Poll: Your all-time greatest players

[eminence]

DSmok1 mentioned preferring Value Over Average vs Over Replacement for a GOAT style list, I'd probably split the difference between the two, as I do think there's value to a career in being an average player (essentially a low end starter), but replacement probably overvalues longevity a bit too much for my taste. Mostly theoretical, as players rank quite similarly with either approach.

[schtevie]

Just to better address the point of Kareem's notional BPM in the second and third years of his career (those of max. Win Shares) I was kind of hoping that I could get away with being lazy, and someone would have filled in the few missing BPM variables with reasoned estimates, thereby possibly confirming that my eyeballing of the data/formula was correct. Alternatively, perhaps Mike's BPM conjecture of 12.2 to 13.2 or Eminence's suggestion that it was at least +11 (possible 12) would prove more accurate.

First the caveat. In the complicated formula that defines BPM, there are team adjustment factors necessary to establish levels (what I can't do) as well as an adjustment of Points for Team Context (what I don't understand how to calculate, even if possible for 1971 and 1972). However, one has the coefficients, and in 1972 (choosing that as the supposed peak season) one can come up with either the actual values or very good estimates for all the box score variables. And as such, one can calculate the change on this margin between then 1977 (Kareem's career BPM peak).

And what does this exercise show?

Well, if you assume Kareem's blocks and steals per 100 possessions were the same as in 1974, the last year before joining the Lakers (and I would strongly argue that given the fundamentals involved, that this should be considered a practical upper bound) there is a low and high number generated - the range depending on whether one believes that Kareem was a "Creator" or "Receiver" on offense. (I am supposing it's much closer to the latter, given that he was a classical Center, but Daniel could speak on that point, if he wishes.)

On the low end (assuming Kareem was a pure "Creator") the marginal change to BPM, from 1977 to 1972 is -0.16, and assuming that he was a pure "Receiver", the decrease is -0.46.

If my math was correct, and the never-to-be-known team factor adjustments wouldn't offset this, then it implies that Kareem's BPM peak was in 1977, being more consistent with an "expected" aging curve (and then yay Oscar!)

And then of course, these adjustments would have to be much larger still get from this notional 1972 BPM range of 8.9 to 9.2 to 11, never mind 12 or 13.

Finally, just to bring it all back to my original point, I await Daniel's revision of the statistic with interest, to see whether it pulls Jokic's peak (to date) of 13.7 down below whatever the revised peak of Kareem's (again, currently 9.4 in 1977).

What was the hit that ol' Russell Westbrook took in the previous revision? Hadn't he been up in the 15 range? And now that season is a "lowly" 11.1.

(Re the claim that aging curves quite commonly peak pre age 26 - holding years of experience constant, blah, blah, blah - I would like to slightly qualify my previous remarks. I was not attempting to deny that one couldn't find any such players. My point related to expectations for the greatest of the greats, where one would think that the underlying fundamentals that predicts a curve peaking somewhat later than 26 would be in play. Again, peak physical maturity, on average, occurs 25ish to 26ish? All the while skills and experience are being gained. And post physical peak, on average, decrepitude sets in, and ultimately wills out. So, I was not talking in the instance about knuckleheads with boffo physical skills who fail to adequately learn the finer points of the game that would afford them better and longer careers.)

[Crow]

Jokic at #1 on qualified career BPM with a 10% on Jordan is interesting. Probably little known. Not that interested in riling some or debating but I've known it for awhile.

Jabari at 18th is interesting too. Again not that interested in debating but I don't think I noticed til today.

I have both in my top 10 picks and so does the average of these votes fwiw.

Exact rank? I don't immediately have. No compelling need to for me.

I am interested in how long Jokic sticks around and at what future levels.

Jokic with 4 straight seasons of +13 and slightly higher. Jordan peak at 13.0. Jokic slightly above that right now.


Jokic's ultimate rank overall will certainly depend partly on number of titles for many. 1 was pretty much essential for possible number 1. Another would help a lot. If it is just 1 title, few would argue for #1.
2 probably isn't enough for most. 3 and it is more conceivable. Top 5? Titles are a factor still but I'd say less so.

Should BPM or VORP or a composite of metrics and considerations be the leading determinant? I'd say use everything, if you want to spend the time and strive to impress or convince. Metrics heavily but not necessarily exclusively.

[Mike G]

A formula that "predicts" BPM based on WS/48, covering all positions is:
bpm = (WS/48-.100)*55

This produces an avg error of 1.24 for the 100 (or 200) players who have gone the most minutes this season.
Applying the same formula for Kareem's 1st 10 seasons with BPM, we get similar results.

Code: Select all

yr   BPM  WS/48   bpm?  diff
74   7.5   .250    8.2    .7
75   7.6   .225    6.9   -.7
76   8.5   .242    7.8   -.7
77   9.4   .283   10.0    .6
78   9.3   .257    8.6   -.7
79   7.6   .219    6.5  -1.1
80   7.2   .227    7.0   -.2
81   6.1   .230    7.1   1.0
82   4.7   .192    5.1    .4
83   4.2   .205    5.8   1.6

avg  7.2   .233    7.3    .1

When he was still blocking lots of shots, his actual BPM (2nd column) tended to be better than the formula predicts.
When his team was better (the Magic '80s), his WS rate was rather inflated and overestimates his BPM.

[Crow]

Technically Jordan didn't reach +13 BPM, but did with rounding. They are the only 2 recorded over +12. Curry just short in 2015-16. James a bit short twice. David Robinson in the mix. So is SGA. Then Giannis, then Westbrook. Nobody else over +11 but Chris Paul would be with rounding. Harden right on rounding edge.

James close to +11 a few more times. Erving was closest to +11 before Jordan. McGrady, M Johson, Durant and Garnett got to +10. Doncic with rounding up from +9.95.

Abdul-Jabbar, 42nd best ranked season. Kawhi 43rd, Bird 44th. Shaq 49th, Barkley 50th. And so on. No real surprises except maybe Kevin Love at 71 and / or J Butler at 84. Duncan's best on this metric was 92nd.

https://www.basketball-reference.com/le ... eason.html

120 seasons of +8 or over. An average of about 2.4 per season. 3 each this season and last, 5 in 2022-23. 6 in 2019-20 is the highest. Erving was first to do it, followed by Abdul-Jabbar. Then Walton and Bird and Magic Johnson, Jordan and Stockton. Grant Hill, K Malone, Notwitski, Wade, Ginobli, Giannis and Drexler are the others not previously mentioned. Satisfying to have seen all of these. Different ways to do it.

Others to go over +8? Time will tell.

KAT, Pippen, Billups, Olawujon, Kobe and V Carter fairly close. Haliburton has had a +7 and a near but +5.2 this season. Wemby +6.4 in partial yr 2.

[Mike G]

From (I believe) that very link that Crow posted ^ of career-best BPM, I looked up 20+ players and found the age at which their career BPM peaked. This is typically not the season with their highest BPM; but just before their seasonal BPM dropped below their career BPM and brought it down to "now" (or latest year of play) level.
Current and retired players are mixed in, and so present "now" career BPM is final for some and not for others.

Code: Select all

BPM:     peak   age   @29  last  now   >29
Jokic    10.2   29   10.2   29  10.2   --
Giannis   7.3   30    7.2   30   7.3    .1
Davis     6.7   26    6.4   31   6.0   -.4
Kawhi     6.9   29    6.9   33   6.5   -.4
Kirilenko 6.5   24    5.3   33   4.9   -.4

Bird      7.4   31    6.9   35   6.9    .0
Harden    7.0   31    6.7   35   6.3   -.4
Magic     7.6   31    7.2   36   7.5    .3
Curry     6.6   34    6.4   36   6.5    .1
Durant    7.6   28    7.6   36   6.5  -1.1

BPM:     peak   age   @29  last  now   >29
Erving    7.2   31    7.1   36   6.2   -.9
Barkley   6.8   29    6.8   36   6.1   -.7
Robinson  8.7   30    8.7   37   7.5  -1.2
Shaq      6.5   29    6.5   38   5.1  -1.4
Paul      8.1   31    8.0   39   6.7  -1.3

Jordan   11.0   27   10.9   39   9.2  -1.7
Duncan    6.3   30    6.2   39   5.6   -.6
Garnett   6.8   29    6.8   39   5.6  -1.2
Olajuwon  5.4   32    4.8   39   4.6   -.2
LeBron   10.2   28   10.1   40   8.6  -1.5

BPM:     peak   age   @29  last  now   >29
Stockton  7.3   32    7.0   40   6.8   -.2
Malone    5.4   37    4.0   40   5.1   1.1
Kareem    8.4   30    8.3   41   5.7  -2.6

averages  7.5  29.9   7.2  36.4  6.6   

This is a Joker-centric exercise, hence the @29 column. The avg career-BPM peak is at 30 for these players.
Does not include anything for Kareem pre-1974.

A linear formula to predict BPM decline for this group is -.12 per year past 29
And this is with the freakish Karl Malone, maxing at age 37!

[Mike G]

Striking to me is that Kareem seems to be the only player with a top-250 WS/48 season both before and after the 1973-74 break, aka pre- and post-BPM.
https://www.basketball-reference.com/le ... eason.html
https://www.basketball-reference.com/le ... eason.html

These 2 lists of 250 highest BPM and WS/48 for a season, I pared down to 212 who appear in both lists OR in just the WS list but pre-1974. In the course of "predicting" unknown BPM from known (?) WS/48, those old-timers are still just ranked by their WS rates.

The early '70s were a clear changing of the guard, with some brief appearances in the list; and the later '70s/early '80s rather desperate for superstars.
* - ABA stars.

Code: Select all

#  year   alltime top 250
4  1952   VMikkelsen  GMikan  PArizin  EMacauley
6  1953   GM  EM  DSchayes  VM  HGallatin  NJohnston
4  1954   DS  NJ  GM  EM
4  1955   LFoust  NJ  CShare  DS
3  1956   LF  NJ  BPettit
1  1957   NJ
1  1958   FRamsey
2  1959   KSears  BP

1  1960   WChamberlain
3  1961   WC  BP  EBaylor
2  1962   WC  WBellamy
2  1963   WC  ORobertson
4  1964   WC  OR  BRussell  JWest
3  1965   JW  OR  BR
3  1966   WC  JW  OR
2  1967   WC  OR
3  1968   CHawkins* WC  JW
3  1969   CH* JJones* WReed

3  1970   WFrazier  JW  WR
2  1971   KAbdulJabbar  ZBeaty*
4  1972   KAJ  CWalker  AGilmore* ZB*
3  1973   KAJ  AG* WC
3  1974   JErving* BMcAdoo  BLanier
2  1975   JE* BM
2  1976   JE* KAJ  
1  1977   KAJ
0  1978
0  1979

1  1980   KAJ
1  1981   JE
1  1982   JE
0  1983
0  1984
1  1985   LBird
1  1986   LB
3  1987   MJordan  LB  Magic
4  1988   MJ LB  JStockton  CBarkley
4  1989   MJ  Magic  JS  CB

5  1990   MJ  Magic  JS  CB  DRobinson
5  1991   MJ  CB  Magic  DR  TPorter
2  1992   MJ  DR
4  1993   MJ  CB  HOlajuwon  KMalone
2  1994   DR  SO'Neal
2  1995   DR  JS
6  1996   MJ  DR  TBrandon  AHardaway  KM  ASabonis
2  1997   MJ  KM
3  1998   MJ  DR  KM
3  1999   SO  KM  DR

2  2000   SO  KM
1  2001   SO
2  2002   SO  TDuncan
4  2003   TMcGrady  TD  DNowitzki  SO
2  2004   KGarnett  TD
4  2005   KG  TD  DN  MGinobili
6  2006   LJames  DN  KG  DWade  EBrand  MG
3  2007   DN  MG  TD
5  2008   LJ  CPaul  MG  KG  CBillups
3  2009   LJ  CP  DW

2  2010   LJ  KDurant
2  2011   LJ  CP
4  2012   LJ  CP  DW  KD
3  2013   LJ  KD  CP
4  2014   KD  KLove  LJ  CP
4  2015   SCurry  ADavis  JHarden  CP
6  2016   SC  KD  KLeonard  LJ  CP  RWestbrook
7  2017   KL  KD  JH  CP  NJokic  JButler  SC
5  2018   JH  SC  DLillard  CP  AD
3  2019   JH  GAntetokounpo  AD

3  2020   GA  JH  AD
5  2021   NJ  GA  JB  JEmbiid  KL
3  2022   NJ  GA  JE
3  2023   NJ  JE  JB
3  2024   NJ  GA  SGilgeousAlexander
2  2025   NJ  SGA

1952: Mikkelsen edged out Mikan due to his astronomical .419 FG%
That is a good batting average, after all.

[DSMok1]

One big factor in those older seasons is that we are measuring versus league average. Clearly the number of players and the number of teams in the league have a huge impact on the level of League average. It's hard for a team to be very many points above League average if League average is the fourth best team.

Conversely you could say that the talent pool has also changed. It gets very difficult to adjust appropriately.

I attempted to adjust for this using a Delta method here:
viewtopic.php?p=26131#p26131

[DQuinn1575]

Striking to me is that Kareem seems to be the only player with a top-250 WS/48 season both before and after the 1973-74 break, aka pre- and post-BPM.
https://www.basketball-reference.com/le ... eason.html
https://www.basketball-reference.com/le ... eason.html

These 2 lists of 250 highest BPM and WS/48 for a season, I pared down to 212 who appear in both lists OR in just the WS list but pre-1974. In the course of "predicting" unknown BPM from known (?) WS/48, those old-timers are still just ranked by their WS rates.

The early '70s were a clear changing of the guard, with some brief appearances in the list; and the later '70s/early '80s rather desperate for superstars.
* - ABA stars.

Code: Select all

#  year   alltime top 250
4  1952   VMikkelsen  GMikan  PArizin  EMacauley
6  1953   GM  EM  DSchayes  VM  HGallatin  NJohnston
4  1954   DS  NJ  GM  EM
4  1955   LFoust  NJ  CShare  DS
3  1956   LF  NJ  BPettit
1  1957   NJ
1  1958   FRamsey
2  1959   KSears  BP

1  1960   WChamberlain
3  1961   WC  BP  EBaylor
2  1962   WC  WBellamy
2  1963   WC  ORobertson
4  1964   WC  OR  BRussell  JWest
3  1965   JW  OR  BR
3  1966   WC  JW  OR
2  1967   WC  OR
3  1968   CHawkins* WC  JW
3  1969   CH* JJones* WReed

3  1970   WFrazier  JW  WR
2  1971   KAbdulJabbar  ZBeaty*
4  1972   KAJ  CWalker  AGilmore* ZB*
3  1973   KAJ  AG* WC
3  1974   JErving* BMcAdoo  BLanier
2  1975   JE* BM
2  1976   JE* KAJ  
1  1977   KAJ
0  1978
0  1979

1  1980   KAJ
1  1981   JE
1  1982   JE
0  1983
0  1984
1  1985   LBird
1  1986   LB
3  1987   MJordan  LB  Magic
4  1988   MJ LB  JStockton  CBarkley
4  1989   MJ  Magic  JS  CB

5  1990   MJ  Magic  JS  CB  DRobinson
5  1991   MJ  CB  Magic  DR  TPorter
2  1992   MJ  DR
4  1993   MJ  CB  HOlajuwon  KMalone
2  1994   DR  SO'Neal
2  1995   DR  JS
6  1996   MJ  DR  TBrandon  AHardaway  KM  ASabonis
2  1997   MJ  KM
3  1998   MJ  DR  KM
3  1999   SO  KM  DR

2  2000   SO  KM
1  2001   SO
2  2002   SO  TDuncan
4  2003   TMcGrady  TD  DNowitzki  SO
2  2004   KGarnett  TD
4  2005   KG  TD  DN  MGinobili
6  2006   LJames  DN  KG  DWade  EBrand  MG
3  2007   DN  MG  TD
5  2008   LJ  CPaul  MG  KG  CBillups
3  2009   LJ  CP  DW

2  2010   LJ  KDurant
2  2011   LJ  CP
4  2012   LJ  CP  DW  KD
3  2013   LJ  KD  CP
4  2014   KD  KLove  LJ  CP
4  2015   SCurry  ADavis  JHarden  CP
6  2016   SC  KD  KLeonard  LJ  CP  RWestbrook
7  2017   KL  KD  JH  CP  NJokic  JButler  SC
5  2018   JH  SC  DLillard  CP  AD
3  2019   JH  GAntetokounpo  AD

3  2020   GA  JH  AD
5  2021   NJ  GA  JB  JEmbiid  KL
3  2022   NJ  GA  JE
3  2023   NJ  JE  JB
3  2024   NJ  GA  SGilgeousAlexander
2  2025   NJ  SGA

1952: Mikkelsen edged out Mikan due to his astronomical .419 FG%
That is a good batting average, after all.

On Kareem, Russell, Wilt, Oscar, West, Baylor - maybe the 5 greatest players ever at that point - all came into the league within 5 years of each other. Then we really had a "slump" of all-time greats, with no one at that super elite level entering until Kareem 9 years later, and then no one at that level (other than Walton) until late 70s (you can debate Walton, Doctor J, Moses). Some of this would be due to the baby boom, as the 1946 births wouldn't be hitting the NBA until 1968.
One of the results of this I believe is that it really helped the Boston dynasty. From 1962-69 no all-time great entered the league to change the balance, and the best player that did come was Havlicek, who went to Boston. So they started 1961 with the best teams, and over the next 8 years they got the best player to join the league.

[Mike G]

I see Willis Reed and Walt Frazier in the list. No Havlicek, Moses, or Walton.
Havlicek once led the league in FGA/min (and 4th in FG made), and never among the top 20 in FG%.
His last year as 1st team all-NBA, he shot .509 TS% vs league avg .503. Last year on 2nd team, .508 vs .504.
He took lots of shots for teams that needed someone to shoot. He stayed on the floor also due to his versatility.

Julius Erving's numbers just cratered for 3 years upon merging with the Sixers, sharing the floor with McGinnis. Then he was as good or better than his ABA glory, if you discount ABA numbers.
And Kareem had arguably about 4 'peaks', each less profound, unless it's amazing to be all-NBA 1st team at 38.
Note the 9 year stretch when there were none but these 2 among the 250 BPM/WS best.

[schtevie]

Many thanks to Daniel for providing a reference to his work a decade ago. What a nice trip down memory lane. And he's quite right that in these GOATiness discussions, era adjustments matter...a lot. (Maybe he'll have interest someday in updating the graph?)

Typically, people's views on these matters seem to be quite rigidly held, but I would like to refer to the graph presented, just to add another comment on my questioning when Kareem's career peak actually occurred, or if perhaps he had a multiple of these.

It so happens that the alleged early first peak coincided with the low point of league strength. Comparing '71/'72 to the documented BPM peak in 1977, you see what looks like a half point should be subtracted from the former to make the comparison era-compatible.

And then, of course, one should also take heed of Daniel's final two sentences of that post: "Statistically, there is a fair amount of error possible, not in the overall shape of the adjustment curve, but in the overall slope--in other words, it is quite possible the left end of the line should be moved up or down a point or two do to some slight systematic bias. So take it with a grain of salt."

If one believes that an average NBA team from the early to mid-50's would be even better than one from the early Naughts, then Kareems notional BPM from 1972 would need to be adjusted the aforementioned, approximate 0.5. Of course, one really shouldn't believe that, and altering the slope of the curve would make the downward adjustment of 1972 (relative to 1977) a little bit greater still.

[Mike G]

Yeah, it's very speculative whether you could just time-machine the 1955 Syracuse Nationals into today's NBA and have an average team.
Equally uncertain is whether the 2025 Bulls would do very well if they teleported to 1955. The shoes, the courts, the fans, the travel, the food, the racism...,

One can imagine Bob Pettit being born in 2000 and having better shoes and trainers and nutrition and exercise and coaching. It's literally endless uncertainty, one on top of another.

In Kareem's breakout 2nd season of 1971, there were 3 expansion teams. The NBA went from 14 teams to 17, so 18% of teams were made up of subs and scrubs. Many of these guys had the best season of their lives, statistically.

But the league did a strange experiment with the scheduling: These 3 teams each played the other two 12 times. The other 14 teams only met them 4 times each, while playing one another 5-6 times. So while the players for Por, Buf, and Cle had relatively easy schedules, everyone else benefitted much less.

The brand-new Blazers finished a respectable 29-53, thanks to going 16-8 vs the Cavs and Braves; 13-45 vs everyone else.

The bigger "inflation" issue, I believe, is that the ABA was gobbling up most of the premier draft talent. Hawkins, Haywood, Issel, Gilmore, Erving, McGinnis ...

[DSMok1]

Many thanks to Daniel for providing a reference to his work a decade ago. What a nice trip down memory lane. And he's quite right that in these GOATiness discussions, era adjustments matter...a lot. (Maybe he'll have interest someday in updating the graph?)

Typically, people's views on these matters seem to be quite rigidly held, but I would like to refer to the graph presented, just to add another comment on my questioning when Kareem's career peak actually occurred, or if perhaps he had a multiple of these.

It so happens that the alleged early first peak coincided with the low point of league strength. Comparing '71/'72 to the documented BPM peak in 1977, you see what looks like a half point should be subtracted from the former to make the comparison era-compatible.

And then, of course, one should also take heed of Daniel's final two sentences of that post: "Statistically, there is a fair amount of error possible, not in the overall shape of the adjustment curve, but in the overall slope--in other words, it is quite possible the left end of the line should be moved up or down a point or two do to some slight systematic bias. So take it with a grain of salt."

If one believes that an average NBA team from the early to mid-50's would be even better than one from the early Naughts, then Kareems notional BPM from 1972 would need to be adjusted the aforementioned, approximate 0.5. Of course, one really shouldn't believe that, and altering the slope of the curve would make the downward adjustment of 1972 (relative to 1977) a little bit greater still.

I do hope someday to update that graph. I intend when I complete my next rework of BPM to extend it to older eras, which would be an additional data point. (Understanding it would not really be exactly the same BPM but a modified version that uses fewer stats.)

It's difficult to overstate how important it is to understand the limitations of the delta method.

The delta can absolutely create the shape of a curve, by looking at how players shift from one season to the next, and thus identify high and low points in NBA talent.

However, it is statistically impossible to estimate the first order slope of the overall curve from the 1950s to now. Because, in order to generate the year to year delta offsets for each player, we must subtract out aging. And aging is itself estimated by looking at the average aging behavior over a large set of players over many years.

So, the aging curve will itself be sloped upward or downward systematically depending on how it was developed, and again it also cannot be established without the potential for systematic bias, slope over the entire curve. There is no way to evaluate this from the data set itself.

The only way to evaluate long-term behavior of the NBA data set as a whole is to use exterior data to inform what we are looking at. I have an approach in mind based on the approximate size of the overall NBA talent pool and population from which it is drawn, assuming the NBA talent is the right tail of the populations overall talent distribution. We should be able to get moderately close to estimating the size of the base source population by evaluating state or country populations and distribution of birth places.

[Mike G]

... estimating the size of the base source population by evaluating state or country populations and distribution of birth places.

This would help to size up NBA talent concentration?

Re: Aging, it seems there's a range around 24-28/29 in which about half of players will be better toward the end of that span, the other half will be worse, and there is nominal curvature. So that's a good size database between any 2 consecutive seasons.

Would it make sense to use the ABA curve to give ABA player-seasons their own BPM which do not sum to zero?

[schtevie]

Yeah, it's very speculative whether you could just time-machine the 1955 Syracuse Nationals into today's NBA and have an average team.
Equally uncertain is whether the 2025 Bulls would do very well if they teleported to 1955. The shoes, the courts, the fans, the travel, the food, the racism...,

One can imagine Bob Pettit being born in 2000 and having better shoes and trainers and nutrition and exercise and coaching. It's literally endless uncertainty, one on top of another.

It's really not speculative or uncertain. Of course you need some basic conceptual ground rules for this "average field of dreams" counterfactual. What you are looking for is a comparison of individual skill level, not the evolution of the game itself (i.e. improvements in offensive and defensive organization and strategy). So, in fairness to the old-timers, you would at least have to grant them a "training camp," where all that was learned over the ensuing half century would be available to them as well.

Looking at Daniel's era-adjustment chart, the implication is that the average players in 1952 (what looks to be the start of the data) were (marginally) better than their grandchildren in 2005. However, below are the comparisons for various scoring propensities for these two years.

Year 3P% 2P% TS% Ortg Pace
1952 0.367 0.438 86.9 95.1
2005 0.356 0.47 0.529 106.1 90.9

Difference 0.103 0.091 19.2

Unless one is of the belief that there was some secret defensive sauce from the '50s that was mysteriously lost over time, the only speculation based on these numbers is the exact dimension of the massacre. For every 5 two-pointers taken in a game between average teams of the two eras, the moderns' lead on the scoreboard would increase by a point. To make the game "fair" they'd pretty much have to only be taking shots from three-point land.

And there's no reasonable accommodation of the ground rules that could overturn this margin of superiority. Players having to worry about being lynched is out of bounds. And if the argument hangs on the moderns losing because they would have to wear Chuck Taylors, yeah, no.

In Kareem's breakout 2nd season of 1971, there were 3 expansion teams. The NBA went from 14 teams to 17, so 18% of teams were made up of subs and scrubs. Many of these guys had the best season of their lives, statistically.

But the league did a strange experiment with the scheduling: These 3 teams each played the other two 12 times. The other 14 teams only met them 4 times each, while playing one another 5-6 times. So while the players for Por, Buf, and Cle had relatively easy schedules, everyone else benefitted much less.

The brand-new Blazers finished a respectable 29-53, thanks to going 16-8 vs the Cavs and Braves; 13-45 vs everyone else.

The bigger "inflation" issue, I believe, is that the ABA was gobbling up most of the premier draft talent. Hawkins, Haywood, Issel, Gilmore, Erving, McGinnis ...

This is an interesting historical tidbit. I guess it would imply that the era-adjustment would be player/team dependent, with appropriate weights. But still negative.