2017-18 stats and underlying metrics thread [Mod: updated season]

Status
Not open for further replies.

Ippenator

Registered User
Jan 6, 2016
5,667
4,435
Espoo
Do you realize you're arguing with someone that studies stats for a living and is paid for doing so by NHL teams?

On a side note I'd love to sit with you at a poker table.
Yes I know that exactly. Proves to me that he knows how to talk his way through. Noticed it many times here too. Nothing bad about that of course. It should make you successful in anything that you do with good knowledge on the subject. Still doesn’t make all that stats stuff really so much needed as some people want to believe, or want others to believe.

Your poker insinuation is honestly weird. What does bluntly speaking in this kind of a forum have anything to do with how people would play poker? I will anyway make it easier for you. I don’t play card games at all, and I also don’t do any other form of gambling at all. Well, unless investing to the stock market is not included. Gambling and card games just are not at all my cup of tea. I will rather play a good RTS -game on my computer if I want some nice strategic challenge.
 
Last edited:

winnipegger

Registered User
Dec 17, 2013
8,092
6,275
People buy tabloid magazines and spinner toys like crazy too. Doesn’t make either of those things really needed or important. Paid money is mostly nowadays more the proof of very well made marketing and sales than how good or needed the products really are. Sure part of it comes from the product often too, but still marketing and sales can be really amazing in selling whatever mumbojumbo nowadays.

Stats guys seems to be pretty good sales people and of course they have pretty much the ”what if”-factor that practically any NHL coach can’t neglect while they want to utilize even the smallest benefits they can get in the battle for the place in the sun

Do you honestly think that NHL owners would give money to statisticians just out of good will? Or that they have been taken in by savvy marketing and lies? NHL owners are in the business of winning and getting advantages over their opponents. Having more information than your competitor = advantage.

If a co-worker came up to me and said "put down that calculator I have a good eye" I would not listen to said person.
 
  • Like
Reactions: Mathmew Purrrr Oh

Ippenator

Registered User
Jan 6, 2016
5,667
4,435
Espoo
Do you honestly think that NHL owners would give money to statisticians just out of good will? Or that they have been taken in by savvy marketing and lies? NHL owners are in the business of winning and getting advantages over their opponents. Having more information than your competitor = advantage.

If a co-worker came up to me and said "put down that calculator I have a good eye" I would not listen to said person.
No, but it’s the ”what if”-factor that makes them grasp even some straws. I can’t claim that stats can’t be at least somehow used to the benefit of coaches. But the meaning is still exaggerated way over what it really should be. And that is because of some very talented marketing by the stats companies.
 

garret9

AKA#VitoCorrelationi
Mar 31, 2012
21,738
4,380
Vancouver
www.hockey-graphs.com
People buy tabloid magazines and spinner toys like crazy too. Doesn’t make either of those things really needed or important. Paid money is mostly nowadays more the proof of very well made marketing and sales than how good or needed the products really are. Sure part of it comes from the product often too, but still marketing and sales can be really amazing in selling whatever mumbojumbo nowadays.

Stats guys seems to be pretty good sales people and of course they have pretty much the ”what if”-factor that practically any NHL coach can’t neglect while they want to utilize even the smallest benefits they can get in the battle for the place in the sun

Stats people are not sales people at all... and that definitely shows you haven't met them... because the ability to market and sell is something they severely lack.

What they have done is stood the test of time, had their work constantly scrutinized, and come out on top.

Difference. :)
 
  • Like
Reactions: Mathmew Purrrr Oh

garret9

AKA#VitoCorrelationi
Mar 31, 2012
21,738
4,380
Vancouver
www.hockey-graphs.com
No, but it’s the ”what if”-factor that makes them grasp even some straws. I can’t claim that stats can’t be at least somehow used to the benefit of coaches. But the meaning is still exaggerated way over what it really should be. And that is because of some very talented marketing by the stats companies.

There is no "BIG CORSI" thing going on.

Stats are not exaggerating their importance, because the whole results with measuring and testing the meaningfulness of these things are is what made them grow in popularity.

They have been constantly retested by multiple people. The results were what they are. They are as meaningful as they have been tested to be.

For fun, 4 of the past and current Hockey-Graph writers literately got into hockey stats to try and prove the "Corsi nerds" wrong, and found out that the nerds were not...
 
  • Like
Reactions: Jack722

garret9

AKA#VitoCorrelationi
Mar 31, 2012
21,738
4,380
Vancouver
www.hockey-graphs.com
The whole hockey analytics thing grew out of people constantly testing and retesting other peoples work.

You can see multiple blog posts about people thinking things must be different and then studying it and finding out it wasn't the case.

It grew out of people ACTUALLY scrutinizing other peoples analytical work appropriately... as opposed to waiving their hands in the air and making some pretty crazy accusations.

This whole conspiracy angle is just asinine.
 
Last edited:

Mathmew Purrrr Oh

#meowmeowmeowmeow
Apr 18, 2013
5,660
145
meow
The whole hockey analytics thing grew out of people constantly testing and retesting other peoples work.

You can see multiple blog posts about people thinking things must be different and then studying it and finding out it wasn't the case.

It grew out of people ACTUALLY scrutinizing other peoples analytical work appropriately... as opposed to waiving their hands in the air and making some pretty crazy accusations.

This whole conspiracy angle is just asinine.


You sold me! I'll take 2.
 
  • Like
Reactions: garret9 and ps241

Whileee

Registered User
May 29, 2010
46,061
33,043
One of my concerns with stats in hockey is that as it is becoming more and more commercialized, the incentives for publishing methods and results in peer-reviewed publications are diminishing. It's becoming more and more difficult to know whether a particular analysis is sound or reproducible because the underlying models and assumptions aren't fully disclosed. Also, we are going to get a lot of flim-flam with profits involved. As an example, Chayka's Stathletes claims that their methodology offers more than 100x the "statistical resolution" of other methods. Should we just take that at face value?

I'm all for greater use of statistical methods, but I think that now we are going to get less and less open-sourced science.
 
  • Like
Reactions: Halberdier

winnipegger

Registered User
Dec 17, 2013
8,092
6,275
The game of checkers can be "solved", i.e. you can win every time because you have 100% information. I can't see the game of hockey ever being solved in the same way so I think we're good to keep watching.
 

YWGinYYZ

Registered User
Jul 3, 2011
28,480
7,117
Toronto
The game of checkers can be "solved", i.e. you can win every time because you have 100% information. I can't see the game of hockey ever being solved in the same way so I think we're good to keep watching.

I don't think anyone has ever said you need to stop watching. Stats are supplemental - people are completely free to ignore them if they'd like.
 

Sperss1997

Registered User
Oct 29, 2015
636
312
Aarhus
1) Desjardins pointed out that even if NHL games were decided by a 50/50 coin toss, there would be some teams better than others in the standings. No team controls their destiny as hockey is a tug-a-war. Probability is just a weighted coin toss (albeit moving weighting) coin toss. So, we expect some percentage of the standings to be pushed by the coin flip as well.

A English soccer statician ran 10,000 simulations over a full Premier League season. I do not have a link, sorry. All 10,000 were made with 20 teams playing 38 games and all games were 50/50.
On average there would be 30 points between the champion and the bottom placed team each season and the table "lied" about 6 points on average for each team.
Hey, found the article - here is the link Does the league table lie?
 

JetsUK

Registered User
Oct 1, 2015
6,723
14,075
A English soccer statician ran 10,000 simulations over a full Premier League season. I do not have a link, sorry. All 10,000 were made with 20 teams playing 38 games and all games were 50/50.
On average there would be 30 points between the champion and the bottom placed team each season and the table "lied" about 6 points on average for each team.
Hey, found the article - here is the link Does the league table lie?

That was a fun read, thanks. I may run a version of this at my lab later on this year when we have some spare capacity. If I do, and it works, I’ll try and report back.
 

garret9

AKA#VitoCorrelationi
Mar 31, 2012
21,738
4,380
Vancouver
www.hockey-graphs.com
One of my concerns with stats in hockey is that as it is becoming more and more commercialized, the incentives for publishing methods and results in peer-reviewed publications are diminishing. It's becoming more and more difficult to know whether a particular analysis is sound or reproducible because the underlying models and assumptions aren't fully disclosed. Also, we are going to get a lot of flim-flam with profits involved. As an example, Chayka's Stathletes claims that their methodology offers more than 100x the "statistical resolution" of other methods. Should we just take that at face value?

I'm all for greater use of statistical methods, but I think that now we are going to get less and less open-sourced science.

The good news is that type of stuff isn't the stats/research we are currently talking about, and rarely is the stats/research we usually talk about on this forum.

The closest would be expected goal and WAR models, which at least have their methodology given, and people have recreated very similar outcomes with similar results and testing performance.

The blackbox stuff is more of a problem for me than the forum posters here hahah. I see coaches and GMs (not just NHL level) falling for blackbox stuff or simply not understanding some statistical foundational stuff and having that being taken advantage of.

A English soccer statician ran 10,000 simulations over a full Premier League season. I do not have a link, sorry. All 10,000 were made with 20 teams playing 38 games and all games were 50/50.
On average there would be 30 points between the champion and the bottom placed team each season and the table "lied" about 6 points on average for each team.
Hey, found the article - here is the link Does the league table lie?

Hahah ya that's a similar experiment to Desjardins.
 

PhilJets

Winnipeg is Good
Jun 24, 2012
10,297
7,783
Somewhere nice
Like stats same as eye test.
It just sometimes not Garrett. Mostly in the main board uses it as be all end all data to figure out a player or a team.
 

Aavco Cup

"I can make you cry in this room"
Sep 5, 2013
37,630
10,440
Athletic piece on Hendricks (paywall of course)

How the Jets can get the most out of Matt Hendricks

IMG_0632.png


There are a lot of stories here. First, most players do better in terms of creating shots without Hendricks than he does without them. In terms of shots allowed, it’s 50/50 – a few players give up fewer with Hendricks than he does without them but a few players also give up more.
Do you know what I think the most striking thing about this chart is?
It’s that every single “player with 15” box is in the positive half of the chart.
If Hendricks is the terrible drag on possession that he’s supposed to be, it’s not showing in these numbers.
 

Whileee

Registered User
May 29, 2010
46,061
33,043

Myers and Kulikov have struggled with extra icetime and tougher match-ups. They actually had good shot metrics before Trouba and Enstrom went down.

Buff and Enstrom have been terrific this season.

Morrissey and Trouba have held their own against the toughest match-ups.
 

mcpw

WPG
Jan 13, 2015
10,024
2,072
5v5 sh% leaders

2016-17 Laine 12.6%
2017-18 Matthews 13.1%

i can't believe the jets drafted laine over matthews
 
Status
Not open for further replies.

Ad

Upcoming events

Ad

Ad

-->