Per

[DunkinDerozan]

I know the acronym stands for Player Efficency Rating but not much else. I know a long and crazy formula is used to formulate the number. Is it even possible to explain in layman terms what the final number represents? I find it hard to beleive this stat is meaningful when you can't understand its usefulness. Anyone care to explain how the components of PER are used / weighted and how looking at the final number helps one evaluate a players game.

[funkie]

PER really stands for i don't know crap about basketball, but i know a lot about stats and i might be able to score a chick if i'm a sports analyst.

[Brandon]

PER is basically all of the good things a player does - all of the bad things he does. It is a measure of production on the court, so a player who does things that don't show up in stats isn't going to have a good PER. We don't have a stat that tells us about great screen-setters at the moment, so we can't tell who sets great screens and who doesn't. That ability will not show up as clearly in a stat like PER (it might result in better team scoring efficiency while that player is on the court).

As a general PER guideline, i think ~15 is average, ~20 is all-star level, and ~25 is mvp level.

I think PER is a useful guide in helping to judge how productive a player is on the court. It does seem to predict all-star/mvp status reasonably well. If you look at the list of the top career PER ratings, it's a who's-who of the very greatest players in NBA history. It also seems to predict team wins, because putting three players with a PER greater than 20 together seems to lead to deep playoff runs/championships (Boston, Lakers, Miami), while teams without such players tend to do poorly. Miami, for example, has 3 players in the top 50 career PER in NBA history -- #42 (Bosh), #6 (Wade), and #2 (LBJ). Miami has little or nothing else on their team at this time. But they don't seem to need anything else, do they?

There are other advanced stats, such as Dave Berri's WP48, 82games.com's Simple Rating, Dean Oliver's Win Shares, Adjusted Plus/Minus. There is no one advanced stat that tells us everything we need to know about a player. I think Berri's stat is flawed because he thinks shots are taken, not created (wrong, great shots are created by great shot makers). Berri has badmouthed PER because he doesn't think it predicts wins as well as WP48. Well, by Berri's stat, Kevin Love is the most productive player in the league, and his team doesn't seem to be winning a lot of ball games (Berri's stat says Love's teammates are so awful they're subtracting a lot of wins from the team).

Career PER: http://www.basketball-reference.com/...er_career.html
Adjusted Plus/Minus: http://www.82games.com/ilardi1.htm
K-Love produces more wins than Lebron or Superman: http://dberri.wordpress.com/2011/04/...network-style/

[WhatWhat]

(According to WP48, Bargnani is by far the least productive player in the league. Lol. They also believe that Hump should have been an All-Star. I think they said that Rodman>Jordan, too.)

Basically, what Brandon said. But you should use multiple stats and metrics rather than just one, and see if it matches your eyes.

[Nine New Faces]

PER's biggest weakness is that it rewards volume shooting by not taking into account efficiency. That's basically the only reason Bargs has a PER above 10.

WP48's biggest weakness is that it doesn't account for who you're playing against and rewards you for being a bit selfish.

Basketball is a team game, any stat attributed to individual players is destined to be flawed in some way.

[Joey]

Wiki
This first quote expands on what Nine was saying. Quote from David Berri, 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."

PER takes into account positive accomplishments, such as field goals, free throws, 3-pointers, assists, rebounds, blocks and steals, and negative ones, such as missed shots, turnovers and personal fouls. The formula adds positive stats and subtracts negative ones through a statistical point value system. The rating for each player is then adjusted to a per-minute basis so that, for example, substitutes can be compared with starters in playing time debates. It is also adjusted for the team's pace. In the end, one number sums up the players' statistical accomplishments for that season.

Hollinger has set up PER so that the league average, every season, is 15.00, which produces sort of a handy reference guide:

A Year For the Ages: 35.0
Runaway MVP Candidate: 30.0
Strong MVP Candidate: 27.5
Weak MVP Candidate: 25.0
Bona fide All-Star: 22.5
Borderline All-Star: 20.0
Solid 2nd option: 18.0
3rd Banana: 16.5
Pretty good player: 15.0
In the rotation: 13.0
Scrounging for minutes: 11.0
Definitely renting: 9.0
The Next Stop: DLeague 5.0

A 30+ PER has only been done 13 times. 4 by MJ.

[Joey]

(According to WP48, Bargnani is by far the least productive player in the league. Lol. They also believe that Hump should have been an All-Star. I think they said that Rodman>Jordan, too.)

Basically, what Brandon said. But you should use multiple stats and metrics rather than just one, and see if it matches your eyes.

Hahahahaha! As if Bargs is at the VERY bottom. Literally. Last out of EVERYONE in the league. Rough.

And I remember near the beginning of the season, I made a thread comparing Bargs and Kevin Love, saying Love is actually a far better player. Low and behold, he's at the TOP of the list, and Bargs is at the Bottom.

[Brandon]

Hahahahaha! As if Bargs is at the VERY bottom. Literally. Last out of EVERYONE in the league. Rough.

And I remember near the beginning of the season, I made a thread comparing Bargs and Kevin Love, saying Love is actually a far better player. Low and behold, he's at the TOP of the list, and Bargs is at the Bottom.

Well, that's WP48, not PER. Most other advanced stats say Bargnani has an average impact on a game. WP48 is heavily weighted in favour of rebounding, so the fact that Bargnani has never heard of it hurts him a lot (it also hurts his team a lot).

[GarbageTime]

I know the acronym stands for Player Efficency Rating but not much else. I know a long and crazy formula is used to formulate the number. Is it even possible to explain in layman terms what the final number represents? I find it hard to beleive this stat is meaningful when you can't understand its usefulness. Anyone care to explain how the components of PER are used / weighted and how looking at the final number helps one evaluate a players game.

Just out of curiousity were you expecting others to not understand it to, and therefore justify why people shouldn't use it and therefore it shouldn't apply to Bargnani?

Of all the advanced stats, I personally think that PER has the least meaning. It is more or less a fantasy basketball metric, where if you take players with good PERs you are going to have a relatively good fantasy team. Thats not to say there isn't some overlap with real basketball, but, as many have pointed out, there are some serious flaws with it (namely efficiency, which can not be overvalued in basketball).

I also find it hilarious that of all the advanced metrics, this is the only one that makes Bargnani look "average", and almost exclusively due to it valueing volume or effectiveness.

But Nine New Faces said it best:

"Basketball is a team game, any stat attributed to individual players is destined to be flawed in some way"

that however does not mean you ignore them. When no advanced says a player is any better than "average", and the vast majority say they are one of the "worst".... chances are they fall, at the very least, somewhere between those two categories.

PS "(Berri's stat says Love's teammates are so awful they're subtracting a lot of wins from the team)." Can anyone honestly say this is impossible to be true?

[Tim W.]

There's obviously no perfect formula, so you really have to look at as many advanced stats as possible to reach a conclusion.

As for the PER, I don't entirely agree with Hollinger's reference guide. If 15 is the average, how does that mean that he's a pretty good player. By Hollinger's own account, he's an average player. Not average starter, but average player. In other words, half the league is better than he is. I like Amir, and he's a lot better than a small minority seem to think, but there's no way he's close to being a solid 2nd banana. I think the whole 2nd option, 3rd banana thing is flawed. Obviously that's meant to be on an average team, but even then it doesn't make much sense. Philly is a perfectly average team (.500), yet they don't have any All Star, yet New York has two perennial All-Stars. Atlanta has three possible All-Stars yet they only slightly better.

Other than that, PER, like the other advanced stats, are at least a good indication of how good a player is.

[Brandon]

There's obviously no perfect formula, so you really have to look at as many advanced stats as possible to reach a conclusion.

As for the PER, I don't entirely agree with Hollinger's reference guide. If 15 is the average, how does that mean that he's a pretty good player. By Hollinger's own account, he's an average player. Not average starter, but average player. In other words, half the league is better than he is. I like Amir, and he's a lot better than a small minority seem to think, but there's no way he's close to being a solid 2nd banana. I think the whole 2nd option, 3rd banana thing is flawed. Obviously that's meant to be on an average team, but even then it doesn't make much sense. Philly is a perfectly average team (.500), yet they don't have any All Star, yet New York has two perennial All-Stars. Atlanta has three possible All-Stars yet they only slightly better.

Other than that, PER, like the other advanced stats, are at least a good indication of how good a player is.

By "average", Hollinger means halfway between 0 and the top PER possible -- 30-ish. He doesn't mean half the players in the league are better than a 15. Actually the overwhelming number of players in the league these days are far under 15. Additionally, many players between 15-20 are garbagemen, who do things that look great statistically but don't necessarily make their teams win a lot of games (Amir Johnson et al).

If someone were to create a Gaussian function (bell curve) I think the expected number would be lower than 15. In other words, most of the players would be at 12, plus or minus 2 (just a guess). Once you get significantly above 15, the players become elite all-stars, perennial all-stars, mvp-candidates, hall-of-famers, ie. the guys you need if you want to win it all.

number of players above 15: 96
number of players at 15.xx: 37
number of players below 15: 206

source.