Should we take old PER less seriously due compound of stat error?

Moderator: Doctor MJ

Borut
Ballboy
Posts: 24
And1: 0
Joined: Jul 16, 2013

Should we take old PER less seriously due compound of stat error? 

Post#1 » by Borut » Wed May 20, 2020 12:22 pm

We assume in the modern game there is not mistakes in the counting of stats.

However in the past there might have well be wrongly counted rebounds, block, field goals ...

As these are stats that are included in the PER aggregate, would this mean the error from them gets compounded in the PER. Should we therefore take PER stats from older seasons less seriously?
s0ciety
Ballboy
Posts: 11
And1: 20
Joined: Feb 26, 2020

Re: Should we take old PER less seriously due compound of stat error? 

Post#2 » by s0ciety » Thu May 21, 2020 2:17 pm

You shouldn't take PER seriously at all, 60's or modern game, don't matter..
Borut
Ballboy
Posts: 24
And1: 0
Joined: Jul 16, 2013

Re: Should we take old PER less seriously due compound of stat error? 

Post#3 » by Borut » Thu May 21, 2020 7:02 pm

s0ciety wrote:You shouldn't take PER seriously at all, 60's or modern game, don't matter..


What is your argumentation? I am aware some people have this view about this stat.
s0ciety
Ballboy
Posts: 11
And1: 20
Joined: Feb 26, 2020

Re: Should we take old PER less seriously due compound of stat error? 

Post#4 » by s0ciety » Fri May 22, 2020 7:26 pm

Borut wrote:
s0ciety wrote:You shouldn't take PER seriously at all, 60's or modern game, don't matter..


What is your argumentation? I am aware some people have this view about this stat.



The name itself is misleading "Player Efficiency Rating", the stat as nothing to do with efficiency.
All it really does really is spit out an aggregate number of your boxscore, so if you want a quick glance at the best sheet-stuffer I guess there is some use.
But really you learn nothing meaningful from PER, you're better of looking at PER100 stats and TS% if you want a better idea of a player offensive output.
I guess you could call it a Fantasy stat, the weight assigned to each components of the formula is totally arbitrary.

If you want to know more about how the stat is calculated: https://www.basketball-reference.com/about/per.html
Borut
Ballboy
Posts: 24
And1: 0
Joined: Jul 16, 2013

Re: Should we take old PER less seriously due compound of stat error? 

Post#5 » by Borut » Fri May 22, 2020 8:26 pm

s0ciety wrote:
Borut wrote:
s0ciety wrote:You shouldn't take PER seriously at all, 60's or modern game, don't matter..


What is your argumentation? I am aware some people have this view about this stat.



The name itself is misleading "Player Efficiency Rating", the stat as nothing to do with efficiency.
All it really does really is spit out an aggregate number of your boxscore, so if you want a quick glance at the best sheet-stuffer I guess there is some use.
But really you learn nothing meaningful from PER, you're better of looking at PER100 stats and TS% if you want a better idea of a player offensive output.
I guess you could call it a Fantasy stat, the weight assigned to each components of the formula is totally arbitrary.

If you want to know more about how the stat is calculated: https://www.basketball-reference.com/about/per.html


Furthermore, I remember I was once heard Steve Kerr say on a podcast PER is not a good stat.

On the other hand I was only using PER as an example to ask about compound of stat error?
User avatar
Dupp
RealGM
Posts: 103,856
And1: 54,808
Joined: Aug 16, 2009
Location: Billy Crystals lap
 

Re: Should we take old PER less seriously due compound of stat error? 

Post#6 » by Dupp » Sun May 24, 2020 3:42 am

They dont even get the stats 100% now days so id say you're absolutely correct about theres a high chance of mistakes the further you go back.
Neato wrote:I'd also like to take this opportunity to apologize to everyone for being a neato
Borut
Ballboy
Posts: 24
And1: 0
Joined: Jul 16, 2013

Re: Should we take old PER less seriously due compound of stat error? 

Post#7 » by Borut » Sun May 24, 2020 7:23 pm

Dupp wrote:They dont even get the stats 100% now days so id say you're absolutely correct about theres a high chance of mistakes the further you go back.


Do they still count the stats through the eye test?

Return to Statistical Analysis