PDA

View Full Version : Per36 vs. Per48?



GunFactor187
12-01-2012, 04:02 AM
Hey guys, I'm relatively new to the world of Basketball Advanced Statistics and just wanted to see your guys' opinions on which Per Minute Stat do you think is more accurate/precise in terms of overall production. Per36 or Per48? Why? Discuss. Thanks! :)

Chronz
12-01-2012, 02:48 PM
same

Gators123
12-01-2012, 04:02 PM
^ agree.

I use PER 36 more just because guys usually don't play 48 minutes.

ChiSox219
12-01-2012, 05:12 PM
There is also Per 40 Pace Adjusted:

http://www.draftexpress.com/stats.php?sort=&q=eff&league=NBA&year=2012%2F13&per=per40pace&min=20&pos=all&qual=all

Patman
12-03-2012, 06:09 AM
doesn't make a difference, the Per 48 will just be higher, but it's exactly the same, because it's minutes based.

Raps18-19 Champ
12-04-2012, 05:27 PM
^ agree.

I use PER 36 more just because guys usually don't play 48 minutes.

This.

DenButsu
12-07-2012, 03:01 AM
Agreed with all above, but would just add - and hopefully this goes without saying - that the most important thing would be to maintain consistency when comparing players, ie. don't compare one's per 36 numbers to anothers per 48 numbers.

Raidaz4Life
12-08-2012, 10:06 PM
I would say a per 36 would be more reliable data personally....

DenButsu
12-09-2012, 12:53 AM
I would say a per 36 would be more reliable data personally....

What do you mean by "reliable"? Because both are equal in terms of showing production per minutes. It's just that time is a variable. If player A scores 15 points per 36 and playeer B scores 21, then player A scores 20 per 48 and player B scores 28. Their comparative efficiency maintains the same proportion.

But if you mean, as others have said above, that 36 minutes more accurately reflects real playing time, then I'd agree.

Baller1
12-09-2012, 11:01 PM
I personally would prefer looking at their per36. Realistically, players aren't going to play a full 48 minutes and they sure as hell aren't going to average that.

Raidaz4Life
12-10-2012, 07:18 AM
What do you mean by "reliable"? Because both are equal in terms of showing production per minutes. It's just that time is a variable. If player A scores 15 points per 36 and playeer B scores 21, then player A scores 20 per 48 and player B scores 28. Their comparative efficiency maintains the same proportion.

But if you mean, as others have said above, that 36 minutes more accurately reflects real playing time, then I'd agree.

I did, maybe reliable wasn't the best word to convey what I meant. I think honestly you want to keep the sample size as universally close to an average playing time as possible because the further you exceed that, the more variables can skew the information. I mean nobody really plays 48 minutes on a nightly basis so to suggest two player's that actually only played 24 minutes a game statistics would remain directly proportionate would more than likely be inaccurate in conveying an actual stat and efficiency spread if they played an entire game.


I mean I already know a lot of people can find the per 36 unreliable for that reason...

Chronz
12-12-2012, 02:20 AM
I did, maybe reliable wasn't the best word to convey what I meant. I think honestly you want to keep the sample size as universally close to an average playing time as possible because the further you exceed that, the more variables can skew the information. I mean nobody really plays 48 minutes on a nightly basis so to suggest two player's that actually only played 24 minutes a game statistics would remain directly proportionate would more than likely be inaccurate in conveying an actual stat and efficiency spread if they played an entire game.


I mean I already know a lot of people can find the per 36 unreliable for that reason...

They are the SAME. What exactly are they skewing?

Ebbs
12-15-2012, 04:54 AM
They aren't the same. Can a player who plays 25-30 minutes maintain production for 36 probably.

But I don't think it's the same per 48. Exhaustion sets in the physicality of the game takes its toll and efficiency will go down.

Per 48 minutes is a dumb scale

Patman
12-18-2012, 10:36 AM
They aren't the same. Can a player who plays 25-30 minutes maintain production for 36 probably.

But I don't think it's the same per 48. Exhaustion sets in the physicality of the game takes its toll and efficiency will go down.

Per 48 minutes is a dumb scale

But in the end it doesn't matter, because some players would even get bigger numbers with more time and some would drop before they reach 36 minutes (allthough there is nearly no correlation between higher minutes and efficiency). They both show per minute stats, maybe we should just change to per minute stats.

if you compare players on a per 48 basis it isn't different then comparing them on a per 36 basis. It only matters that you use the same numbers.... you could also use per 1 minute per 10/12/24 or whatever floats your boat, the stat tells the same story every time.

RyanStorm
12-31-2012, 10:09 PM
The 36 is just showing what they could make if they played a good nights game with 36 mins on the floor. Its good for estimating if your bench player could become a starter or not if given the right minutes.

Per 48 is more for comparing a lot of people and if you don't care on the numbers. Like 36 is meant to be a starters minutes. Across the board the 48 to me is a good basis to estimate a players overall greatness. Although I like the 36 more for the types of things I look into.

JordansBulls
07-29-2013, 07:52 PM
PER 48 is good, but I like PER 40 more than PER 36.

DenButsu
07-30-2013, 11:39 AM
PER 48 is good, but I like PER 40 more than PER 36.

Why?

Chronz
08-01-2013, 04:12 PM
Why?

aesthetics

my guess

DenButsu
08-02-2013, 01:34 AM
FWIW, the 106 players who started 60+ games last season averaged 32.1 minutes per game, so *if* you're trying to use per/minutes stats to approximate starters' minutes, per36 would be the closest among the ones discussed above. Not that it makes any difference statistically.

JordansBulls
08-02-2013, 08:34 AM
Why?

Most players in the playoffs or season play closer to 40 minutes a game than 36 minutes a game.

WadeKobe
08-07-2013, 02:37 AM
Most players in the playoffs play closer to 40 minutes a game than 36 minutes a game.

Yes.


Most players in the season play closer to 40 minutes a game than 36 minutes a game.

No.

DenButsu
08-08-2013, 06:13 AM
The bottom line of all of this is that it doesn't matter. One reason to do per minute stats is to create a normalized baseline in order to make comparisons. And as long as that baseline (in this case number of minutes) is the same for all the players you're comparing, you're good.

I suppose there's a reasonable argument to be made that using per36 for regular season and per40 for the postseason is the way to go when comparing starters. But in the case of starters, they're most likely playing comparable minutes in the first place which diminishes the need for going per minute at all.

Its greater utility is projecting what a 15 or 20 mpg player might accomplish as a starter if his role was expanded. And in that case it makes more sense to use the better approximation for the regular season.

Gators123
08-08-2013, 05:30 PM
PER 48 is good, but I like PER 40 more than PER 36.


Why?


Because MJ averaged closer to 40 MPG than 36 MPG ;)

KwameBeanJordan
03-12-2014, 06:13 PM
Firstly, I don't think you should be comparing players with per minute stats for the most part. There are really only three reasons to use per minute stats.

1. You are looking at bench players and want to see if they should get more time. In this case per 36 is probably best.
2. You are comparing clutch (4th qtr/ot, no team ahead by more than 5 points) play. Using a per 48 for comparing clutch play lets you see changes in shot behavior more easily.
3. You are worried that players stay in the game to pad their stats. However, I don't think this is really ever the case. If the game is already over, the coaches will pretty much ALWAYS pull out the star players and put in the bench. When Kobe scored 63 points or so in three quarters, Phil Jackson kept him out of the entire fourth quarter.