My first instinct is to think the Kraken are massive frauds.
But then I think a little longer and I come to the conclusion that publicly available analytics are good at fronting some things, but not everything. They can’t measure heart.
It makes sense if you watch the games. You pretty instantly realize the expected goals model doesn’t track well for this team (mostly on the against side) due to how their playing and inherent flags in both the data and expected goals model. The goalies are also being artificially lifted in xGA due to this. I can tell the author is, like Don at the Athletic, too much into interpreting data and not really into wondering if they’re missing something
How dare. They’re not wrong but like….. rude.
Would be interesting to isolate to see if there’s been an actual uptick in our statistical profile over the last 10 games or so. The article is using season long data (which makes sense from their perspective), but this team hasn’t just been winning lately – they actually *look* better.
When we were in the playoff mix just before Thanksgiving we still were kind of a brutal watch. Combine that with how bad we looked during the month of December and this statistical profile backs up what I was watching. Lately though, this team actually looks a lot more competitive and consistent. I wonder if the xGF% is still in the same neighborhood if you isolate this stretch.
Not that it means anything if the numbers are better or not. Teams can figure it out and lose it again multiple times in an 80+ game sample.
What I find interesting is that none of this is accidental. The Kraken are looking at the same analytics and are intentionally deploying their resources this way. The article gets at this — Stephenson’s analytics look awful but he’s somewhat of a sacrificial lamb. Playing Stephenson less or deploying him differently would not make us better.
A lot of the other comments make me question whether these stats are truly capturing what they say that they are. Is Grubauer truly playing out of his mind? Is our defense truly “subpar”?
Would love to see the numbers pre-Marchment trade and post-Marchment trade because when we do make the playoffs I think the story will be that the pre-Marchment trade Kraken did indeed get lucky and win more points than they deserved to, but the post-trade Kraken (especially with Kakko and McCann playing well) are actually a good hockey team.
This is my fear is that it is year 2 all over again. This is such a strange year in the Western Conference. Don’t let this down year for every team, make you think that this team is more than they are.
‘Nothing about the Seattle Kraken makes sense’
Lol on this part of the article
It’s more fun to win than to lose, but it’s honestly frustrating to watch this happen again. If we make the playoffs as a bubble team based on essentially non repeatable luck, it gives Ron Francis more validation and support to continue whatever it is that we are doing to be eternally a mid-bad team with no real path to contention on the horizon. We really should be playing young players almost exclusively and losing games until there is a path to contention.
You can’t argue with the facts, but that doesn’t mean it isn’t fun as hell to watch.
I dont see us being able to sustain this. We are still fourth from the bottom in scoring. Even if we make the playoffs, we wont survive the first round.
That being said- are we frauds? No, I dont think so. This team does play with heart, we just dont play very good.
I don’t think it makes sense to say that Stephenson is the statistical sacrificial lamb, and then go on to say that the Kraken look better without him. I mean, *someone* has to take defensive zone starts. Someone has to go up against top lines. It doesn’t have to be Stephenson as *much* as it has been, but these data don’t show us if Stephenson is uniquely bad in those situations or if he’s just being put in them more. Maybe we’d look even worse giving those minutes to another center, who knows.
I don’t generally mind posts like this because I think the curiosity is real: the Kraken are managing to win games despite having relatively low offensive power. How? It’s a genuinely interesting question.
But I think the data work is generally sloppy. It’s throwing up a few numbers without thinking about what they measure and pointing at them like, “look, see? This is lower/higher than this!” There’s non enough digging into the *why* of it all. Do the stats look better for a win vs a loss? What about over time? Against certain opponents? Or in certain situations? With certain players in or out of the roster? It ultimately makes for a story that’s not especially satisfying. Just sort of a shrug, and then luck and goaltending. Sure.
11 comments
My first instinct is to think the Kraken are massive frauds.
But then I think a little longer and I come to the conclusion that publicly available analytics are good at fronting some things, but not everything. They can’t measure heart.
It makes sense if you watch the games. You pretty instantly realize the expected goals model doesn’t track well for this team (mostly on the against side) due to how their playing and inherent flags in both the data and expected goals model. The goalies are also being artificially lifted in xGA due to this. I can tell the author is, like Don at the Athletic, too much into interpreting data and not really into wondering if they’re missing something
How dare. They’re not wrong but like….. rude.
Would be interesting to isolate to see if there’s been an actual uptick in our statistical profile over the last 10 games or so. The article is using season long data (which makes sense from their perspective), but this team hasn’t just been winning lately – they actually *look* better.
When we were in the playoff mix just before Thanksgiving we still were kind of a brutal watch. Combine that with how bad we looked during the month of December and this statistical profile backs up what I was watching. Lately though, this team actually looks a lot more competitive and consistent. I wonder if the xGF% is still in the same neighborhood if you isolate this stretch.
Not that it means anything if the numbers are better or not. Teams can figure it out and lose it again multiple times in an 80+ game sample.
What I find interesting is that none of this is accidental. The Kraken are looking at the same analytics and are intentionally deploying their resources this way. The article gets at this — Stephenson’s analytics look awful but he’s somewhat of a sacrificial lamb. Playing Stephenson less or deploying him differently would not make us better.
A lot of the other comments make me question whether these stats are truly capturing what they say that they are. Is Grubauer truly playing out of his mind? Is our defense truly “subpar”?
Would love to see the numbers pre-Marchment trade and post-Marchment trade because when we do make the playoffs I think the story will be that the pre-Marchment trade Kraken did indeed get lucky and win more points than they deserved to, but the post-trade Kraken (especially with Kakko and McCann playing well) are actually a good hockey team.
This is my fear is that it is year 2 all over again. This is such a strange year in the Western Conference. Don’t let this down year for every team, make you think that this team is more than they are.
‘Nothing about the Seattle Kraken makes sense’
Lol on this part of the article
It’s more fun to win than to lose, but it’s honestly frustrating to watch this happen again. If we make the playoffs as a bubble team based on essentially non repeatable luck, it gives Ron Francis more validation and support to continue whatever it is that we are doing to be eternally a mid-bad team with no real path to contention on the horizon. We really should be playing young players almost exclusively and losing games until there is a path to contention.
You can’t argue with the facts, but that doesn’t mean it isn’t fun as hell to watch.
I dont see us being able to sustain this. We are still fourth from the bottom in scoring. Even if we make the playoffs, we wont survive the first round.
That being said- are we frauds? No, I dont think so. This team does play with heart, we just dont play very good.
I don’t think it makes sense to say that Stephenson is the statistical sacrificial lamb, and then go on to say that the Kraken look better without him. I mean, *someone* has to take defensive zone starts. Someone has to go up against top lines. It doesn’t have to be Stephenson as *much* as it has been, but these data don’t show us if Stephenson is uniquely bad in those situations or if he’s just being put in them more. Maybe we’d look even worse giving those minutes to another center, who knows.
I don’t generally mind posts like this because I think the curiosity is real: the Kraken are managing to win games despite having relatively low offensive power. How? It’s a genuinely interesting question.
But I think the data work is generally sloppy. It’s throwing up a few numbers without thinking about what they measure and pointing at them like, “look, see? This is lower/higher than this!” There’s non enough digging into the *why* of it all. Do the stats look better for a win vs a loss? What about over time? Against certain opponents? Or in certain situations? With certain players in or out of the roster? It ultimately makes for a story that’s not especially satisfying. Just sort of a shrug, and then luck and goaltending. Sure.