Gar and xGar difference

[Snowsii]

Saw Evolving Hockey's tweet for rookies, and lundells stats are interesting... (10 Gar / 1,5 xGar)

Is It good or bad?

Click to read more...

 

[hatterson]

Without seeing the specific tweet I can't confirm, but typically GAR is Goals Above Replacement a type of all-encompassing value of a players contributions on the ice. xGAR would be expected GAR which is basically saying "if you ran these situations over and over and over again, you'd expect to get this many goals on average from that"

Theoretically if your GAR is above your xGAR it means that you are getting more goals than expected out of the situations you've been involved in. That can indicate you're a skilled player and thus able to convert chances better than expected or it can indicate that you're getting lucky over the sample.

 

[Hockey Outsider]

Without seeing the specific tweet I can't confirm, but typically GAR is Goals Above Replacement a type of all-encompassing value of a players contributions on the ice. xGAR would be expected GAR which is basically saying "if you ran these situations over and over and over again, you'd expect to get this many goals on average from that"

Theoretically if your GAR is above your xGAR it means that you are getting more goals than expected out of the situations you've been involved in. That can indicate you're a skilled player and thus able to convert chances better than expected or it can indicate that you're getting lucky over the sample.

Good explanation.

Looking at the numbers - the PDO when he's on the ice (the sum of his team's shooting percentage and save percentage) is 107.0 at even strength. This is the 6th best result out of 6,996 player-seasons (minimum 40 games) going back to 2010. In other words, Lundell's team has been shooting and saving the puck at the 99.9th percentile when he's been on the ice. Obviously, that's unsustainable. The most likely result is, for the remainder of the season, as the team's performance regresses towards the mean, he'll be on the ice for fewer goals for and/or more goals against.

 

[Snowsii]

Good explanation.

Looking at the numbers - the PDO when he's on the ice (the sum of his team's shooting percentage and save percentage) is 107.0 at even strength. This is the 6th best result out of 6,996 player-seasons (minimum 40 games) going back to 2010. In other words, Lundell's team has been shooting and saving the puck at the 99.9th percentile when he's been on the ice. Obviously, that's unsustainable. The most likely result is, for the remainder of the season, as the team's performance regresses towards the mean, he'll be on the ice for fewer goals for and/or more goals against.

As today; Panthers have 21 players with PDO 100 or more.. (highest being 113)

 

[JaegerDice]

Basically, xGAR accounts for sh% and GAR doesn't.

The gap between the two can, reductively, be considered 'luck'.

If your GAR is above your xGAR, then you got a bit lucky. If it's below, you got a bit unlucky. If it's pretty close to even, you scored about what would be expected given the volume of shots and quality of chances you were creating.