Can we create an NHL equivalent to PER?

[BlueBull]

PER is Player Efficiency Rating and is mostly used in basketball. If you get a PER above 25 you might win the mvp, but if you get a PER above 30 you might be one of the best players in nba history. Naturally Jordan is on top with 27.9 with Lebron just behind at 27.6.
What does this have to do with NHL statistics?
Well, i want to try and make an nhl version of PER, but i don't know how to make one
Does anyone want to try and make a consistent and fair formula?

(ps. NBA league average PER is 15.00

 

[morehockeystats]

You should post the formula for PER.

 

[BlueBull]

Starting with the formula unadjusted per (this requires nba knowledge and a good iq, which i think i have both. anyone else can help if you want)
uPER = (1 / MP) *
[ 3P
+ (2/3) * AST
+ (2 - factor * (team_AST / team_FG)) * FG
+ (FT *0.5 * (1 + (1 - (team_AST / team_FG)) + (2/3) * (team_AST / team_FG)))
- VOP * TOV
- VOP * DRB% * (FGA - FG)
- VOP * 0.44 * (0.44 + (0.56 * DRB%)) * (FTA - FT)
+ VOP * (1 - DRB%) * (TRB - ORB)
+ VOP * DRB% * ORB
+ VOP * STL
+ VOP * DRB% * BLK
- PF * ((lg_FT / lg_PF) - 0.44 * (lg_FTA / lg_PF) * VOP) ]
Most is obvious but this is the simplifier to people not that smart yet i.e. high schoolers (also me)


factor = (2 / 3) - (0.5 * (lg_AST / lg_FG)) / (2 * (lg_FG / lg_FT))
VOP = lg_PTS / (lg_FGA - lg_ORB + lg_TOV + 0.44 * lg_FTA)
DRB% = (lg_TRB - lg_ORB) / lg_TRB
Now to adjust
pace

pace adjustment = lg_Pace / team_Pace

However pace adjustments are almost not needed so dont bother (adjustments range from *0.97 to *1.03, depending on the Teams ability and Strength of schedule)

You should post the formula for PER.

 

[BlueBull]

It's actually way more complicated then i thought because then you have to get everyone else adjusted PER (pace adjustment times unadjusted PER) then do an equation to adjust league average to 15

 

[BlueBull]

Well i have decided to create much more simple versions of PER
two versions
Offensive PER and Defensive PER.
I think i want to make a new thread to introduce it properly

 

[bambamcam4ever]

PER is a terrible metric for basketball, so why would you want to make a hockey equivalent of it?

 

[BlueBull]

So I can say that I made a new statistic for hockey.

 

[MVP of West Hollywd]

PER is about as useful as Pts is in hockey. Reasonably close, but there's a lot more to it to get real value, and doesn't capture defense well. Maybe including face-offs won and Corsi to goals and assists in a relatively basic way would do it.

 

[Captain Bowie]

I made up a new stat for hockey once. I took a players height in inches and subtracted it from their weight in pounds, I called it a players "Hweigth". For some reason it never really caught on.

Moral of the story, coming up with a new stat just for the sake of it is a bad idea and an incredibly waste time. Lucky for me mine took 10 seconds.

 

[lanky]

I made up a new stat for hockey once. I took a players height in inches and subtracted it from their weight in pounds, I called it a players "Hweigth". For some reason it never really caught on.

Moral of the story, coming up with a new stat just for the sake of it is a bad idea and an incredibly waste time. Lucky for me mine took 10 seconds.

I disagree. There's clearly interest in a player rating stat that factors in offensive and defensive impacts. I'd love to see 2017/18 GAR or WAR but nobody seems to be sharing their hard work for free online.