Marcus Camby first day news thread

I used a PER of 15 to support it, not to justify it, but I stand by it. What is the PER for an average NBA player? According to Hollinger, it is 15 for players, not for the starters.

http://sports.espn.go.com/nba/columns/story?columnist=hollinger_john&id=2850240

Please correct me if I am wrong. It's just my opinion!

 

Anyone have a link to the 95.5 interview? I can't seem to find it anywhere.

 

We have had this discussion before, on this very site, and I gave you a link to the way PER is calculated and to Hollinger's own definitions.

Anyway, here is Hollinger's initial definition:

He has since, many times in his columns, referred to PER of 15 as "average starter" which is what you would assume is what a pretty good player is. The NBA has 450 players and 150 starters.

Again, "average NBA player PER" is pretty meaningless because PER is averaged by minute, not per player, again, we go to Hollinger:

Since PER is an efficiency rating (not totals, so it has to have some kind of "balance" in it - and in PER it is balanced by minutes and pace) and is normalized to the league as a whole - it makes no sense to normalize something that measures efficiency to "number of players", it has to be normalized to something else that is efficiency based as well - thus the use of per minute).

Specifically from the article about calculating PER:

The normalization is for minute. With 15 set as the "average NBA minute played". (Notice that 15 is just an arbitrary number, it is just there to inter-season comparisons somewhat reasonable to make).

In other words - you do not average efficiency (how much work per hour a machine can do) with the number of machines in the factory - the 2 are not related.

Miles per gallon is an efficiency measure for work performed by car engines and the car they are installed it - what constitute it is not calculated by the number of miles driven or the average of all the cars on the road - these things are just not directly related.

Same with PER. It's a per-minute efficiency rating - and as such - the way you normalize it is "per minute", not "per player".

It only makes sense that the better players play more minutes, not all NBA players play the same minutes. So - by finding the "league efficiency" and normalizing for it - you normalize per minute, not per player.

 

Hollinger explained it before this season. I can't find "average starter" in the column.

http://sports.espn.go.com/nba/columns/story?columnist=hollinger_john&id=2850240

Unless you have a direct quote from Hollinger contradicting his quote in this article, I will have to remain unconvinced.

 

http://www.955thegame.com/pages/6031105.php

Go down to Feb 16th hour 2 of Wheels at Work - starts about a minute in.

 

The Google search PER average NBA starter 15 shows nothing support this "average starter" theory outside of your own post in this thread.

Again, no big deal, and I'm only offering my opinion of Travis Outlaw being an average NBA player.

 

He is telling you that league average (per season) is set to 15 in the very link you provided, not that the player average is 15. You yourself provided the direct link.

and the specific of how to calculated it, which is based on his book (and expanded to prior years in the link above) that I provided above shows you that the math specifically normalizes to 15 to match that.

It's math, there is no interpretation of the formula. It is pretty clear that this is exactly what he means.

 

This just tells you that there are tons of people that have no real concept of statistics.

 

By Hollinger's own ranking Blake has a PER of 11.42 which makes him a barely servicable backup.

So he gave a team that traded away the league's second best rebounder and their best defender for a barely servicable back up point guard and an injured small forward a B+?


He has officially lost his damn mind.

 

Where does he say that 15 is the average for an NBA starter? If that's the case, does that mean that Dante Cunningham and his 14.9 PER in scrub minutes should be an All-Star level starter?

I'm sorry, but I think you're the one misunderstanding Hollinger. "League average" is "league average". I have read nothing about him ever stating it applies on a "starter scale" as opposed to an "average" scale.

 

I just emailed Hollinger to find out the answer.

I'll post the reply when I get it!

 

No, it means that his efficiency is about the same as the per minute efficiency of all the players played - which roughly translates to the same as the production an average starter will give you. Of course, in small sample sizes - this is not a reliable measure - but I think that everyone is happy with Dante giving us "quality minutes" and the PER measure shows that it is pretty reasonable. Dante is giving you quality production per minute he is there. Of course, given that he is an inexperienced rookie and has a lot of learning to do - one expects that if he played more minutes and some of it as a starter - his production might dip. But, overall, it shows you that the gut feeling you get that Dante is a good player and plays well - is proven by this.

Average by what? The moment you talk about efficiency, it means that it is not "by player", but "by minute".

The formula does not have any division by "number of players" - thus it is not average per player.

 

As I just posted, I emailed Hollinger to find out what the '15' represents in terms of rating/comparing players. My question was more in depth, and I will post the reply when I receive it. I still don't see anything from him that states that 15 is an "average starter", so I'm curious to find out from him directly what the answer is.

 

His definition in print is 15 = "Pretty Good Player". You can take that to the bank.

http://www.alleyoop.com/prates.shtm

 

I'm not saying you are wrong. I'm just wondering what the number '15' represents as a qualitative value. Since I don't see an "average" player on that list, I anxiously await John's reply!

 

Given the player rates around 15, I'd call 15 the 4th or 5th option on offense, likely a minor starter or talented 6th man. Dante and Travis both fit this description pretty much spot on.

 

Here is a direct quote from Hollinger in one of his past chats:

This is from - http://espn.go.com/sportsnation/chat/_/id/22977

Notice that Hollinger does a good job of using the word "mean" (as mathematical average) - specifically he tells that "in any given year there are far more players below the mean PER than above it." - if there are more player below the mean PER (15) than above it - again, it is all "per minute", not "per player".

 

I found Camby's interview intersting when it came to Miller. Initially when discussing playing for the Blazers, he mentioned Roy and Aldrdige. He never mentioned Miller which I found odd since he is the vet starting PG for the Blazers and a player Camby personally has played with in Denver.

Camby was professional and when asked directly about Miller, he said some nice things about Miller. But in the end, when asked who he was looking forward to playing with (or something like that, I was listening in the car) he mentioned Roy and Aldrdige but not Miller.

I wonder if there is a history with those two because I was surprised how he never on his own mentioned Miller and only spoke positively about him when directly asked.

Probably not a big thing, but got a funny feeling (about Miller) when it came to his thoughts about this team.

 

I'm fairly sure andalusion is right. 15.0 PER is the average per-minute production of the league. But since the better players, the starters, play far more minutes than the reserves, the vast majority of that 15.0 PER average comes from their play. Thus 15.0 is not precisely "average starter," but it's closer to that than to "average player" and therefore 15.0 is a fine proxy for average starter.

If you're putting up a 15.0 PER, you're meeting a standard set mostly by other starters in the league. Also by the reserves, but their contribution to that standard is much less.