GDT: It's Beginning To Look A lot Like Hockey Season. STL vs DAL. Don't make me break out the trains.

[sbet1998]

Meh


Game 1 of 82

This is the correct amount of analysis needed after one game.

 

[BlueDream]

It's going to be a long season if this is what we can expect from our top forwards.

Well, fortunately, it would be absurd to assume that after one game.

 

[Reality Czech]

Screen cap of Krug/Faulk's advanced stats last night:

View attachment 752353

Now I sorted by xGF%, but no matter what stat you sort by, they are at the bottom of the list and it's ugly.

Sunny's numbers were pretty ass too. This is why the eye test is in many ways misleading. I get it with Sunny though, he's a fan favorite and he's back so we're all happy about that. But make no mistake, he was statistically pretty terrible last night.

Good thing games are determined by goals actually scored instead of xGF%

 

[BlueDream]

Screen cap of Krug/Faulk's advanced stats last night:

View attachment 752353

Now I sorted by xGF%, but no matter what stat you sort by, they are at the bottom of the list and it's ugly.

Sunny's numbers were pretty ass too. This is why the eye test is in many ways misleading. I get it with Sunny though, he's a fan favorite and he's back so we're all happy about that. But make no mistake, he was statistically pretty terrible last night.

I find it hard to place credibility on someone who makes an entire post analyzing the team with nothing but advanced stats. You didn’t post anything about what YOU saw with your own two eyes. You are just explaining who was good and bad based on what a stat sheet is telling you.

It’s my biggest pet peeve, especially after just one game.

 

[542365]

Good thing games are determined by goals actually scored instead of xGF%

But generating more xGF over a big enough sample size(say like 82 games for an arbitrary number off the top of my head) tends to lead to more “goals actually scored” and more “games actually won”. It’s not that hard to understand.

 

[Reality Czech]

But generating more xGF over a big enough sample size(say like 82 games for an arbitrary number off the top of my head) tends to lead to more “goals actually scored” and more “games actually won”. It’s not that hard to understand.

Sure but it doesn't mean that much in the end. As someone else mentioned, Vegas was around league average and they managed to win the Cup. Most good teams were towards the top of the table, but you also had teams like Pittsburgh and Calgary in the top 5 league wide. That tells me the stat should be taken with a grain of salt, like most advanced stats should be.

 

[542365]

Sure but it doesn't mean that much in the end. As someone else mentioned, Vegas was around league average and they managed to win the Cup. Most good teams were towards the top of the table, but you also had teams like Pittsburgh and Calgary in the top 5 league wide. That tells me the stat should be taken with a grain of salt, like most advanced stats should be.

They didn’t have their full team until the playoffs and they were still at 55% xGF%, well above league average. In the playoffs, with their full team against the other best teams in the league they were over 60% xGF% all situations.

 

[Majorityof1]

I find it hard to place credibility on someone who makes an entire post analyzing the team with nothing but advanced stats. You didn’t post anything about what YOU saw with your own two eyes. You are just explaining who was good and bad based on what a stat sheet is telling you.

It’s my biggest pet peeve, especially after just one game.

I get what your saying. But is that any worse than someone making a post just stating the conclusion of their eye test without any justification. "Krug and Faulk were great, I juse my eye test. So that is the only answer." I mean that tells us nothing. I'd much rather someone use only advanced stats than just give an eye-test based opinion, provide no support and treat it like its gospel.

Using only advanced stats gives a bunch of concrete information. I'd love it if the poster added some eye-test info to add context. But it says something about the underlying reasons they have their opinion. He did give his opinion of what he saw at the start of the post too, albeit briefly, and then clearly labeled it "by the numbers" after,

 

[Thallis]

Sure but it doesn't mean that much in the end. As someone else mentioned, Vegas was around league average and they managed to win the Cup. Most good teams were towards the top of the table, but you also had teams like Pittsburgh and Calgary in the top 5 league wide. That tells me the stat should be taken with a grain of salt, like most advanced stats should be.

They weren't quite league average, they were definitely better even on public xGF models. My previous comment was not to say we can just toss public xGF. There still has not been a Stanley cup winner and only ~4 conference finalists under 50% public xGF in the years it has been tracked. It was to say the public model is clearly lacking in the quality calculation compared to the ClearSight and Sportlogiq. Calgary and Carolina were the 2 big outliers that point to that. Public xGF overemphasizes volume, which does matter too, but most NHL teams have begun preferring quality.