RealGM 2023 Top 100 Project - #2 (Kareem Abdul-Jabbar)

Moderators: Clyde Frazier, Doctor MJ, trex_8063, penbeast0, PaulieWal

OhayoKD
Head Coach
Posts: 6,042
And1: 3,932
Joined: Jun 22, 2022

Re: RealGM 2023 Top 100 Project - #2 (Kareem Abdul-Jabbar) 

Post#221 » by OhayoKD » Fri Jul 7, 2023 7:27 pm

rk2023 wrote:
Doctor MJ wrote:Jordan dropping below Kareem, and dropping so decisively is almost certainly going to be the big event of this project even presuming he just drops to 3.

I think we're definitely going to want to further analyze what factors resonated to shift the ground so dramatically.


Perhaps it could be more of an appreciation for longevity, based on what I’ve seen from voting similarities across posters. Unsure how much to consider it a personal chicken-egg effect, where fans are seeing some of this gens’ players (eg. Paul Curry Durant James) do impressive things well past the age of 30 threshold. If people might be fans/admirers of those players, perhaps longevity resultedly becomes more of a emphasis amidst criterion. Perhaps admiration for players doing that (inversely) gives people a greater sense of appreciation for the sport. There’s a few different ways to interpret it, but a higher consensus on Kareem than usual and more open-minded conversation I feel would be driven by longevity.

Would say Kareem as a goat-tier peak has gotten more traction, though many of the regulars here who have kareem as or high or higher at his apex are not participating in this project
rk2023
Starter
Posts: 2,265
And1: 2,270
Joined: Jul 01, 2022
   

Re: RealGM 2023 Top 100 Project - #2 (Kareem Abdul-Jabbar) 

Post#222 » by rk2023 » Fri Jul 7, 2023 7:31 pm

OhayoKD wrote:
rk2023 wrote:
Doctor MJ wrote:Jordan dropping below Kareem, and dropping so decisively is almost certainly going to be the big event of this project even presuming he just drops to 3.

I think we're definitely going to want to further analyze what factors resonated to shift the ground so dramatically.


Perhaps it could be more of an appreciation for longevity, based on what I’ve seen from voting similarities across posters. Unsure how much to consider it a personal chicken-egg effect, where fans are seeing some of this gens’ players (eg. Paul Curry Durant James) do impressive things well past the age of 30 threshold. If people might be fans/admirers of those players, perhaps longevity resultedly becomes more of a emphasis amidst criterion. Perhaps admiration for players doing that (inversely) gives people a greater sense of appreciation for the sport. There’s a few different ways to interpret it, but a higher consensus on Kareem than usual and more open-minded conversation I feel would be driven by longevity.

Would say Kareem as a goat-tier peak has gotten more traction, though many of the regulars here who have kareem as or high or higher at his apex are not participating in this project


Access to new information may change views as well. For example, how nuanced 70s (and others) Kareems’ analyses and deep dives have been. Agree.
Mogspan wrote:I think they see the super rare combo of high IQ with freakish athleticism and overrate the former a bit, kind of like a hot girl who is rather articulate being thought of as “super smart.” I don’t know kind of a weird analogy, but you catch my drift.
Doctor MJ
Senior Mod
Senior Mod
Posts: 53,227
And1: 22,236
Joined: Mar 10, 2005
Location: Cali
     

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#223 » by Doctor MJ » Fri Jul 7, 2023 7:50 pm

OhayoKD wrote:
Doctor MJ wrote:
OhayoKD wrote:But you're not a big fan of raw "impact" so why don't we use AUPM? A combination of on/off and...BPM

Let me say that again: ON/OFF and...

...BPM

And you know whose 3-year playoff peak scores higher? In the metric which literally thinks Micheal Jordan's weakside-help is more valuable than Duncan's paint-deterrence?


I must point out that you're confusing BPMs - which is quite understandable of course since it's so hard to keep all of these similarly named things straight.

Ben's assessment of his own BPM is that it weights big man blocks considerably more strongly than Daniel Myers' BPM (used on bkref).

Oh wow. That's a pretty significant misstep on my part. Will not include that then on my duncan explanation next thread.

Thank you for clarifying. Is there a place where ben explains the specific methodology with his own BPM?


As I say, understandable. I know I've made mistakes like this before.

To your question: Looking and not seeing it. I'll ask.
Getting ready for the RealGM 100 on the PC Board

Come join the WNBA Board if you're a fan!
iggymcfrack
RealGM
Posts: 11,740
And1: 9,239
Joined: Sep 26, 2017

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#224 » by iggymcfrack » Fri Jul 7, 2023 9:26 pm

Doctor MJ wrote:Initially Induction tally yields:

Kareeem 10 (trelos, rk, ceoofk, Dr P, Ohayo, hcl, eminence, falco, Colbinii, LA B)
Jordan 6 (ltj, trex, iggy, DQuinn, Clyde, f4p)
Russell 5 (beast, AEnigma, Lou, ShaqA, Doc)
Duncan 1 (OaD)
Wilt 1 (ZPage)

Re-allocating 2nd votes:

Kareem adds 4 (OAD, AEnigma, ShaqA, ZPage)
Jordan adds 3 (beast, Lou, Doc)

Kareem wins 14-7 over Jordan.

Kareem Abdul-Jabbar Inducted as #2 on Top 100, up one spot from the 2020 project.

Image

Nomination count:
Hakeem 6 (rk, AEnigma, Ohayo, ShaqA, f4p, LA B)
Shaq 4 (OaD, ceoofk, Dr P, iggy)
Mikan 3 (beast, eminence, Moonbeam)
Garnett 1 (Lou)
Magic 1 (Doc)

Hakeem Olajuwon becomes a Nominee.

Image


14-7 is nuts. This is something I’m actually going to bring up to people IRL (especially one particular Jordan stan I know).
:lol:
User avatar
OldSchoolNoBull
General Manager
Posts: 9,078
And1: 4,468
Joined: Jun 27, 2003
Location: Ohio
 

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#225 » by OldSchoolNoBull » Fri Jul 7, 2023 9:38 pm

iggymcfrack wrote:
Doctor MJ wrote:Initially Induction tally yields:

Kareeem 10 (trelos, rk, ceoofk, Dr P, Ohayo, hcl, eminence, falco, Colbinii, LA B)
Jordan 6 (ltj, trex, iggy, DQuinn, Clyde, f4p)
Russell 5 (beast, AEnigma, Lou, ShaqA, Doc)
Duncan 1 (OaD)
Wilt 1 (ZPage)

Re-allocating 2nd votes:

Kareem adds 4 (OAD, AEnigma, ShaqA, ZPage)
Jordan adds 3 (beast, Lou, Doc)

Kareem wins 14-7 over Jordan.

Kareem Abdul-Jabbar Inducted as #2 on Top 100, up one spot from the 2020 project.

Image

Nomination count:
Hakeem 6 (rk, AEnigma, Ohayo, ShaqA, f4p, LA B)
Shaq 4 (OaD, ceoofk, Dr P, iggy)
Mikan 3 (beast, eminence, Moonbeam)
Garnett 1 (Lou)
Magic 1 (Doc)

Hakeem Olajuwon becomes a Nominee.

Image


14-7 is nuts. This is something I’m actually going to bring up to people IRL (especially one particular Jordan stan I know).
:lol:


Shouldn't it be 14-9 though? Jordan had 6 + 3 re-allocated 2nd votes.

In any case, I in fact do think it's nuts either way, but as an MJ fan I'm just going to let it go and not get worked up about it.
f4p
Sixth Man
Posts: 1,716
And1: 1,731
Joined: Sep 19, 2021
 

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#226 » by f4p » Fri Jul 7, 2023 9:50 pm

OhayoKD wrote:
f4p wrote:
so let me ask, why do you think considering playoff risers and fallers can't be done? in your 25 year RAPM thread, in your first post you had a section for playoff risers and fallers. so you don't seem disinclined to acquiesce to the idea that they exist. i certainly didn't come up with the concept and am not the only one who discusses it on this board. why shouldn't they exist on a team basis? if you have 82 games to prove yourself as a vastly superior team to another and then lose to them, why should i say you were actually just worse all along. at the very least, if we're ascribing lift to these teams, then we would seem to need to lower that lift a lot when you end up losing as a 4.7 SRS favorite. it would be fair to say that whatever lift you are ascribing to kareem and jordan, it takes a substantial shift in the direction of jordan once the postseason is factored in.

And why are you assuming I don't? The problem here is kareem was not "consistently" losing as a +4.7 favorite.


but he was. this is the list of people i have tracked for playoff SRS differential purposes:

Spoiler:
Hakeem Olajuwon
Michael Jordan
Lebron James
Shaquille O'neal
Larry Bird
Magic Johnson
Tim Duncan
Kareem Abdul-Jabbar
Wilt Chamberlain
Bill Russell
Kobe Bryant
James Harden
Isiah Thomas
Kevin Garnett
Charles Barkley
Dirk Nowitzki
Patrick Ewing
Julius Erving
Moses Malone
Karl Malone
Stephen Curry
Kevin Durant
Dwyane Wade
Steve Nash
Reggie Miller
Dominique Wilkins
Allen Iverson
Carmelo Anthony
Russell Westbrook
Paul Pierce
Tracy McGrady
Jimmy Butler
Giannis Antetokounmpo
David Robinson
Jerry West
Elgin Baylor
Oscar Robertson
John Havlicek
Jason Kidd
Gary Payton
Chris Paul


kareem is the only one with 3 losses as +4. and only 3 have lost even 2 series as a +4 favorite:

Gary Payton and the sonics epic back to back 1994 and 1995 1st round losses
Karl Malone
Oscar Robertson

and both of the Big O's were with Kareem. so basically only two other players have lost even 2 such series. about 70% of the players don't have even a single loss as a +4 favorite. kareem is literally the GOAT of losing heavily favorited series. now kareem played in the 2nd most behind duncan, but as mentioned, his losing percentage is still more than double the average.



As I outlined in my first post on this thread, the Bucks got better in 71 and 72 using sans's psrs.


is that the sans list that had the 2018 rockets as the 95th best team of all time? either way, as an aside, i think (and will maybe make a thread) postseason SRS needs to be calculated slightly differently. as far as i can tell (maybe i am wrong), it traditionally it gets calculated for a given series as:

Postseason SRS = Opp SRS + Series MOV

this assumes your opponent has played at their normal level (and you overperformed/underperformed them), with no possibility for the idea that you played at your normal level and your opponent overperformed/underperformed you. as such, it produces different results for the 2 teams in a series depending on which team is your reference team.

i think it should be (unless it already is):

Postseason SRS = (Opp SRS + Tm SRS + Series MOV) / 2

this produces the same result when looking at either team. back to kareem...

All three of the 71, 72, and 74 Bucks were better than the 91 and 90 Bulls(compared to the latter 2) respectively. The 77 Lakers were on-par with the 88 Bulls after a better regular season. And I have Jordan having more help all those postseasons. Moreover you seem inclined to give Jordan all the credit for his "lift", but in 1990 do you know what increased from the regular season? The Bulls defensive rating. Do you know who purely by box-aggregations(which miss a bunch of things Jordan was helped with) like BPM saw the biggest rise? Pippen.


the 71 and 72 bucks are excellent, i agree. one of the few +10 SRS teams to ever lose, but only because they ran into an even bigger juggernaut. kareem certainly shot out of the gates quickly and dominantly. then he started losing +4 series, missing the playoffs, having 44.7 TS% series, getting outplayed by moses twice (losing to a 40-42 team one of those times), and picking up most of his titles while not being the best player on his team any more. you can't play with magic (and then worthy) and win 5 out of 9 years and then say playing with scottie pippen is unfair on the way to 6 out of 7.

Do you know what the 94 Bulls did in the playoffs while their best players were at each other's throats?


played well but still lost? i have them with 0.668 expected series win and they got 1, for a 2.0 SRS rise.


In 74 with even weaker help, the Bucks did just as well against a better opponent after being by far the best regular season team. As far as I'm concerned, Kareem was flat-out better, regular season and playoffs, and consequently achieved comparable or better playoff results against better opponents without a system equivalent to the triangle(refer back to blackmill's notes on suboptimal deployment), or a team so good they could **** off with all the chemistry and still conceivably win by themselves. Even if I gave jordan all the credit for the team-wide elevation(obviously undeserved), he still looks like a worse riser than Hakeem(who we're just nominating) and Lebron(who you dont have at #1 even though the Cavs playoff lineups which were better lebron and stars, lebron and no stars, and worse with stars and no lebron)


jordan is certainly a worse playoff riser than hakeem. no one except healthy kawhi compares to hakeem. and you tend to give a lot of credit to the triangle. but when jordan left after '93, the bulls rORtg fell by 7. then, perfectly balanced, it went up by +7 when he returned. whatever magic the triangle may have held was gone after a few years (i have to say, just watching bulls games, it does not strike me as nearly helpful to a team as anything like SSOL or an adelman motion offense or a kerr motion offense. easy buckets and backdoor layups seem few and far between, with a lot of possessions involving a lot of meandering passes around the perimeter before someone realizes the shotclock is under 10 and just gives it to scottie or michale, whoever is closer).

....

can we really discount the regular season that much? it's not as if jordan was a lazy regular season player who made it easy on himself to overperform.

Lazy or not lazy, he is shooting guard who never outputted significant defensive impact(at least by "lift) as shooting guards do not do. He also was not a lebron/magic/nash type guy who could monopolize team-wide offense to excellent success, and while his biggest advantage was ball-handling, he only achieved impressive offenses as a secondary ball-handler.


i feel like this is part of the issue with your jordan comments. you give him credit for almost literally nothing. there's always an adjustment over here, a full-strength over there, a lift, a helpful teammate, a system. something that means it was never jordan. ever. even lower-end lebron seasons will get held up as possibly above peak jordan. i think you've even written sentences akin to "the bulls offense didn't pick up until the triangle and then the bulls defense didn't fall off when jordan retired". the obvious implication being that jordan had no impact on offense or defense. it's amazing he could tie his shoes in the morning without phil or scottie around to tell him how to do it. the guy with the dominant stats, who everyone who played or coached against him said was the best, who then won more than anyone since russell, only it wasn't him. he would have to have the most unimaginably good luck ever for it to all be everyone else and not really have much to do with him, and maybe even more luck to have no one alive at the time notice it was everyone else. i mean maybe if this board had put forth a great "scottie pippen is top 10" argument over the years, i could see it. but even here he's in the 30's.

In the year you see as his peak, he was mostly shooting over single coverage while a "unpolished" offensive player like say 2009 Bron was mantaining similar "end-of-possession" efficiency while doing more through the duration of possessions against significantly more defensive coverage.


illegal defense certainly helped jordan's offense, though i do think it limited his defensive impact as he was never allowed to roam the way a modern player can. but then bill russell's interior defense in a league that can't shoot is almost inarguably a better era boost and kareem playing in by far the most watered down, expanded league in nba history was also a huge boost he took very little advantage of.

So yeah, I do not think laziness is required here. He is not forcing game-winning play-calls or subsitutions on one-end. He is not a top-tier passer which means his looks are not as high quality(passer-rating) and he has no hope of even dreaming of the defensive influence from a guy like Lebron in his 30's never mind a proper **** big like Kareem. He was way worse than Kareem in his first 6 years of basketball. He took a break and was still fell off quicker. He also demonstrated he didn't understand his own limitations in washington and where Kareem just moved on from Magic trying to upsurp him as a rookie, Jordan blew-a-team up because he didn't always get his way.


kareem presumably was better comparing only their first 6 years. but i don't see how jordan fell of quicker. as someone pointed out, jordan got a POY win at age 34, with kareem only getting one at 32. and by the time the 1998 playoffs rolled around, jordan was 35 because his birthday is just after the BBRef cut-off. kareem was not dominating at 35 like jordan, even if he kept playing past 40.

Kareem should be better, and the results tell me he is better, so why do I care about what the "expected championship differential" is when he's leading better or comparable regular season and playoff teams with less help over and over and over again?


he's not leading comparable playoff teams, though. one guy won 6 as the best playing while going 90-26 in those playoffs. kareem has nothing like that stretch as a leader. if you want to claim it's because he had less help, ok, but the results certainly weren't better or even on par.

for the 70's, kareem's regular season SRS said he should win 10.9 playoff series, or 68.3%. he won 9, or 56.3%. to change odds like that for a 7 game series, that's a shift of 1.5 SRS points underperformance.

for pre-1991, jordan regular season SRS said he should win 3.4 playoff series, or 31.1%. he won 5, or 45.5%. that's a shift of 1.8 SRS points overperformance.

so that's a shift of 3.3 SRS points in total between the two, or about 9 wins. based on the playoffs. for basically smack dab in the middle of his prime kareem compared to jordan. is the difference between them before the playoffs really more than 9 wins?

Doesn't seem unreasonable, especially since that included 84-87 MJ who probably wasn't as good at basketball as pre-nba Kareem. For the "peaks", 72 and 71 saw improvement per San's calc. Compare that to 90/91 and I think the playoffs significantly bolster Kareem's case. Also M.O.V here allows for us to compare losses. 1972 may not have ended in a win, but i'd say in terms of "lift" that playoff outcome is more impressive than anything from Jordan. Ditto with 1974 where the help was weaker than 1972.


MOV obscures the actual winning and losing though. especially give the small sample of a playoff series. did kareem really win the 1973 warriors series since his team had a +2.5 MOV? of course not. he played horribly and lost to a lesser team, even if both victories by his team were blowouts to swing the MOV to his team. this goes back to my "jordan never gets any credit" thing. even by MOV, 1974 is kareem losing by almost 5 ppg to a team he had a +4 SRS advantage over. that seems like a lot of underperformance by winning or MOV.

and it's hard to do kareem in the 80's because of the conflation with magic being the signal, but for jordan from 91-98, even including 95:

jordan's regular season SRS said he should win 20.3 playoff series, or 78.0%. he won 25, or 96.2%. that's a shift of 4.8 SRS points, or 13 wins. jordan is just massively outperforming what it seems like his team should have done.

The Bulls are, yes, but again, they didnt even need Jordan to do that. And like I said, using "converting chances" instead of mov forces us to reduce the sample


once again, jordan's team has the best playoff winning percentage over a sustained stretch, improved by 13 wins, but it was really the others (and i only have the '94 bulls as +2 SRS compared to +4.8).

and since we may never agree, why are you so low on converting chances into titles? the nba is maybe the most predictable league in the world. the winning team has the best player or 2nd best player, a lot. the best players tend to win even more than things like SRS suggest. i don't value these things just to help jordan, i value them because they seem some of the clearest signals from nba history. here is the SRS favorite record i posted in one of my earlier duncan posts (with curry updated for 3 losses now):

Because it reduces what can be used as evidence. Again, by this metric 1972 and 1974 and 1977 are all "negatives" when if we compare opponent and m.o.v(curve for proportionality) they are outright better or more impressive than what peak Jordan did. I would need to see what specifically you're comparing for the Lakers because beyond 80(another team that rose in the playoffs), Jordan should look better.


what do you mean by "comparing for the lakers"? before magic or after? in his 3 playoffs before magic, he had 2.33 expected series wins and got 2, so give or take pretty close to expected. maybe 1 SRS off.

and i'll go back to my question. you seem to focus a lot on jordan in his early days, constructing a case he wasn't really valuable. but i'll ask again, what else could he have done from 1991-98?

Match the 71 Bucks? Outperform the Best Spurs sides more than once? Not get pushed by a worse variant of the 2012 Thunder b2b? Not get pushed by the Reggie's Pacers? Win with a loaded deck in 94? Replicate even late Lebron or late Kareem performance in 1995?. Jordan did not have the most dominant 6-year stretch. That was obviously bill. As 70's pointed out, even Magic has a case. What about Jordan here is special?


how did he not match the 71 bucks? the 96 bulls have the same SRS and more win and a 15-3 playoffs. he didn't ever outperform the best spurs? by what measure is that? he has 4 seasons with a higher SRS than any duncan season except his very last as a role player, and finished 3 of them with 4 or less losses in the playoffs, only topped once by the spurs with 2 losses in 1999 to 4 opponents with a combined 9.6 SRS ('91 bulls played a similarly weak set of 4 teams).

You assume the Bulls did not have a chance in 1990 or before that. I'd guess peak kareem or lebron with era-translation have the Bulls at 50 starting in 84 and potentially winning from 88-90.


many have 88-90 as park of MJ's absolute peak with 1991. and you have other guys who outperform him by so much that a 2nd round loss in 88, with the celtics/lakers/pistons still existing, gets you to a title. or so much they can boost a 6th seed in '89 to the title, with the bulls the only team that actually took a game off the pistons that year? keep in mind, kareem was capable of straight up missing the playoffs. and after lebron got through with superhuman lift in 2009 and 2010, seemingly the best ever, he got to play with dwade and bosh and actually managed to start winning fewer games, certainly calling into question if the cleveland type lift is possible for lebron on an actual good team.

But I don't much like this framing as it gives(imo) undue preference for timing. What I do think is Kareem needed less to compete, less to win, and less to dominate(all things he actually proved imo). So if you add that to Kareem also being proven outside of specific circumstances and then you add him just being way better at basketball at the start, and you add him sustaining his excellence longer...

Yeah, very clear-cut to me who deserves to be ranked higher.

And again, whatever arguments Jordan does have, Russell's are better.


russell is certainly the jordan case on steriods. but the significantly weaker league, unrepeatable era boost, and the fact that one team in 8 having a coach/GM advantage with no free agency to break up a team makes repeated winning easier than in a 25-30 team league with free agency, is why i would ultimately hold him back, even behind kareem.
f4p
Sixth Man
Posts: 1,716
And1: 1,731
Joined: Sep 19, 2021
 

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#227 » by f4p » Fri Jul 7, 2023 9:55 pm

LA Bird wrote:
trex_8063 wrote:Oh dear. I think we can expect contact from an angry Jordan-contingent:

Spoiler:
Image


I'll be hiding under my bed for awhile (even though I voted for him).

All hell would break loose if Jordan drops to 4, which is possible considering he was only one vote ahead of Russell this round.


i kinda think it would be a bad look for the project if he drops to 4, making anyone from the outside looking in want to dismiss the whole thing. i suppose we can say we don't care about such things, but i'm not sure it furthers discussion if that happens. but we shall see i guess.
f4p
Sixth Man
Posts: 1,716
And1: 1,731
Joined: Sep 19, 2021
 

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#228 » by f4p » Fri Jul 7, 2023 10:26 pm

AEnigma wrote:
f4p wrote:after a title with an incredible 94.6% championship odds at the start of the playoffs, [Kareem] then hit a lull where he had 36%, 29%, and 69% chances to win and didn't (average champion is at 34%). that 69% chance in 1974 is the 3rd worst non-title, trailing only the 1958 celtics (70.7%) and, of all teams, the 1976 warriors (72.9%). then he missed the playoffs altogether for the next 2 years. in a decade with massive expansion (more than jordan), even more dilution to the ABA, where his biggest competitor in walton was ruined by injuries, and where great teams were rare in the 2nd half of the decade, we get one dominant title, a series of disappointments, and missed playoffs.

he lost 2 series in the 70's as a +4 SRS favorite (1973 warriors, 1974 boston). when comparing him to someone like jordan with no losses as a favorite at all, this is a huge deal. and he tacked on another +4 loss to 1986 houston. kareem was losing +4's at over twice the rate of the other top 30-40 guys i looked at (15.8% to 7.2%, with 7.2% showing you how rare it is). and he even has a near miss with a +3.5 loss.

the standard for jordan seems to be that if he has the most dominant 6 season stretch in nba history, it's still not enough. what even would be enough if the best stretch of results ever isn't enough? it's not enough because his teammates are too good, the odds too stacked in his favor, too much expansion. but the world was kareem's oyster in the 70's. he had all those things and more at various times in the 70's. were all his rosters contender worthy? no. but you can't lose a 69% title and miss the playoffs twice as by far the best player in the league and tell me you maxed out. and early in the 80's, we already have him getting outplayed by his counterpart moses malone and losing to a 40-42 team as a defending champion. these things are simply not on the resume of jordan.

yes, kareem was amazing right from the start. and probably lost out on a season or two of being very good by not getting to leave college early. and put up massive numbers. and has a few years where he dragged a team as far as possible, even if he didn't win. and has scoring numbers and efficiency numbers, even in the playoffs, that are amazing. i'm still voting for him #3 after all. he was very good. and he has lots of championships.

but of course part of kareem's 6 titles are that he got 2 as very clearly not the best player on his team, barely having above average stats (though i think others have indicated possibly good impact numbers). playing next to a top 10 player at his peak. kareem's age 36-38 seasons are 1984-1986, next to essentially peak magic johnson. if MJ came back to play his age 36/37/38 seasons and told the bulls to trade him to the lakers for kobe so he could play with peak shaq (or maybe slightly below peak shaq to make it more fair), he's probably climbing up to 7/8/9 titles. but he retired on top instead of dragging it out.

is it really just all about continually adding seasons together? picking the guy who was most willing to play it out to the very end just to add that last little bit of career value? in the last project, jordan finished #2 and kareem #3 and my notes indicate that neither of them have played any NBA games since then. the "portability/scalability" fever that gripped this place during the peaks project last year seems to have thankfully broken, but now we seem to have a longevity fetish building. is the lebron case on longevity bringing other longevity cases to the fore now?

As someone voting Kareem ahead of Jordan, I think a lot of the Jordan arguments ahead of him make a fair amount of sense and I am surprised to see it be so mainstream among this panel for Kareem to have the edge.

To your hypothetical, and with the acknowledgment that there is an extent to which I think this view should be more niche than it is: yes, that would reflect better for Jordan in my eyes, even if the weight given would be tempered. The hypothetical “Lakers Jordan” would be a more valuable player. The literal sense is obvious, in that playing is (generally) adding value. Winning titles is also adding value, even if it is as a sidekick and thus lesser value. But that is where I think your hypothetical breaks a bit. Is that Jordan to you? Knowing everything we know about him, is he handling the shift into a secondary role next to Shaq? I kind-of doubt it, but if he had, that would increase my valuation of him, yes.


do i think he would take a backseat? arguably not. almost certainly not without some butting of heads. but i look at the acrimony between kobe and shaq and the fact they still won 3 and i'm not sure, short of jordan beating shaq to death with a crowbar at practice, that shaq/jordan don't win 1 or 2 titles just by accident. also, shaq is kind of weird about respect and rings. he talks about immediately respecting phil in a way he didn't do with his other coaches. i could see a perpetually sensitive, image conscious shaq who is ringless bending the knee to jordan (though crowbar is always a possibility).

I understand why people go with the “no worlds left to conquer” angle. But I also do not think it is especially accurate. I mean, for one, he did not stay retired. For another, he announced his retirement in the middle of January 1999. This was not some planned result after that sixth title. There was an article on Grantland or maybe The Ringer or something like that… cannot find it now, but maybe someone knows my point of reference here. Anyway, it went into detail about how Jordan was to some extent looking for a team in the offseason, but for various reasons, nothing materialised or made much sense. So he retired.


admittedly, we're all guessing. i think if scottie and phil loved the front office, and scottie didn't want to break from jordan's shadow, and everyone was coming back, the path of least resistance is probably for jordan to return. but the "no worlds left to conquer" comes into play more when it looks like coming back is going to be a hassle, which it apparently was. maybe even partly caused by jordan himself. i do think if he had won less to that point and only had 4, especially if it was 3 from the first 3-peat and maybe only one recent title, that he probably comes back. but who knows.

He was also tired and drained; not disputing that. And he has the right to be. But that is Jordan, who constantly maintains a psychotic and evidently exhausting mindset in all levels of competition. Is Kobe retiring after 2010 if he had won in 2008? Nah. Would Lebron have retired by now? Well, maybe, but I think Lebron has always been good at setting new goals for himself. He said he wanted to play at the same time as his kids a ways back, but I think even with six titles, the scoring record would be an attractive accolade to him, and once you have the scoring record, hey, just need to tough it out for another year and then maybe your kid can be drafted somewhere. Kareem did not. Duncan might have, but he seemed to be more a case of, “I cannot provide enough value to my team for another season of wear to be worth it.” We can understand Jordan’s decision without celebrating it.


lebron is arguably a one-off. the immense pressure, far beyond jordan, that he shouldered from day 1, the amount of hate he had to get through compared to jordan who was basically beloved and celebrated for his whole career minus a few detractors, the amount of dedication to keeping his body right, and still seemingly being able to do all that outside stuff at the same level of commitment now into year 21, even with a few down years recently where he could have packed it in, is possibly unmatchable. he would have even more right to be burned out than jordan. i sometimes feel worn out by lebron's career, and i didn't have to do anything. maybe jordan really just couldn't maintain the maniacal drive for an unbroken stretch of seasons. but then his dad being killed is obviously something others have not dealt with so who knows how it plays out.

This is not a new narrative either. I think after 2016, Bill Simmons — only mentioned because of the cultural cachet he carries — wrote about how that retirement may ultimately let someone (Lebron) dethrone him. And look, I have been very clear that the extra years are ultimately irrelevant to me with Lebron versus Jordan. And Russell played roughly the same amount of time so it is pretty irrelevant there.

It is relevant for Kareem. I think prime Kareem can stand with Jordan, but if you asked me whether I would rather have ~930 games of Jordan (his entire Bulls regular season tenure) or Kareem (his entire regular season career 1970-81), I am taking Jordan. Add a year. Big year because Kareem adds another title. Still, Jordan has an extra seventy playoff games and an extra two postseasons (albeit with conference seeding luck not extended to 1975/76 Kareem), and I would continue to take Jordan. 1983? Kareem adds another Finals run. Getting closer now, but maybe we still take Jordan. 1984 is a first-team all-NBA season nearly ending in a title. 1985 he wins Finals MVP. 1986 is another first-team all-NBA spot (almost certain to remain the oldest ever for at least another decade). You can talk about the disappointments in some of these seasons, but they are not negative value seasons. Again, I recognise a lot of people look at primes only, and there, sure, Jordan will probably never fall behind Kareem even in this community. But those extra superstar years add up.

Idk, in some sense, yeah, it feels weird to say that Kareem was the best player in the world for longer than Jordan was, yet he seemed to disappoint in a way Jordan never really did. However, there is that Duncan idea where it also feels weird to punish guys for being so incredible that they exaggerate their team’s real playoff quality.


there will always be a problem with trying to talk about playoff rising/falling. in theory we should only care about the absolute. but i have possibly lower confidence than others that we can ever truly bulls-eye in on this guy being worth 17.2 wins and this guy being worth 14.7 wins, or whatever (and i'm not sure the differences are that huge if we're comparing something like everyone's peak seasons or their 8th best season to everyone else's 8th best season). you'll get a WOWY number over here that looks great, an RAPM that is meh, a box score that is great, a signal in one situation that says you can lift anyone to great heights but a signal from another that says you aren't much to write home about (let's call that the steve nash 2005 suns and 2005/6 mavs situation).

however, i think we can be very confident of playoffs over/underperformance to the regular season. from what i can tell, the box score, the impact metrics and team performance tend to line up very well. after all, you are the same type of player all season. so if your numbers get better, you probably played better, and vice versa. if your +7 team loses to a +3 team, it certainly feels safe to say your team underperformed the regular season. now Ohayo will say that there is far greater difference in players and that team performance in the regular season can be heavily attributed to a great player and not so much the supporting cast, whereas i see a guy like KG miss 3 straight playoffs and then win 66 with a new team and say your regular season fortunes rise and fall a lot with teammates.

beyond being confident in playoff over/underperformance, i see the huge winning percentages that players get as favorites, especially as heavy favorites, and it informs my rankings such that i heavily punish losing as a favorite (especially a big favorite) and heavily reward winning as an underdog (especially a big underdog). right or wrong, i see the fine lines of greatness showing up most in pushing close series one way or the other. in having the guy who you know will get a good shot or get a teammate good shot as becoming significantly more important in the playoffs, if only because there seems to be so much separation in a skill like that between the best player in the game and, say, the 10th best player. guys who routinely come through in those moments, with jordan being the apex certainly for offense, seem to affect winning in the biggest moments, where series often come down to a few bounces here or there. and maybe bill russell can be considered the defensive equivalent of that, even if i have very little footage to go off of to know if it's true.

anyway, there's no way to know if it's fair, but i'm probably going to be a big playoff riser/faller guy for most of this project (unless i need to be inconsistent to rank someone how i want :D ).



I saw 1974 Cowens brought up. He outscored Kareem in Game 7, yes. Great player, great game. Would we say the same thing about 2016 Lebron if Kyrie had missed that shot, or if Steph or Klay had managed to connect on some garbage look at the end of the game? Outscored by Draymond, here dies his legacy.


certainly it's tough to say. so much of nba history is based on one play here or there. intuitively we know that a lot of it just straight up comes down to luck in these one possession situations. i love probabilistic thinking, but ultimately for this, i tend to think we have to go with what happened, because who knows if i'm just missing the secret sauce that made that guy win that close game on that close play. a different f4p in a different universe is probably holding lebron back right now because the warriors won that game 7.

hell, bill russell won 7 of his famous 10 game 7's in the following ways:

double OT
OT
OT
regulation by 1
regulation by 2
regulation by 2
regulation by 2

there's another universe where he has like 5 titles and no one seriously pushes him for GOAT status. but here, he had a warehouse full of horseshoes and rabbit's feet so it is what it is.



2016 is what best kept him competitive in this race despite the two title deficit, so maybe that is consistent in a way. “Kareem needed 1974 to be above Jordan.” I see the angle. To me though, the Celtics had the better team. They were better than the version of themselves which won 68 games the year before, and the Bucks were not the version of themselves that established a 4 SRS advantage.


maybe the celtics were. but we would probably need more time (a big upsets project to dissect which were really that bad and which had mitigating circumstances?) to do it for everyone to keep it fair. at best, we can say kareem lost 3 as a +4, so it certainly seems to be part of a trend, not just an unfortunate one-off like russell in 1958.

For me the bad loss is 1973. Decent Warriors squad with Barry back, but no real way to excuse it. Like with how 2011 affected Lebron, I think that helped develop Kareem into the nearly unstoppable half-court scorer he would become, but it is a clear failure. And when someone is essentially looking for the fewest number of failures, I recognise Kareem is going to be in a tougher spot. But for myself, I have always believed players can make up for it, and by sometime in that 1985-87 range, I personally think he had… even if I am surprised by the extent to which people here have slowly come to agree.


saying guys can make up for it is certainly a valid approach. i have 2016 lebron almost getting back 2011. i again tend to look at the ridiculous win percentages some guys have as favorites and say that it's not entirely impossible to get through a career with very limited numbers of failures. and indeed jordan seems to have done it. and as mentioned before, i have a harder time penalizing lack of longevity for non-basketball related things like jordan just choosing to retire or magic losing several prime years to something literally no one else has ever lost years to. as opposed to injuries or skill decline making you a much lesser player.
User avatar
Moonbeam
Forum Mod - Blazers
Forum Mod - Blazers
Posts: 10,313
And1: 5,096
Joined: Feb 21, 2009
Location: Sydney, Australia
     

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#229 » by Moonbeam » Sat Jul 8, 2023 1:21 am

OhayoKD wrote:This is going to be blunt but, I think you need to use actual direct winning as a check for this to be useful. As in, check what these stats say relative to actual team results and how they correlate with what happens indiciually. Jordan's Bulls are not too affected by his depature defensively(as is typically the case with all-d guards). Russell sees better defensive results than duncan throughout his prime and then sees a much bigger defensive drop-off after his retirement than anything from anyone ever(and that overall drop-off is comparable to the increase we would get if we gave jordan all the credit for the bulls improvement from 84 to 88.

Writ large, big defense consistently rates lower in ws relative to impact(if you want more certainty you can check databall if you want more certainity) and guard defense sees the opposite effect.

You can make all these tiny corrections here and there but the fundamental issue is reliance on box-score which, as I've said a bunch before, only really tracks the end of a possession. A player handling the ball less, coordinating teammates(russell's biggest edge over everyone else listen), facing single-coverage as opposed to doubles(this was part of what occured with the triangle), gambling for steals and missing, the difference beween weaksidde help and being the primary deterrent, ect, ect ,ect.

All is completely missed here.

FWIW, by available "impact" data for the time period(its all raw more or less), I'd say kareem and russell are the standouts. Duncan/MJ seem like peers, and Wilt seems worse than advertised(69 is an all-time terrible singal both ways)

Second note. +5 in one period of time does not necessarily = +5 in another. This is most prevalent in the 60's and probably the mid-late 70's, but "best in the league" is different during different time periods and so is the value(at least towards chanpionships) of different scores


I'm not sure I fully understand. The direct winning is factored in due to the total team OWS/DWS, though, right? A player joining a good offensive/defensive team from one that isn't as good would expect to see an OWS/DWS increase per 48 by virtue of there being more team Win Shares to be allocated among the players, but that's kind of what we want in this, don't we? For example, the fact that the Spurs defense was good is testament to Duncan being good, so players joining the team (or leaving to another team) are most likely to have a net positive effect teaming up with Duncan (and a net negative effect when leaving him) due to the chance that they will be coming from (or going to) a team with a worse defense. Am I missing something here?
User avatar
Moonbeam
Forum Mod - Blazers
Forum Mod - Blazers
Posts: 10,313
And1: 5,096
Joined: Feb 21, 2009
Location: Sydney, Australia
     

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#230 » by Moonbeam » Sat Jul 8, 2023 1:25 am

trex_8063 wrote:
Moonbeam wrote:
Spoiler:
So I found a pretty big flaw with my age curves which were used to reallocate Win Shares. In short, I discovered that age curves had Bill Sharman and Bob Cousy peaking as defenders in their mid-30s. :lol: I should have seen it coming, but it's easy to get lost in code and results that "feel" right. The general framework of the approach I am working on is still promising, but it needs some tweaks.

That said, I've written some code to examine the impact on assigned Win Shares to teammates based on changes in minutes played, and have some reasonably compelling graphs to indicate whether certain players tend to be overcredited or undercredited in their own Win Share totals.

As an example, let's consider Russell's potential* impact on the 1957 Celtic D.

In 1956, Bob Cousy (without Russell as a teammate) had 0.045 DWS/48 in 2767 minutes, and in 1957, he had 0.096 DWS/48 in 2364 minutes with rookie Russell as a teammate. That's a jump of 0.051 DWS/48 associated with an increase in 1695 minutes with Russell on the team. In 1957, Togo Palazzi had 0.086 DWS/48 in 233 minutes with Boston and rookie Russell as a teammate, but dropped to 0.048 DWS/48 in 1001 minutes the following year without Russell as a teammate. That's a drop of 0.038 DWS/48 associated with a reduction in 233 minutes with Russell as a teammate. We can do this with all teammates across Russell's career to get a glimpse of his potential impact on teammate DWS/48 (and OWS/48 too).

*I say potential here as there is serious multicollinearity with the Celtics given how stable their rosters were. 1957 also saw rookie Tom Heinsohn. I'm battling to separate Heinsohn's impact from Russell's in regression approaches so far. Not sure if ridge or lasso will help, but I'll keep working away.

I'll present career graphs now of changes in teammate minutes played for the 5 eligible players in question against the corresponding changes in OWS/48 and DWS/48. The size of the dots is related to the minimum MP in consecutive seasons being considered. In the earlier example, Cousy's would be 2364, and Palazzi's would be 233. I also have a weighted correlation I can compute across their careers with weights corresponding to these minimum MP.

Bill Russell:

Image

Weighted correlation for OWS/48: -0.158

Image

Weighted correlation for DWS/48: 0.790

This suggests that Russell may be somewhat overcredited in OWS, but also suggests (or screams!) that he is massively undercredited in DWS. That graph is stunning!

Wilt Chamberlain:

Image

Weighted correlation for OWS/48: -0.206

Image

Weighted correlation for DWS/48: 0.236

This suggests Wilt may be overcredited in terms of OWS, but perhaps surprisingly, undercredited by roughly the same amount in DWS.

Kareem Abdul-Jabbar:

Image

Weighted correlation for OWS/48: -0.068

Image

Weighted correlation for DWS/48: 0.487

This suggests Kareem may be slightly overcredited in terms of OWS, but quite undercredited in terms of DWS.

Michael Jordan:

Image

Weighted correlation for OWS/48: 0.127

Image

Weighted correlation for DWS/48: 0.132

This suggests MJ may be somewhat undercredited in both OWS and DWS.

Tim Duncan:

Image

Weighted correlation for OWS/48: -0.155

Image

Weighted correlation for DWS/48: 0.704

This is nearly identical to Russell!

These graphs and measures give a lot of weight to rookie and final years (and in MJ's case, many final and "rookie" years, haha), but I thought it would be interesting to share.



Super-interesting stuff, MB, thanks for sharing.

Just to make sure I'm understanding correctly: what you're calling "weighted correlation" is the correlation coefficient, yes?

If so, that is a strikingly high correlation seen in the DWS change for both Russell and Duncan (I would say supports the credibility of one poster's statement regarding Duncan's case as defensive GOAT, too, fwiw).
Marginally weakens the GOAT [or #2] case for Wilt (relative to the others). I would say strengthens Kareem's candidacy relative to Jordan for this spot, albeit only marginally.


Out of curiosity, are there other players would you be willing/able to run some of the other great defensive bigs through similar methodology: Robinson, Hakeem, Mutombo, Ewing, maybe Dwight Howard, too. Would love to see how theirs shape out. Obviously, would love to see other all-timers, too: LeBron, Magic, Shaq, Garnett, Bird, Kobe, etc.


Also, is there any way to take this information and quantify an "adjustment/correction" to the individual WS/48 of the players analyzed? That would be cool, but I could see where any such effort would come with an EXTREME amount of noise.

Anyway, thanks again for sharing. Great work.


Weighted correlation is Pearson's correlation coefficient weighted by the minutes played of the player. We see all sorts of players who log less than 10 minutes in a season, so their numbers vary wildly from season to season, so I don't want to give their changes in OWS/48 or DWS/48 the same weight as a player who plays 2000+ minutes.

Yep, the intention is to try to come up with an adjusted Win Shares for players to use as inputs into models of team regular season and postseason performance. I already have a nice framework in place to do this, but after discovering issues with the age curves for teams with little roster turnover, I've had to re-think how to come up with a sort of expected Win Shares for players.

My computer is running some code right now on doing just that, but when it's done, I'd be happy to share similar graphs for other players. Maybe it should be in its own thread so as not to derail?
User avatar
LA Bird
Analyst
Posts: 3,606
And1: 3,365
Joined: Feb 16, 2015

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#231 » by LA Bird » Sat Jul 8, 2023 1:45 am

f4p wrote:
LA Bird wrote:
trex_8063 wrote:Oh dear. I think we can expect contact from an angry Jordan-contingent:

Spoiler:
Image


I'll be hiding under my bed for awhile (even though I voted for him).

All hell would break loose if Jordan drops to 4, which is possible considering he was only one vote ahead of Russell this round.


i kinda think it would be a bad look for the project if he drops to 4, making anyone from the outside looking in want to dismiss the whole thing. i suppose we can say we don't care about such things, but i'm not sure it furthers discussion if that happens. but we shall see i guess.

Pretty sure people from the outside looking in already dismissed this whole project as garbage the second Jordan didn't end up #1. Also, this idea that we should make our list closer to the general public's to further discussion seems very counter-intuitive to me. If creating an uncontroversial list was the objective, we could just average all mainstream media lists and be done with it right away. The PC board coming up with a different list after a year of debates is more likely to generate interest and drive discussion among the people who actually want to learn something new.

Now you could argue that we are alienating 95% of the public by not ranking Jordan higher but so what? This is the type of post that get you the most and1s on the general board:

6/6

Compare that to Doctor MJ's post on Russell that got the most and1s on this board. The PC board top 100 list reflects what the voters here believe in, nothing more. If people outside of this board disregard our list and think we are a bunch of lunatics for having Jordan 3/4, that's their problem not ours.
1993Playoffs
Assistant Coach
Posts: 4,085
And1: 4,247
Joined: Apr 25, 2017

Re: RealGM 2023 Top 100 Project - #2 (Kareem Abdul-Jabbar) 

Post#232 » by 1993Playoffs » Sat Jul 8, 2023 7:58 am

Definitely didn’t expect Kareem to overtake MJ.

Having MJ 3rd or 4th at best is pretty unusual. Definitely not saying it’s wrong though
OhayoKD
Head Coach
Posts: 6,042
And1: 3,932
Joined: Jun 22, 2022

Re: RealGM 2023 Top 100 Project - #2 (Kareem Abdul-Jabbar) 

Post#233 » by OhayoKD » Sat Jul 8, 2023 9:20 am

f4p wrote:
OhayoKD wrote:And why are you assuming I don't? The problem here is kareem was not "consistently" losing as a +4.7 favorite.


but he was. this is the list of people i have tracked for playoff SRS differential purposes:

Yeah, and "lost the most as x"(over the second longest career) =/ "consistently losing". A total of 3 series(without any sort of specific comparison) does not prohibit Kareem at his best to being worth more than Jordan at his best in the playoffs
As I outlined in my first post on this thread, the Bucks got better in 71 and 72 using sans's psrs.


is that the sans list that had the 2018 rockets as the 95th best team of all time? either way, as an aside, i think (and will maybe make a thread) postseason SRS needs to be calculated slightly differently. as far as i can tell (maybe i am wrong), it traditionally it gets calculated for a given series as:

Postseason SRS = Opp SRS + Series MOV

this assumes your opponent has played at their normal level (and you overperformed/underperformed them), with no possibility for the idea that you played at your normal level and your opponent overperformed/underperformed you. as such, it produces different results for the 2 teams in a series depending on which team is your reference team.
From what I recall San's is rolling and heavily playoff-weighted(3/4). Unsure about the specifics beyond that. Feel free to suggest alternate calculations. I'm looking at the standard deviations and then the rs, psrs, and full-season srs as well as full-strength ratings per what Ben provides.

All three of the 71, 72, and 74 Bucks were better than the 91 and 90 Bulls(compared to the latter 2) respectively. The 77 Lakers were on-par with the 88 Bulls after a better regular season. And I have Jordan having more help all those postseasons. Moreover you seem inclined to give Jordan all the credit for his "lift", but in 1990 do you know what increased from the regular season? The Bulls defensive rating. Do you know who purely by box-aggregations(which miss a bunch of things Jordan was helped with) like BPM saw the biggest rise? Pippen.


the 71 and 72 bucks are excellent, i agree. one of the few +10 SRS teams to ever lose, but only because they ran into an even bigger juggernaut. kareem certainly shot out of the gates quickly and dominantly. then he started losing +4 series, missing the playoffs, having 44.7 TS% series, getting outplayed by moses twice (losing to a 40-42 team one of those times), and picking up most of his titles while not being the best player on his team any more.

His first "+4 loss" is literally a comparably competitive performance to a better opponent than who Jordan faced in 1990(no you do not need to use the prior year if you dont want to) with both his co-stars further diminished from what he had in 1972 which you agree was "great". Do you see the issue here? This is a 2009 analogue, except losing to the magic is actually a worse result(relative to 90) and the bucks performance is a better one. The Bucks were the best regular season team by a margin when they had no-buisness being that, and then fall to a stronger opponent than who consensus and statistical peak jordan kept getting felled by before they broke down...with weaker help than the year you call great. If you really want to criticize, you can point out this is where the ABA had peaked and talent was most diluted, but strictly league-relative, marking this as a negative for Kareem relative to Jordan makes no sense. Also, fwiw 71/72 and 77 do not really have that same asterix.

You also need to tell me what seasons you're specifically comparing for Kareem to Jordan. I am not interested in uneven comparisons or "disappointment". I specifically chose what should be high-marks for Mike(and even artificially inflated those marks) when I made the "lift" comparisons considering both the regular season and the postseason and for the 71/72 vs 90/91 comp I also looked at what was going on at a granular level for the co-stars and Kareem and Jordan. You can dispute it if you want, but I've got 4 separate years within a 8-year span where Kareem led similar or better playoff-teams with what I would consider less help than 88, 90 or 91(so the likely stat peak and the two years everyone likes the most). That is a big gap to make up prime to prime and unless you can establish it is likely that prime Kareem folds, I am still taking his peaks(where he did the opposite).

if you're telling me his prime was more valuable(Anenigma cares more about the "accomplishment" bit, I just want the "goodness") because of "playoff elevation" then haphazardly chucking Kareem's "dirt" at me isn't going to work. On top of those 4 separate seasons, there's also the 3-years he was just way better at basketball pre-nba. There are years where he missed the playoffs due to injury or teammate injuries putting jordan+ floor-raising(1975 ~ (juiced)1988, 1976>1986), and there is his obviously far stronger rookie campaign. 1980 is also great for a "late-prime" year leaving not many seasons for Jordan to make up the gap unless we are disingenuously counting Kareem's post-prime.

And even if Kareem is Icarus, when Icarus can get closer to the sun than you can 4 times in 8-years, it's hard for me to care. I do not treat champions and non-champions as fundamentally different (unless you are someone who actually always wins regardless of context(russell) in which case I have to start asking if you transcend mortal proxies like "srs"). Lebron did not have to win in 2009 to be better than champion! MJ to me. From an era-relative perspective, I'd say 1974, 1977 and 1972 have decent cases of being as good or stronger than 09(assuming you compare teams relative to the league). I'm not gonna automatically put a guy with 3 potential 2009's lower because he didn't win just because your metric doesn't differentiate between a 2009 and a 1990/1988.
you can't play with magic (and then worthy) and win 5 out of 9 years and then say playing with scottie pippen is unfair on the way to 6 out of 7.

Do rosters have 2 players? Kareem was obviously much worse on the Lakers and as is, he ended up the more successful player. And I still don't see why this stops Kareem from having a higher peak.
Do you know what the 94 Bulls did in the playoffs while their best players were at each other's throats?


played well but still lost? i have them with 0.668 expected series win and they got 1, for a 2.0 SRS rise.

They got better and one of the series they overperformed in doesn't count to your calculation at all because it was a loss. Just like in 90 the defense went up by 2-points from the regular season, and then by 5 points for the knicks and pistons series. I know where you "have them", but a calculation that does not see any difference in losses or wins beyond "srs gap" is not very useful to me, and is especially not that useful with a player when comparing seasons where players both lose or both win. If Kareem is not being credited for when his team plays better against better opponents, I don't really care what "you have it" at.
In 74 with even weaker help, the Bucks did just as well against a better opponent after being by far the best regular season team. As far as I'm concerned, Kareem was flat-out better, regular season and playoffs, and consequently achieved comparable or better playoff results against better opponents without a system equivalent to the triangle(refer back to blackmill's notes on suboptimal deployment), or a team so good they could **** off with all the chemistry and still conceivably win by themselves. Even if I gave jordan all the credit for the team-wide elevation(obviously undeserved), he still looks like a worse riser than Hakeem(who we're just nominating) and Lebron(who you dont have at #1 even though the Cavs playoff lineups which were better lebron and stars, lebron and no stars, and worse with stars and no lebron)


jordan is certainly a worse playoff riser than hakeem. no one except healthy kawhi compares to hakeem. and you tend to give a lot of credit to the triangle. but when jordan left after '93, the bulls rORtg fell by 7. then, perfectly balanced, it went up by +7 when he returned.

It did not fall by 7-points when you compare full-strength lineups(pippen and grant barely missed time during the first-three peat).

In 1994, Pippen assumed the lead dog role and upped his scoring and creation with essentially no loss in efficiency. Given the makeup of Chicago’s roster and its implementation of the triangle, this was probably Pippen’s maximum offensive output, as he peaked in offensive load at 43 (96th percentile) while the Bulls posted a respectable +2.2 rORtg when healthy, better than any offense Michael Jordan led before Phil Jackson arrived.

In 1995, with key cog Horace Grant lost to Orlando (and Ron Harper aboard), a healthy Bulls team still played at a 52-win pace (3.8 SRS) with an rORtg of +1.1 before Michael Jordan returned.

Hell we can very generously compare the Bulls best offensive rating(1992) to what the Bulls were doing without Grant and it does not hit 7-points:
In 1992, Chicago peaked at +7.4, the 12th-best offense ever, and one of three attacks in history with a raw efficiency above 115.5 points per 100. After three consecutive top-100 offensive seasons, Jordan took a baseball sabbatical and the Bulls scoring efficiency dropped to 2 points above league average. In 1995, before his return, Chicago chugged along at +1.2 for 63 games (playing at a 52-win pace)

Moreover, 7-points would not be some unprecedented offensive drop-off. Presumably you have Jordan as the best offensive player ever but that does not bear out in the results. Cleveland's offenses were average(net rating, worse if we used wowy) without Lebron and still hit higher playoff highs. Oscar turned the worst offense into the best offense as a rookie. Per usual there is no real comparison when it's time to do something other than churn out conventional box-aggregations.
Lazy or not lazy, he is shooting guard who never outputted significant defensive impact(at least by "lift) as shooting guards do not do. He also was not a lebron/magic/nash type guy who could monopolize team-wide offense to excellent success, and while his biggest advantage was ball-handling, he only achieved impressive offenses as a secondary ball-handler.


i feel like this is part of the issue with your jordan comments. you give him credit for almost literally nothing.

Friend, you literally tried to use Duncan winning regular season games against him. I am not giving him the degree of credit you want me to, and when the rubber meets the road, you don't offer much for me to give him that credit beyond vaguely arguing i'm unfair, appealing to consensus, and then acting like he's actually as much of a winner as bill russell before citing subjectively filtered "production" which is the biggest source of that aforementioned consensus in the first place.

Here's an idea. You want Jordan to be ranked higher? Then instead of telling everyone it's a "bad look", argue for it. Because right now you're basically throwing out seasons on the basis of "rings" while also voting for 6 over 11, and then acting like health is a piece of context we can't and shouldn't adjust for.

I have Jordan as a top-10/possibly best of his era level player. That is what "lift" suggests he is, and as this discussion illustrates that doesn't change in the playoffs unless you assume players who lose cannot be compared to players who win(72,74). If I used PER maybe he would be higher, but of course if I used "goat-points" or the stupid rodman stat from the 90's, he'd be lower. Jordan is not entitled to be ranked higher because people say he is. And if you really felt this strongly about "missing chances", would you really be voting him above someone who basically never missed(even when his help "wasn't" what you seem to think is neccesary for contention), and kept not missing without the "contend-without-him" superteam he started winning with?

You said it's not fair to rank curry higher by putting him in the 90's. And yet when the rubber meets the road, your push vs the likely more dominant(indivodually and team-wide) candidate is "weak league". I'm sure you've seen the graphs. The 90's were not the peak of basketball. Pick a lane and stick with it.

"the bulls offense didn't pick up until the triangle and then the bulls defense didn't fall off when jordan retired"

No. I said Jordan never led a great offense without the triangle, just like Steph Curry never led a great offense without Kerr's system. I quite like Curry(you've even called me one of his fanboys), but I am consistent. And if you're going to +1 me when I call-out Steph for never proving he could lead great offenses outside of that system, then I'm not sure why you argue that me pointing out the same with Jordan is unfair.
. the obvious implication being that jordan had no impact on offense or defense.

The implication is he does not have the offensive impact of goat-tier offensive players, yes. And as I have said he has "potentially game-changing, but not season-definding" defensive lift which in a scale with the likes of Lebron and Kareem who can spike season-wide d-ratings by multiple points(on occasion closer to 10) respectively, that can rightly be called "not significant", even dpoy voters disagree. And even with that, in-spite of "impact" consistently disagreeing, I still place Jordan ahead of a better offensive player in Magic because of his defense. I give credit where I think it's due. But when you have to conjure up something like "Duncan won too much", I am not going to be moved.
i mean maybe if this board had put forth a great "scottie pippen is top 10" argument over the years, i could see it. but even here he's in the 30's.

How is "scottie pipen top 10" more relevant than "the bulls contended without Jordan". I'd say the latter is quite strong given the grounds for dismissal are...
-> ignore 1995
-> focusing on healthy lineups(for on and off) is unfair
-> the +5.8 net-rating rotation(below average comp tbf) was "injury ravaged"
-> Outscoring the knicks was a fluke

Mind you I wanted to put that last one in all-caps but I have been told to not be so boisterous when disagreeing :(
In the year you see as his peak, he was mostly shooting over single coverage while a "unpolished" offensive player like say 2009 Bron was mantaining similar "end-of-possession" efficiency while doing more through the duration of possessions against significantly more defensive coverage.


illegal defense certainly helped jordan's offense, though i do think it limited his defensive impact as he was never allowed to roam the way a modern player can. but then bill russell's interior defense in a league that can't shoot is almost inarguably a better era boost and kareem playing in by far the most watered down, expanded league in nba history was also a huge boost he took very little advantage of.

That was not a time-machine case. The point was that before the triangle he had to deal with doubles alot more frequently and crucially his "not lebron, nash, or magic-level decision-making" was far more of a factor. Jackson and Pippen(much like kerr and draymond) also were also the primary and secondary decision-makers for the team over the duration of each offensive and defensive possession. Again, for a player who isn't lebron that's can be a very nice boost(see: 2014 steph vs 2015 steph).

None of those factor into the box-score, but they do manifest in results where Jordan is peaking(at his stat peak) at a +4 offense(+2.9 if i use bens' derivation but I'm counting 30-games post trade to give Jordan the best score) while a merely "comparable" offensive player like Lebron posts a +6.5(56-win overall) offense at 21 and is still more valuable than a prime mj-analog on nearly 65k minutes(in the regular-season where he only played a few games on a torn tendon) and **** spacing alongside players like Westbrook. lebron and jordan by box are the same offensively, but lebron does way more before the end of a possession, and hence is the more impactful offensive player. I do not have Jordan at that top-level because the results do not justify it. That is all there is to it.

If you want to see the effects on tape, pull up a 90 or 91 pistons game vs an 88 or 89 pistons game and count how often MJ is doubled(and look at how often said doubles arrive). I think there's actually some tracking for that lying around I can find, but maybe you should look for yourself.
kareem presumably was better comparing only their first 6 years. but i don't see how jordan fell of quicker. as someone pointed out, jordan got a POY win at age 34, with kareem only getting one at 32. and by the time the 1998 playoffs rolled around, jordan was 35 because his birthday is just after the BBRef cut-off. kareem was not dominating at 35 like jordan, even if he kept playing past 40.

Fair enough. Even by season-count jordan holds up better later though I will again note Kareem never took a break(and I'd point to hakeem here as an example of how hard that can be). I'd still say Kareem's last prime-mj level year was 1980 which is still a couple seasons short. Would bet on the bigger player with better defense lasting long beyond that, but I can't confidently extrapolate that from the Wizards stint. Do not really care about POY voting though(where Jordan apparently got 2 top 2 dpoy finishes...), I have Hakeem supplanting Mike as "best player" by 1993 and better or in contention till 1989 with Magic in contention until 1991.

Kareem should be better, and the results tell me he is better, so why do I care about what the "expected championship differential" is when he's leading better or comparable regular season and playoff teams with less help over and over and over again?


he's not leading comparable playoff teams, though.

He is relative to three of jordans four consensus best years. And the other year 1989, would just look worse with how I'm doing things because for 1988 I literally gave Micheal all the credit for the team improvement. 1993 does not look good. A reasonable extrap for 1996(taking from 94 where they had a rodman-like player) looks good, but not better than 1988-1984. You can say 'Jordan won more", but Kareem led a team on the level and to me looked more valuable, he floor-raised another team higher or on par with the 88 bulls and looked more valuable and led multiple teams to higher rs and playoff heights than the 90 bulls and look smore valuable with plenty of regular season corraboration(i acknowledge he fluctuates more in the playoffs, but again, I quite like Icarus). Do you have some reason to assume Jordan was better or more situationally valuable in 97, 98, or 92? Or to think when I do **** like "hey lets give jordan all the credit for the bulls post 84 improvement", that actually undersells him. Because if not, then the "playoffs" do not help Jordan beyond "he won more".

Doesn't seem unreasonable, especially since that included 84-87 MJ who probably wasn't as good at basketball as pre-nba Kareem. For the "peaks", 72 and 71 saw improvement per San's calc. Compare that to 90/91 and I think the playoffs significantly bolster Kareem's case. Also M.O.V here allows for us to compare losses. 1972 may not have ended in a win, but i'd say in terms of "lift" that playoff outcome is more impressive than anything from Jordan. Ditto with 1974 where the help was weaker than 1972.

MOV obscures the actual winning and losing though. especially give the small sample of a playoff series. did kareem really win the 1973 warriors series since his team had a +2.5 MOV? of course not. he played horribly and lost to a lesser team, even if both victories by his team were blowouts to swing the MOV to his team. this goes back to my "jordan never gets any credit" thing. even by MOV, 1974 is kareem losing by almost 5 ppg to a team he had a +4 SRS advantage over. that seems like a lot of underperformance by winning or MOV.

There are trade-offs, but I'm pretty sure the Warriors series would be seen as an underperformance anyway in something that adjusts for competition. And I do not see much justification for equating 73 with 74 where he ran into a better opponent than the guys who beat Micheal, went the **** off, had worse help in the series you think was "great" in 1972, was similarly competitive as the Bulls, was not being optimized schematically, but his team won too many regular season games so "choke!"

Maybe if Kareem had never won, I might ask if there is something there. But he hit the "best team ever" note in his 2nd-year, so I know Kareem can win. And just like I wouldn't say 2017 Lebron can't win championships, I'm not going to let strong performances against stronger opponents be construed as things that "call that into question" Kareem's chops as a winner. 1973 is a choke. But it is not the trend and lumping 1974 with 1973 doesn't make sense. If Jordan had a 1974 it would be hard for me to argue he doesn't have a decent GOAT(era-relative) peak case, even against someone like 2009 Lebron.

As for sample size, again, I think the regular-season stuff is very strongly in Kareem's favor so...
and it's hard to do kareem in the 80's because of the conflation with magic being the signal, but for jordan from 91-98, even including 95:

jordan's regular season SRS said he should win 20.3 playoff series, or 78.0%. he won 25, or 96.2%. that's a shift of 4.8 SRS points, or 13 wins. jordan is just massively outperforming what it seems like his team should have done.

The Bulls are, yes, but again, they didnt even need Jordan to do that. And like I said, using "converting chances" instead of mov forces us to reduce the sample


once again, jordan's team has the best playoff winning percentage over a sustained stretch, improved by 13 wins, but it was really the others (and i only have the '94 bulls as +2 SRS compared to +4.8).
[/quote]
Cool. That is not the same thing as "lift" which is what you're telling me gets cooked if I look at the playoffs. You can say Jordan won more frequently, but if the help is good enough(or the competition weak enough), that does not necessitate that Jordan was more valuable than Kareem because he won. By a more inclusive method, the Bulls srs in 1994 increased to +8 from +4.7, and I've outlined why I think it was cast-improvement that played the larger factor in 90. I think Jordan is a playoff elevator(again, I have him ahead of Magic), but I do not support equating team with best player in this case. With Lebron, we can look at the star and no star lineups and isolate for what teammates were doing before. I did a cruder version of that for 72 Kareem. If you want to say Jordan elevates more frequently, sure. But i see no playoff-outcomes that match the four-years I've highlighted as far as "Lift" is concerned and I have Jordan starting from a lower base instead of value, so just throwing out "winning percentages" where being pressed by outmatched teams like the Knicks boosts Jordan's score isn't going to sway me much. I would rather just more directly compare team success to help/context for the regular season and then do the same for the playoffs while using "production"/context to internally scale where the elevation/drops are coming from.
and since we may never agree, why are you so low on converting chances into titles? the nba is maybe the most predictable league in the world. the winning team has the best player or 2nd best player, a lot. the best players tend to win even more than things like SRS suggest. i don't value these things just to help jordan, i value them because they seem some of the clearest signals from nba history. here is the SRS favorite record i posted in one of my earlier duncan posts (with curry updated for 3 losses now):

Because it reduces what can be used as evidence. Again, by this metric 1972 and 1974 and 1977 are all "negatives" when if we compare opponent and m.o.v(curve for proportionality) they are outright better or more impressive than what peak Jordan did. I would need to see what specifically you're comparing for the Lakers because beyond 80(another team that rose in the playoffs), Jordan should look better.


what do you mean by "comparing for the lakers"? before magic or after? in his 3 playoffs before magic, he had 2.33 expected series wins and got 2, so give or take pretty close to expected. maybe 1 SRS off.
[/quote]
After. But specifically I would like to see what Jordan seasons you think are analogous.
and i'll go back to my question. you seem to focus a lot on jordan in his early days, constructing a case he wasn't really valuable. but i'll ask again, what else could he have done from 1991-98?

Match the 71 Bucks? Outperform the Best Spurs sides more than once? Not get pushed by a worse variant of the 2012 Thunder b2b? Not get pushed by the Reggie's Pacers? Win with a loaded deck in 94? Replicate even late Lebron or late Kareem performance in 1995?. Jordan did not have the most dominant 6-year stretch. That was obviously bill. As 70's pointed out, even Magic has a case. What about Jordan here is special?


how did he not match the 71 bucks? the 96 bulls have the same SRS and more win and a 15-3 playoffs.

He did it in 96, he did not do it anywhere else despite having multiple "chances" to(91, 92, and 93 all were good enough help I think) while Kareem only really had 1 and actually "exceeded 1971" when his team was healthy despite a diminished co-star in 1972. It is not a big knock, but again, as I have it, Kareem is a more valuable rs and playoff player several times than the years you point out make up Jordan's "consensus peak". You are pulling up a bunch of years outside of that do not have similarly high "lift" compared to the inflated **** I marked 88 MJ at when he is by consensus worse and telling me he was more valuable those years because he won.
he didn't ever outperform the best spurs? by what measure is that?
point out
By standard deviation(which is what matters for championship winning) the 1999 and 2007 spurs were better than all but two Bulls teams and the 2003 Spurs were close behind. Not sure why I said one, but again(you can check the post i replied to you regarding duncan for justification). I'd guess Duncan had less help for 3 of his championships including 1999 based on what the Bulls did in 94 and 95. Did not look at 2007. 2005 and 2007 were tougher gauntlets and the 2005 suns>any of mj's wins per san's whatever. Before looking at san's thing I also felt the 2005 Pistons and the 2007 suns as better than any team Jordan beat but statistically(by san's method) that does not bear out(though personally I think it underrates the 05 pistons but I'm going to stay emperically grounded here) but the 2007 suns come closer to the 97 jazz than the 97 jazz do to the 2005 suns.
You assume the Bulls did not have a chance in 1990 or before that. I'd guess peak kareem or lebron with era-translation have the Bulls at 50 starting in 84 and potentially winning from 88-90.


many have 88-90 as park of MJ's absolute peak with 1991.

I actually have them higher but in case you noticed, it is in comparisons regarding those specific seasons where I am concluding Kareem was more valuable. 1988 is likely the situational value peak imo. Again, I juiced up 1988 and suppressed 77 and Kareem came up on top. 71 to 91 is much murkier but I stand by it being more likely that Jordan had more help than Kareem and 72 scales up very naturally both by impact(team improves(full-strength) despite teammate falling, team is still great without player(62 wins!), team elevates and outscores another all-time team when already diminished co-star has a second major drop-off because of injury).
and you have other guys who outperform him by so much that a 2nd round loss in 88, with the celtics/lakers/pistons still existing, gets you to a title.

I said potentially(90 I'd say is [b]likely[b]), but again, assuming era-translation? I wouldn't rule it out. The Bulls were not blown out by the Pistons, and if you beat the pistons, you have a shot. They're also probably going to have a higher-seed(potentially home court vs detroit) and easier early round competition. I'm more confident with Lebron than Kareem just on the availablity of data(and him basically just being a bigger, stronger, faster, more versatile and smarter jordan in 2009). Magic is injured in 89 so if you take the pistons you're probably champions. 1990, healthy opponent, but the east was the "real nba finals" arguably and by the 1990 playoffs Jordan had very good help. You might want to make a "fit" adjustment replacing pippen with someone like mark-price with Lebron but as I have it, Lebron already beat a pistons+ team in the 2013 Spurs with less and Kareem outscored a pistons+ team in the 72 Lakers(going by standard deviations too if you're worried about expansion) with less. Lebron anchored a -5 defense with his 2nd best defender missing half the season and put up better box-score stuff than Micheal while facing more defensive attention, running his team on both ends, and destroying "impact" as a concept. Put aside MJ's limited impact portfolio, what do you think Jordan was offering in terms of basketball to offset all that? Lebron even had similar turnover economy while handling the ball way more.
or so much they can boost a 6th seed in '89 to the title, with the bulls the only team that actually took a game off the pistons that year? keep in mind, kareem was capable of straight up missing the playoffs.

He was in seasons marred by injury. At full-strength though 75 Kareem "missing the playoffs" is offering comparable lift to a juiced 1988 MJ(And to be clear, I used "full-strength" for both players with the with and the without). That would also be an "off-year" not a "peak" by my book.
But I don't much like this framing as it gives(imo) undue preference for timing. What I do think is Kareem needed less to compete, less to win, and less to dominate(all things he actually proved imo). So if you add that to Kareem also being proven outside of specific circumstances and then you add him just being way better at basketball at the start, and you add him sustaining his excellence longer...

Yeah, very clear-cut to me who deserves to be ranked higher.

And again, whatever arguments Jordan does have, Russell's are better.


russell is certainly the jordan case on steriods. but the significantly weaker league, unrepeatable era boost, and the fact that one team in 8 having a coach/GM advantage with no free agency to break up a team makes repeated winning easier than in a 25-30 team league with free agency, is why i would ultimately hold him back, even behind kareem.

Sure, but then you should be fine with people glossing Steph and Duncan the same way. The "winning" gap is much smaller.

After all, his two most dominant MVP wins, best 2 regular-season teams, best overall team, and 3 championships comes with expansion. It is also in that watered down league(competition: Karl Malone) he looks like a league-best player by "impact". And within 6-years of 1998 the number of foreign nba-talent pool has already doubled.
OhayoKD
Head Coach
Posts: 6,042
And1: 3,932
Joined: Jun 22, 2022

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#234 » by OhayoKD » Mon Jul 10, 2023 11:24 am

Moonbeam wrote:
OhayoKD wrote:This is going to be blunt but, I think you need to use actual direct winning as a check for this to be useful. As in, check what these stats say relative to actual team results and how they correlate with what happens indiciually. Jordan's Bulls are not too affected by his depature defensively(as is typically the case with all-d guards). Russell sees better defensive results than duncan throughout his prime and then sees a much bigger defensive drop-off after his retirement than anything from anyone ever(and that overall drop-off is comparable to the increase we would get if we gave jordan all the credit for the bulls improvement from 84 to 88.

Writ large, big defense consistently rates lower in ws relative to impact(if you want more certainty you can check databall if you want more certainity) and guard defense sees the opposite effect.

You can make all these tiny corrections here and there but the fundamental issue is reliance on box-score which, as I've said a bunch before, only really tracks the end of a possession. A player handling the ball less, coordinating teammates(russell's biggest edge over everyone else listen), facing single-coverage as opposed to doubles(this was part of what occured with the triangle), gambling for steals and missing, the difference beween weaksidde help and being the primary deterrent, ect, ect ,ect.

All is completely missed here.

FWIW, by available "impact" data for the time period(its all raw more or less), I'd say kareem and russell are the standouts. Duncan/MJ seem like peers, and Wilt seems worse than advertised(69 is an all-time terrible singal both ways)

Second note. +5 in one period of time does not necessarily = +5 in another. This is most prevalent in the 60's and probably the mid-late 70's, but "best in the league" is different during different time periods and so is the value(at least towards chanpionships) of different scores


I'm not sure I fully understand. The direct winning is factored in due to the total team OWS/DWS, though, right? A player joining a good offensive/defensive team from one that isn't as good would expect to see an OWS/DWS increase per 48 by virtue of there being more team Win Shares to be allocated among the players, but that's kind of what we want in this, don't we? For example, the fact that the Spurs defense was good is testament to Duncan being good, so players joining the team (or leaving to another team) are most likely to have a net positive effect teaming up with Duncan (and a net negative effect when leaving him) due to the chance that they will be coming from (or going to) a team with a worse defense. Am I missing something here?

No. Unless I'm missing something, the "winning" is taken from bbr's individual defensive rating and [/b]individual offensive rating[/b]:
Crediting Defensive Win Shares to players is based on Dean Oliver's Defensive Rating. Defensive Rating is an estimate of the player's points allowed per 100 defensive possessions (please see Oliver's book for further details). Here is a description of the process (once again using LeBron James in 2008-09 as an example):

Offensive Win Shares are credited to players based on Dean Oliver's points produced and offensive possessions. The formulas are quite detailed, so I would point you to Oliver's book Basketball on Paper for complete details. The process for crediting Offensive Win Shares is outlined below (using LeBron James of the 2008-09 Cleveland Cavaliers as an example):

That defensive or offensive rating is derived from a formula tied to what is observable in the box-score:
Just as Oliver's Offensive Rating represents points produced by the player per 100 possessions consumed, his Defensive Rating estimates how many points the player allowed per 100 possessions he individually faced while on the court.

The core of the Defensive Rating calculation is the concept of the individual Defensive Stop. Stops take into account the instances of a player ending an opposing possession that are tracked in the boxscore (blocks, steals, and defensive rebounds), in addition to an estimate for the number of forced turnovers and forced misses by the player which aren't captured by steals and blocks.

The formula for Stops is:

Stops = Stops1 + Stops2
where:

Stops1 = STL + BLK * FMwt * (1 - 1.07 * DOR%) + DRB * (1 - FMwt)
FMwt = (DFG% * (1 - DOR%)) / (DFG% * (1 - DOR%) + (1 - DFG%) * DOR%)
DOR% = Opponent_ORB / (Opponent_ORB + Team_DRB)
DFG% = Opponent_FGM / Opponent_FGA
Stops2 = (((Opponent_FGA - Opponent_FGM - Team_BLK) / Team_MP) * FMwt * (1 - 1.07 * DOR%) + ((Opponent_TOV - Team_STL) / Team_MP)) * MP + (PF / Team_PF) * 0.4 * Opponent_FTA * (1 - (Opponent_FTM / Opponent_FTA))^2
Also necessary is the calculation of Stop%, which is the rate at which a player forces a defensive stop as a percentage of individual possessions faced (essentially the inverse of Floor%, but for defenders):

Someone else more qualified can decipher the math better, but from eye-ball, it looks heavy emphasis is put on "steals+blocks" and some sort of estimate for "shots contested" and "forced turnovers".

They also cut out a bunch for players form the 70's and earlier including(I think?) # of possessions which might prove especially problematic for the Celtics who played at a high-pace. By "direct winning", 69 Wilt(who has twice as high a score as 69 Russell in your metric) left his team and they were still really good for a team that was really good before he arrived. Russell then beats Wilt+that really good team in 69(posts a higher srs during the regular season too), leaves and with Hondo's scoring jumping up and a 6th man Sam Jones leaving the team experiences a 7-point drop-off with no health adjustment and an 8-point one just looking at games with the 4 starters.

Yet by your metric 69 Wilt is way better than 69 Russell? That doesn't seem like you've solved the box-score problems there.

Russell also scores higher in WOWY, wowyr(should not be taken seriously imo, but posters here use it so whatever) for his career, and his teams are not really affected by teammate departures through-out. Wilt doubling Russell's "win-shares" seems like a big contradiction with what the actual results suggest so I personally can't really justify putting that much stock on it.

I do appreciate the effort though.
User avatar
Moonbeam
Forum Mod - Blazers
Forum Mod - Blazers
Posts: 10,313
And1: 5,096
Joined: Feb 21, 2009
Location: Sydney, Australia
     

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#235 » by Moonbeam » Tue Jul 11, 2023 1:30 am

OhayoKD wrote:
Moonbeam wrote:
OhayoKD wrote:This is going to be blunt but, I think you need to use actual direct winning as a check for this to be useful. As in, check what these stats say relative to actual team results and how they correlate with what happens indiciually. Jordan's Bulls are not too affected by his depature defensively(as is typically the case with all-d guards). Russell sees better defensive results than duncan throughout his prime and then sees a much bigger defensive drop-off after his retirement than anything from anyone ever(and that overall drop-off is comparable to the increase we would get if we gave jordan all the credit for the bulls improvement from 84 to 88.

Writ large, big defense consistently rates lower in ws relative to impact(if you want more certainty you can check databall if you want more certainity) and guard defense sees the opposite effect.

You can make all these tiny corrections here and there but the fundamental issue is reliance on box-score which, as I've said a bunch before, only really tracks the end of a possession. A player handling the ball less, coordinating teammates(russell's biggest edge over everyone else listen), facing single-coverage as opposed to doubles(this was part of what occured with the triangle), gambling for steals and missing, the difference beween weaksidde help and being the primary deterrent, ect, ect ,ect.

All is completely missed here.

FWIW, by available "impact" data for the time period(its all raw more or less), I'd say kareem and russell are the standouts. Duncan/MJ seem like peers, and Wilt seems worse than advertised(69 is an all-time terrible singal both ways)

Second note. +5 in one period of time does not necessarily = +5 in another. This is most prevalent in the 60's and probably the mid-late 70's, but "best in the league" is different during different time periods and so is the value(at least towards chanpionships) of different scores


I'm not sure I fully understand. The direct winning is factored in due to the total team OWS/DWS, though, right? A player joining a good offensive/defensive team from one that isn't as good would expect to see an OWS/DWS increase per 48 by virtue of there being more team Win Shares to be allocated among the players, but that's kind of what we want in this, don't we? For example, the fact that the Spurs defense was good is testament to Duncan being good, so players joining the team (or leaving to another team) are most likely to have a net positive effect teaming up with Duncan (and a net negative effect when leaving him) due to the chance that they will be coming from (or going to) a team with a worse defense. Am I missing something here?

No. Unless I'm missing something, the "winning" is taken from bbr's individual defensive rating and [/b]individual offensive rating[/b]:
Crediting Defensive Win Shares to players is based on Dean Oliver's Defensive Rating. Defensive Rating is an estimate of the player's points allowed per 100 defensive possessions (please see Oliver's book for further details). Here is a description of the process (once again using LeBron James in 2008-09 as an example):

Offensive Win Shares are credited to players based on Dean Oliver's points produced and offensive possessions. The formulas are quite detailed, so I would point you to Oliver's book Basketball on Paper for complete details. The process for crediting Offensive Win Shares is outlined below (using LeBron James of the 2008-09 Cleveland Cavaliers as an example):

That defensive or offensive rating is derived from a formula tied to what is observable in the box-score:
Just as Oliver's Offensive Rating represents points produced by the player per 100 possessions consumed, his Defensive Rating estimates how many points the player allowed per 100 possessions he individually faced while on the court.

The core of the Defensive Rating calculation is the concept of the individual Defensive Stop. Stops take into account the instances of a player ending an opposing possession that are tracked in the boxscore (blocks, steals, and defensive rebounds), in addition to an estimate for the number of forced turnovers and forced misses by the player which aren't captured by steals and blocks.

The formula for Stops is:

Stops = Stops1 + Stops2
where:

Stops1 = STL + BLK * FMwt * (1 - 1.07 * DOR%) + DRB * (1 - FMwt)
FMwt = (DFG% * (1 - DOR%)) / (DFG% * (1 - DOR%) + (1 - DFG%) * DOR%)
DOR% = Opponent_ORB / (Opponent_ORB + Team_DRB)
DFG% = Opponent_FGM / Opponent_FGA
Stops2 = (((Opponent_FGA - Opponent_FGM - Team_BLK) / Team_MP) * FMwt * (1 - 1.07 * DOR%) + ((Opponent_TOV - Team_STL) / Team_MP)) * MP + (PF / Team_PF) * 0.4 * Opponent_FTA * (1 - (Opponent_FTM / Opponent_FTA))^2
Also necessary is the calculation of Stop%, which is the rate at which a player forces a defensive stop as a percentage of individual possessions faced (essentially the inverse of Floor%, but for defenders):

Someone else more qualified can decipher the math better, but from eye-ball, it looks heavy emphasis is put on "steals+blocks" and some sort of estimate for "shots contested" and "forced turnovers".

They also cut out a bunch for players form the 70's and earlier including(I think?) # of possessions which might prove especially problematic for the Celtics who played at a high-pace. By "direct winning", 69 Wilt(who has twice as high a score as 69 Russell in your metric) left his team and they were still really good for a team that was really good before he arrived. Russell then beats Wilt+that really good team in 69(posts a higher srs during the regular season too), leaves and with Hondo's scoring jumping up and a 6th man Sam Jones leaving the team experiences a 7-point drop-off with no health adjustment and an 8-point one just looking at games with the 4 starters.

Yet by your metric 69 Wilt is way better than 69 Russell? That doesn't seem like you've solved the box-score problems there.

Russell also scores higher in WOWY, wowyr(should not be taken seriously imo, but posters here use it so whatever) for his career, and his teams are not really affected by teammate departures through-out. Wilt doubling Russell's "win-shares" seems like a big contradiction with what the actual results suggest so I personally can't really justify putting that much stock on it.

I do appreciate the effort though.


You're right that individual Win Shares are tied to individual ORating/DRating, but these individual measures are basically partitions of team-level Win Shares and ORating/DRating to the team's players. On a team-level, Offensive and Defensive Win shares are nearly perfectly correlated to a team's Offensive and Defensive Rating. Team Offensive and Defensive Ratings (points scored/allowed per 100 possessions) are not disputed as reasonable measures of team performance on offense and defense as far as I know. That a team's total Offensive and Defensive Win Shares are so strongly correlated with team-level Offensive and Defensive Ratings is a pretty convincing selling point of the team totals, IMO.

I've paused fully writing up the metrics I'm developing while I explore some issues with age curves, but I'll post this start to an initial draft as the beginning of it showcases what I mean about the team-level stuff. Note that the sections about Adjusting Win Shares and so forth are still in development. The general framework is one I think has a lot of promise, but the details of how to determine expected Win Shares for teammates is what is the current holdup. Here it is.
OhayoKD
Head Coach
Posts: 6,042
And1: 3,932
Joined: Jun 22, 2022

Re: RealGM 2023 Top 100 Project - #2 (Deadline 7/6 11:59pm) 

Post#236 » by OhayoKD » Tue Jul 11, 2023 7:25 pm

Moonbeam wrote:
OhayoKD wrote:
Moonbeam wrote:
I'm not sure I fully understand. The direct winning is factored in due to the total team OWS/DWS, though, right? A player joining a good offensive/defensive team from one that isn't as good would expect to see an OWS/DWS increase per 48 by virtue of there being more team Win Shares to be allocated among the players, but that's kind of what we want in this, don't we? For example, the fact that the Spurs defense was good is testament to Duncan being good, so players joining the team (or leaving to another team) are most likely to have a net positive effect teaming up with Duncan (and a net negative effect when leaving him) due to the chance that they will be coming from (or going to) a team with a worse defense. Am I missing something here?

No. Unless I'm missing something, the "winning" is taken from bbr's individual defensive rating and [/b]individual offensive rating[/b]:
Crediting Defensive Win Shares to players is based on Dean Oliver's Defensive Rating. Defensive Rating is an estimate of the player's points allowed per 100 defensive possessions (please see Oliver's book for further details). Here is a description of the process (once again using LeBron James in 2008-09 as an example):

Offensive Win Shares are credited to players based on Dean Oliver's points produced and offensive possessions. The formulas are quite detailed, so I would point you to Oliver's book Basketball on Paper for complete details. The process for crediting Offensive Win Shares is outlined below (using LeBron James of the 2008-09 Cleveland Cavaliers as an example):

That defensive or offensive rating is derived from a formula tied to what is observable in the box-score:
Just as Oliver's Offensive Rating represents points produced by the player per 100 possessions consumed, his Defensive Rating estimates how many points the player allowed per 100 possessions he individually faced while on the court.

The core of the Defensive Rating calculation is the concept of the individual Defensive Stop. Stops take into account the instances of a player ending an opposing possession that are tracked in the boxscore (blocks, steals, and defensive rebounds), in addition to an estimate for the number of forced turnovers and forced misses by the player which aren't captured by steals and blocks.

The formula for Stops is:

Stops = Stops1 + Stops2
where:

Stops1 = STL + BLK * FMwt * (1 - 1.07 * DOR%) + DRB * (1 - FMwt)
FMwt = (DFG% * (1 - DOR%)) / (DFG% * (1 - DOR%) + (1 - DFG%) * DOR%)
DOR% = Opponent_ORB / (Opponent_ORB + Team_DRB)
DFG% = Opponent_FGM / Opponent_FGA
Stops2 = (((Opponent_FGA - Opponent_FGM - Team_BLK) / Team_MP) * FMwt * (1 - 1.07 * DOR%) + ((Opponent_TOV - Team_STL) / Team_MP)) * MP + (PF / Team_PF) * 0.4 * Opponent_FTA * (1 - (Opponent_FTM / Opponent_FTA))^2
Also necessary is the calculation of Stop%, which is the rate at which a player forces a defensive stop as a percentage of individual possessions faced (essentially the inverse of Floor%, but for defenders):

Someone else more qualified can decipher the math better, but from eye-ball, it looks heavy emphasis is put on "steals+blocks" and some sort of estimate for "shots contested" and "forced turnovers".

They also cut out a bunch for players form the 70's and earlier including(I think?) # of possessions which might prove especially problematic for the Celtics who played at a high-pace. By "direct winning", 69 Wilt(who has twice as high a score as 69 Russell in your metric) left his team and they were still really good for a team that was really good before he arrived. Russell then beats Wilt+that really good team in 69(posts a higher srs during the regular season too), leaves and with Hondo's scoring jumping up and a 6th man Sam Jones leaving the team experiences a 7-point drop-off with no health adjustment and an 8-point one just looking at games with the 4 starters.

Yet by your metric 69 Wilt is way better than 69 Russell? That doesn't seem like you've solved the box-score problems there.

Russell also scores higher in WOWY, wowyr(should not be taken seriously imo, but posters here use it so whatever) for his career, and his teams are not really affected by teammate departures through-out. Wilt doubling Russell's "win-shares" seems like a big contradiction with what the actual results suggest so I personally can't really justify putting that much stock on it.

I do appreciate the effort though.


You're right that individual Win Shares are tied to individual ORating/DRating, but these individual measures are basically partitions of team-level Win Shares and ORating/DRating to the team's players. On a team-level, Offensive and Defensive Win shares are nearly perfectly correlated to a team's Offensive and Defensive Rating. Team Offensive and Defensive Ratings (points scored/allowed per 100 possessions) are not disputed as reasonable measures of team performance on offense and defense as far as I know. That a team's total Offensive and Defensive Win Shares are so strongly correlated with team-level Offensive and Defensive Ratings is a pretty convincing selling point of the team totals, IMO.

I've paused fully writing up the metrics I'm developing while I explore some issues with age curves, but I'll post this start to an initial draft as the beginning of it showcases what I mean about the team-level stuff. Note that the sections about Adjusting Win Shares and so forth are still in development. The general framework is one I think has a lot of promise, but the details of how to determine expected Win Shares for teammates is what is the current holdup. Here it is.

Does that correlation hold for teams which data is cut out for?(50-74)
User avatar
homecourtloss
RealGM
Posts: 11,375
And1: 18,774
Joined: Dec 29, 2012

Re: RealGM 2023 Top 100 Project - #2 (Kareem Abdul-Jabbar) 

Post#237 » by homecourtloss » Fri Jul 21, 2023 3:06 am

This was a great post from the 2017 top 100 project. It’s too bad lot of the clips or actually most of them aren’t working now.

Blackmill wrote:Because of the early start, there is a lot I couldn't include, so here's a much abbreviated post on my thoughts. I may add some more later this evening.

_______________________________Film Study: Kareem's Defense_______________________________

    Image

I'm starting with Kareem's defense because opinions on it vary from he was a lazy defender to he was an all time great defender. I think Kareem, at his best, was an all time great defender. Take a look at the play above. There's a couple components to it.

    1. Kareem shows on the PnR so Bing can't pull up.
    2. Kareem quickly recovers to his man (Hayes).
    3. On the second PnR, Kareem strongly contests Havlicek's pull up, forcing a pass.
    4. Kareem moves to the strongside, but because of miscommunication, Havlicek is left open.
    5. Kareem runs Havlicek off his jumper, and though Kareem seems beat on the drive, his length lets him to alter the shot.

What Kareem does in this play is nothing short of exceptional. He has the quickness to show on the PnR and recover, something that has become all the more important with the three point line, but many bigs can't do. And while smaller centers could, they wouldn't be able to salvage the miscommunication, stopping both the spot up and the drive through sheer length and size.

You may notice that this is not a regular season game. This is the 1976 NBA All Star Game, which I will tell you, is one of the most important games we have of Kareem's career. From the mid- to late-70's, we only have four full games, which are of Kareem's 1977 playoff run and the 1976 All Star Game. Comparatively, the 1976 All Star Game is especially important for several reasons:

    1. We see Kareem guarding the PnR more in this game than any other.
    2. We see Kareem defending in a much more spaced offensive environment than usual.
    3. With capable teammates, Kareem was able to be much more active on help defense.

The last point is particularly important. It can be observed in the 1977 playoffs that the Lakers routinely failed to help the helper when Kareem rotated. Samurai, a poster who you may know for his insights and having watched much of Kareem's career, made this observation several years ago:

[...]Washington gave LA the big power forward to battle someone like Lucas and give Kareem an enforcer in much the same way that Lucas helped Walton. KAJ could be more active in help defense, switch off, and chase bigs who could shoot outside because he knew that Kermit was there to grab the rebound.[...]


Hopefully it's clear now why the 1976 All Star Game is of particular importance. I will now go through several of Kareem's of defensive plays this game. First, let's look at four clips which show Kareem's understanding of passing angles, a crucial trait of any good help defender.

    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


In the first play we see Kareem sag back as he sees the cutter about the flash across the lane. Then we see several clips of Kareem anticipating the pass and deflecting it. The last clip is an example where there's no player to help the helper, which Kareem is very aware of, and positions his body to prevent the pass to his man while using his great length to disrupt the drive.

Kareem's defensive awareness, when in position to be a help defender, was often superb. Now, let's look at his PnR defense, which is special for a player his size.

    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


One thing we see in all these clips is Kareem's willingness to show or switch on the PnR and that he can effectively recover. In the first clip we see Havlicek trying to attack Kareem when he shows, since Kareem had been stopping him from pulling up off the screen, but Kareem has the speed to stay with him and the block the layup. In the last clip Kareem does a lot:

    1. He shadows Frazier, who has a step on his defender, forcing the ball to kicked out.
    2. He closes out on Hayes, stays in front on the drive attempt, and forces a tough jumper.
    3. The East gets the offensive rebound, and Kareem seems out of the play, but his length lets him get an improbable block.
    4. Kareem also gets the rebound.

Now, I do want to spend some time on Kareem's defense during the 1977 playoffs, even though he had limited freedom to help. One thing I've mentioned about some of the previous clips is how Kareem can appear out of the play but still alter or block the shot because of his length and mobility. This carried into the 1977 playoffs with plays like these:

    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


The last play really show off Kareem's length. He blocks Walton's dunk, standing to the right of the basket, while Walton attempted it from left of the basket. Kareem still displayed a willingness to show or switch on the PnR too (thought not as often):

    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


You'll find many other possessions like the ones I've shown in the 1976 All Star Game. These are just a couple. As Kareem exited his defensive prime, he was distinctly less active defensively, but still a force at the rim and able to briefly switch onto smaller players. For lack of time, I only have footage from the 1980 finals, but that should paint a fair picture. At 32, Kareem was still a great rim protector:

    Spoiler:
    Image


    Spoiler:
    Image


Kareem's interior presence was enough to influence shots without needing to contest. Often, layups attempted near him were done hastily, and missed as a result. Even wide open ones.

    Spoiler:
    Image


    Spoiler:
    Image


He prevented many shots from even being attempted as driving players would kick out the ball if he was close.

    Spoiler:
    Image


    Spoiler:
    Image


And, as mentioned, he still would switch onto smaller defenders though far less frequently:

    Spoiler:
    Image


Finally, I want to address what I'm guessing causes some people to think Kareem was a lazy defender. Occasionally he would make no effort to contest the outside shot. For instance,

    Spoiler:
    Image


Now, while this looks bad, it is very much the opposite. We regularly see how poor three point shooters are sometimes allowed the shot in order to cover more dangerous players. The Spurs just about gave Lebron any open jumper in Game 7 of the 2013 Finals. Sometimes it works, sometimes it doesn't, but what matters is the dice roll is in your favor. In three games of the 1980 Finals, and three games of the 1977 Playoffs, I counted Kareem making no effort to contest the outside shot 21 times. The opposition converted 33%. When the jumper was attempted closer to the basket, Kareem was much more focused on contesting it, as we see Kareem go over a screen to contest Walton's short jumper:

    Spoiler:
    Image


Kareem not contesting long jumpers, by players not known for shooting, was neither uncommon nor incorrect.

What I hope's apparent from all this is that Kareem, in his defensive prime and with the liberty to help, was constantly aware, moving, and disrupting the defense with his combination of length, height, and quickness. He was nothing short of a tremendous defender who could compete with the all time greats. After his defensive prime, he was still exceptional at preventing shots near the rim, and more mobile than your typical seven-footer.

_______________________________Film Study: Kareem's Offense_______________________________

    Image

I think we're all familiar with Kareem's sky hook. You can see it above. What do you think Kareem shot on sky hooks? Below 50%? Better? In the games I've recorded, Kareem shot better than 65%, with a sample of over 60 sky hooks. This shot is one of the primary reasons I consider Kareem the GOAT, and thus, I will spend some time discussing it.

I would argue the sky hook makes Kareem the most resilient scorer to have played in the NBA. By most resilient, I mean defense and team composition have the least affect on his efficiency and volume. Just consider how efficient remained as he aged. Despite becoming slower and weaker, which in effect is like the defense become stronger and faster, his efficiency endured.

The only shots of comparable efficiency are at the rim or behind the arc. To get to the rim, though, is much harder than to receive the ball at the low or mid post. This is especially true on team's lacking shooters to keep the defense honest. As for the three, it's harder to stop than a drive, but often relies on off-ball movement. While this isn't bad, holding players off-ball has always gone uncalled, particularly in the playoffs, and that's precisely why the Warriors abandoned their vaunted, off-ball movement for the Durant & Curry PnR in Game 5. We see this in the next few clips how the defense can have no holes and Kareem can still take the sky hook.

    Spoiler:
    Image


    Spoiler:
    Image


Kareem could also take the sky hook from the mid post where aggressive double teams would be unacceptably compromising to the defense:

    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


The sky hook was practically a cheat. There's no other shot where defense matters so little. Aside from being a scorer that few can compare to, Kareem was also an exceptional passer, and could operate from both the low and mid post. His mid post passing was particularly impressive and, at times, reminiscent of Walton's play:

    Spoiler:
    Image


    Spoiler:
    Image


From the low post Kareem could would often pass to the baseline:

    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


    Spoiler:
    Image


but was also capable of passing to players cutting down the middle:

    Spoiler:
    Image


(It's too bad I didn't have time to go through my footage of his later years. He had some extraordinary bounce passes against the Celtics and Mavs in the mid 80s.)

Kareem's offense is an unusual combination of superb passing and scoring. We rarely see these players, and when we do they tend to be guards, but Kareem was a center with the mind and body to warrant comparison with the all time great defensive players. I think he's singular in having a case as a top ten offensive player and defensive player. He gets my vote as the GOAT.

First Vote: Kareem
Second Vote: Duncan


Blackmill wrote:I don't know if this should be posted here, since Kareem has been voted in, but last time I posted in the old voting thread my post was moved. And, after all, it's discussion which everyone might be interested in.

drza wrote:
Doctor MJ wrote:Real quick:

Someone made a statement implying that Kareem's impact could be approximated by Shaqs impact. I disagree.

Shaqs attacks are rapid. The defense must prepare before he gets the ball because afterward there's little to be done. This fact is what gives him gravity, and it is a gravity optimized for giving space to perimeter teammates.

Kareem is methodical. The defense has time to adjust optimally regardless of their set up before he gets the ball. He probably has more in common with Dantley than Shaq in terms of the shape of his impact.


Sent from my iPhone using RealGM Forums


You just made the same point I did, but in a much quicker, more concise method. And to reiterate, I agree...Kareem didn't have Shaq's gravity. If he did, his impact would have been larger than it was


If you don't mind, I'll also disagree. Shaq attacked quickly because he needed to. He's only allowed three seconds in the paint, and unlike Kareem, his shots from 3-10 feet were not good enough to carry an offense.

If you read my earlier post, one of the reasons I chose Kareem was that he could get a great shot, despite being methodical and the defense being set up. And this did manifest itself in a very noticeable gravity, because once Kareem got the ball, knowing what he'd do didn't matter nearly so much as for other players. As such, a second defender often fronted him, to the point of giving up the drive baseline or an open jumper. In 1984, when Magic developed an outside shot, the majority (yes, I'm betting over half) of his open jumpers were because of Kareem's gravity. And this is well past his offensive peak.

If there's a reason this didn't produce more impact, it's because many of his teammates weren't great shooters, or the weakside spacing was too poor for the baseline drive to be attempted. But this has less to do with Kareem than the personnel surrounding him and the coaching.

Here's some videos of open drives and jumpers resulting from Kareem's gravity:

Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Spoiler:
Image


Much of Kareem's attacks came a bit further away from the rim. If you had a solid post defender you could keep him from being close and thus "collapse" didn't mean the perimeter was open.

In the case of the classic series against Portland the Blazers just put Walton on him and let the rest of the offensive players just do nothing as Kareem spent many second working to get his shot.


Collapsing the defense further isn't necessarily better. As I've mentioned above, fronting Kareem often gave up the open baseline drive, which wouldn't be available if Kareem operated closer to the basket. And he did generate plenty of open jumpers once he had shooters comfortable taking shots from further out.

About the Portland series, I don't think your assessment is fair, since Kareem's teammates consistently struggled to advance the ball past half court. They couldn't dribble against Portland's ball pressure and they weren't good enough shooters to regularly exploit the times Kareem was doubled. A lot of players won't do anything offensively in a series if they can't dribble or shoot well from distance. Kareem did receive off-ball doubles during the series so saying he was played only by Walton is not entirely true.
lessthanjake wrote:Kyrie was extremely impactful without LeBron, and basically had zero impact whatsoever if LeBron was on the court.

lessthanjake wrote: By playing in a way that prevents Kyrie from getting much impact, LeBron ensures that controlling for Kyrie has limited effect…

Return to Player Comparisons


cron