led the NHL. In other words, of 155 defencemen from 2019-22 with at least 2000 minutes at 5-on-5, no one had a lower goals-against rate compared to his team than Carson Soucy. He and Jamie Oleksiak made a very fine defensive pair, even if Soucy played at least 110 minutes at 5-on-5 with four different defencemen in 2021-22.
Now, we have three years in a row with good-to-great defence, across two different teams, in slightly varied roles, with fluctuating ice time and line mates. He had cemented this area of his game by the end of his first year in Seattle.
At the other end of the ice, his expected goals-for impacts fell from his second year in Minnesota, and the goals-for fell with it. Of course, playing for the team that was 28th by goals and 30th by expected goals certainly did not help in this respect.
There were a lot of offensive improvements, though. In his final season with Minnesota, he was a lot better with controlled zone entries, but the Wild had a tough time generating scoring chances off of them (it was a below-average rate). That did not persist in Seattle as he maintained very good controlled zone entry rates, but the team around him generated a lot more scoring chances off of them. He was far, far ahead his fellow Kraken d-men in terms of scoring chances created off his entries:
The funny part of this is Soucy rated very closely to a plethora of other young blue liners: Timothy Liljegren, Rasmus Sandin, Noah Dobson, and Cam York. It is an interesting list, to say the least.
Not only was he staying consistently good with the zone entries, but his team was now actually creating scoring opportunities with them. Perhaps there was a reason Seattle scored more with him and Oleksiak on the ice at 5-on-5 than any of their other rearguards.
Lastly, he improved his individual shot quality again. Across his first two seasons, he was good but not great. That changed in his first year with Seattle as his individual expected goal rate at 5-on-5 was in the 78th percentile, or back to a top pair rate (as did Oleksiak). It helped lead Soucy to tie for the league lead in goals per minute by a defenceman at 5-on-5. No, seriously, he tied Aaron Ekblad and out-paced a few notable names:
Spanning his first three seasons in the NHL, here is the list of defencemen who outscored Soucy on a per-minute basis at 5-on-5 (min. 2000 minutes):
- Jakob Chychrun
- Tony DeAngelo
- Cale Makar
- Mike Matheson
- Dougie Hamilton
- Zach Werenski
- John Carlson
- Jared Spurgeon.
/end list
Impressive, yeah?
The one thing that was lagging was his playmaking. It was not a strength in Minnesota, and his scoring chance assist rate at 5-on-5 in 2021-22 was 0.64 per 60 minutes (that measures his rate of assists on a teammate’s scoring chance). For comparisons, Nashville’s Jeremy Lauzon was at 0.65/60 and Carolina’s Brendan Smith was at 0.6/60. That is the level of playmaking we’re talking about, which tells us there’s a reason he had as many assists (7) as goals in 2019-20 and why he had just one more assist (11) than goals in 2021-22.
Heading into the 2022-23 season, Soucy had 169 games played and had shown top-pair defensive impacts – or even higher – the entire way. It was done playing largely third-pair minutes, but it’s certainly a lot better than showing bottom-pair defensive impacts in bottom-pair minutes. He had to learn to adapt offensively to the NHL, but he made strides in Year 2 and was well on his way to being proficient in Year 3. To be really valuable, he didn’t need to be elite offensively. Just being average, to go with top-end defensive play, would lead to a very valuable NHL defenceman.
Seattle 2022-23
For Soucy’s second year with the Kraken, his ice time declined. That would normally be a concern, but it probably has more to do with Vince Dunn than it does Soucy; by Evolving Hockey’s Wins Above Replacement method, Dunn led all defencemen this past season. Like, not per minute, or at even strength or whatever, he led the whole fucking thing. Vince Dunn added three minutes of TOI to his game from the year prior and it’s really hard to argue with Seattle’s reasoning here.
As for Soucy’s role, it actually got easier. His current playoff partner, Justin Schultz, was often his partner in the regular season. They were kept away from top competition and Soucy set a career-high by starting in the offensive zone 58.6% of the time; he had never had a full season over 50%. Though he got some tougher minutes in his first year with Seattle, that relaxed in Year 2.
He didn’t relax his impacts, though. The expected goals against impacts fell to a second-pair rate, but still in good company with names like Matheson, Darnell Nurse, and Jeff Petry. His expected goals-for impacts improved and were the highest for him in any non-Bubble season. A positive year by those measures, but it’s elsewhere the key to his value reveals itself.
Once again, the good defensive impacts led to the lowest goals against rate on his team at 5-on-5. He has done that in three of his four seasons, and the one he didn’t was his first year in Seattle, and his defence partner led the team while Soucy was second. He has never had the super-elite expected goals against metrics, but they have always been good-to-great, and the goals against rates have been super-elite. We won’t want to rely solely on goal data in small samples, but once we get to 250 games, it’s pretty hard to ignore them.
It wasn’t a case of his partner carrying him to defensive success, either. Schultz is a fine blue liner and a guy I would want on my team, but Soucy has kept doing the things on defence that led to success early in his career, but only better, and more often. We can see this visually as he had the highest blue-line denial rate among all Kraken d-men, which means opposing teams had the hardest time entering the zone with control against Soucy than anyone else on that roster. He denied entries at a 14.6% clip; Miro Heiskanen was at 13.6% and Hampus Lindholm at 14.7%. They are highlighted in the red box (along with Derek Forbort and Vince Dunn):
In an area where he was once good, he is now legitimately great. If teams can’t enter the zone cleanly, generating offence is hard, and Soucy is making life difficult for the opposition more than almost anyone. Teams only generated 2.8 entries that led to scoring chances on a 60-minute basis against Soucy; the league average is 3.7.
It wasn’t only defending the blue line, though. If the team dumped the puck on Soucy, he was legitimately one of the best in the league at retrieving it and getting out of the zone. His retrieval rate was 12.1 per 60 minutes at 5-on-5. Again for comparisons, Shea Theodore was at 12.4, Adam Fox was at 12.2, Rasmus Dahlin at 12.0, and Hampus Lindholm at 11.5. When you’re smack in the middle of that kind of list, by any metric, it is a very good sign.
Here’s the big change: playmaking. Earlier, we discussed how he had poor playmaking metrics every season, and it was why his assist totals were low, even for his ice time. He was good at jumping in the play for his own chances, but he was the engine that got plays started instead of the steering wheel guiding the offence.
For the 2022-23 season, Soucy saw a colossal jump in scoring chance assists. His first year with Seattle was 0.64 such assists per 60 minutes at 5-on-5; his second season saw that number jump to 1.92 per 60 minutes. He literally tripled his SCA/60 in one year. Again, there’s a reason his assist rate jumped over 70% from the previous year! It didn’t put him among the elite or anything – the next tier down with Dunn, Bowen Byram, Morgan Rielly, and the like – but it was certainly a monumental jump in the one area that was his big weakness.
There was a trade-off as Soucy’s 5-on-5 goal scoring plummeted. He scored just three such goals, having nine the year before in 14 fewer games played. Natural shooting percentage regression explains part of this, but his individual expected goal rate declined even as his shot rate climbed. He was taking more shots, but a lot of them much worse than before. Soucy finished the year still above-average by scoring chances among d-men per minute, but going from easily a top-end rate to just above average is a steep drop.
The last drop was in zone entries with control, as he was above average in 2021-22 but fell below average to where he was earlier in his career. That may have been at the behest of the team, though, as Dunn led the Kraken blue line in carry-in percentage at a rate that was still lower than names like Urho Vaakanainen, Ilya Lyubushkin, and David Savard. Soucy’s drop here would be a concern if his teammates were excelling in this respect, but they’re not, and given the team’s success, I would bet on Seattle coaching that into the team.
And now, we’re back to where we started, which is the 2023 Playoffs where Soucy and Schultz are beating the brakes off the opponents. Even in a season where his scoring chances declined, Soucy improved in a number of other areas, particularly ones that would help the guys on the ice with him. He may be turning 29 in the summer but he’s really hitting his stride in the NHL, and he’s getting better every year.
Not all metrics will love Soucy. Despite not having good expected goals-for impacts at any point of his career, his teams have been very good at generating goals with Soucy on the ice. Over the last four years, his overall expected goals metrics have been relatively even, when balancing modest offence with great defence, but the actual goals impacts have been decidedly in his favour. In fact, the difference between his goals and his expected goals at 5-on-5 in those four years is sandwiched between two other guys who’ve been breakout players over the last few years: Ryan Graves and Brandon Montour. Found that a bit interesting.
This is not to disparage expected goals models in any way. We’ve used them heavily in this piece and I use them every day. They are wonderful at illumating most of what is happening on the ice and is great for quick-referecing comparable players. But they don’t catch everything, and players have been known to consistently out-perform those metrics; Roman Josi and Justin Faulk fit that bill. It isn’t to say Soucy is or isn’t one of those hidden players, but I think after 250 games and a great playoff run (so far), we have to wonder if him putting up fantastic results pretty much year after year is an example of that model-breaking behaviour.
It all sets up an interesting summer. Soucy is turning 29 and will be a free agent, as mentioned. He is coming off a deal signed in Minnesota that got him three years and $2.75M a season (perhaps influenced by the pending expansion draft but again, not here to debate that). He has had consistently great results year after year, and looks to be a tremendous two-way bottom-pair blue liner that is incoporating more types of offence. At the same time, he has never played 18 minutes a game in any season and his career-high is 21 points. He has never played top pair minutes, or even second pair minutes, and he has never played against top competition. At least not extensively. If someone asked me to make a decision today, I would say he has every look at being the next Jake Walman; a guy perennially under-utilized in a minimal role that can really break out somewhere else, even if it’s just 19-20 minutes a game. How much a team is willing to pay to find that out, and what Soucy looks like when/if it does, is the final chapter to this saga.
Cliffy's hell-hole 