I'm confused how on a forum where advanced stats are used in basically every thread, that it's a Johannes Kinnvall thread where people are all of a sudden confused by advanced stats and don't know what they mean.
And yes, Colorado and Carolina are two teams specifically that have really took a heavy-analytics approach the last few seasons, becoming top teams in the process.
Look, I'm not criticizing you posting stuff and getting excited. In fact, I really appreciate it. But logically speaking, stats have flaws at times. Forgive the collective eyebrow raising when the stats saying he's near league best. We want to be excited, but also cautiously optimistic.
That's why when you posted it, a few eyebrows are raised. I honestly think most of us posting here want to be as excited as you, but we aren't willing to do it based on a single stat line. We want more data to prove we should be excited. Like, is Kinnvall the equivalent of us picking up a free Myers or Zamula calibre talent or something? Or is this a lucky and hot streak? Then I see another page that has Kinnvall's NHLe at like 67 vs Valimaki's at 39. This further confuses me. He's near SHL league top and likely to basically replicate most of what he's doing is what I get if I combine the two.
I understand the context if they are reasonable with reasonable comparable to match the eye test, or long term sustainability of the stats (ie: If he had similar stat lines last season). Didn't Burke say something along the lines of, "Stats are great for illuminating the dark alley, but not as a lamp post to support you from falling over." or something along those lines? Didn't we go through something similar with Gulutzan where we had great advanced stats, but eye test showed something really wonky and rickety and seemingly high risk against the Flames?
I'm an auditor by trade and one of the key underlying things we do is that we derive the same conclusion using different paths and methods as well as sources. That's what many try to do with combining advanced stats and eye test. If the stats say good, but eye test says bad, there's something wrong, or you have to keep looking to see what you're missing. The part that drives many bonkers is that many rely on the data and refuse to keep using alternate methods to see what is missed. This is what is meant about the story of the blind men describing the elephant. All of them are correct, yet all of them are wrong at the same time.
Stats are great for quick glances, but they don't take certain aspects into account. What many of us are wondering is what Kinvall is doing to be league top and whether he is legitimately good enough to be league top, or are we dealing with some weird out of context stat. Someone leaned on heavily to shoulder heavy burdens on a bad team is going to get wrecked advanced stats wise. A sheltered player on a great team will have good stats, but unlikely to replicate it in different circumstances.
I've always believe that with many advanced stats, it should be two line. Base stat, then stat comparison to team. If you can get the average team stat line of the one you posted for Kinnvall and he's still wrecking his next closest teammates, then I'll get more excited even without eye test (based on the understanding that the team he plays on is near league bottom and it's kinda hard to maintain good stats on a team that is often losing).
So what you’re saying is that you know nothing about advanced stats either, because those stats are used quite commonly.
Or this is another advanced stats whoosh moment.
Or, this is another situation like +/- and corsi where the real story isn't told (ie: Quality of competition etc.). Not to mention, we had another poster reveal that the Flames track significantly different stats that the average fan isn't privy to. One that for instance shows Brodie as "meh" vs Andersson as "damn Daniel!".
We literally have stats that say Kinvall is basically the top of the league. What we are trying to figure out is if the league sucks, if he's actually good, does he have a great partner he's leeching off of etc.
What absolutely drives many who actually try to understand or question stats up the wall is that those who tout them absolutely go nuts if you try and audit the methods or audit the validity of the stat. A good audit shows that several paths can be used to come to the same conclusion. If there are mismatches in the information or there is an odd logic to the data, you have to investigate. Otherwise, those are the holes by which someone may use to commit fraud (or there is an error that will just keep rolling forward indefinitely because no one questioned it).
It's not that we just go brain dead on stats, it's that Kinnvall's stats say he's league top, there's a recent article on Reddit that shows Kinnvall's NHLe is 67 vs Valimaki is 39 which means that Kinnvall is more likely to replicate what he's doing now in the NHL than Valimaki is? But again, what is he actually doing in the SHL that is giving him dmen wrecking advanced stats? Are they outliers in stats or are they legit data points?
