2019/2020 Advanced Stats / Analytics Tracker

[jc17]

By the numbers: The significance of the season's first five...

Summary:
Basically it is statistically unlikely for a good team to have a 5 game bad stretch of xGF% and unlikely for a bad team to have a good 5 game stretch. So while the number itself is not necessarily a sign of things to come, it can tell you if the team will be a positive xGF% team or negative one with some confidence

Of course, it’s important to remember that not all five-game samples are created equally – far from it – as there will be wild differences in venue, rest and opponent quality, but even with that caveat in mind, extreme results are still very uncommon for teams at opposite ends of the spectrum.

 

[Snippit]

We rank in the top 5 in HDCF% (2nd), SCF% (5th), xGF% (4th) and rank 7th in CF%.

However I think our ranking is influenced heavily by that Pens game where these stats were really heavily tilted in our favour.

 

[EichHart]

Keep posting in this thread. Don't really know too much about advanced stats but remember everyone saying last year during the 10 game winning streak that it was unsustainable and getting extremely lucky due to how bad they were. This year they look to have improved somewhat, but still getting lucky from some solid goaltending.

 

[Chainshot]

We rank in the top 5 in HDCF% (2nd), SCF% (5th), xGF% (4th) and rank 7th in CF%.

However I think our ranking is influenced heavily by that Pens game where these stats were really heavily tilted in our favour.

What happens if we toss that game out? Still looks pretty good, especially if we look at last year to this right?

 

[jc17]

What happens if we toss that game out? Still looks pretty good, especially if we look at last year to this right?

Without that game we're at 52% CF, 49% SC, 58% hdsc, and 51% xGF

 

[Irving Zisman]

By the numbers: The significance of the season's first five...

Summary:
Basically it is statistically unlikely for a good team to have a 5 game bad stretch of xGF% and unlikely for a bad team to have a good 5 game stretch. So while the number itself is not necessarily a sign of things to come, it can tell you if the team will be a positive xGF% team or negative one with some confidence

Very cool stuff. Thanks for sharing this!

 

[valet]

i was never one for advanced stats but over the summer i've become quite interested. its great to have all this data available. we're looking good so far. i'd expect at least some regression but i think we'll be in the wildcard hunt this year even with the limited sample size. and when we shed some of the dead weight off of the roster... man, who knows what will happen

 

[Chainshot]

Without that game we're at 52% CF, 49% SC, 58% hdsc, and 51% xGF

Which is all very good given where they were. Of note is the difference in HDSC.

 

[jc17]

Fancy stats took a big old hit last last night

 

[Chainshot]

 

[jc17]

 

[Fezzy126]

Pretty fascinating thread. I've noticed some abnormalities myself; On Friday's game we led in all major metrics through the 1st 2 periods yet our xGF% was still under 50%, just seemed odd. In any event, I'll be interested to see what the NHL has actually changed in their data.

 

[Aladyyn]

lmfao this league will literally f*** EVERYTHING up if given the chance

 

[Tatre]

Who thought changing that was a good idea?

 

[Gabrielor]

Not exactly ‘advanced’ stats, but through 6 games, here’s the 4 lines gf and ga counts at 5v5:

Eichel line gf-5 ga-3
Mojo line gf-5 ga-1
Log line. gf-3 ga-2
Mitts line gf-1 ga-0

 

[Chainshot]

 

[Chainshot]

Six games in, things haven't changed much since the first couple games.

 

[vcv]

What I'd like to do is plot how the Sabres move on these charts over the season. Not all of them, but a few of them. Update every 10 games or so. See how they are trending.

 

[Dingo44]

This is all sustainable. Ralph is golden.

 

[Fezzy126]

What I'd like to do is plot how the Sabres move on these charts over the season. Not all of them, but a few of them. Update every 10 games or so. See how they are trending.

Tierney does a 5 game rolling average of xG% that I found interesting to follow last season:

Expected Goal Share Over Time

 

[Dubi Doo]

This is all sustainable. Ralph is golden.

Still a small sample size, but it's hard not to be a bit excited be the results so far.

 

[Fezzy126]

NHL maybe fixed its data

 

[Fezzy126]

Defense pairings - last 4 games (I'm really enjoying this tool):

Takeaways:

• Dahlin struggling, despite very beneficial deployment; high o-zone starts and lower competition
• Miller looking good in brief stints away from Dahlin
• Risto & McCabe starting to fall back (SCF% still good though)
• I can't believe I'm saying this, but is it time to give Marco & Joker more responsibility?

 

[jc17]

Tierney does a 5 game rolling average of xG% that I found interesting to follow last season:

Expected Goal Share Over Time

Do you have last years? I was looking for it but couldn't find it. Made my own in a spreadsheet but not as fun to look at.

 

[Chainshot]

Defense pairings - last 4 games (I'm really enjoying this tool):

View attachment 265585

Takeaways:

• Dahlin struggling, despite very beneficial deployment; high o-zone starts and lower competition
• Miller looking good in brief stints away from Dahlin
• Risto & McCabe starting to fall back (SCF% still good though)
• I can't believe I'm saying this, but is it time to give Marco & Joker more responsibility?

Very much the eye matching the stats - 6-10 needs more ToI.