PDA

View Full Version : Is the PER system accurate?



JoeyBoy718
01-04-2011, 11:28 PM
Player Efficiency Rating. They went back to 1978 and the 3 highest rated PERs of all time are 1. Jordan, 2. LeBron, 3. Shaq. It's hard to argue with that. Wilt Chamberlin is surprisingly lower than I expected (still in top 10, I think 7). Oscar Robertson is way lower than I expected (like 12 or 15 or something). Anyway, some of the top 5 in the league right now are Amare, LeBron, Durant. Hard to argue with that. Kobe is lower than guys like Zach Randolph and Monta Ellis. The top 2 in the league, however, might surprise you: Kevin Love is #1 with a PER of 27.94, and Pau Gasol is #2 with a PER of 26.26. It isn't hard to see with Love because his stats are ridiculous. Gasol is kind of surprising because his stats don't look nearly as impressive as some other guys with high PERs. Anyway, my question is: is the PER system the real deal? Jordan is #1 all time. That sounds pretty accurate. If you look at the all time list, none of the names are surprising. So, is Kevin Love the best player in the league (or close to it)? Is Pau Gasol the second best player in the league and the first best player on his team? Is Blake Griffin one of the top 6 players in the league and the best rookie of all time? How legit is the PER?

Doogolas
01-04-2011, 11:34 PM
No, it's not a very good stat. It's got efficiency in its name, yet it doesn't care about efficiency at all.

D Roses Bulls
01-04-2011, 11:35 PM
No the PER system is not accurate or even a real stat. I could write a paper on why PER is not accurate.

SteveNash
01-04-2011, 11:53 PM
2 things.

1. PER is horrible.

2. Kevin Love doesn't lead the NBA in PER, you're thinking of the NBA's efficiency stats which are completely different than PER.

kArSoN RyDaH
01-05-2011, 12:00 AM
No.

Chronz
01-05-2011, 12:01 AM
I havent checked out the leagues leaders in awhile but there is no ****ing way Love leads the league. To answer the Q its a tool that puts a players per minute production (skewed towards shot creation) in league context.

The reason Big O isnt as high as you would think is because it accounts for pace. Its been great at opening peoples eyes that per game #'s dont mean jack.


No, it's not a very good stat. It's got efficiency in its name, yet it doesn't care about efficiency at all.
It values efficiency, just not as much as usage. Its a very good stat tho


No the PER system is not accurate or even a real stat. I could write a paper on why PER is not accurate.
I highly doubt that considering how you started this post.

D Roses Bulls
01-05-2011, 12:03 AM
I havent checked out the leagues leaders in awhile but there is no ****ing way Love leads the league


It values efficiency, just not as much as usage. Its a very good stat tho



I highly doubt that

I could, I've read enough books, did my homework on the system. not hard to find faults with that joke of a system

asandhu23
01-05-2011, 12:03 AM
no.

Chronz
01-05-2011, 12:05 AM
I could, I've read enough books, did my homework on the system. not hard to find faults with that joke of a system
You read books on PER and its validity?

beasted86
01-05-2011, 12:06 AM
Yes, it is.

It's a fairly decent tool to see "who has the better stats" between two players.

That's the most important thing people overlook. That's all it does, tell us who has better stats for the most part. But some people really think it says who is the better player, when it doesn't really.

D Roses Bulls
01-05-2011, 12:07 AM
I havent checked out the leagues leaders in awhile but there is no ****ing way Love leads the league. To answer the Q its a tool that puts a players per minute production (skewed towards shot creation) in league context.

The reason Big O isnt as high as you would think is because it accounts for pace. Its been great at opening peoples eyes that per game #'s dont mean jack.


It values efficiency, just not as much as usage. Its a very good stat tho


I highly doubt that considering how you started this post.

I missed this part, but no it doesn't the system doesn't really care about whether you make or miss the shot, but about how many shots you throw up pretty much.

Chronz
01-05-2011, 12:09 AM
I missed this part, but no it doesn't the system doesn't really care about whether you make or miss the shot, but about how many shots you throw up pretty much.
Umm no

A player who doesnt make a shot gets negative credit a player who makes the shot gets credited.

D Roses Bulls
01-05-2011, 12:09 AM
You read books on PER and its validity?

not PER in particular, but there are some good books that mention PER in a chapter and talk about it's flaws.

beasted86
01-05-2011, 12:09 AM
I missed this part, but no it doesn't the system doesn't really care about whether you make or miss the shot, but about how many shots you throw up pretty much.

:facepalm:

He was spot on. It does value efficiency, but it values usage more. You seem to be saying it doesn't value efficiency at all which is way wrong.

JoeyBoy718
01-05-2011, 12:10 AM
2 things.

1. PER is horrible.

2. Kevin Love doesn't lead the NBA in PER, you're thinking of the NBA's efficiency stats which are completely different than PER.

You're right. Disregard my mention of PER. I'm talking about the efficiency rating you see on NBA.com. However, the rankings I mentioned that included Jordan, LeBron, Shaq, Wilt and Oscar were the PER rankings.

D Roses Bulls
01-05-2011, 12:10 AM
Umm no

A player who doesnt make a shot gets negative credit a player who makes the shot gets credited.

To quote Dave Berri, the author of The Wages of Wins:

"Hollinger argues that each two point field goal made is worth about 1.65 points. A three point field goal made is worth 2.65 points. A missed field goal, though, costs a team 0.72 points. Given these values, with a bit of math we can show that a player will break even on his two point field goal attempts if he hits on 30.4% of these shots. On three pointers the break-even point is 21.4%. If a player exceeds these thresholds, and virtually every NBA player does so with respect to two-point shots, the more he shoots the higher his value in PERs. So a player can be an inefficient scorer and simply inflate his value by taking a large number of shots."

D Roses Bulls
01-05-2011, 12:12 AM
:facepalm:

He was spot on. It does value efficiency, but it values usage more. You seem to be saying it doesn't value efficiency at all which is way wrong.

you know what they say about people that assume things.

ManRam
01-05-2011, 12:12 AM
I think it's a perfectly fine stat. More telling the a lot of things. Is it perfect? No...but no stat is. It has it flaws, but it does tell you a whole lot more than any simple counting stat...If you know what exactly it's telling you, and remain aware of that, then of course it's a good stat. Problem is, people don't usually know what exactly it does mean and what exactly it does tell.

That's all I'll say here.

Chronz
01-05-2011, 12:15 AM
To quote Dave Berri, the author of The Wages of Wins:

"Hollinger argues that each two point field goal made is worth about 1.65 points. A three point field goal made is worth 2.65 points. A missed field goal, though, costs a team 0.72 points. Given these values, with a bit of math we can show that a player will break even on his two point field goal attempts if he hits on 30.4% of these shots. On three pointers the break-even point is 21.4%. If a player exceeds these thresholds, and virtually every NBA player does so with respect to two-point shots, the more he shoots the higher his value in PERs. So a player can be an inefficient scorer and simply inflate his value by taking a large number of shots."
Dave Berri doesnt realize the importance of shot creation, he lives in a world where Dennis Rodman was the REAL reason the Bulls won. Youd be better off not quoting him.


not PER in particular, but there are some good books that mention PER in a chapter and talk about it's flaws.
Really which ones cuz from the sounds of it, seems like you just read a blog.


you know what they say about people that assume things.
Hes right tho, you missed the part where I laid down the context of how it values efficiency. It does in fact value efficiency.

D Roses Bulls
01-05-2011, 12:20 AM
Dave Berri doesnt realize the importance of shot creation, he lives in a world where Dennis Rodman was the REAL reason the Bulls won. Youd be better off not quoting him.


Really which ones cuz from the sounds of it, seems like you just read a blog.


Hes right tho, you missed the part where I laid down the context of how it values efficiency. It does in fact value efficiency.

actually david berri is right and he didn't say rodman was the reason why the bulls won, but he was a big reason why they kept winning especially in 97-98 when pippen was starting to break down along with ron harper with a weaker bench. even bill simmons who I know most of you love even mentions david berri all the time and how he in his opinion is one of the best basketball minds out there which I wouldn't go that far, but he is good.

Chronz
01-05-2011, 12:31 AM
actually david berri is right and he didn't say rodman was the reason why the bulls won, but he was a big reason why they kept winning especially in 97-98 when pippen was starting to break down along with ron harper with a weaker bench. even bill simmons who I know most of you love even mentions david berri all the time and how he in his opinion is one of the best basketball minds out there which I wouldn't go that far, but he is good.
Read Page 144" of The Wages of Wins.

Sorry bro but according to Berri and his methodology, rebounding is the most important aspect of basketball and Rodman was the MAIN reason the Bulls won. You should really look into his stats, Rodman does in fact score higher.

He doesnt value shot creation one bit. If its efficiency you value then look at WinShares. PER provides the contrast. Berri's metrics provide absolutely nothing and is by far the least credible in the community.

But when did BS ever mention him?

Chronz
01-05-2011, 12:36 AM
Every statistical formula is littered with flaws, these are the inherent nature of statistics, but there is something called a laugh test, and Berri's far and away doesnt pass it. Its one thing to have guys like Iverson ahead of Nash or whatever efficiency vs usage debate you want to bring up, its quite another when you have a guy like Jerome Williams, a career journeyman, as one of the leagues better players.

DODGERS&LAKERS
01-05-2011, 12:44 AM
Something I would like to know about PER is how much better is a PER of 28, over a PER of 26. Does one less turnover and one less foul a game give a player those 2 extra PER points?

JordansBulls
01-05-2011, 12:54 AM
http://www.prosportsdaily.com/forums/showthread.php?t=535699

D Roses Bulls
01-05-2011, 12:54 AM
Read Page 144" of The Wages of Wins.

Sorry bro but according to Berri and his methodology, rebounding is the most important aspect of basketball and Rodman was the MAIN reason the Bulls won. You should really look into his stats, Rodman does in fact score higher.

He doesnt value shot creation one bit. If its efficiency you value then look at WinShares. PER provides the contrast. Berri's metrics provide absolutely nothing and is by far the least credible in the community.

But when did BS ever mention him?

Well you have your opinion, and I have mine, but Hollinger freely admits that two of the defensive statistics it incorporates blocks and steals can produce a distorted picture of a player's value and that PER is not a reliable measure of a player's defensive acumen. I think PER is a terrible stat, but I'm not going to tare down peoples opinions. I just do not like it used when people are trying to compare players. I mean really I could sit here and rant and rave for a long time on what is wrong with the system, but it's really not going to be worth it cause people will still use it on here.

anyways BS mentioned him in two of chris sherridans pod cast? or is it chad ford who has the pod cast, anyways one of those two guys pod casts and in his mail bag over the summer. also a couple interviews with him which I would have to google cause they are at least a year old, but two interviews he has mentioned him.

JordansBulls
01-05-2011, 12:55 AM
Something I would like to know about PER is how much better is a PER of 28, over a PER of 26. Does one less turnover and one less foul a game give a player those 2 extra PER points?

Depends on minutes played and how well everyone else is playing in the game and in the league.

Chronz
01-05-2011, 12:56 AM
Something I would like to know about PER is how much better is a PER of 28, over a PER of 26. Does one less turnover and one less foul a game give a player those 2 extra PER points?
I dunno bout that but there are external factors, assuming your talking about players in the same season that one turnover could be more costly to a players PER depending the pace of his team and his individual minutes, if your talking about different eras that value of that turnover depends on league averages as a whole. Ive always hated that PER includes Fouls but its not that impactful.

Personally PER is at its best when you focus on its offensive components.

Chronz
01-05-2011, 01:03 AM
Well you have your opinion, and I have mine,
Not sure what your talking about but Im just saying, this is Berri's belief and methodology. With regards to PER, its my opinion that its very useful in providing context to a players statline. If you feel its garbage thats fine too, just dont hold it down for not measuring something its not meant to as your about to.


but Hollinger freely admits that two of the defensive statistics it incorporates blocks and steals can produce a distorted picture of a player's value and that PER is not a reliable measure of a player's defensive acumen.
What in my post made you assume I felt otherwise, I dont use PER to measure a players defensive value. Atleast not a players individual PER but Im interested in the PER of his counterpart to a degree.


I think PER is a terrible stat, but I'm not going to tare down peoples opinions. I just do not like it used when people are trying to compare players. I mean really I could sit here and rant and rave for a long time on what is wrong with the system, but it's really not going to be worth it cause people will still use it on here.

If PER is a terrible stat then all stats are terrible. Comparing players is something you do all the time, PER assists in that.

rhino17
01-05-2011, 01:04 AM
Nope, its a completely useless stat

Chronz
01-05-2011, 01:15 AM
Nope, its a completely useless stat
Your GM doesnt agree

Hawkeye15
01-05-2011, 01:26 AM
PER is a very usefull statistic to measure a players offensive worth, but its big fault is it favors shot creation, so chuckers get a small bump. But if factors in percentages, pace, and uses a projection system as a base, that I agree with.

Is PER the end all of stats? Um, no way, no stat is. But is it a telling stat? Yep, as good as any

zn23
01-05-2011, 01:30 AM
PER is useful and it is accurate. I prefer looking at PER rather than regular per game numbers.

Btw the player with the highest PER currently is Chris Paul, 2nd I believe is Dwade and then leBron...

rhino17
01-05-2011, 01:31 AM
Your GM doesnt agree

Still creating "facts" yourself I see

zn23
01-05-2011, 01:35 AM
Well you have your opinion, and I have mine, but Hollinger freely admits that two of the defensive statistics it incorporates blocks and steals can produce a distorted picture of a player's value and that PER is not a reliable measure of a player's defensive acumen. I think PER is a terrible stat, but I'm not going to tare down peoples opinions. I just do not like it used when people are trying to compare players. I mean really I could sit here and rant and rave for a long time on what is wrong with the system, but it's really not going to be worth it cause people will still use it on here.



well you've been getting consistently owned by Chronz in this thread, so you might wanna stop..

Hawkeye15
01-05-2011, 01:37 AM
Still creating "facts" yourself I see

Really? Morey has been involved with the advanced statistics community for a long time. You better believe he values PER

D Roses Bulls
01-05-2011, 01:42 AM
Not sure what your talking about but Im just saying, this is Berri's belief and methodology. With regards to PER, its my opinion that its very useful in providing context to a players statline. If you feel its garbage thats fine too, just dont hold it down for not measuring something its not meant to as your about to.


What in my post made you assume I felt otherwise, I dont use PER to measure a players defensive value. Atleast not a players individual PER but Im interested in the PER of his counterpart to a degree.


If PER is a terrible stat then all stats are terrible. Comparing players is something you do all the time, PER assists in that.

I'm just saying if you feel like PER is a good stat, thats fine. thats all I meant, you have your opinion and I don't use PER to measure a player or his counterpart, I think it's a waste of time. It's not a good system to use in evaluating talent and to be quite honest, you never made clear what you use PER to measure. the reason why i mentioned the blocks and steal was because of the berri stat you used.

D Roses Bulls
01-05-2011, 01:43 AM
well you've been getting consistently owned by Chronz in this thread, so you might wanna stop..

:confused: please explain...... cause honestly I think your just pulling **** out your ***

SteveNash
01-05-2011, 01:49 AM
You're right. Disregard my mention of PER. I'm talking about the efficiency rating you see on NBA.com. However, the rankings I mentioned that included Jordan, LeBron, Shaq, Wilt and Oscar were the PER rankings.

Wilt and Oscar's PER ratings aren't reliable because of the lack of stats back then. As long as they weren't historically bad in the turnover department, then there PER should be higher than the estimate.

Hawkeye15
01-05-2011, 01:53 AM
I'm just saying if you feel like PER is a good stat, thats fine. thats all I meant, you have your opinion and I don't use PER to measure a player or his counterpart, I think it's a waste of time. It's not a good system to use in evaluating talent and to be quite honest, you never made clear what you use PER to measure. the reason why i mentioned the blocks and steal was because of the berri stat you used.

what stat(s) do you use to evaluate a player?

Hellcrooner
01-05-2011, 02:00 AM
stats are fine but they cant measure the intangibles .
Watch the games not the scoreboards!!!!!!!!!!

Hawkeye15
01-05-2011, 02:04 AM
stats are fine but they cant measure the intangibles .
Watch the games not the scoreboards!!!!!!!!!!

ca'llate Hellcrooner! :)

D Roses Bulls
01-05-2011, 02:06 AM
what stat(s) do you use to evaluate a player?

Well like I have said countless times before, stats aren't everything and you can't only use stats to judge a player because stats dont also include the intangibles like leading ship, court vision, toughness, hustle and ect. I think you actually have to watch and study players which is mostly what I do and have done ever since I started watching basketball 20 years ago, but the stats i do use are just the basic one's but like i said, I dont usually use stats all that much when judging talent, and if anyone wants to challenge my knowledge of the game, go look at the archive history on psd in my profile it's self because things people use to call me crazy on were actually proven right and 99 percent of the stuff I said or the predictions I made were right on the nose.

Hawkeye15
01-05-2011, 03:00 AM
Well like I have said countless times before, stats aren't everything and you can't only use stats to judge a player because stats dont also include the intangibles like leading ship, court vision, toughness, hustle and ect. I think you actually have to watch and study players which is mostly what I do and have done ever since I started watching basketball 20 years ago, but the stats i do use are just the basic one's but like i said, I dont usually use stats all that much when judging talent, and if anyone wants to challenge my knowledge of the game, go look at the archive history on psd in my profile it's self because things people use to call me crazy on were actually proven right and 99 percent of the stuff I said or the predictions I made were right on the nose.

I can hold my own against you dawg.

And of course stats need to be taken in context. But there are plenty out there that paint a picture the eyes don't see/

Chronz
01-05-2011, 03:52 AM
Still creating "facts" yourself I see
LOL still dont know a damn thing about your own GM I see. Still havent read the book he wrote a foreword for I see.