All-Purpose Advanced Stats, Player Usage, etc. Thread

[jamers]

Wait, the most deployed center on the worst possession team in the league (with no other full-time pivots for the second year in a row) has seen a lot of shots against? Better buy him out

Coller is a tool with zero actual comprehension of advanced stats.

To be fair, there's basically zero context on Twitter. Will be curious to see how he frames that stat in his article.

 

[struckbyaparkedcar]

His last article involving Hodgson was unreadable. That's enough context for me.

 

[stokes84]

Wait, the most deployed center on the worst possession team in the league (with no other full-time pivots for the second year in a row) has seen a lot of shots against? Better buy him out

Coller is a tool with zero actual comprehension of advanced stats.

You act as though he is a victim of circumstance and not a major contributor to the problem.

 

[struckbyaparkedcar]

I readily invite anyone who believes that to name the players behind him who will spend every healthy game at top 9 C. Don't worry, I'll wait.

 

[stokes84]

I readily invite anyone who believes that to name the players behind him who will spend every healthy game at top 9 C. Don't worry, I'll wait.

What does being a center have to do with anything? We are talking about any forward.

 

[joshjull]

What does being a center have to do with anything? We are talking about any forward.

Because Hodgson is the only proven top 9 center on the roster at the moment. As such he plays in more situations than he should and it exacerbates his d-zone issues thus leading to the stat being mentioned.

EDIT: In the context of Coller's tweet and struck's response to it. Its hardly an earth shattering revelation that the center getting the most ice time on the worst puck possession team in the league is at the top of the charts in shots allowed 5 on 5. Add in his well known struggles in our zone and it further adds to the no **** nature of Coller's pronouncement.

 

[stokes84]

Because Hodgson is the only proven top 9 center on the roster at the moment. As such he plays in more situations than he should and it exacerbates his d-zone issues thus leading to the stat being mentioned.

Does he? It doesn't matter who is playing below him. 5v5, he is playing every 3rd or 4th shift, just like everyone else.

 

[stokes84]

Because Hodgson is the only proven top 9 center on the roster at the moment. As such he plays in more situations than he should and it exacerbates his d-zone issues thus leading to the stat being mentioned.

EDIT: In the context of Coller's tweet and struck's response to it. Its hardly an earth shattering revelation that the center getting the most ice time on the worst puck possession team in the league is at the top of the charts in shots allowed 5 on 5. Add in his well known struggles in our zone and it further adds to the no **** nature of Coller's pronouncement.

Of course not, but it doesn't excuse him from being part of the problem. One of the reasons we are that team is because he's not good enough.

 

[joshjull]

Does he? It doesn't matter who is playing below him. 5v5, he is playing every 3rd or 4th shift, just like everyone else.

Sigh........

Are we really going to argue again over situational use?

You also aren't grasping that I'm not defending his play in our end.

 

[stokes84]

Sigh........

Are we go to argue again over situational use?

You also aren't grasping that I'm not defending his play in our end.

Yes. It's a common theme that there is no situational rhyme or reason to what Rolston does.

 

[joshjull]

Yes. It's a common theme that there is no situational rhyme or reason to what Rolston does.

Now that I can't really argue with, at least for this season. But that furthers makes my point.

EDIT: No situational rhyme or reason like Grigs getting 47.8% o-zone starts , Ennis getting 47% but Girgs gets 53.2%

 

[tsujimoto74]

Just wanted to pop in and see if anyone besides me noticed that Grigorenko is winning over 50% of his draws. Only other guy besides Ott who's doing that is McCormick. Not really an advanced stat, I know, but there isn't a thread for real-time statistics.

 

[26CornerBlitz]

‏@coreypronman

Silly Chemmy. The interval starts at 1. RT @felixpotvin: Corsi isn't as hard as you think. Start with this equation

 

[gallagt01]

‏@coreypronman

Silly Chemmy. The interval starts at 1. RT @felixpotvin: Corsi isn't as hard as you think. Start with this equation

Thanks for sharing.

 

[kenfury]

NHL shot heatmaps are up.

http://www.sportingcharts.com/nhl/icetrack/

 

[jc17]

The default player that comes up is crosby. Compare him to Stafford and its not mystery why Drew's Shot% is so low. He's no crosby, but could benefit by crashing the net.

 

[Zip15]

Of Sabres forwards with at least 300 minutes of ice time, here are the leaders in pts/60 at 5-on-5 (this may surprise some):

(1) Drew Stafford: 1.44 pts/60
(2) Marcus Foligno: 1.36 pts/60
(3) Zemgus Girgensons: 1.31 pts/60
(4) Cody Hodgson: 1.20 pts/60
(5) Matt Moulson: 1.14 pts/60

http://stats.hockeyanalysis.com/rat...rs&minutes=300&disp=1&sort=HARTp&sortdir=DESC

NOTE: If you reduce it to a minimum of 100 minutes, Cody McCormick is the leader of the pack at 1.64 pts/60 at 5-on-5.

NOTE 2: Other than the guys listed, the Sabres have no other forwards above 1 pt/60 mins at 5-on-5. That is terrible.

 

[Zip15]

Of the seven skaters who've played at least 72:58 with Girgensons at 5-on-5, all have a higher Corsi For% while playing with him than they do when playing without him. Girgensons showing good possession indicators at an early age.

 

[tsujimoto74]

Of the seven skaters who've played at least 72:58 with Girgensons at 5-on-5, all have a higher Corsi For% while playing with him than they do when playing without him. Girgensons showing good possession indicators at an early age.

I think he's been maybe our best defensive forward this year, not that there's stiff competition there or anything. It definitely says something about the kid that at 19, as a rookie, coaches are trusting him to handle a lot of PK minutes, including as the lone forward out on 5-on-3s.

 

[struckbyaparkedcar]

He's generally been pretty sheltered though (he almost never saw Kronwall vs. the Wings, for example). Not that his stats aren't impressive for a first year NHLer, but some of it is the byproduct of getting the easiest defensive/possession matchups on the team. Another point to his credit, he's managed to maintain this despite Ennis-Leino-Ott turning into the offensive zone start line, a role Girgensons had been enjoying under Rolston.

Also, Hodgson is getting strength on strengthed a lot. Even more than under Ron. Weird life.

 

[Manny*]

If anyone's interested, I track zone entries for Atlantic division teams. I share all my game reports and findings on Twitter if you want to stay up to date on how the Sabres are doing.

I'm working on WZER (Weighted Zone Entry Rating), a possession metric like Corsi or Fenwick. It's an event differential with separate values assigned to entries with and without possession. Entries without possession are curbed by a constant factor defined by the ratio of shot attempts generated on average by each type of entry.

Some stuff on the Sabres:

Progression in weekly increments:

Comparison with the other Atlantic division teams:

Feedback and questions are more than welcome.

 

[Paxon]

If anyone's interested, I track zone entries for Atlantic division teams. I share all my game reports and findings on Twitter if you want to stay up to date on how the Sabres are doing.

I'm working on WZER (Weighted Zone Entry Rating), a possession metric like Corsi or Fenwick. It's an event differential with separate values assigned to entries with and without possession. Entries without possession are curbed by a constant factor defined by the ratio of shot attempts generated on average by each type of entry.

Some stuff on the Sabres:

Progression in weekly increments:

Comparison with the other Atlantic division teams:

Feedback and questions are more than welcome.

Thanks for sharing. Interesting work on the whole. It confirms what we all know about how bad the team is in the possession game overall and at gaining the zone with the puck.

 

[Sean McG]

Does anyone know of a site that has a glossary of the stats, but also examples of what is a good/average/bad team or player's numbers would look like? There are plenty that explain what they are or how they're calculated, and while I sorta understand some of the more popular statistics, when I look at a site like behindthenet, I don't really know what I'm looking at in terms of whether they're doing good, bad, having an outlier season, etc. Just looking for a basic explanation of how to look at the numbers.

I am pretty fascinated by advanced statistics in all sports, specifically in baseball where I understand them fairly well, and I'm trying to grasp the hockey stats. While, like I said, I don't completely understand every one, from what I've read the game certainly seems to be headed in the way of advanced statistics with the correlation some of these show with winning teams, and I certainly look at games differently now.

 

[Manny*]

Does anyone know of a site that has a glossary of the stats, but also examples of what is a good/average/bad team or player's numbers would look like? There are plenty that explain what they are or how they're calculated, and while I sorta understand some of the more popular statistics, when I look at a site like behindthenet, I don't really know what I'm looking at in terms of whether they're doing good, bad, having an outlier season, etc. Just looking for a basic explanation of how to look at the numbers.

I am pretty fascinated by advanced statistics in all sports, specifically in baseball where I understand them fairly well, and I'm trying to grasp the hockey stats. While, like I said, I don't completely understand every one, from what I've read the game certainly seems to be headed in the way of advanced statistics with the correlation some of these show with winning teams, and I certainly look at games differently now.

Context is important when employing analytics to evaluate teams or players. Much of the unwarranted criticism directed towards advanced statistics is born out of misinterpretation and misunderstanding. Statistics, by nature cannot lie. They can, however, mislead. It's a dangerous practice to look at such things in a vacuum. You're taking a snapshot of a team/player's performance, so be sure to include the proper backdrop.

If you're looking at Corsi/Fenwick% from a team's perspective, think of it as a close approximation for zone time. Buffalo, near 40%, is one of the very worst possession teams in the BTN era. If you're evaluating players, Rel Corsi% is a good place to start. That number will tell you how much better the team's possession game is with a player on the ice than without. Another thing you may want to look at is the quality of a player's teammates, or the opponents he faces. On Extra Skater, TotTm% QoT is a rating for quality of on-ice teammates, weighted for time played. TotTm% QoC is the same for on-ice competition. Here, a larger number represents higher quality. Acknowledging player deployment is important if you're to properly evaluate a player's statistics. A player who takes an unproportionately large number of defensive-zone face-offs will typically have a lower Corsi rating.

Hope that helps.

 

[26CornerBlitz]

Just heard Rob Vollman on WGR. His website is http://www.hockeyabstract.com/home