Sean Tierney's GAR projection model

[GranteedEV]

Originally Posted by getbak

Which means what?

It basically factors in things like:

- Penalties taken
- Penalties drawn
- Shot locations + Passing data for expected goals for and against
- individuals' goal scoring history

etc for each individual player, with a "replacement" level player.

Here were Calgary's forwards' GAR last season, for instance:



A few areas where it might be flawed include:

- players grow (Tkachuk, Bennett could reduce their penalties, Monahan could and should improve his defensive play)
- historical sample sizes may be lacking for some of these younger players
- prospects join teams (Jankowski, Foo, etc)
- it weighs quality of teammate based on past season data, but sometimes teammates improve too

However, like I said earlier, it does have some demonstrable predictive value. We're not talking about rico stat here.

[looooob]

Originally Posted by GranteedEV

It basically factors in things like:

- Penalties taken
- Penalties drawn
- Shot locations + Passing data for expected goals for and against
- individuals' goal scoring history

etc for each individual player, with a "replacement" level player.

Here were Calgary's forwards' GAR last season, for instance:



A few areas where it might be flawed include:

- players grow (Tkachuk, Bennett could reduce their penalties, Monahan could and should improve his defensive play)
- historical sample sizes may be lacking for some of these younger players
- prospects join teams (Jankowski, Foo, etc)
- it weighs quality of teammate based on past season data, but sometimes teammates improve too

However, like I said earlier, it does have some demonstrable predictive value. We're not talking about rico stat here.

Re: penalties can it account for reversing the wideman effect 😁

[djsFlames]

The Jets suck and Vegas will be better than them.

[Enoch Root]

GAR as an individual stat is one thing, but trying to apply it to teams is ridiculous. Way too many variables, and way too much noise.

#### like this puts advanced stats back two years.

[jemjey]

GARbage

[starseed]

Within their division, Flames get to beat up on the predicted 3 worst teams in the league, and still place that low?

VGK VAN EDM and COL will give away a tonne of free points in the Pacific next year.

[edslunch]

Originally Posted by Enoch Root

GAR as an individual stat is one thing, but trying to apply it to teams is ridiculous. Way too many variables, and way too much noise.

#### like this puts advanced stats back two years.

It's like when people use PDO for individual players - some stats make sense for a team, others not.

[ForeverFlameFan]

Let the doubters doubt and the believers believe.

Loving the high Edmonton odds, they have quite a lot to prove.

[SebC]

Originally Posted by New Era

Another moron who thinks MS Excel can predict anything other than how little you get laid.

Nice anti-intellectualism you have there.

[SebC]

I believe this is the stat being used explained. https://hockey-graphs.com/2016/10/28...final-remarks/

[Textcritic]

Originally Posted by SebC

I believe this is the stat being used explained. https://hockey-graphs.com/2016/10/28...final-remarks/

I am still far from convinced based on this explanation. I agree with Enoch Root: there is for too much noise in this metric as an evaluation of team performance to be taken seriously.

[CaptainCrunch]

Originally Posted by starseed

Within their division, Flames get to beat up on the predicted 3 worst teams in the league, and still place that low?

VGK VAN EDM and COL will give away a tonne of free points in the Pacific next year.

[Lanny_McDonald]

Originally Posted by Textcritic

I am still far from convinced based on this explanation. I agree with Enoch Root: there is for too much noise in this metric as an evaluation of team performance to be taken seriously.

This ranks right up there with the ricardodw GRIT index.

[DJones]

Isn't GAR the stat that says Klefbom is the best defender in the league because he doesn't take penalties?

[GGG]

What really needs to happen with all of these stats is for error bars to be attached to them. Run them through previous seasons and measure the standard deviation of the results and post the 2SD bars along with the predictions.

I suspect in this case it would predict most teams as being 4th - 12th once error bars are accounted for

[CampbellsTransgressions]

And if it had the Flames in the top 5 and the Penguins in 25th, it would be the greatest model ever devised.

Yawn.

[Gaudfather]

GAR - General Avoidance of Rationality!

[Textcritic]

Originally Posted by CampbellsTransgressions

And if it had the Flames in the top 5 and the Penguins in 25th, it would be the greatest model ever devised.

Yawn.

Vroom! Vroom!