Greatest Playoff Scaled PER and WS/48 Player Seasons all-time

Post #1 » by **trex_8063** » Wed Aug 31, 2016 2:43 am

This is piggy-backing on the work outlined in this post:

viewtopic.php?f=64&t=1474454

Was curious to see who would populate the list of greatest ever scaled playoff PER’s and WS/48’s, so those are listed below.

I used the same standard deviations derived from the regular season numbers, as I felt that would be more accurately representative of the actual amount of talent parity or disparity in the league within a given year (just due to larger sample size and presence of complete player pool). So I just applied those standard deviations to the playoff numbers.

Small sample of games in many instances has produced some spurious results (especially with WS/48, which I think takes a little longer to gel).

Anyway, here’s what it yielded (EDIT: have placed an * by all values the result of 7 games or fewer (i.e. generally just a single series))…..

EDIT2: Have updated to include ABA player seasons (highlighted in blue-violet within lists).

Top 100(ish) scaled playoff PER’s of all-time

Spoiler:

Top 100(ish) scaled playoff WS/48’s

Spoiler:

2klegend · Post #2 » by **2klegend** » Wed Aug 31, 2016 3:13 am

It's important to know how many games they played that get them those number. That's why I'm generally more impress with Lebron '09 than Hakeem '88. Lebron did it for 21 games. Hakeem only 4. Larger game sample will normalize those stat. For most players, that enough games to win an NBA championship. Lebron may have the most incomplete GOAT season. The most complete GOAT season belong to '91 Jordan. Had Lebron won a championship in '09, then it wouldn't be debatable anymore.

Post #3 » by **Quotatious** » Wed Aug 31, 2016 3:15 am

2klegend wrote:That's why I'm generally more impress with Lebron '09 than Hakeem '88. Lebron did it for 21 games. Hakeem only 4.

LeBron played 14 games in 2009 playoffs. He played 21 this year (2016).

2klegend · Post #4 » by **2klegend** » Wed Aug 31, 2016 3:26 am

Quotatious wrote:
2klegend wrote:That's why I'm generally more impress with Lebron '09 than Hakeem '88. Lebron did it for 21 games. Hakeem only 4.

LeBron played 14 games in 2009 playoffs. He played 21 this year (2016).

My bad. Then those numbers seem skew a bit. Though, still very impressive to put up those stat for 14 games.

I'm consider redoing my peak project which based on playoff game play coefficient rather than round play. This will help normalize every players. Though it may hurt the older player.

Post #5 » by **trex_8063** » Thu Sep 1, 2016 12:52 am

2klegend wrote:It's important to know how many games they played that get them those number.

Sure, and I'd mentioned in OP that the small sample sizes can create some wonky results in some instances. fwiw, I just edited OP to put an "*" by any value that came from a sample that's 7 games or less (i.e. one series).

Post #6 » by **trex_8063** » Sun Jan 8, 2017 5:58 pm

Bumping because I've updated to include ABA player seasons. Of note are '76 Erving and '68 Connie Hawkins (both monsters in playoffs in respective years). McGinnis ('75) makes his first appearance in these lists, as dose Issel ('71), and Jimmy Jones ('69). '71 Zelmo Beaty once again showing up, too.

70sFan · Post #7 » by **70sFan** » Sun Jan 8, 2017 6:05 pm

Wow, Hagan was a real beast back then.

Post #8 » by **trex_8063** » Thu May 23, 2019 4:42 pm

Bump. Upon request I have updated this [since '16] to include playoff runs from '17, '18, and ONLY those '19 runs that are fully completed.

Four playoff runs in the last nearly-three seasons made the cut for playoff Scaled PER:
'18 Lebron James (16th all-time)
'17 Kawhi Leonard (18th)
'17 Lebron James (36th)
'19 Nikola Jokic (70th)
**Kawhi, Giannis, and KD all have a fair chance of squeaking on to the list for their '19 runs, depending upon how they finish out the playoffs.

Four playoff runs from last <3 seasons have made the cut for playoff Scaled WS/48:
'17 Kawhi Leonard (27th)
'17 Kevin Durant (tied for 69th)
'17 Lebron James (85th)
'17 Stephen Curry (88th)

An Unbiased Fan · Post #9 » by **An Unbiased Fan** » Thu May 23, 2019 5:53 pm

PER is already scaled to era, so not sure this is the right approach. Weaker seasons get overrate,d hence why the 70's is all over the Top 10.

More than that.....SRS of opponent's faced would be critical.

Post #10 » by **trex_8063** » Thu May 23, 2019 6:26 pm

An Unbiased Fan wrote:PER is already scaled to era, so not sure this is the right approach. Weaker seasons get overrate,d hence why the 70's is all over the Top 10.

I'd refer you to post #14 of this thread (and read some speculations in OP of that one, too, if you haven't; fwiw, I'm starting to think my 3rd postulation [regarding playing time] might be the most relevant; EDIT: recent edit to 2nd postulation possibly very relevant too).
This study is not to suggest you shouldn't still factor in era-strength considerations; it's addressing the relative parity/disparity seen in certain years, nothing more.

Post #11 » by **penbeast0** » Fri May 24, 2019 12:52 am

70sFan wrote:Wow, Hagan was a real beast back then.

And look at Frank Ramsey in the WS/48 list! Not even a starter (though he was usually 5th in minutes over nominal starter Jim Lotscutoff). What a magnificent playoff performer, which the Celtics needed as Bob Cousy's shooting during this era fell off the table in the playoffs a lot.

Odinn21 · Post #12 » by **Odinn21** » Fri May 24, 2019 3:22 am

An Unbiased Fan wrote:PER is already scaled to era, so not sure this is the right approach. Weaker seasons get overrate,d hence why the 70's is all over the Top 10.

More than that.....SRS of opponent's faced would be critical.

No. PER isn't scaled to era. At least not the ratings.

PER uses iterative methords to achieve 15.0 average across the league but only for that season. The idea behind the stat is, the average is always 15.0. But they don't disclose what would be the averages of player with 15.0000 PER.
Imagine a player with 15.0000 PER in 2017-18, and he had the exact statline for 2018-19 season and his PER in the season won't be 15.0 any more.
We don't know a single thing about the average 15.0 PER player from statistical pov. I know it can sound ridicilous but NBA 2K games are the only PER calculator I know. If you simulate a season with all-time teams in there, you'll see none of the players would break 24.0 PER even though their EFF and GmSc numbers would look good. ( https://i.imgur.com/LZVVFHS.jpg )

That was about PER only. Your point isn't valid IMO. Larry Bird led the league in PER with 25.6 in '85-86 season. The next season Jordan led the league with 29.8. What, the league became that weak over a season?..

RCM88x · Post #13 » by **RCM88x** » Fri May 24, 2019 5:28 am

Win shares are a weird stat.

iggymcfrack · Post #14 » by **iggymcfrack** » Fri May 24, 2019 5:47 pm

“Scaled PER” has to be my least favorite stat ever. PER is already a league adjusted stat! If there weren’t a lot of stars in the 70s and you didn’t have to be much better than average to lead the league, that doesn’t mean the stars who did play in that weak era should get extra credit.

Post #15 » by **trex_8063** » Fri May 24, 2019 11:25 pm

iggymcfrack wrote:“Scaled PER” has to be my least favorite stat ever. PER is already a league adjusted stat! If there weren’t a lot of stars in the 70s and you didn’t have to be much better than average to lead the league, that doesn’t mean the stars who did play in that weak era should get extra credit.

To each his own, but seriously: if we note that whereas we routinely have multiple players with PER's >29 (if not >30) in recent years, and yet see an entire "era" in which no one [or at least: no one outside of Kareem] manages a PER >25 (sometimes not even 24).......red flags should be popping up in your mind.

If past eras are so much weaker in terms of average competition (an opinion you've frequently championed), and there were truly few real "stars", BOTH of those factors would lower the effective quality of the mean, and if anything should make it easier for the rare uniquely talented individual to distance himself from that mean.

imo, there are factors unidentified contributing to greater/lesser parity in some eras compared to others. If you haven't, I'd read the hypotheses #2 and #3 in this thread; the recent edit I just put in #2 can hardly be denied as a potential factor, imo.

This isn't aimed at giving "extra credit", but rather giving some accounting to factors which are clearly [imo] at play wrt parity. NOTE [as has been indicated in multiple responses to others]: this metric does NOT give any particular accounting to strength of era said "Scaled PER's/WS48's" were achieved in (i.e. a Scaled PER of 31.00 obtained in '69 is NOT necessarily superior to a Scaled PER of 28.00 obtained in '19).

feyki · Post #16 » by **feyki** » Mon Jan 4, 2021 8:26 pm

Excellent thread, work. I just want to note that Russell had %38 of his team win shares in the 1962 Playoffs. Only 00 Shaq,03 Duncan,50 Mikan,67 Wilt and 65 Russel have over of %35 TmWS's.

Post #17 » by **penbeast0** » Mon Jan 4, 2021 9:29 pm

Also fun to see Steve Mix in the top 10 for WS/48, without an * even. More one of those small sample size things than an indication of his true abilities but he was a player I really liked back in the day.

Post #18 » by **trex_8063** » Mon Jan 4, 2021 9:34 pm

feyki wrote:Excellent thread, work. I just want to note that Russell had %38 of his team win shares in the 1962 Playoffs. Only 00 Shaq,03 Duncan,50 Mikan,67 Wilt and 65 Russel have over of %35 TmWS's.

Yeah, '62 Russell just barely made the cut-off in scaled ps WS/48 (#111 on the list above), but did so while averaging 48 mpg.....pretty beastly.
'65 Russell is 25th on the list (and while averaging nearly 47 mpg).

'50 Mikan unfortunately we cannot say as WS/48 didn't exist yet (because minutes were not yet tracked). Safe bet that he'd be up there if they had been.

Haven't updated for '20, fwiw.

Post #19 » by **penbeast0** » Mon Jan 4, 2021 9:35 pm

feyki wrote:Excellent thread, work. I just want to note that Russell had %38 of his team win shares in the 1962 Playoffs. Only 00 Shaq,03 Duncan,50 Mikan,67 Wilt and 65 Russel have over of %35 TmWS's.

Another, equally interesting, thread if you want to post it. Minutes are again going to be a big part of this. Wilt is running 47 mpg, Russell 46, Shaq and Duncan around 43, Mikan likely similar.

feyki · Post #20 » by **feyki** » Wed Jan 6, 2021 8:17 am

penbeast0 wrote:
feyki wrote:Excellent thread, work. I just want to note that Russell had %38 of his team win shares in the 1962 Playoffs. Only 00 Shaq,03 Duncan,50 Mikan,67 Wilt and 65 Russel have over of %35 TmWS's.

Another, equally interesting, thread if you want to post it. Minutes are again going to be a big part of this. Wilt is running 47 mpg, Russell 46, Shaq and Duncan around 43, Mikan likely similar.

I had created a thread in a Turkish Bball forum five years ago. I was calling it as ring shares, how much load ATG's take on their championship runs. There are top 10 championship runs:

Spoiler:
1-1962 Bill Russell, 1965 Bill Russell - %38
3-1950 George Mikan,2000 Shaquille O'Neal,2003 Tim Duncan - %35
6-1967 Wilt Chamberlain %34
7-1960 Bill Russell,1963 Bill Russell,1984 Larry Bird,1954 George Mikan %33

.