RealGM 2017 Top 100 List #2

pelifan · Post #61 » by **pelifan** » Thu Jun 22, 2017 6:11 am

I dont have a vote but my mount Rushmore of the NBA is Jordan, Kareem, Russell, Lebron, Wilt in that order.

I'd love to hear more about Wilt when his name finally comes around. Some of the numbers around him on and off the court (gross, I mean his track and field numbers) are insane. 6-6 high jump 10.9 100 meters 40 inch vertical 4.6 40, grabbing coins off the top of the backboard, shotput 53 feet :-?

. Arguably the greatest athlete who ever lived, just maybe not the best basketball player. I'd like to see all the best clips and stories.

Post #62 » by **penbeast0** » Thu Jun 22, 2017 6:15 am

ardee wrote:
penbeast0 wrote:1. Modern
2. 2000s
3. 60s (concentration of talent)
4. 90s
5. 80s (serious expansion weakening)
6. 70s (extreme expansion issues)
7. 50s (segregation)

Why 90s over 80s? IMO the main expansion happened in the latter half of the 90s where the league got severely diluted.

First, the talent pool had not caught up with the expansion of the 70s yet where the league more than doubled from 9 teams in 66 to 22 in 77. The later expansions were:
1 in 80
2 in 88
2 in 89
2 in 95
a 20%+ expansion in 7 years from 88 to 95.
Still serious watering down of league wide talent but not as extreme as the effects of the mass expansion where the NBA was trying to freeze out the ABA in the late 60s/70s.

drza · Post #63 » by **drza** » Thu Jun 22, 2017 7:25 am

By now, those reading this thread/project know that I wrote up a peak comparison of Kareem, Walton (the unicorn of healthy Walton), and Duncan using some scouting notes, WOWY scores during near-peak seasons for all three, and impact trends established from the databall era to suggest that Kareem's game of all-history scoring, strong defense and strong passing lacked the impact of Walton's all-history defense, all-history point big man passing, and only adequate scoring or Duncan's all-history defense, strong scoring and strong passing. That Kareem's prime, despite the gaudy boxscore numbers, falls prey (when compared to other GOAT impact bigs) to the notion that a big man almost always has his best impact on defense, and even on offense a big might have more impact as a distributor even with lesser scoring than as a high volume, high efficiency scorer. Post here: http://forums.realgm.com/boards/viewtopic.php?p=56474291#p56474291

However, there has been some pushback. TRex questioned whether ElGee properly used all of Duncan's absence data, which could potentially affect his score. Blackmill did an impressive simulation to show that there is noise in WOWY data that could account for some of the difference between Kareem's scores and Duncan's. And upon further review, I think ElGee may have flipped the sign of Kareem's 1978 while-playing SRS. The pushback has validity, because I indeed recognize that we don't have a huge amount of impact data for the pre-databall era. By definition, I'd expect there to be some noise here.

But with that said...the difference between Kareem's peak impact and Walton's peak impact, in era at simultaneous times, does seem to be a significant effect and not really subject to any of the push-back or potential sign flips. For me, at least, the notion that this is Kareem at his best, in all of his boxscore glory, but that he had a contemporary that was clearly making a bigger impact on his team's fortunes by utilizing the very horizontal/active defense and big-man-floor-general principals that show up as clear big impact indicators in the databall era...that's a red flag for me. It makes me question just how valuable his boxscore glory really is, in the big scheme of things. (And again, I feel I should emphasize, I mean this only with respect to other GOATs. Obviously Kareem was having a big impact, and it shows up in the numbers. Just not as well as perhaps the other GOATs).

Kareem vs Russell, impact observations and rough estimates

Anyway, tonight I've been taking a different crack at impact analysis, comparing Kareem with a different GOAT big man: Bill Russell. And again, right off the bat, I'll stipulate that this is not granular, copious data like what we have access to in the databall era. This won't be a mathematically rigorous proof. And frankly...I'm hesitant to lean too hard on this type of analysis, because it relies on team unit ratings, and anyone that's ever read my posts knows that I'm all about isolating an individual's impact from the team's impact as much as possible.

But, with that said, it's the data that we have available. Plus, by looking at trends over long periods of time with the players in different situations, I'm hoping that some interesting tidbits might fall out. (Thanks to ElGee, who pointed out to me that bball-ref now has the estimated team offense/defense ratings back to the shotclock and thus made this analysis angle possible for me). And, this analysis isn't in a vacuum, as I still have the results of the more WOWY-type approach in my mind as well.

1) Russell.
Years ago, on BackPicks, ElGee made a cool chart and article, showing how the Celtics' defensive rating tracked perfectly with Russell's career: https://elgee35.wordpress.com/2010/12/31/bill-russells-defensive-impact/ . The defense became great Russell's rookie season, improved into ridiculous territories during Russell's peak, stayed great until Russell retired, then fell off a cliff as soon as he left.

I replicated that chart, tonight, using the BB-ref data. I also did the same for the Celtics' offense. As has been pointed out, the Celtics' offense during the Russell era was routinely below average, sometimes significantly below. Thus, it seems clear that the Celtics won those 11 championships purely on the strength of their all-time defense. And the defense tracked perfectly with Russell, staying consistently great no matter which teammates came and went. On offense, the ratings barely changed at all when Cousy retired, nor did either rating really change when Hondo joined the squad. Looking at averages of relative offensive and defensive ratings:

Celtics (2 years pre-Russell): ORtg-rel:: +1.7, DRtg-rel: +1.5 (- is good)
Celtics (13 years of Russell): ORtg-rel: -1.6, DRtg-rel: -6.9
Celtics (2 years post-Russell): ORtg-rel: -1.0, DRtg-rel: -1

Kareem
Kareem entered the league the year that Russell retired, so they were very close to being contemporaries. The year before he arrived, the Bucks were poor offensively (-1.6) and defensively (+2.7). In Kareem's first year, they became a solid offense (3.1) and improved to a not-terrible defense (-0.9). I'd say that was strong impact for rookie Kareem. A similar but attenuated thing happened when he went to the Lakers, as they went from a similar offense/defense to the pre-Kareem Bucks to a neutral offense (+0.6) and a neutral defense (+0.5, which was an improvement). Kareem obviously had an impact.

But, Russell was able to define a dominant unit, in a clear way, that showed up very clearly in the team unit ratings regardless of the coming and going of his biggest name teammates. Kareem played on some dominant units through the years, but that dominance didn't track with his presence, as much as it did the presence of two particular teammates...

Bucks (2 years of Kareem, no Oscar): ORtg-rel: +1.7, DRtg-rel: -0.4
Bucks (4 years of Kareem & Oscar): ORTg-rel: 4.2, DRtg-rel: -4.8
Royals (Oscar's entire career): ORtg-rel: +3.3, defense sucked

Lakers (4 years of Kareem, no Magic): ORtg-rel: 1.8, DRtg-rel: -0.2
Lakers (10 years of Kareem & Magic): ORtg-rel: +4.9, DRtg-rel: -1.0
Lakers (2 years of Magic, no Kareem): ORtg-rel: +5.1, DRtg-rel, -2

Let me unpack this. In 6 years of prime/peak career when he wasn't playing with Oscar or Magic, Kareem had monster boxscore numbers and multiple MVPs, but his team offenses were only +1.7 or +1.8. The offenses only got elite when the legendary point guards were around. And those offenses with Kareem + point guard played more similar to the offenses of those point guards without Kareem, than they did like Kareem's offenses without them.

On defense, Kareem had a four-year streak while playing with Oscar on the Bucks where his defenses were really strong. In the two Bucks years without Oscar, and essentially his entire Lakers career, the defenses were meh. I've seen it argued that Oscar (who clearly didn't have anything to do with the defensive results directly) motivated Kareem in those Bucks years in ways that no one else ever did before or after, and that thus this helped explain Kareem's dominant defensive stretch, if you give any credence at all to that. Perhaps more importantly, by the 1972 - 1976 period, the NBA had more than twice as many teams AND also an ABA, compared to Russell's hey-day. Things were very watered down in the early 70s NBA. The ABA collapsed early in Kareem's Lakers career, though. So, another factor in Kareem's dominant defensive run could be that the league was weak, right when he was most motivated.

Bottom line: This is another attempt at gauging Kareem's impact stats. Again, vs another All-time great, using the same method.

Russell's impact translated faithfully and obviously in his team's dominant defense regardless of teammate turnover, which was clearly the unit that made them dominant as a team in a concentrated league.

Kareem's historical boxscore dominance didn't seem to translate to great offenses in 6 prime/peak years. It was only during the years that he had an all-history point guard whose teams minus Kareem had similar caliber offenses, that Kareem's team offenses looked great. Kareem had one four-year stretch where he led a dominant team defense, but it also came at the most watered down competitive period in NBA history.

Again...nothing here is set in stone. This is not a mathematical proof, and I'm not typing QED. But this is now two different impact approaches, compared against several other all-time great bigs, where Kareem just doesn't show up as well. Walton and Russell seemed, to the extent that I can examine the evidence, to have significantly larger impacts on their teams' fortunes than Kareem did. It certainly seemed that Duncan may have, as well.

Someone in the last thread opined whether Kareem's actual impact may have been more on the level of a Barkley than a Jordan. That he may have been great, but not GOATish as far as how much he was impacting the game. I once made a similar analogy, but to Karl Malone instead of Charles Barkley. And the Malone comp may be more apt, because both had the boxscore accolades and the absurd longevity. But, just like Malone, I wonder if absurd longevity at great-but-not-GOAT level is worth more than a shorter career that routinely hit GOAT level.

I don't think that it is.

Right now, not only would I vote Russell ahead of Kareem...I'd say that (in no order) Magic, Duncan, LeBron, Garnett, Wilt, Shaq, Bird, and Olajuwon all have a heck of an argument as well. At this moment, I feel strongly that each of those others have spent time on that GOAT level that Walton and Russell seemed to reach but that I'm less and less convinced that Kareem ever did.

Vote: Bill Russell
2nd: will edit in later

Gibson22 · Post #64 » by **Gibson22** » Thu Jun 22, 2017 7:30 am

RSCD3_ wrote:
Texas Chuck wrote:I'm casting my official vote for Mr. William Russell.

My primary reasoning is really simple--he had the single greatest impact any player has ever had in-era and he did it for a length of time that completely validates his candidacy for GOAT. This impact was most largely seen through his combination of defensive intelligence, ingenuity, and physical ability. He simply thought and played defensive basketball at a level never seen before and unlikely we will ever seen again. And I think its a near certainty that we will never see someone that much further ahead of anyone else in any element of basketball.

We could stop there and he's a GOAT candidate imo. But then we look at his rebounding which belongs in a discussion of the greatest rebounders of all-time and almost certainly the most valuable rebounder of all-time because of what happened next. He didn't just hold the ball and wait for the PG to make himself available. No he jump-started their high-pace offense through his own ability to handle the ball, his intelligent pushing of the ball through the pass. And of course he didn't just pass the ball and glide his way up the court like most big men--he was a weapon in transition.

He gets killed for his offense here for reasons I still don't comprehend. This was an era of inefficient scoring. Nobody was shooting the ball well. But Russell was leading or second on his team in FG% nearly every season, he was in the top five in FG% in the entire league 4 years in a row. He wasn't a poor scorer by the standards of his day. And scoring was well down the list of what he brought to the team offensively. This "horrible" offensive player was putting up 20 points and 5 assists in the playoffs on 4 separate occasions.

His rap as a defense only player is undeserved. And while its subjective, I feel like it matters too and that's his unprecedented leadership as a player and then as a player-coach.

I went into this feeling like 5 players deserved my serious consideration for #1. Mike was one of them and he is in at #1. Will edit back in my vote for my 2nd choice at #2 after reading some more thoughts. Surprised Lebron isn't getting much traction yet, but I'm listening hard to the Duncan and Kareem stuff and am leaning Timmy over Kareem right now, but not ruling out Lebron just yet either.

Compared to the list of 2014
1. Michael Jordan
2. Kareem Abdul-Jabbar
3. Bill Russell
4. Wilt Chamberlain
5. Tim Duncan
6. Shaquille O'Neal
7. LeBron James

It's crazy how these 3 years by LeBron may have only moved him up 1-2 spots

What do you mean? I think it's clear that he's going to be #3

lorak · Post #65 » by **lorak** » Thu Jun 22, 2017 8:12 am

drza,

Celtics in 1957 already improved a lot (probably also on defense) before Russell joined them:

Code: Select all

SRS   ORTG   DRTG   YEAR
0,72   1,9   1,4   1956
3,60   1,2   -2,7   1957 w/o Bill
5,40   -1,3   -6,1   1957 w/ Bill
5,02   -0,8   -5,2   1958
5,84   -0,7   -5,7   1959
7,62   -0,1   -6,2   1960
4,94   -3,4   -7,6   1961

It seriously should be questioned, if during first 4-5 years (I'm not sure what to think about 1960) his impact was at GOAT level. He basically improved 3-4 SRS team to 5-6 SRS.* No doubt on defense he was great, but at the same time on offense he was negative player - 1957 shows it perfectly (and keep in mind, that Boston before 1957 basically every year was among the best offensive teams in the NBA, so it isn't like he had bad offensive supporting cast).

* Or maybe his impact was even smaller if we control for games missed by other starters (data from Elgee):

Cousy-Sharman-Heinsohn-Luscatoff: +5.8 SRS in 19g
Cousy-Sharman-Heinsohn-Luscatoff-Russell: +6.3 SRS in 48g

PS
If anyone would like to use sample of combined Russell's games missed after 1957, then keep in mind, that Celtics didn't have even decent replacement for him - for example in 1962 (when he missed 4 games) 2nd tallest player in Boston had 6-7 (Heinsohn). Besides such data (2-3 games missed from each season during -10 years span) is less valuable than big sample from one year. Of course Bill definitely improved after rookie campaign, but first several years don't show much improvement as Celtics stayed at ~the same level.

kayess · Post #66 » by **kayess** » Thu Jun 22, 2017 8:44 am

This KAJ vs TD debate feels like a prelude to a "KG better than both" hammer drop. Not that it can't be true - but I feel like allowances are being made in those two's favor that isn't being made for Kareem.

Impact's the thang, but if a player has the skill-set for it and is unable to express the impact it would, we shouldn't be so quick to conclude that he didn't because he couldn't - his environment obviously plays a huge role into that as well.

Can we get opponent O/DRTG adjusted WOWY numbers for those KAJ stretches? Blackmill's analysis was great, but it does feel like a SSS - TheRegulator, do you have some articles that will illuminate what KAJ's defense was truly like?

BasketballFan7 · Post #67 » by **BasketballFan7** » Thu Jun 22, 2017 10:26 am

Joao Saraiva wrote:
What separates KAJ from Bill Russell? (since I think he will be the other main candidate along LeBron and I've already discussed LBJ)

- Offensive impact. Sure Russell has the edge on D, but I don't feel it's gigantic. Clear one. On the other end of the floor, I think Russell doesn't even belong in the same page as KAJ. I can understand the defensive argument for Russell against players who don't come near him in defensive impact like Magic or Bird, but not against KAJ; (I see a good case for LeBron here)

This line of thinking is repeatedly used to the detriment of Russell. Are there any data to back it up? To me it seems more like an intuitive thought than anything else. As if we are saying, "there is no possible way that Russell's defense can be impactful enough to overcome the advantage that KAJ has over Russell's pitiful offense (by GOAT standards)". Which, to me, doesn't seem like a trustworthy foundation to form an opinion on when there is evidence to the contrary. And, let's not forget, Russell retired just a year before KAJ came into the league, and a 34 year old Russell was still leading the best defense in the NBA. The league Russell dominated was stronger than the drug laden, divided environment that KAJ dominated in the 1970s.

Xherdan 23 · Post #68 » by **Xherdan 23** » Thu Jun 22, 2017 10:30 am

I have a feeling I'll have a lot of questions during this project.

I'd be interested to know if anyone voted Jordan for#1 and Russell at #2 and if so what's your criteria?

It seems to me that any consistent criteria that would favor MJ over Russell should also favor LeBron over Russell.

Same question for guys voting Duncan over KAJ.

Joao Saraiva · Post #69 » by **Joao Saraiva** » Thu Jun 22, 2017 10:51 am

BasketballFan7 wrote:
Joao Saraiva wrote:
What separates KAJ from Bill Russell? (since I think he will be the other main candidate along LeBron and I've already discussed LBJ)

- Offensive impact. Sure Russell has the edge on D, but I don't feel it's gigantic. Clear one. On the other end of the floor, I think Russell doesn't even belong in the same page as KAJ. I can understand the defensive argument for Russell against players who don't come near him in defensive impact like Magic or Bird, but not against KAJ; (I see a good case for LeBron here)

This line of thinking is repeatedly used to the detriment of Russell. Are there any data to back it up? To me it seems more like an intuitive thought than anything else. As if we are saying, "there is no possible way that Russell's defense can be impactful enough to overcome the advantage that KAJ has over Russell's pitiful offense (by GOAT standards)". Which, to me, doesn't seem like a trustworthy foundation to form an opinion on when there is evidence to the contrary. And, let's not forget, Russell retired just a year before KAJ came into the league, and a 34 year old Russell was still leading the best defense in the NBA. The league Russell dominated was stronger than the drug laden, divided environment that KAJ dominated in the 1970s.

Look into the rankings of KAJ's teams on defense and on offense.

Look at Russell's teams ranks on offense and defense.

The Celtics had some pretty bad ORTG as a unit. KAJ's teams DRTG wasn't constantly really low.

I know this is about team stats and I don't like to use it a ton...

And some part of it is how I feel. That's why I use terms as I don't feel it's gigantic. That's how I feel about their defensive gap... I don't rely as much on statistical evidence for defensive purposes. From what I've observed I feel like KAJ is a definite plus on defense, and the only area he's not strong is 1vs1 (and he's not bad, I'd say above average). Given that even if Russell excels in all defennsive areas, I don't think his rim protection, ground covered, etc etc can have tremendous margins over what KAJ did. Clear ones, but not gigantic.

Idk if my post is written in a way that's easy to understand what I'm telling in the last part, if you wish let me know.

Jaivl · Post #70 » by **Jaivl** » Thu Jun 22, 2017 11:11 am

Some stuff on some of the favorites. I changed PPG and oPPG to ORtg and DRtg, but I was too lazy to look up the pace for the games missed and just went with the average pace for the season. So the results can be misleading between offense and defense (the net rating and the SRS should be ok).

Something I think I didn't note yet: I try to only use multiple-game samples in the "missed games" to, in a sense, "give time" to the team to adapt to the missing player.

Spoiler:

Vote: LeBron James
Second option: TBD / will edit

Why LeBron over Russell:

Spoiler:

LeBron "doesn't work well with other stars":

Spoiler:

eminence · Post #71 » by **eminence** » Thu Jun 22, 2017 11:27 am

Jaivl wrote:
Spoiler:
Some stuff on some of the favorites. I changed PPG and oPPG to ORtg and DRtg, but I was too lazy to look up the pace for the games missed and just went with the average pace for the season. So the results can be misleading between offense and defense (the net rating and the SRS should be ok):

[spoiler]
Duncan's 04 looks funny having the +/- data we have. A reminder that with/without can be imprecise.

LeBron looks superior to anybody else, as usual.

Rookie season Russell. Seems like Boston played a bit slower without him, so instead of 0.5/-0.5 the net ratings could be -0.5/-1.5... still, seems like Boston did prety well without him.

Seems great, but doesn't look like a juggernaut on defense, to be honest. Quite big sample size, too (n > 30).

Vote: LeBron James
Second option: TBD / will edit[/spoiler]

Oofta, that '04 vs '05 Duncan gap is ridiculous. Makes it tough to buy into the approach as too definitive. I think I would expect LeBron to be towards the top of a thing like this partailly due his playstyle and how teams are built around him (and him being really damn good of course).

Post #72 » by **penbeast0** » Thu Jun 22, 2017 12:13 pm

lorak wrote:... (and keep in mind, that Boston before 1957 basically every year was among the best offensive teams in the NBA, so it isn't like he had bad offensive supporting cast)....

Yes, but the best offensive team from 56 would be among the worst by 61. NBA offenses changed more than in any similar stretch in league history. Russell's defensive impact actually improved over that time, Cousy and Sharman didn't adapt to the changing league, they stayed 40% fg shooters (Sharman a little better, Cousy a little worse) . . . then add Cousy's ridiculously bad playoff shooting and it's hard to buy him as still an outstanding scorer.

Russell was never a very good offensive player, early on he was a finisher with good fg% (for the era), top 5 in the league 4 of his first 5 years, sort of Tyson Chandler and probably not the passer he developed into; but with terrible FT shooting in an era where that was a major key. He did have some nice playoff performances though. Later he was a bad shooting high post passing hub, still with the lousy FT shooting but his offensive game didn't develop much either and moving him to the high post hurt his efficiency as well.

Joao Saraiva · Post #73 » by **Joao Saraiva** » Thu Jun 22, 2017 12:14 pm

I'd like to see how the vote counting goes. I guess there is a 3 man battle with KAJ, LeBron and Bill Russell right now.

BasketballFan7 · Post #74 » by **BasketballFan7** » Thu Jun 22, 2017 1:22 pm

Joao Saraiva wrote:
BasketballFan7 wrote:
Joao Saraiva wrote:
What separates KAJ from Bill Russell? (since I think he will be the other main candidate along LeBron and I've already discussed LBJ)

- Offensive impact. Sure Russell has the edge on D, but I don't feel it's gigantic. Clear one. On the other end of the floor, I think Russell doesn't even belong in the same page as KAJ. I can understand the defensive argument for Russell against players who don't come near him in defensive impact like Magic or Bird, but not against KAJ; (I see a good case for LeBron here)

This line of thinking is repeatedly used to the detriment of Russell. Are there any data to back it up? To me it seems more like an intuitive thought than anything else. As if we are saying, "there is no possible way that Russell's defense can be impactful enough to overcome the advantage that KAJ has over Russell's pitiful offense (by GOAT standards)". Which, to me, doesn't seem like a trustworthy foundation to form an opinion on when there is evidence to the contrary. And, let's not forget, Russell retired just a year before KAJ came into the league, and a 34 year old Russell was still leading the best defense in the NBA. The league Russell dominated was stronger than the drug laden, divided environment that KAJ dominated in the 1970s.

Look into the rankings of KAJ's teams on defense and on offense.

Look at Russell's teams ranks on offense and defense.

The Celtics had some pretty bad ORTG as a unit. KAJ's teams DRTG wasn't constantly really low.

I know this is about team stats and I don't like to use it a ton...

And some part of it is how I feel. That's why I use terms as I don't feel it's gigantic. That's how I feel about their defensive gap... I don't rely as much on statistical evidence for defensive purposes. From what I've observed I feel like KAJ is a definite plus on defense, and the only area he's not strong is 1vs1 (and he's not bad, I'd say above average). Given that even if Russell excels in all defennsive areas, I don't think his rim protection, ground covered, etc etc can have tremendous margins over what KAJ did. Clear ones, but not gigantic.

Idk if my post is written in a way that's easy to understand what I'm telling in the last part, if you wish let me know.

RIght, but we can't simply look at the rankings in a vacuum (1st, 2nd, 3rd, etc.). The disparity between 1st and 2nd isn't consistent. Russell's "1st" ranked defenses had massive advantages over other defenses of the era.

drza · Post #75 » by **drza** » Thu Jun 22, 2017 2:22 pm

eminence wrote:
Jaivl wrote:
Spoiler:
Some stuff on some of the favorites. I changed PPG and oPPG to ORtg and DRtg, but I was too lazy to look up the pace for the games missed and just went with the average pace for the season. So the results can be misleading between offense and defense (the net rating and the SRS should be ok):

[spoiler]
Duncan's 04 looks funny having the +/- data we have. A reminder that with/without can be imprecise.

LeBron looks superior to anybody else, as usual.

Rookie season Russell. Seems like Boston played a bit slower without him, so instead of 0.5/-0.5 the net ratings could be -0.5/-1.5... still, seems like Boston did prety well without him.

Seems great, but doesn't look like a juggernaut on defense, to be honest. Quite big sample size, too (n > 30).

Vote: LeBron James
Second option: TBD / will edit[/spoiler]

Oofta, that '04 vs '05 Duncan gap is ridiculous. Makes it tough to buy into the approach as too definitive. I think I would expect LeBron to be towards the top of a thing like this partailly due his playstyle and how teams are built around him (and him being really damn good of course).

It's very interesting. Jaivi's 2004 data and ElGee's are almost diametrically opposed. And neither used all of the games that Duncan actually missed...ElGee sited 10 games, and a 3.2 SRS improvement with Duncan in the game. Jaivi sited 9 games, and a 1.0 SRS decrease with Duncan in the game. Duncan actually missed 13 games, and didn't start a 14th. I'm unsure why either didn't use all 13 games, and it seems clear that they didn't use the same games, so...that 2004 data point, for sure, seems in question.

The one positive is that, being in the databall era, we have full 5-man and +/- info for 2004 Duncan, so we don't have to rely upon WOWY to evaluate his impact that year.

kayess · Post #76 » by **kayess** » Thu Jun 22, 2017 3:20 pm

Jaivl wrote:Some stuff on some of the favorites. I changed PPG and oPPG to ORtg and DRtg, but I was too lazy to look up the pace for the games missed and just went with the average pace for the season. So the results can be misleading between offense and defense (the net rating and the SRS should be ok).

Something I think I didn't note yet: I try to only use multiple-game samples in the "missed games" to, in a sense, "give time" to the team to adapt to the missing player.

Spoiler:

Duncan's 04 looks funny having the +/- data we have. A reminder that with/without can be imprecise.

LeBron looks superior to anybody else, as usual.

Rookie season Russell. Seems like Boston played a bit slower without him, so instead of 0.5/-0.5 the net ratings could be -0.5/-1.5... still, seems like Boston did prety well without him.

Seems great, but doesn't look like a juggernaut on defense, to be honest. Quite big sample size, too (n > 30).

Vote: LeBron James
Second option: TBD / will edit

Why LeBron over Russell:
Spoiler:
It's not only the fact that at this point he has more longevity, or that his game is more between portable between different contexts/eras.

I think LeBron has superior impact, each on their own era.

"LeBron has never lead a team as good as xxxxx"
This year's Cavs topped as one the best offenses of all time in the playoffs. Remember how good they looked when LeBron sit? Yeah, not very much. That team had a -5 relative ORtg without LeBron on court (on the RS). And on the Playoffs... We've seen it! Like a week ago. Those stretches of bad play when LeBron was out. A FINALS GAME lost because LeBron had to rest for 2 minutes. And some people have the nerve of calling this a superteam.

LeBron offensive plus/minus +19.9
Irving/Love offensive plus/minus +2.1/+1.5 (and those are the second and third best players!)

His total +/- in the postseason tops +30, +17 on the regular season. That's on an offensive ATG team, +12 SRS on the playoffs. That's higher quality than any of Russell's teams. And when that load isn't supported on LeBron's shoulders, it falls to the ground, badly. The same happened in the last years of the Heat, by the way.

Russell took a team that without him was around a -2 defensively (1957) and took them to -4... and even to -10 GOAT defenses. That's massive: a team built mostly on his shoulders (I'm not gonna enter the narrative of "Russell was surrounded by HOF's!!" because I don't buy it: most of those are HOF's because of Russell). But... that's not enough when you compete against something that just made more with less.

LeBron "doesn't work well with other stars":
Spoiler:
If you have LeBron and Wade in the same team (similar roles), you play through LeBron (the better player) and Wade has to adapt to another role. Just like if you had Russell and Ben Wallace in the same team... yeah, Wallace better start learning to shoot 3s or something, because he is not gonna play center.

Love isn't worse than before: Love has changed his role. In fact he is better with LeBron, he just shoots less:

Code: Select all
min TSA PTS TSA/36 PTS/36 PPS Love with LeBron 6402 2811 3209 15.81 18.04 1.142 Love w/o LeBron 1726 968 1026 20.19 21.40 1.060

Boxscore numbers don't make you a better player.

Is this just pure WOWY? Or does the "with" sample account for just oncourt differential?

I'm sure you had mentioned this in another thread - but do you have a stats background?

colts18 · Post #77 » by **colts18** » Thu Jun 22, 2017 3:31 pm

lebron3-14-3 wrote:
RSCD3_ wrote:
Texas Chuck wrote:I'm casting my official vote for Mr. William Russell.

My primary reasoning is really simple--he had the single greatest impact any player has ever had in-era and he did it for a length of time that completely validates his candidacy for GOAT. This impact was most largely seen through his combination of defensive intelligence, ingenuity, and physical ability. He simply thought and played defensive basketball at a level never seen before and unlikely we will ever seen again. And I think its a near certainty that we will never see someone that much further ahead of anyone else in any element of basketball.

We could stop there and he's a GOAT candidate imo. But then we look at his rebounding which belongs in a discussion of the greatest rebounders of all-time and almost certainly the most valuable rebounder of all-time because of what happened next. He didn't just hold the ball and wait for the PG to make himself available. No he jump-started their high-pace offense through his own ability to handle the ball, his intelligent pushing of the ball through the pass. And of course he didn't just pass the ball and glide his way up the court like most big men--he was a weapon in transition.

He gets killed for his offense here for reasons I still don't comprehend. This was an era of inefficient scoring. Nobody was shooting the ball well. But Russell was leading or second on his team in FG% nearly every season, he was in the top five in FG% in the entire league 4 years in a row. He wasn't a poor scorer by the standards of his day. And scoring was well down the list of what he brought to the team offensively. This "horrible" offensive player was putting up 20 points and 5 assists in the playoffs on 4 separate occasions.

His rap as a defense only player is undeserved. And while its subjective, I feel like it matters too and that's his unprecedented leadership as a player and then as a player-coach.

I went into this feeling like 5 players deserved my serious consideration for #1. Mike was one of them and he is in at #1. Will edit back in my vote for my 2nd choice at #2 after reading some more thoughts. Surprised Lebron isn't getting much traction yet, but I'm listening hard to the Duncan and Kareem stuff and am leaning Timmy over Kareem right now, but not ruling out Lebron just yet either.

Compared to the list of 2014
1. Michael Jordan
2. Kareem Abdul-Jabbar
3. Bill Russell
4. Wilt Chamberlain
5. Tim Duncan
6. Shaquille O'Neal
7. LeBron James

It's crazy how these 3 years by LeBron may have only moved him up 1-2 spots

What do you mean? I think it's clear that he's going to be #3

Since the last project, LeBron has played 3 seasons where he was the best playoff player in each of those seasons. He averaged 30-10-8, 12 BPM, 28 PER in that span. He had 2 top 3 MVP finishes and likely a top 5 MVP this season. He's added basically the equivalent to 3 peak seasons since the last project. I can't see how Shaq, Duncan, or Wilt have an argument against LeBron anymore.

Post #78 » by **trex_8063** » Thu Jun 22, 2017 3:39 pm

lorak wrote:
trex_8063 wrote:

But otherwise, considering the sample size, you don't see this as kinda result oriented thinking?

You talk about sample size when I analyzed KAJ, but you didn't say a word, when Blackmill did pro KAJ case using cherry picked plays...

See the above post by Blackmill (that I copied here): not cherry-picked, but rather EVERY SINGLE defensive possession (whether Kareem or his man was involved or not) broken into categories.

lorak wrote:And except of that game (which, keep in mind, Blackmill called the most important, because of talent around him, KAJ could finally play the best defense he was capable of) I also provided full seasons defensive data, which is consistent with that game - in both cases KAJ looks much below ATG level on defense and great on offense.

I really would like to see explanation (other than with/without can't be trusted) of at least these numbers:

Code: Select all
1975, 17 G without with 99.3 ORTG, 97.8 DRTG w/o 92.8 ORTG, 98.4 DRTG net +7.4 (+6.5 ORTG, -0.8 DRTG)

Code: Select all
1978, 20 G without with 111.9 ppg, 107.9 opp ppg w/o 105.3 ppg, 106.7 opp ppg net +5.4 (+6.6 offense, +1.2 defense)

Why with/without "sees" how great KAJ was on offense, but at the same time doesn't see his ATG defense?

On/off takes a whole lot longer to gel than just about any other stat one could come up with, which is why playoff results (even within a finals run) often have some bizarre and counter-intuitive results. It's why the impact metrics based upon on/off splits (RAPM) have spawned prior-informed versions, because people were finding that even results based on entire seasons often still had a lot of noise.

It's still worthwhile looking at these----it's certainly more useful or relevant than drawing broad conclusions based on the recovery-rate of four blocked shots in a single game, for example----but I'd be cautious about taking them at face value.
btw, I like how you preemptively stated you'd disregard any "with/without can't be trusted" suggestion, I presume because you know it's a probable retort (not because it's a dodge, but because it's quite simply valid concern with this size of sample, as it's biggest proponent itt has even acknowledged).

And fwiw, I'd note some context, such as significant roster shake-ups in both of these years (especially in '78), and that in '78 Kermit Washington (excellent defensive role player big man) was around for all the games Kareem missed, but was absent for nearly all of the games Kareem played.

Anyway, I'm not going to argue the point further, as I've been in these debates with you before, and I know from past experience that once you're dug in I'd be wasting my breath, as it were.

lorak · Post #79 » by **lorak** » Thu Jun 22, 2017 3:48 pm

penbeast0 wrote:
lorak wrote:... (and keep in mind, that Boston before 1957 basically every year was among the best offensive teams in the NBA, so it isn't like he had bad offensive supporting cast)....

Yes, but the best offensive team from 56 would be among the worst by 61. NBA offenses changed more than in any similar stretch in league history. Russell's defensive impact actually improved over that time, Cousy and Sharman didn't adapt to the changing league, they stayed 40% fg shooters (Sharman a little better, Cousy a little worse) . . . then add Cousy's ridiculously bad playoff shooting and it's hard to buy him as still an outstanding scorer.

Look at numbers from 57 - Russel affected offense in a negative way, so your explanation doesn't look plausible. However I do agree, that NBA evolved offensively at the time, but it doesn't change the fact, that Russell was negative on O. And BTW, even at the end of his career Cousy had positive impact on offense as Celtics without him in 1964 regressed a lot (r_ortg) despite increased role of Hondo.

Jaivl · Post #80 » by **Jaivl** » Thu Jun 22, 2017 4:15 pm

kayess wrote:Is this just pure WOWY? Or does the "with" sample account for just oncourt differential?

Pure WOWY. Only way to fairly compare with earlier eras.

kayess wrote:I'm sure you had mentioned this in another thread - but do you have a stats background?

I'm working on it (applied statistics into civil engineering).

drza wrote:It's very interesting. Jaivi's 2004 data and ElGee's are almost diametrically opposed. And neither used all of the games that Duncan actually missed...ElGee sited 10 games, and a 3.2 SRS improvement with Duncan in the game. Jaivi sited 9 games, and a 1.0 SRS decrease with Duncan in the game. Duncan actually missed 13 games, and didn't start a 14th. I'm unsure why either didn't use all 13 games, and it seems clear that they didn't use the same games, so...that 2004 data point, for sure, seems in question.

I made the calcs again -used 04/11/2003 through 08/11/2003 (3 games) and 28/02/2004 through 09/03/2004 (6 games) and 12/03/2004 to 15/03/2004 (3 games)-. I missed 3 of the games before.

So I get the following:

Spoiler:

Deleting 2 of the games (08/11/2003 & 28/02/2004) I think I'm able to get the same results as ElGee. I think he controlled for starting lineups or something like that. When in doubt, I would use his data (mine is a lazier approach, in more than a way).

Spoiler: