Thks in advance: why Jokic's PER was higher in 21-22 than this season?

[Rubios]

I thought I understood reasonably good PER, but probably it's a case of Dunning-Kruger.

As far as I know, it's a comparison bw the whole League per-minute adjusted for pace and possessions.

So, was it because the competence was a tad "weaker" in 21-22? Was it because he played 3 minutes less? Or because he played 4 games more?
A combo of some of them?

Or -most likely- am i missing something?

Thks for the insight!

[UcanUwill]

games played has no impact on PER. PER is basically production stat, no matter how much you play, if you put good stats in 30 seconds, you could have highest PER. Of course during the season they have minimum minute played qualifiers and such, but just looking at per minute basis, let say per 36 minutes.

21-22 season : 27.5 ppg. 14.8 rpg. 8.5 apg. 1.4 spg. 0.7 bpg. 66.1 TS%
24-24 season : 29.0 ppg. 12.5 rpg. 10.0 apg. 1.8 spg. 0.6 bpg. 66.3 TS%

So, very similar, 25 season is a bit better, but as you pointed out, your PER is impacted by whole league, not just your own stats, and the difference between 32.8 PER and 30.0 PER is very little. in 22, league had lower TS%, offensive rating and scored less points, so I think that swings it.

[Bloodbather]

First off, PER is a garbage stat. EPM and LEBRON are the best publicly available catch-all stats.

I think the reason is that PER overstates the importance of rebounding and Jokic averaged 1.1 RPG more in 3.2 fewer minutes in 21-22.

[Mavrelous]

@JohnHollinger

[Rubios]

Thks!

The "problem" (for what I'm looking for, I mean) with LEBRON, vORP, EPM, BPM (...) are that they depend on your team.
As PER depends on the level of the whole League field that given year.

Is there any strictly individual -understand me- advanced stat? PIE, perhaps?
"Understand me" means i.e. assisting is easier if the ones receiving them are better, of course.

Or the closer I'd get would be "just take a good, deep look at the non-advanced box score and... you know, watch the games"?

Thanks again.

[Blame Rasho]

First off, PER is a garbage stat. EPM and LEBRON are the best publicly available catch-all stats.

I think the reason is that PER overstates the importance of rebounding and Jokic averaged 1.1 RPG more in 3.2 fewer minutes in 21-22.

PER isn’t a garbage stat. How people think of it and use PER is. You just got to put it into context. At its most basic level, it is a measurement of how you good you are compared to the avg player in that year. It is why you shouldn’t use PER by years to year comparisons. As you pointed out what was the cause. One rebound more in less mins is significant. PER also fails to account for defensive ability as it will never be able to tell you if a player knows how to hedge a screen correctly. It* will know if you gamble and get a steal though.

Catch all stats by its nature will never determine who is better. I don’t need to see some players PER in the 25 to 35 range to know that yeah… x player is pretty good.

It is simple minded to just use one stat to evaluate players. Several advanced stats would say that Manu Ginobili was better than Kobe Bryant but is that really the truth? They are evaluating tools, and each advanced stat need to be put into context before using it. Each has it pros and cons.

[tsherkin]

PER isn’t a garbage stat. How people think of it and use PER is.

Well, and it also favors certain stats pretty specifically, while not incorporating any kind of plus-minus data. It loves rebounds, it deflates the utility of assists, and it most certainly loves volume shooters and doesn't understand the value of defense outside of produced stocks.

"Garbage" is wrong, but it's certainly of only rudimentary value, which is generally what people tend to mean when saying that anyway. As you say, context is required, but ultimately, the context for PER is mostly that the player produces a lot in the box score. If you happen to be particularly efficient, then so much the better. So it doesn't really tell us much that just looking at box score averages and percentages doesn't already indicate, which makes it of very little value.

[Blame Rasho]

PER isn’t a garbage stat. How people think of it and use PER is.

Well, and it also favors certain stats pretty specifically, while not incorporating any kind of plus-minus data. It loves rebounds, it deflates the utility of assists, and it most certainly loves volume shooters and doesn't understand the value of defense outside of produced stocks.

"Garbage" is wrong, but it's certainly of only rudimentary value, which is generally what people tend to mean when saying that anyway. As you say, context is required, but ultimately, the context for PER is mostly that the player produces a lot in the box score. If you happen to be particularly efficient, then so much the better. So it doesn't really tell us much that just looking at box score averages and percentages doesn't already indicate, which makes it of very little value.

I am of* the opinion on/off for an example is a horrible metric given that for years Matt Bonner led the Spurs in that stat. It is reflective of who you are replacing, what your other teammates impact is in games. It is beyond horrible in small sample sizes.

I don’t go off of PER as a be all end all stats but rather it is good to supplement an argument that X player is playing better than Y player at times with combinations of other stats along with actual observations.

Using any one catch all stat is by all indications is flawed.

[tsherkin]

I am of* the opinion on/off for an example is a horrible metric given that for years Matt Bonner led the Spurs in that stat. It is reflective of who you are replacing, what your other teammates impact is in games. It is beyond horrible in small sample sizes.

Raw on/off is extremely messy, yes. And it isn't generally my favorite thing because it describes efficacy in role more than anything else, so it isn't great for comparison, either. But it does tend to highlight defenders a little more effectively, particularly when they don't accumulate stocks. RAPM and the like have their limitations like anything else, naturally.

I don’t go off of PER as a be all end all stats but rather it is good to supplement an argument that X player is playing better than Y player at times with combinations of other stats along with actual observations.

It really isn't, it offers basically nothing we don't know from looking at the box score. It's been like twenty-ish years, we know it isn't good for much other than a one-number summary of their box score line at this point.

Using any one catch all stat is by all indications is flawed.

Assuredly, we agree there.

[MarcusBrody]

The other thing with PER is that it is normalized. If the league gets more efficient overall, a player can drop in per while putting up the same (or in theory even better) stats.

[Rubios]

The other thing with PER is that it is normalized. If the league gets more efficient overall, a player can drop in per while putting up the same (or in theory even better) stats.

Copy/pasting here.

The "problem" (for what I'm looking for, I mean) with LEBRON, vORP, EPM, BPM, DARKO (...) are that they depend on your team.
As PER depends on the level of the whole League field that given year.

Is there any strictly individual -understand me- advanced stat? PIE, perhaps?
"Understand me" means i.e. assisting is easier if the ones receiving them are better, of course.

Or the closer I'd get would be "just take a good, deep look at the non-advanced box score and... you know, watch the games"?

Thanks again.

[Bloodbather]

First off, PER is a garbage stat. EPM and LEBRON are the best publicly available catch-all stats.

I think the reason is that PER overstates the importance of rebounding and Jokic averaged 1.1 RPG more in 3.2 fewer minutes in 21-22.

PER isn’t a garbage stat. How people think of it and use PER is. You just got to put it into context. At its most basic level, it is a measurement of how you good you are compared to the avg player in that year. It is why you shouldn’t use PER by years to year comparisons. As you pointed out what was the cause. One rebound more in less mins is significant. PER also fails to account for defensive ability as it will never be able to tell you if a player knows how to hedge a screen correctly. It* will know if you gamble and get a steal though.

Catch all stats by its nature will never determine who is better. I don’t need to see some players PER in the 25 to 35 range to know that yeah… x player is pretty good.

It is simple minded to just use one stat to evaluate players. Several advanced stats would say that Manu Ginobili was better than Kobe Bryant but is that really the truth? They are evaluating tools, and each advanced stat need to be put into context before using it. Each has it pros and cons.

Sure, you can't just use one stat. That's a general rule, you also have to contextualize everything.

But I call PER garbage because it doesn't tell you anything the simple boxscore doesn't. It's garbage because it's useless, you can just stick to the boxscore and you haven't lost much. It inflates the value of rebounding, devalues assists, doesn't have the right approach to at least approximate defensive impact, etc. It's a bad stat among other catch-all metrics, in my opinion.

[bledredwine]

PER isn’t a garbage stat. How people think of it and use PER is.

Well, and it also favors certain stats pretty specifically, while not incorporating any kind of plus-minus data. It loves rebounds, it deflates the utility of assists, and it most certainly loves volume shooters and doesn't understand the value of defense outside of produced stocks.

"Garbage" is wrong, but it's certainly of only rudimentary value, which is generally what people tend to mean when saying that anyway. As you say, context is required, but ultimately, the context for PER is mostly that the player produces a lot in the box score. If you happen to be particularly efficient, then so much the better. So it doesn't really tell us much that just looking at box score averages and percentages doesn't already indicate, which makes it of very little value.

The opposite. PER favors point scorers as it favors production, over emphasizing efficiency over volume scorers. See Kobe, AI, Melo, even Hakeem and then compare to players like Paul (sorry way overrated), Harden, Lebron, Jokic and so on.

[tsherkin]

The opposite. PER favors point scorers as it favors production, over emphasizing efficiency over volume scorers. See Kobe, AI, Melo, even Hakeem and then compare to players like Paul (sorry way overrated), Harden, Lebron, Jokic and so on.

Lebron largely scored in similar volume to Kobe, rebounded better at either end, produced more assists and drew more free throws. It's not surprising his PER was higher.

Paul was a high-efficiency guy who also got a bunch of steals and rarely turned it over, which was also relevant. And it did like volume shooters. It just liked efficient volume shooters more than crap volume shooters.

At the absolute apex of PER, you have to be efficient, but there are many ways to get it to love you.

[Rubios]

The "problem" (for what I'm looking for, I mean) with LEBRON, vORP, EPM, BPM, DARKO (...) are that they depend on your team.
As PER depends on the level of the whole League field that given year.

Is there any strictly individual -understand me- fairly good advanced stat? PIE, perhaps?
"Understand me" means i.e. assisting is easier if the ones receiving them are better, of course.

Or the closer I'd get would be "just take a good, deep look at the non-advanced box score and... you know, watch the games"?

Thanks again.

[dhsilv2]

games played has no impact on PER. PER is basically production stat, no matter how much you play, if you put good stats in 30 seconds, you could have highest PER. Of course during the season they have minimum minute played qualifiers and such, but just looking at per minute basis, let say per 36 minutes.

21-22 season : 27.5 ppg. 14.8 rpg. 8.5 apg. 1.4 spg. 0.7 bpg. 66.1 TS%
24-24 season : 29.0 ppg. 12.5 rpg. 10.0 apg. 1.8 spg. 0.6 bpg. 66.3 TS%

So, very similar, 25 season is a bit better, but as you pointed out, your PER is impacted by whole league, not just your own stats, and the difference between 32.8 PER and 30.0 PER is very little. in 22, league had lower TS%, offensive rating and scored less points, so I think that swings it.

32.8 and 30 is a massive massive massive difference.

[dhsilv2]

Thks!

The "problem" (for what I'm looking for, I mean) with LEBRON, vORP, EPM, BPM (...) are that they depend on your team.
As PER depends on the level of the whole League field that given year.

Is there any strictly individual -understand me- advanced stat? PIE, perhaps?
"Understand me" means i.e. assisting is easier if the ones receiving them are better, of course.

Or the closer I'd get would be "just take a good, deep look at the non-advanced box score and... you know, watch the games"?

Thanks again.

LEBRON et al also are relative to the league. No good stat won't be.

[dhsilv2]

PER isn’t a garbage stat. How people think of it and use PER is.

Well, and it also favors certain stats pretty specifically, while not incorporating any kind of plus-minus data. It loves rebounds, it deflates the utility of assists, and it most certainly loves volume shooters and doesn't understand the value of defense outside of produced stocks.

"Garbage" is wrong, but it's certainly of only rudimentary value, which is generally what people tend to mean when saying that anyway. As you say, context is required, but ultimately, the context for PER is mostly that the player produces a lot in the box score. If you happen to be particularly efficient, then so much the better. So it doesn't really tell us much that just looking at box score averages and percentages doesn't already indicate, which makes it of very little value.

Biggest value to created was allowing us to better compare players with lower minute counts. That alone makes the stat fairly handy even if other metrics have come about since that do a better job.

[dhsilv2]

First off, PER is a garbage stat. EPM and LEBRON are the best publicly available catch-all stats.

I think the reason is that PER overstates the importance of rebounding and Jokic averaged 1.1 RPG more in 3.2 fewer minutes in 21-22.

PER isn’t a garbage stat. How people think of it and use PER is. You just got to put it into context. At its most basic level, it is a measurement of how you good you are compared to the avg player in that year. It is why you shouldn’t use PER by years to year comparisons. As you pointed out what was the cause. One rebound more in less mins is significant. PER also fails to account for defensive ability as it will never be able to tell you if a player knows how to hedge a screen correctly. It* will know if you gamble and get a steal though.

Catch all stats by its nature will never determine who is better. I don’t need to see some players PER in the 25 to 35 range to know that yeah… x player is pretty good.

It is simple minded to just use one stat to evaluate players. Several advanced stats would say that Manu Ginobili was better than Kobe Bryant but is that really the truth? They are evaluating tools, and each advanced stat need to be put into context before using it. Each has it pros and cons.

Sure, you can't just use one stat. That's a general rule, you also have to contextualize everything.

But I call PER garbage because it doesn't tell you anything the simple boxscore doesn't. It's garbage because it's useless, you can just stick to the boxscore and you haven't lost much. It inflates the value of rebounding, devalues assists, doesn't have the right approach to at least approximate defensive impact, etc. It's a bad stat among other catch-all metrics, in my opinion.

You would have to take it the entire box score for the entire year by every team to actually capture how far from an average player someone is and do it year to year adjusted. PER resolves this. That's extremely valuable.

[dhsilv2]

PER isn’t a garbage stat. How people think of it and use PER is.

Well, and it also favors certain stats pretty specifically, while not incorporating any kind of plus-minus data. It loves rebounds, it deflates the utility of assists, and it most certainly loves volume shooters and doesn't understand the value of defense outside of produced stocks.

"Garbage" is wrong, but it's certainly of only rudimentary value, which is generally what people tend to mean when saying that anyway. As you say, context is required, but ultimately, the context for PER is mostly that the player produces a lot in the box score. If you happen to be particularly efficient, then so much the better. So it doesn't really tell us much that just looking at box score averages and percentages doesn't already indicate, which makes it of very little value.

The opposite. PER favors point scorers as it favors production, over emphasizing efficiency over volume scorers. See Kobe, AI, Melo, even Hakeem and then compare to players like Paul (sorry way overrated), Harden, Lebron, Jokic and so on.

As always wrong. PER's bias is toward's volume in terms of usage as defined by John. Guys like kobe are perfectly examples of players who PER is biased towards and increases their numbers. The metric that has a huge bias to TS% and efficiency is winshare which thinks Reggie Miller was a god among men.

John was of the belief that coaches knew more than a database of stats. As such he built in a bias for volume shooting because he assumed all else equal it was likely that the coach knew something. That isn't to say being more efficient isn't better...because that's obvious.

Remember PER was made with the assumption Jordan must be the greatest and everything started with him as the perfect player.