2023-24 Roster Thread #7: A shave and a haircut, two bits

[JojoTheWhale]

There’s a difference bt GF and xGF. And Charlie loves to use this stat by Evolving Hockey called RAPM, regularly adjusted plus/minus. It attempts to place individual responsibility of certain metrics on players. Basically, how did the individual player affect a certain metric. They run it on xG, GF, CF.

Frost didn’t grade out particularly well with RAPM, which is what Charlie hangs his hat on. But that same website has another metric, GAR and xGAR. Frost grade it out extremely well for xGAR in ES offensive impacts, and pretty well in ES defensive impacts.

He had a good season last year, and Charlie doesn’t want to admit it because he’s never liked Frost.

Sigh. Charlie should (and does) know why playmaking Forwards specifically are significantly better measured by GAR/xGAR than CF RAPM or xGF RAPM.

 

[iceman42]

Will he even play?

Already scratched

 

[CanadianFlyer88]

Will he even play?

They said they were going to call a forward up for the road trip. He may get in if Torts wants to punish one of the 12 regulars...

 

[Curufinwe]

Slightly Surprised it wasn’t Laczynski

Although I read he wasn’t taking faceoffs in a recent game and might be banged up

 

[Cody Webster]

They said they were going to call a forward up for the road trip. He may get in if Torts wants to punish one of the 12 regulars...

12 regulars? Since when does this team have 12 regulars? It's more like 9

 

[chadateit]

Sigh. Charlie should (and does) know why playmaking Forwards specifically are significantly better measured by GAR/xGAR than CF RAPM or xGF RAPM.

Could you please explain it to an analytics newby like myself?

 

[Beef Invictus]

Charlie feels the need to buy into the team's guiding philosophy:

 

[Magua]

Charlie feels the need to buy into the team's guiding philosophy:

I like how you think the Flyers are as evolved as the apes. I would trust Caesar running the Flyers over Tortorella and Jones.

 

[freakydallas13]

Atkinson to anywhere for whatever we can get should be a high priority for Danny at the TDL.

You're fooling yourself if you don't think this team is a buyer if they keep up in the standings.

 

[JojoTheWhale]

Could you please explain it to an analytics newby like myself?

Anytime at all.

Up thread a bit, you can see a bit of info Magua posted about how Defensemen don’t meaningfully impact 5v5 Shooting Percentages for their teams. He specified Defensemen because Forwards decidedly do. This is because they're often making higher value passes as well as completing passes directly to the individual shooter much more often. Someone like Patrick Kane obliterated normality for Shooting Impacts because he not only shot the puck well himself, he was HoF level at directly creating prime chances for others.

What RAPM actually is doesn't matter here. Think of it as normalized. The key is the first word in each group -- CF RAPM and xGF RAPM. No matter what you do with Corsi or Expected Goals, they're still just a simple counting of Shots and Shot Locations. Whether the net was wide open or not will never change a bit of data. GAR and xGAR are the opposite. They assign credit to the players on the ice for increases or decreases in On Ice Shooting Percentage.

Over a large enough sample size, you would expect a Forward who passes and sees the game well to make it more likely than average that the result of his pass is a Goal. So when we have that archetype and the major difference between metrics is that one credits a player for that difference and the other doesn't, choose the former.

The flip side of this is that I would choose CF RAPM and xGF RAPM for Defensemen before I would look at GAR or xGAR. Because we know we don't want to credit them for higher shooting percentages.

 

[Beef Invictus]

But, apes together strong

 

[CanadianFlyer88]

12 regulars? Since when does this team have 12 regulars? It's more like 9

Since Cates got himself injured.

 

[freakydallas13]

Anytime at all.

Up thread a bit, you can see a bit of info Magua posted about how Defensemen don’t meaningfully impact 5v5 Shooting Percentages for their teams. He specified Defensemen because Forwards decidedly do. This is because they're often making higher value passes as well as completing passes directly to the individual shooter much more often. Someone like Patrick Kane obliterated normality for Shooting Impacts because he not only shot the puck well himself, he was HoF level at directly creating prime chances for others.

What RAPM actually is doesn't matter here. Think of it as normalized. The key is the first word in each group -- CF RAPM and xGF RAPM. No matter what you do with Corsi or Expected Goals, they're still just a simple counting of Shots and Shot Locations. Whether the net was wide open or not will never change a bit of data. GAR and xGAR are the opposite. They assign credit to the players on the ice for increases or decreases in On Ice Shooting Percentage.

Over a large enough sample size, you would expect a Forward who passes and sees the game well to make it more likely than average that the result of his pass is a Goal. So when we have that archetype and the major difference between metrics is that one credits a player for that difference and the other doesn't, choose the former.

The flip side of this is that I would choose CF RAPM and xGF RAPM for Defensemen before I would look at GAR or xGAR. Because we know we don't want to credit them for higher shooting percentages.

I'm sure I speak for other advanced stats scrubs like myself when I say I always appreciate breakdowns like this from you, @VladDrag, and @Magua

 

[Magua]

I'm sure I speak for other advanced stats scrubs like myself when I say I always appreciate breakdowns like this from you, @VladDrag, and @Magua

I don’t do shit. I just try to make it sound like I know what I’m talking about.

 

[ponder719]

You're fooling yourself if you don't think this team is a buyer if they keep keep up in the standings.

That doesn't preclude the relocation of Atkinson to make roster space for that short-sighted acquisition.

 

[freakydallas13]

That doesn't preclude the relocation of Atkinson to make roster space for that short-sighted acquisition.

You know this, and I know this, but this team's brain trust does not know this.

 

[BernieParent]

Will he even play?

Unfortunately, the Flyers are heading west to play the Coyotes and Lycksell is only allowed to move north/south.

 

[Flyerfan4life]

You're fooling yourself if you don't think this team is a buyer if they keep up in the standings.

these idiots will wait till its way to late to sell..

its just how they work.

 

[chadateit]

Anytime at all.

Up thread a bit, you can see a bit of info Magua posted about how Defensemen don’t meaningfully impact 5v5 Shooting Percentages for their teams. He specified Defensemen because Forwards decidedly do. This is because they're often making higher value passes as well as completing passes directly to the individual shooter much more often. Someone like Patrick Kane obliterated normality for Shooting Impacts because he not only shot the puck well himself, he was HoF level at directly creating prime chances for others.

What RAPM actually is doesn't matter here. Think of it as normalized. The key is the first word in each group -- CF RAPM and xGF RAPM. No matter what you do with Corsi or Expected Goals, they're still just a simple counting of Shots and Shot Locations. Whether the net was wide open or not will never change a bit of data. GAR and xGAR are the opposite. They assign credit to the players on the ice for increases or decreases in On Ice Shooting Percentage.

Over a large enough sample size, you would expect a Forward who passes and sees the game well to make it more likely than average that the result of his pass is a Goal. So when we have that archetype and the major difference between metrics is that one credits a player for that difference and the other doesn't, choose the former.

The flip side of this is that I would choose CF RAPM and xGF RAPM for Defensemen before I would look at GAR or xGAR. Because we know we don't want to credit them for higher shooting percentages.

Thank you. This is great. A lot to chew on, but a good intro. Is there a good place you can recommend to see a lot of these definitions side by side? or are they all basically proprietary things that everyone does differently?

 

[tymed]

Just watching this Wild game and they threw down some pairings stats for pairs over 200 minutes. Seelsy/Walker ranked 4th best in the league for expected goals against, who'd a thunk?

 

[scumpup]

Interesting perspective from Zamula. Everything tracks that it seems that AHL exists for players to rot away in. It's an afterthought.

 

[CanadianFlyer88]

View attachment 778363

Interesting perspective from Zamula. Everything tracks that it seems that AHL exists for players to rot away in. It's an afterthought.

Depends on the team. Some teams connect the parent and farm club systems, some hire Ian Laperriere as their farm team's head coach.

 

[scumpup]

Depends on the team. Some teams connect the parent and farm club systems, some hire Ian Laperriere as their farm team's head coach.

You would think that in a global world of Hockey, you would fill your farm clubs (ECHL,AHL) to the BRIM with local and foreign prospects vying for shots at the big club, and to be traded to other clubs for assets and not washed up veterans who will never ever play in the NHL.

 

[Beef Invictus]

View attachment 778363

Interesting perspective from Zamula. Everything tracks that it seems that AHL exists for players to rot away in. It's an afterthought.

You can only learn to play in the NHL by playing in the NHL. This organization once understood that until Hextall broke development priorities.

 

[Flyerfan4life]

You can only learn to play in the NHL by playing in the NHL. This organization once understood that until Hextall broke development priorities.

remember when Ghost flat out said theres nothing to learn down there once youve been in the NHL.