Playoff vs Regular Season PER

[mook]

Career averages:

Miller:
Reg season 18.0
Playoffs 15.7

Wallace:
Rs: 17.6
Po: 16.3

Aldridge:
Rs: 19.1
Po: 18.1

Camby:
Rs: 18.0
Po: 15.7

I didn't include anyone else. Batum has been in about as many playoff games as Aldridge, but he had a pretty bad injury last year. So his numbers aren't going to be significant. Roy probably won't matter. Sample size for everybody else is probably too small to be meaningful (and you may argue the same problem with some of the guys above).

FWIW, looking around the league Kobe & Garnett dropped a point in PER, Duncan and Dirk rose a point, LeBron was flat. Role players like Marion (-2.0), Battier (-3) and Artest (-1.7) all tended to fall, although Fisher was flat.

No real agenda in posting this. I was curious, and thought others might be too.

 

[oldguy]

That kinda makes sense to me, at least for LA, AM and GW. They would be playing better teams, with gameplans tailored to stopping them. Not sure why Marcus's PER would drop.

Go Blazers

 

[Masbee]

Goat: (Jordan):
RS: 27.9 (are you f'ng kidding me!)
PO: 28.6 (ha, ha, ha. Can't be human. includes a high of 32.0 in 90-91 with WinShare per/48 of .333)

 

[Masbee]

Amare:
RS: 22.6
PO: 23.9 (much of this boost on the basis of one incredible run in the 07 playoffs where he was the best player in the world.)

 

[blazerboy30]

Career averages:

Miller:
Reg season 18.0
Playoffs 15.7

Wallace:
Rs: 17.6
Po: 16.3

Aldridge:
Rs: 19.1
Po: 18.1

Camby:
Rs: 18.0
Po: 15.7

I didn't include anyone else. Batum has been in about as many playoff games as Aldridge, but he had a pretty bad injury last year. So his numbers aren't going to be significant. Roy probably won't matter. Sample size for everybody else is probably too small to be meaningful (and you may argue the same problem with some of the guys above).

FWIW, looking around the league Kobe & Garnett dropped a point in PER, Duncan and Dirk rose a point, LeBron was flat. Role players like Marion (-2.0), Battier (-3) and Artest (-1.7) all tended to fall, although Fisher was flat.

No real agenda in posting this. I was curious, and thought others might be too.

Interesting. Maybe PERs drop due to the more physical and grind-it-out nature of playoff basketball?

 

[PapaG]

Interesting. Maybe PERs drop due to the more physical and grind-it-out nature of playoff basketball?

Well, I'll need a full data set of the PER for every player who has played in the playoffs, compared to their regular season PER, as well as a correlation between the two. I may need to add in some variables, once I figure them out. This won't really prove anything about any individual players, but we can at least estimate how each player should perform, regardless of their actual statistics.

Does anybody have that information available?

 

[jlprk]

I've been on it for 4 days. Are you enjoying your vacation in Hawaii? Give me another week, please? Thank you, thank you!

 

[PapaG]

I've been on it for 4 days. Are you enjoying your vacation in Hawaii? Give me another week, please? Thank you, thank you!

Get on it. I can't tell you how irrelevant your statistics are until I can give them a cursory review.

 

[blazerboy30]

Well, I'll need a full data set of the PER for every player who has played in the playoffs, compared to their regular season PER, as well as a correlation between the two. I may need to add in some variables, once I figure them out. This won't really prove anything about any individual players, but we can at least estimate how each player should perform, regardless of their actual statistics.

Does anybody have that information available?

1) Strange that you keep asking for statistics even though you won't understand them or know how to use them.

2) Strange that you think you'll get causation from some stats instead of correlation. Refer back to point (1).

 

[BrianFromWA]

While PER is a great tool for looking at effectiveness over the regular, 82-game season, I wonder about its effectiveness in determining playoff success. My hypothesis is that deviation is a bigger part of playoff success than straight PER.

I mean, in the regular season, if Nic alternates between "amazing" and "total crap", but averages out to a 16 PER, then that's pretty decent and the Blazers would probably be happy with it, even if that means that we beat the Spurs but lost to the Warriors. But game-to-game (and even in-game) everything's different in the playoffs. That's how some in here can say that Steve Blake played a great series for us b/c he had a PER of 15, while looking at the game logs he had 4 below-average game scores (including one horrible one) and that one of his above-average games was the one with the out-of-control 3. Interestingly, that was how Miller performed last year...one great game (Game 1) with a game score of 29 and then 4 average-to-below-average games and a stinker in Game 6...but he had a PER of 15! Or you could correlate that Travis was out of his element the HOU series (though he'd probably be better if exposed to more playoff games) by looking at the game logs and say "yeah, he was consistently bad" (which passes the eye test, and is shown in his PER average).

SO I guess my hypothesis is that, to win in the playoffs, you'd like consistency from your role players (whatever that may be) while looking for your stars to have the ability to put up epic games while not bottoming out. Brandon in the HOU series in 08-09 was able to put up a couple of huge game scores, which brought up his average but also didn't correlate one-to-one for wins, b/c of LMA's rock-bottom performances.

I'll get a lot of tl;dr's from this one, but the bottom line is that I think there's too much at stake each playoff game and too small a sample to make PER overall a tool to use for playoff success--more that you should look to game-score consistency from role players and the ability for your stars to put up big numbers without bottoming out, even if their "off games" are below average.

 

[PapaG]

1) Strange that you keep asking for statistics even though you won't understand them or know how to use them.

Is that number 7 or 8? I'm losing track.

2) Strange that you think you'll get causation from some stats instead of correlation. Refer back to point (1).

I didn't post that. I actually posted the opposite of that. Did they teach English at faux Stanford? I ask, because you seem to have a difficult time comprehending the language.

 

[blazerboy30]

Is that number 7 or 8? I'm losing track.

That's because you seem to have trouble with stats and numbers.

 

[PapaG]

That's because you seem to have trouble with stats and numbers.

Let's call it 9.