Just a couple thoughts about PER

[ardee]

I feel PER is not a stat to go by when you're looking at who has had the best season.

There are two reasons for that.

One is that involved in the calculation of PER, you are adjusting it according to the league average PER. That means that you're not measuring how efficient James is: you're measuring how efficient he is COMPARED to the average player. So for example, LeBron's 2012 season is not necessarily better than any of Jordan's seasons from 88-91. They are merely better than the 2012 average player by a greater margin than Jordan was over the average player in those times.

Two, I think the pace adjusting part of PER makes no sense. The creators of PER claim that it is so that those who play on faster teams don't have an unfair advantage. Consider this: doesn't it make sense that a team can be playing faster BECAUSE of that player?

Bill Russell for instance, in today's day a 15-24-5 guy in the current pace would have a great PER. When there are maybe 30-40 more possessions, not so much. Consider that it's BECAUSE of Russell's rebounding that his team gets those extra possessions, and has more opportunities to score on a fast-break. If he didn't rebound as much, sure, his team's pace would go down, but would that help them win as much?

PER does not reflect the impact of a player accurately.

I would venture that if pace and the league average are not taken into account, you'd have a more accurate reflection of a player's impact.

[GrangerDanger]

That's why we use win shares and HCA to rank players

[Darain]

I feel PER is not a stat to go by when you're looking at who has had the best season.

There are two reasons for that.

One is that involved in the calculation of PER, you are adjusting it according to the league average PER. That means that you're not measuring how efficient James is: you're measuring how efficient he is COMPARED to the average player. So for example, LeBron's 2012 season is not necessarily better than any of Jordan's seasons from 88-91. They are merely better than the 2012 average player by a greater margin than Jordan was over the average player in those times.

But you should only compare players about how they played with their peers

Like a 1950's basketball player wouldn't do as aswell in today's game

[JordansBulls]

I feel PER is not a stat to go by when you're looking at who has had the best season.

There are two reasons for that.

One is that involved in the calculation of PER, you are adjusting it according to the league average PER. That means that you're not measuring how efficient James is: you're measuring how efficient he is COMPARED to the average player. So for example, LeBron's 2012 season is not necessarily better than any of Jordan's seasons from 88-91. They are merely better than the 2012 average player by a greater margin than Jordan was over the average player in those times.

Two, I think the pace adjusting part of PER makes no sense. The creators of PER claim that it is so that those who play on faster teams don't have an unfair advantage. Consider this: doesn't it make sense that a team can be playing faster BECAUSE of that player?

Bill Russell for instance, in today's day a 15-24-5 guy in the current pace would have a great PER. When there are maybe 30-40 more possessions, not so much. Consider that it's BECAUSE of Russell's rebounding that his team gets those extra possessions, and has more opportunities to score on a fast-break. If he didn't rebound as much, sure, his team's pace would go down, but would that help them win as much?

PER does not reflect the impact of a player accurately.

I would venture that if pace and the league average are not taken into account, you'd have a more accurate reflection of a player's impact.

It means the player distanced himself more so than his peers stats wise, but that doesn't say much if most of the stars are out with injuries during the season much. Now if this were 2003 and he was doing this when guys like Kobe, Tmac, Shaq, Duncan, KG, Dirk were relatively in there primes then yes it would be sensational. But now other than maybe Dwight, who do you have? Wade is 2nd in PER this year at 28 and he is only averaging 22/4/5 on 50% FG. That would be around a 23 PER in 2003 or so.

[Doctor MJ]

One is that involved in the calculation of PER, you are adjusting it according to the league average PER. That means that you're not measuring how efficient James is: you're measuring how efficient he is COMPARED to the average player. So for example, LeBron's 2012 season is not necessarily better than any of Jordan's seasons from 88-91. They are merely better than the 2012 average player by a greater margin than Jordan was over the average player in those times.

You've got something of a valid point here, however:

1) Don't overrate the difference between average players from year to year. The quality and quantity of superstars ebbs and flows a great deal, but the less singular the talent, the more people have it, and thus the more the league is pretty much assured of things only changing significantly with massive changes to the league.

So stable is this that even expansion is not as big of a deal as people think, as there are always players who fail to make the league despite being as good as some NBA players simply because of the finite opportunity presented with the teams involved.

2) I don't know of anyone knowledgeable who really takes PER so seriously that they would not factor in era changes. I mean, for most of us, it's just one more tool we use and has nowhere near the credibility to even dominate our yearly MVP thought.

Two, I think the pace adjusting part of PER makes no sense. The creators of PER claim that it is so that those who play on faster teams don't have an unfair advantage. Consider this: doesn't it make sense that a team can be playing faster BECAUSE of that player?

What you're not understanding is that playing fast is never a goal in and of itself because the other team gets about as many possessions as you do regardless of fast you chuck a shot. Hence any model that penalizes players because they made better use of their full 24 second is TERRIBLE.

'60s offenses sucked - as in they typically ended in failure - and if you watch the film you see a major reason why is because they often settled for the first mediocre shot they could find. The slowing of pace happened not because players could no longer play fast but because coaches got wise and started crafting better offensive possessions.

Incidentally, the amount of missed shots back then, not simply do to pace but inaccuracy, further inflates rebounding totals.

Bill Russell for instance, in today's day a 15-24-5 guy in the current pace would have a great PER. When there are maybe 30-40 more possessions, not so much. Consider that it's BECAUSE of Russell's rebounding that his team gets those extra possessions, and has more opportunities to score on a fast-break. If he didn't rebound as much, sure, his team's pace would go down, but would that help them win as much?

1. Again, Russell is not giving his team extra possession above what the opponent has. Rebounding can give you more opportunities to score in a given possession, but that does not affect pace, and is not what is being adjusted for.

2. You don't seem to be thinking about efficiency at all in your PER estimates. Do you realize how weak the league used to be on this front?

3. Speaking as a guy who has Russell as his GOAT, Russell's never going to look great by PER because his impact is hard to put in the box score. That's a knock on any box score stat like PER, but not for the reasons you claim.

I would venture that if pace and the league average are not taken into account, you'd have a more accurate reflection of a player's impact.

As mentioned, your thoughts on pace are really just wrong on every level. However I'd actually really like to see how not adjusting for league average looks, but it's going to hold some surprises for you. Go take a look at, say, Elgin Baylor's TS%. It's horrendous yet he has a peak PER up there with Shaq's prime. Transplant Baylor into the modern game with that efficiency, and his PER gets hit hard.

Also, Baylor gets his hard as his coach tells him to stop jacking bad shots...

[Rerisen]

Can we put LeBron's stats into the 1988 season and see what his PER would be? Not that this would be the fairest way to make a comparison, but I'd just be curious to see how much the league wide depression in efficiency this year is affecting such a stat.

[mopper8]

Wade is 2nd in PER this year at 28 and he is only averaging 22/4/5 on 50% FG. That would be around a 23 PER in 2003 or so.

You conveniently forgot to mention that Wade is playing under 32 mpg. Per-36 he's averaging 25.3 points, 5 rebounds, 5.5 assists, 1.9 steals, 1.4 blocks, 3 TOs, with a TS% slightly over 56

That would not be a 23 PER in 2003. In 2003, for example, Kobe posted per-36:

26 points, 6 rebounds, 5.1 assists, 1.9 steals, .7 blocks, 3.0 TOs, 55 TS%. Numbers are nearly identical per-minute, which PER takes into account. Main difference between them statistically is that Kobe played 41.5 minutes per game that year, a good 10 more minutes per game than Wade is playing this year.

Now, I'm not arguing Wade right now is as good as Kobe was then; doubtful that Wade keeps up his volume/efficiency numbers playing extended minutes like that. But that's not what PER is measuring, and for PER calculation purposes, they would be nearly statistically identical. Kobe had a PER of 26, not 23, in 2003.

[Rerisen]

Actually a better, current era comparison for LBJ would be 2009. That year and this look pretty close by per 36.

[GreenHat]

I feel PER is not a stat to go by when you're looking at who has had the best season.

There are two reasons for that.

One is that involved in the calculation of PER, you are adjusting it according to the league average PER. That means that you're not measuring how efficient James is: you're measuring how efficient he is COMPARED to the average player. So for example, LeBron's 2012 season is not necessarily better than any of Jordan's seasons from 88-91. They are merely better than the 2012 average player by a greater margin than Jordan was over the average player in those times.

Two, I think the pace adjusting part of PER makes no sense. The creators of PER claim that it is so that those who play on faster teams don't have an unfair advantage. Consider this: doesn't it make sense that a team can be playing faster BECAUSE of that player?

Bill Russell for instance, in today's day a 15-24-5 guy in the current pace would have a great PER. When there are maybe 30-40 more possessions, not so much. Consider that it's BECAUSE of Russell's rebounding that his team gets those extra possessions, and has more opportunities to score on a fast-break. If he didn't rebound as much, sure, his team's pace would go down, but would that help them win as much?

PER does not reflect the impact of a player accurately.

I would venture that if pace and the league average are not taken into account, you'd have a more accurate reflection of a player's impact.

It means the player distanced himself more so than his peers stats wise, but that doesn't say much if most of the stars are out with injuries during the season much. Now if this were 2003 and he was doing this when guys like Kobe, Tmac, Shaq, Duncan, KG, Dirk were relatively in there primes then yes it would be sensational. But now other than maybe Dwight, who do you have? Wade is 2nd in PER this year at 28 and he is only averaging 22/4/5 on 50% FG. That would be around a 23 PER in 2003 or so.

Yeah you're right a 32+ PER would be sensational ONLY if other players were playing better

PER is FAR from a perfect stat but randomly throwing up the greatest season ever in it is sensational no matter what. It isn't totally meaningless because of the rest of the league.

More ridiculous over protection of Jordan in every random stat even when it doesn't mean much of anything. This is just like guard shot blocking thread.

Oh and Wade is second in PER because he is putting up those numbers in ONLY 32 minutes. Also in the "who do you have?" category you conveniently forgot Durant and Chris Paul who are having big years.

If the whole league's efficiency went down because of the compressed schedule (or whatever you want to say it is) and one guy's goes way up completely against the grain, I would say that should be a positive not a negative.

Before someone quotes me out of context and changes my post to something else I don't think PER is 100% accurate nor do I think Lebron is better than Jordan.