PER................. good or bad........... pros and cons....

Welcome to our community

Be a part of something great, join today!

THE HCP

NorthEastPortland'sFinest
Joined
Sep 16, 2008
Messages
73,033
Likes
62,251
Points
113
Is using PER the best way to evaluate a players worth? I like it (although am still a little confused on how it works). It sure does seem to rank the players properly. Are there flaws n the system? What does it miss? Wish there was a person on here that knew the ends and outs!
 
PRO:

Normalizes stats by taking into account team pace and minutes played.

CON:

No defense.
 
Ahh, I didn't realize it didn't take into consideration defense. What about blocks and steals?
 
http://sports.espn.go.com/nba/columns/story?columnist=hollinger_john&id=2850240

From the mouth of the creator:

The player efficiency rating (PER) is a rating of a player's per-minute productivity.

To generate PER, I created formulas -- outlined in tortuous detail in my book "Pro Basketball Forecast" -- that return a value for each of a player's accomplishments. That includes 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.

Two important things to remember about PER are that it's per-minute and is pace-adjusted.

Because it's a per-minute measure, it allows us to compare, say, Jordan Farmar and Derek Fisher, even though there is a disparity in their minutes played.

I also adjust each player's rating for his team's pace, so that players on a slow-paced team like Detroit aren't penalized just because their team has fewer possessions than a fast-paced team such as Golden State.

Bear in mind that PER is not the final, once-and-for-all evaluation of a player's accomplishments during the season. This is especially true for defensive specialists -- such as Quinton Ross and Jason Collins -- who don't get many blocks or steals.

What PER can do, however, is summarize a player's statistical accomplishments in a single number. That allows us to unify the disparate data on each player we try to track in our heads (e.g., Corey Maggette: free-throw machine, good rebounder, decent shooter, poor passer, etc.) so that we can move on to evaluating what might be missing from the stats.

I set the league average in PER to 15.00 every season.

I like PER for what it is, a normalized stat for measuring mostly offensive efficiency. Like Hollinger says, it is not an end all be all evaluation of talent.
 
It's still just a stat.

It doesn't take into account setting screens, picks, keeping the ball moving, team defense, hustle, making players around you better...

So it has value, but it has a lot of holes in it. I'd never use it much in evaluating a player, except in a minor way.
 
So it DOES include defense.

It includes defensive statistics, but a player can get a lot of steals and/or blocks but be terrible as a defender in other ways... PER only looks at those things, so someone like Bruce Bowen's true value is never going to be accurately reflected in PER.

Another weakness of PER is that it tends to overvalue high volume, low-percentage shooters. Someone like Iverson was always overrated in terms of PER because he scored so many points and his inefficiencies were understated.

It's still the best single stat for measuring individual player contributions, though.

Ed O.
 
Does it hurt or help Nic then?
 
Does it hurt or help Nic then?

Well it doesn't hurt Nic that he shot the hell out of the ball last year, but it dings players who are not high usage guys (ie. guys who don't take a lot of shots or actively participate in a lot of play-making) in Nic's particular case it ranked him at 17+ which puts him at 'above average' for a starter.
 
Or to put it more technically, the formula is:

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) ]

Glossary for above:
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


Let me know if you have questions.

Oh, and to answer your question, the value is really to provide a single number that for a range of stats (pts, reb, asst, steals, blks) and then adjust them to pace. It doesn't blend 15 stats, or take into account multitudes of factors such as where guys shoot from, how much they get double-teamed, and of course guys playing against second units vs. starters. But it does take the very basic stats that every boxscore junkie like to read and adjust them to make a guy who plays 35 minutes on a team that walks the ball up the court adjusted from the guy who is playing 38 minutes on a team that pushes every time (thus padding extra stats on everyone's resume who happens to play under that system).

Not perfect, but a great general indicator for starters at the same positions who play 24+ minutes against other starters as you can really compare the blended stats of two "apples". IE - great for really getting a starting point for who would be the top candidates for all-star voting for starting centers, etc.
 
Last edited:
From the horses mouth! Thanks brother........ if you are in fact who you say you are!
 
Another weakness of PER is that it tends to overvalue high volume, low-percentage shooters. Someone like Iverson was always overrated in terms of PER because he scored so many points and his inefficiencies were understated.
I would counter that Iverson has always been under-appreciated by the stats geeks who focus on his efficiency (or lack thereof) more than his status as one of the most difficult guys in the league to stop. In which case, I think PER gets it right by not overstating his inefficiencies.
 
I would counter that Iverson has always been under-appreciated by the stats geeks who focus on his efficiency (or lack thereof) more than his status as one of the most difficult guys in the league to stop. In which case, I think PER gets it right by not overstating his inefficiencies.

PER catches the fact that Iverson got to the line a lot to make up for his inefficiency. The problem is, that the really truly elite players keep their field goal percentage up and get the free throws too. Iverson was a good player. But that doesn't mean he isn't a volume shooter.
 
Most non-volume shooters would love to have inefficiently lead their team to the Finals... You're confusing productivity with your own values of a good/great player.
 
A.I. was not just a good player, he is a HOFer!
 
Another weakness of PER is that it tends to overvalue high volume, low-percentage shooters. Someone like Iverson was always overrated in terms of PER because he scored so many points and his inefficiencies were understated.



Ed O.

I dunno about that. Iverson's career average in PER is 20.9, and during his 2000-01 MVP season, it was only 24, which is the lowest of any MVP winner that I can find. If anything, PER seems to accurately rate Iverson's value, while the media and fans tended to completely overrate Iverson. Even Kobe Bryant, who had a PER of 25 when he finally won an MVP award, is usually in the 24-25 range.

By comparison, Brandon Roy had the same PER of 24 for the 2008-09 season, and was never considered seriously as an MVP candidate.
 
What does PERs look like for players like Brian Grant, Horace Grant, Dennis Rodman, who simply make the team SO much better, but aren't stat sluts?
 
What does PERs look like for players like Brian Grant, Horace Grant, Dennis Rodman, who simply make the team SO much better, but aren't stat sluts?

I'd argue that Rodman, at least to some degree, was a "stat slut". Horace Grant, in his prime, had a PER in the 18ish range, which is well above average. Brian Grant, who is obviously one of my favorite players, never had the impact of the other two you mentioned, yet he still was a 17 PER player in his prime.

Rodman does seem to get hosed by PER, based on his defensive abilities, but he was a completely inefficient player on the offensive end, so that has to be taken into consideration. Plus, Rodman was a fluke, a rebounding/defensive monster who was allowed to be that player regardless of his terrible offense. The other two and their values are adequately reflected by PER, IMO.
 
Is there a list on ESPN of last years or past years PER rankings? Would love to know who had high PER's last year but were not stars.
 
JH, name 1 player your evaluation helps the most and 1 it hurts the most. Somebody has to slip through the cracks right, in a good way and a bad way.
 
HOLY SHIT! Miami has 3 of the top 4! And you don't think they will dominate?
 
I tend to agree with the guys who think Iverson's PER is about right. He not only averaged 9 FTA/game, he averaged > 6 APG and > 2 SPG for his career. His PER is right there with Melo's which is pretty fair, IMO.
 
The problem with PER is it doesn't take into account the opposition. One of the articles quoted talks about comparing Jordan Farmar and Derrick Fisher. Well, Fisher generally played against teams' first stringers and at the end of Q4 with the game on the line, and Farmer put up his stats against 2nd stringers.

You might consider this when talking about Batum and his 24 MPG, some against starters some against 2nd stringers.

Like any stat, you can't take the one as the end-all. You have to consider multiple stats to get a good picture of a guy's skills. Like the 24 MPG and PER taken together says "something."
 
The problem with PER is it doesn't take into account the opposition. One of the articles quoted talks about comparing Jordan Farmar and Derrick Fisher. Well, Fisher generally played against teams' first stringers and at the end of Q4 with the game on the line, and Farmer put up his stats against 2nd stringers.

You might consider this when talking about Batum and his 24 MPG, some against starters some against 2nd stringers.

Like any stat, you can't take the one as the end-all. You have to consider multiple stats to get a good picture of a guy's skills. Like the 24 MPG and PER taken together says "something."

I feel the same way about PER, at least in terms of mpg. It's why I take Oden's 23 PER with a grain of salt. I hope he puts up those numbers over 70 games/32 mpg, but I certainly don't expect it.
 
Or to put it more technically, the formula is:

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) ]

Glossary for above:
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


Let me know if you have questions.

Oh, and to answer your question, the value is really to provide a single number that for a range of stats (pts, reb, asst, steals, blks) and then adjust them to pace. It doesn't blend 15 stats, or take into account multitudes of factors such as where guys shoot from, how much they get double-teamed, and of course guys playing against second units vs. starters. But it does take the very basic stats that every boxscore junkie like to read and adjust them to make a guy who plays 35 minutes on a team that walks the ball up the court adjusted from the guy who is playing 38 minutes on a team that pushes every time (thus padding extra stats on everyone's resume who happens to play under that system).

Not perfect, but a great general indicator for starters at the same positions who play 24+ minutes against other starters as you can really compare the blended stats of two "apples". IE - great for really getting a starting point for who would be the top candidates for all-star voting for starting centers, etc.

Why does Frankie say, "Relax, don't do it?"
 
JH, name 1 player your evaluation helps the most and 1 it hurts the most. Somebody has to slip through the cracks right, in a good way and a bad way.

Obviously it helps the guys with fewer minutes. If you get fewer minutes as a starter (like Oden) you can get a true measure of where you'd be if you get over the foul situation and see about where you really line up with the guys plugging away 35-38 mpg. Then to the extreme it helps the guys with nearly no minutes like Patty Mills, who go in at garbage time and make a few lay ins against some scrubs who aren't playing defense, all in the matter of a couple minutes. It boosts those guys to actually look good, so they slip through the cracks unless you filter.

The guys it hurts are guys that either share a lot of touches (see Boston with Rondo, KG, Piece, etc. all sharing the stats, see Kobe having to share with Pau, while both are likely better than Wade or Bosh, Wade and Bosh being the only decent players on their entire rosters have inflated stats and thus their PER gets pushed way above what it would be if Kobe were to change places with Wade, or Pau to exchange with Bosh, etc. Also defense specialists get the shaft as they literally change the shots of opposing teams, and that doesn't get factored in. So guys like Dwight Howard and Yao are probably much lower as a single "value" put on them compared to guys that really aren't impact defenders and thus have a higher PER but much less value to the W/L record for a team.
 
Obviously it helps the guys with fewer minutes. If you get fewer minutes as a starter (like Oden) you can get a true measure of where you'd be if you get over the foul situation and see about where you really line up with the guys plugging away 35-38 mpg. Then to the extreme it helps the guys with nearly no minutes like Patty Mills, who go in at garbage time and make a few lay ins against some scrubs who aren't playing defense, all in the matter of a couple minutes. It boosts those guys to actually look good, so they slip through the cracks unless you filter.

The guys it hurts are guys that either share a lot of touches (see Boston with Rondo, KG, Piece, etc. all sharing the stats, see Kobe having to share with Pau, while both are likely better than Wade or Bosh, Wade and Bosh being the only decent players on their entire rosters have inflated stats and thus their PER gets pushed way above what it would be if Kobe were to change places with Wade, or Pau to exchange with Bosh, etc. Also defense specialists get the shaft as they literally change the shots of opposing teams, and that doesn't get factored in. So guys like Dwight Howard and Yao are probably much lower as a single "value" put on them compared to guys that really aren't impact defenders and thus have a higher PER but much less value to the W/L record for a team.

Does HCP think you're actually John Hollinger? LOL!
 

Users who are viewing this thread

Back
Top