When we last checked in on macro trends in the Seattle Kraken data, the Kraken were limping home after a calamitous five-game road trip. In a curious bit of historical mirror imagery, this next look comes after a heartening trip that included wins over the Carolina Hurricanes and New York Rangers. After salvaging a point in Tuesday’s home tilt against the reigning Stanley Cup champion Florida Panthers, we are at the 30-game mark in the season and due for another deep dive on team-level information and player-level results.
Big picture, the team’s season sits on the knife’s edge. While the Kraken have been accumulating points in a tough stretch, most playoff prognosticators still put Seattle’s playoff odds below 25 percent. Looking at the standings, it is becoming increasingly apparent that Seattle will need to climb over one or two quality teams if they are going to eke their way into the playoff field in a relatively deep Western Conference. Our own Blaiz Grubic has detailed the task ahead, but, put simply, every standings point is going to be “high leverage” from here on out. The time is now to deploy changes based on the lessons learned to date. Even after some adjustments, the playoffs are still a long shot, but any further delay in addressing problem areas could sink Seattle’s playoff push well before the March trade deadline.
Unless otherwise noted below, the data used in this article is from Natural Stat Trick. Let’s dive in.
Kraken scoring has stabilized but remains below average
Seattle’s scoring offense has found necessary depth over the last few weeks with Vince Dunn’s return to the lineup and with Shane Wright, Eeli Tolvanen, and Oliver Bjorkstrand all improving their production.
The picture isn’t quite as bleak as it seemed during our first update, but there is also little to suggest a playoff run as a 5-on-5 behemoth like two years ago. The shots went in on the road trip, but the team still struggled to create consistent shot quality inside. The chart below shows the five-game rolling average of 5-on-5 goals scored and expected goals scored (i.e. “shot quality”) by game number.
The actual and expected results were progressing and then saw a significant bump with Vince Dunn’s return. But 5-on-5 shot quality regressed again at the end of the trip and in the Florida game. Was that the product of tough competition or simply a regression to the team’s established “average” production this season?
The story is similar on the power play. The team has found its game to an extent; it no longer resides among the basement dwellers in terms of shot quality or production, but there is also little evidence it will prove to be a team strength. (Most notable to me on this chart, though, is the persistence of Edmonton’s struggles.)
Again, looking at how the team’s production has evolved over time, we see a spike in production over the last few games, but little in the underlying shot data to suggest it’s sustainable. (This conforms to my eye test too. Recently, the Kraken have gotten a handful of power-play goals off tips or deflections after stagnant efforts that don’t feel particularly heartening to me.)
Kraken 5-on-5 defense has come around, but the penalty kill is a problem
Over the last 15 games, the team has improved its consistency with defensive-zone coverages, while also benefiting from the return of Dunn.
Since our last update the 5-on-5 defense has been average or better at limiting shot quality and is trending toward average numbers overall. The team struggled to prevent goals over the last couple weeks, but at a macro level, that looks like a run of bad goaltending for the most part. With Joey Daccord playing extremely well against the Panthers on Tuesday, the team is surely hopeful the goaltending is back on track.
I don’t see similar room for optimism on the penalty kill though. The shot quality against numbers are quite poor overall. Indeed, it seems that strong goalie play is the only thing separating Seattle from a bottom-five shorthanded unit.
More troubling, though, is the utter absence of any successful run of play. In a rolling five-game sample over the course of the season, the unit has been below average at suppressing opposing shot quality almost the entire time.
Put simply, the ultra-passive box approach the team has been using so far is not working. In fact, the shot quality results have been getting worse recently. Perhaps other teams have realized the Kraken are unwilling or unable to disrupt possession.
The team used a similar box scheme at the end of last season, but the players were empowered to create significantly more puck pressure. This year’s unit might benefit from a shift back to that prior mindset.
On the personnel side, I’d prefer to see more of Matty Beniers (who brings natural defensive aggressiveness) and less Chandler Stephenson (who is passive to an extreme on the penalty kill).
Kraken skater data shows an average group with a few lagging behind
It’s no surprise to fans at this point that Kraken individual player deployment and production is relatively flat. While the team lacks high-end producers, it has depth that few teams can counter. This creates few true standouts in the data set. Beniers has emerged as the team’s No. 1 even-strength center both in terms of total ice time and quality of competition (more on that in a minute). Brandon Montour continues to be a workhorse pacing the team in 5-on-5 time on ice. From a scoring efficiency standpoint, Jared McCann and Oliver Bjorkstrand lead the way up front, with Vince Dunn rapidly catching up on the blue line.
Losing Jordan Eberle has been a blow to the team’s 5-on-5 offense. No player generated individual shot quality or goals more efficiently than Eberle when he was in the lineup.
Drawing back to examine team performance depending on who is on the ice, we start to get a sense of a group of players who have been struggling to tilt the ice in Seattle’s direction this season.
Using Evolving Hockey’s on-ice Quality of Competition metric, we find that these poor results aren’t necessarily due to being matched with difficult competition. Stephenson and Andre Burakovsky see only slightly more difficult than average matchups, while Borgen has been seeing relatively poor competition.
The team benched Burakovsky previously and saw immediate dividends, with a goal in the first period of his first game back. I believe that option should be on the table with Will Borgen too. It goes without saying the team is better with Borgen in the lineup over Josh Mahura, but Borgen has been fighting it this year. The team’s full health on the blue line gives the Kraken the luxury of getting Borgen “a breather” like it did for Burakovsky, Wright, Bjorkstrand, and Tye Kartye, to varying levels of success.
The situation with Stephenson, who is in Year 1 of a seven-year contract, is exponentially more complicated. In the short term, however, I think the team would be better off by scaling back his even-strength (and penalty kill) exposure a bit.
Turning to the power play, Stephenson has stood out as a key facilitator. On the other hand, I have found myself frustrated with him, Burakovsky, and Beniers to varying degrees for their propensity to lock in on passing when opportunities for a decent shot attempt are there.
On the penalty kill side, the Kraken don’t have a net positive performer in Evolving Hockey’s shorthanded play-driving metric. This brings me back to the scheme because there is no reason players like Jamie Oleksiak or Yanni Gourde shouldn’t be strong penalty killers. Only Brandon Tanev is performing near league average levels in Evolving Hockey’s model.
Is there anything else that you’d like to see? Or anything else that stands out in the data here? Drop us a note in the comments below or on X, formerly known as Twitter, or on BlueSky @deepseahockey or @sound_hockey.
Curtis Isacke
Curtis is a Sound Of Hockey contributor and member of the Kraken press corps. Curtis is an attorney by day, and he has read the NHL collective bargaining agreement and bylaws so you don’t have to. He can be found analyzing the Kraken, NHL Draft, and other hockey topics on Twitter and Bluesky @deepseahockey.
Pretty bleak observations from Curtis sadly but somewhat true. I am more optimistic that the team is working hard and has the talent to turn the season around.
Go Kraken!!!
With the penalty kill, it’s a small sample size: Seattle is the fifth in terms of fewest shorthand minutes per game and tied for seventh on fewest pp goals against. The difference in terms of the PK % being in the top 5 vs the current 19 is 1 goal.
Most of the teams ahead of them play a minute(or more) on the kill each game, so maybe if Seattle’s PK unit had more time to play together they’d bump up the %. Or, you know, maybe analysts could do a better job of putting the numbers into context, particularly since goals are what count, not percentages.
RB… I think you’re missing the point.
“Analysts” use “analytics” to support their “analysis”… it’s got nothing to do with context. I wish those “boneheads” who are running the team would just for once listen to the hobbyists.
Ahem, professional analyst here…and any good analyst will look at factors such as sample size and significance and put information into context, particularly when dealing with small numbers.
But you keep up with your hobby…
I was being sarcastic… I couldn’t agree with you more.
Apologies…it’s been a very long week 😉
Tis the season… for long weeks… and just to be clear, hobbyist was not a reference to you. I, personally, do not even rise to that level.
The ultra-passive penalty kill and the stagnant power play both come through tonight… even though they both looked terrible.
Again… being sarcastic.
I think regardless of expectations and percentages, the difference in play across the board has been dramatically improved this month and I think that’s kind of hard to miss.
A few thoughts:
1. I probably could have done a more in depth job setting up the penalty kill chart, but it’s a challenge to keep something like this short enough that it’s actually readable. Anyway, no percentages in the chart. The chart compares goals against with expected goals against.
2. Agree that the Kraken have done a good job of avoiding going shorthanded, and that’s half the battle. It has kept them middle of the pack so far in goals against (as shown in the chart, and as you note). But I’d say the low TOI underscores the severity of the problem rather than contradicting it. On a rate (per 60 basis), the Kraken penalty kill is 2nd worst in the league in expected goals against per sixty minutes. The reason I think the penalty kill is worth discussing is–what happens when the PK goaltending or the penalties regress toward the mean?
3. It sounds like from the passion of these responses on this specific issue, that you both think the penalty kill is good. Which is awesome. I think Daryl says he thinks it’s been “dramatically improved” recently. Would love to hear your observations on what’s been working well (of late or in general).
4. Thanks for reading!
Power play… I think the power play has been dramatically improved.
I also think the play in general at 5v5 has been much improved since the end of last month. The expected effort per 60 (xE/60) is generating 1.62 more expected chances per opportunity. I don’t think I’ve seen much difference on the penalty kill lately and the “actual” goals per 60 on a comparative basis seems to agree – in the league 17th against this month, 18th against all season.
Thank you Curtis for correcting the poor saps on here who are under the illusion that Stephenson doesn’t suck. He sucks. Maybe you sign a guy, like Montour, who will clearly help in the first few years, and worry about years 5-7 later; but you don’t sign a guy who already sucks in year 1 of the deal. Which is exactly what everyone who analyzed his play last year said would happen. My only hope was that the front office knew about some injury he had last year that had healed; plainly, that is not the case.
The Numbers don’t lie and some of them are not promising after more than a 3/8ths of a season gon and sittin in a position that will require better numbers from some disappointing player stats from some of the common suspects. Brokovsky may have scored a goal but his over all play is going in the wrong direction. Thank you Chris,
Numbers lie all the time…
On the power play…
“Again, looking at how the team’s production has evolved over time, we see a spike in production over the last few games, but little in the underlying shot data to suggest it’s sustainable. (This conforms to my eye test too. Recently, the Kraken have gotten a handful of power-play goals off tips or deflections after stagnant efforts that don’t feel particularly heartening to me.)
I don’t doubt the “underlying shot data” doesn’t support a difference in the power play… but I would propse a detached “eye test” of the recent power play would clearly recognize an obvious difference in the ability to maintain zone control and the “stangnat” observation is suspect.
The point being… grabbing a number and then agreeing with it seems to me one of the biggest failings – and commonalities – of “analytics”.
There is probably no person in all of hockey who I respect more than Ray Ferraro… give today’s Ray & Dregs a listen.
I’ll keep eye out for what you’re seeing in the power play, Daryl. Certainly hope you’re right that they’ve turned a corner with their play and it has just escaped me.
For clarity, though, about the purpose of these “data check in” articles: They’re just to get data out there in a digestible form. Any one of these charts/paragraphs would need a full article to give all necessary context and and analysis. You’re welcome to take or leave my parenthetical editorializing on the “eye test.” (Probably best to leave it!)
I think the parenthetical editorializing on the eye test shouldn’t actually be a take it or leave it proposition… it should absolutely be part of the analysis. This gets at one of the things I find so frustrating about a lot of the discourse around “analytics”. It seems to me most rational folks believe data should be used as a tool to inform opinions and conclusions. Unfortunately, I think too often the data IS the conclusion. To me the most important variable isn’t x or z… it’s why, and that’s where the “editorializing” comes in. While listening to the podcast this past week I got the impression I’m not the only one who was having a different “view” of the power play recently. Obviously, the “eye test” is subjective, but “stagnant” seemed – to me – hard to justify.
As “clear cut” as the numbers may appear, they can, I believe, be misleading… hence the famous Mark Twain quote about lies.
I don’t expect the Kraken will continue to score an actual goal a game on the advantage (that would be top of the league) and any “regression” may be seen as a statistical confirmation of the observations. Given what I’ve “observed”… through the end of November Seattle was scoring half a power play goal per game and a comparison to that benchmark is what I’ll be referencing to compare my “eye test” to the data.
I’d posit that a few medium-to-low danger shots have gone in and that is causing a biased view of the effectiveness of the unit and it’s future prospects. Would certainly agree with you that I’ve heard from others (not just you) that the power play is playing well right now. Vince Dunn is certainly an upgrade, so that is a tangible reason for optimism. However, I don’t see anything in the data or from what I’ve seen in their play to suggest we’re likely to see an above-average unit moving forward. That’s the point that was made in the article.
When I go back and actually watch the six PPGs from the last five games (and the Schwartz goal just seconds after expiration), I’m not seeing “lucky” goals.
If I had to posit how it could be the underlying numbers (xG) are not consistent with the increase in actual goal scoring I would begin what I’m seeing on the ice first. As John mentioned on the Patreon, I think they are doing a better job of establishing the zone. I also think they are maintaining the zone better and moving the puck more both latterly and high-to-low. This to me seems like it’s resulting in better chances. While they’ve had a couple low-danger goals lately, I’m not sure how “lucky” they were.
When I compare the numbers from December to the rest of the season the first significant difference I find apart from goals is a substantial increase in HDCF/60 from 18.18 to 30.42. I think they’ve been doing a much better job of getting the puck to shooters near the net – especially Bjorkstrand – and I think this is, to some degree, being facilitated by Beniers and Stephenson who I believe you’ve said should be shooting more rather than passing.
The same NST data window also shows the team getting more than half-again as many HDCF/60 and getting a similar ratio increase in xG/60 from 5.48 to 7.46… and none of these numbers capture the Schwartz goal. Additionally, all the improvement has come while spending just 3:56 on the advantage, a full minute fifty-two less per game. Without sifting through the numbers, I’m gonna guess you spend less time on the PP when you score. It could be just drawing less penalties… I didn’t bother to dig it out because I don’t think it really matters, just something of note.
The increase in actual goals outstrips the other numbers significantly, but rather chalking it up to “puck luck”, I think the data actually points to a change in play that is consistent with what I’ve observed and may help explain why the power play has actually improved.
Just my thoughts…
I’m curious if your use of a five game rolling average with the power play data is problematic, especially considering the recent decline in PP TOI being inverse to goals scored?
I actually didn’t realize that Borgen was having so many problems, but it does make sense.
Hopefully Evans will get better as he gets more experience. He seems like a keeper.
Contus’s guide on building a WhatsApp-like app is a goldmine for developers! It delves into technicalities like XMPP, Erlang, and MirrorFly SDKs, offering a comprehensive roadmap. Highly recommended for aspiring app creators! 🔧📱🚀