RealGM 2017 Top 100 List: #3

[Tesla]

Some quick thoughts on KG this high.

I dont want to offend anyones opinion here, but I just dont see it, its borderline absurd. This is not a peak ranking (which even then I disagree with, but sure I can see a legit argument for) As a ranking of greatest of all time, at this high where there is a handful of resumes, his argument revolves around his +/- and value over replacement on ridiculous bad teams... so the arguement I suppose is if he had a great team around him we can infer from those numbers that he would have Tim Duncan-like sucess? OK maybe, but Tim Duncan ACTUALLY happened. Its insane to me to give that much credit on a career level to someone over others that actually lived it, did it, proved it, got the results and did it inn extraordinary fashion.
I get it, its unfortunate that he played a majority of his prime with some garbage teams... but that is what we are calling one of the handful greatest of all time careers? I use Duncan just because its a contemporary that he is clearly inferior to in terms of a GOAT ranking, but even if we look at their careers at 31 years old and onward (when KG got a great team), Duncan still outshines and does so clearly. /end rant.

[kayess]

That's the difference being debated. If someone believes that the Cavs actually do have more talent but playing "LeBron's way" limits them, then they could very easily argue that while still being very consistent to credit Garnett's impact-based accomplishments.

Yeah you are right.

I should clarify my comments to state that I'm not surprised by how much the Cavs struggle with Lebron off the court because I don't think Kyrie and Love are really good enough to anchor a team--which as I stated we've seen in their actual careers pre-Lebron.

I also disagree overall with the idea that Lebron should be prioritizing helping his teammates succeed versus helping his team succeed. And while some may wonder if there is truly a difference between those two things I clearly think there is. I also disagree with how much he is actually holding them back. Kyrie in particular is an iso scorer and a brilliant one, but adds nearly nothing defensively and isn't that great of a team offensive player. Not sure what Lebron's supposed to do for him. Love does get marginalized as an individual player because obviously he could do more than he does, but is the team better off putting the ball in Love's hands and playing Lebron off-ball more? No, clearly they aren't. So Love's stats aren't what they were in Minny but its absolutely the best approach for the Cavs as a team and that's what should matter.

And as I stated in a separate line of discussion, I agree KG lifts his team. But he can't do as much with poor rosters as Lebron can. So its frustrating to hear the same posters who continually praise KG for his lift to criticize Lebron who is actually doing more to benefit his teams because he doesn't play in the manner they believe to be best. Fine to suggest Lebron might could do more--but if what he is doing is already at a level like we've almost never seen, it feels unfair.

(1) The bolded is, 100%, what people can't seem to grasp. Some people argue in another direction and go "well yeah - but LeBron is so ball dominant that he will almost always marginalize stars, while Garnett won't" and then cite Bosh, Love as examples, completely forgetting that:

- Early Cavs LeBron didn't have a decent playmaker, so he had to take on the bulk of it
- When he DID have the opportunity to defer to a capable playmaker (Cavs with Mo somewhat, Heat with Wade, Cavs with Delly, Deron), he played off-ball. 2014, which is one of his two greatest seasons on O, had him cutting off-ball a ton, spotting-up, etc. Spamming PNRs with him and Delly led to amazing results (way better on-court ORTG than Kyrie+him), as well as with Deron - heck, he even let Kyrie take on more playmaking duties - but (and this is KEY!) would take those back when it was clear the offense was stagnating. This would suggest that this dominance is based on assessment of what helps the team the most, not because it's the way he wants to play (because God knows there are times when he just wants to stay 35 feet from the basket, completely uninvolved.

To re-cap: if we are testing the hypothesis that he plays this way because he wants to, we would have to see if he never changed the way he played. But he did - and it was always when there was a capable playmaker next to him, and even when there wasn't (like Kyrie sometimes - who still jacks up stupid shots when LeBron is open on a cut, something Delly would never miss). There's a proper control and an experimental sample here - am I missing something? So then, the new hypothesis is: he only did it because he had to [only verifiable by: asking him, to check intent, and developing a model that spits out what the best way to play is given a bunch of inputs, to check if it was really optimal both of which are pretty farfetched at the moment. But we can make best guesses, of course]

(2) Second, on Bosh and Love: I love this example because I know that basically anyone who uses +/- and on/off, skill-set evaluations as a major part of their analytical toolkit, but then proceeds to use Bosh and Love as an example... is clearly not applying their criteria evenly.

Bosh in Toronto was the alpha scorer, the classic 20-10 guy, but modernized, because he had a great jumper.
In Miami, he slowly morphed from that to... a ~17ppg 8 rpg guy, extended his range to 3pt land, highly mobile on the perimeter, blows up PNRs like no one's business, decent rim protection, and good enough in the post that only Roy "Hakeem OlaJabbar O'Chamberlain" Hibbert could punish the Heat for going small.

But yeah, of course he was marginalized because his box score stats went down! When in fact, Bosh was a much, much better player in Miami. And the credit goes 100% to him - he adapted his game around LeBron because his #1 skill at the time - scoring/playmaking - was something they already had in spades. The Bosh example doesn't show you LeBron marginalizes his teammates - it shows you that it's not easy playing with him if your skillset somewhat overlaps with his (applies to any star, but particularly to LeBron), but in the end, you maintain your impact, and the team does better - which is all that matters!

Love's a similar story - and Jaivl has already touched on this, but this one's way easier: they've put up some amazing ORTGs together, and while Love's lost weight and can't really bang inside anymore, his shooting and playmaking have been as utilized as ever (and he's freed up some effort on O so that he can expend more effort on D, on which end he's improved: shame that it wasn't enough against the one team that matters).

(3) Last point: The converse of someone like KG allowing other stars to shine because he shores up the D, is that, like Love and Bosh, they can focus on other things that are additive (shooting and defense) because LeBron can take up so much of the primary scoring/playmaking burden, without taking away from your ability to make plays as well (IF you can). We've seen this with other stars and situations too - Jordan on the Bulls (great D, and the guys had tons of offensive value just by their great OREB%), LeBron himself on the 2007 team (side note: did not realize his RAPM was +7 there. Can someone shed some light on this?? I can't believe this is accurate, tbh), whose calling card was defense and a not-terrible offense courtesy of LeBron, also some Dirk seasons/teams have this. I don't think one is more valuable than the other: it depends on the available talent pool, which are based on factors beyond a player's control, so I don't think punishing/rewarding players for this is fair.

[kayess]

Some quick thoughts on KG this high.

I dont want to offend anyones opinion here, but I just dont see it, its borderline absurd. This is not a peak ranking (which even then I disagree with, but sure I can see a legit argument for) As a ranking of greatest of all time, at this high where there is a handful of resumes, his argument revolves around his +/- and value over replacement on ridiculous bad teams... so the arguement I suppose is if he had a great team around him we can infer from those numbers that he would have Tim Duncan-like sucess? OK maybe, but Tim Duncan ACTUALLY happened. Its insane to me to give that much credit on a career level to someone over others that actually lived it, did it, proved it, got the results and did it inn extraordinary fashion.
I get it, its unfortunate that he played a majority of his prime with some garbage teams... but that is what we are calling one of the handful greatest of all time careers? I use Duncan just because its a contemporary that he is clearly inferior to in terms of a GOAT ranking, but even if we look at their careers at 31 years old and onward (when KG got a great team), Duncan still outshines and does so clearly. /end rant.

I get where you are coming from, and this was a big hurdle for me initially as well: it seems like we are crediting KG for what he COULD HAVE done, over what others ACTUALLY ACCOMPLISHED.

But then I realized two things:
- Team success is a function of not only the star player's impact/abilities, but also: his teammates', his coach's, the system/environment that they function in, the league-wide environment, etc.
- His impact on the court, how much better he made teams - these are things that actually happened. Things that he DID.

These have the following implications:
(1) This means that there are many, many potential reasons for a team not succeeding: bad teammates, bad coach, bad system... But even if you are already great, there might somebody still be greater. Most of these things are beyond the control of any single individual - therefore, why should we credit OR punish an individual for team success? Unless of course, we are able to look at team success, and somehow isolate how much impact an individual has on it...
(2) ...which of course, we are able to do (albeit, not perfectly). We know that the best predictor of success on the basketball court is margin of victory (MOV) - how much you outscore your opponents by. We also know that RAPM tries to isolate an individual's impact on his team's MOV, i.e., which player's presence on the court correlates most highly with his team's performance? By how much is he driving this?
(3) We can't stop there, though. Because it just means he is able to have an impact in that specific context. And we already know that this context (teammates, coach, environment, etc.) is difficult to control for an individual. So then we must ask ourselves: can he replicate this impact across different situations? This is where analyzing an individual's skill-set, and trying to make a best guess of (a) why he is successful, how he makes his teams successful, and (b) can he carry this across to different teams, comes in.
(4) Taking all previous 3 points together, it's easy to see that if an individual has great impact over his career, across different contexts... it's a real, tangible thing. The only thing he has control over, of the factors that affect a team's success.

So, to summarize: this isn't crediting KG for "could have been" results - it's realizing that for an individual, the results that are most indicative of ability aren't rings, it's on-court impact. And KG's influence, his impact on winning was as big as nearly anyone else's in history (Duncan is in the same boat), but Duncan won more simply because he was on better teams (their league context/environment is obviously identical).

[Jaivl]

LeBron himself on the 2007 team (side note: did not realize his RAPM was +7 there. Can someone shed some light on this?? I can't believe this is accurate, tbh)

It isn't, he has a +5.49 that year (by Engelmann's PI RAPM).

[SideshowBob]

(3) Last point: The converse of someone like KG allowing other stars to shine because he shores up the D, is that, like Love and Bosh, they can focus on other things that are additive (shooting and defense) because LeBron can take up so much of the primary scoring/playmaking burden, without taking away from your ability to make plays as well (IF you can). We've seen this with other stars and situations too - Jordan on the Bulls (great D, and the guys had tons of offensive value just by their great OREB%), LeBron himself on the 2007 team (side note: did not realize his RAPM was +7 there. Can someone shed some light on this?? I can't believe this is accurate, tbh), whose calling card was defense and a not-terrible offense courtesy of LeBron, also some Dirk seasons/teams have this. I don't think one is more valuable than the other: it depends on the available talent pool, which are based on factors beyond a player's control, so I don't think punishing/rewarding players for this is fair.

Few things - this is his first truly good defensive year (> +1 by my personal evaluation), this is his first year with a strong prior in 2006 (assuming PI RAPM is what you're talking about), and this is the year in which he debuted the 07-10 athleticism/body, so his finishing at the rim improved (meaning his offensive gravity improved).

Judging him in a vacuum - 07 regressed from 06 slightly on offense due to poorer shooting (likely due to the adjustment curve for the added mass), but got substantially better on defense (great defensive playoff run by him and the Cavs as well - Brown is hugely responsible for his defensive development, and I believe the 07-09 Cavs put up the best postseason DRTG in that 3-year stretch). Overall, 06 and 07 are the same level of player, but given what I described in the paragraph above, the situation was more favorable in 07 for him to have a higher RAPM score.

[SideshowBob]

LeBron himself on the 2007 team (side note: did not realize his RAPM was +7 there. Can someone shed some light on this?? I can't believe this is accurate, tbh)

It isn't, he has a +5.49 that year (by Engelmann's PI RAPM).

No +5.49 is Engelmann's multi-year no prior RAPM. His PI RAPM has 07 Lebron at +7.3.

EDIT: Fixed first link.

[SideshowBob]

Obligatory RAPM Guide:

[#1]APM is Simple OLS, exactly as Spaceman described it. Set up every 5on5 matchup, set equal to the scoring margin and solve for each player across the league (I've run it for a few years, its like ~65000 lines of 5on5 matchups). The resulting coefficients (on each player) are the APM values. This needs a very large sample size to say anything of considerable meaning; a single-year APM has large error terms on each coefficient, multi-year (usually 2-year) studies are preferred.

RAPM is essentially the same thing (OLS) with one exception. It introduces what we'll call a "reference matrix", basically each player is given a baseline value, towards which their coefficient will be pulled. I believe this tries to reduce the multicollinearity problem.

In [#2A]vanilla/basic RAPM, every value in the reference matrix is set to 0. The greater the amount of games played, the less weight that reference of 0 has. It is almost the same as APM [#1], but the regression towards 0 in theory reduces the error within a single-season set. It's still fairly volatile, but it's better than APM [#1] is within a single year. There is also [#2B]multi-year RAPM, which just uses a larger number of seasons, with most weight given to the current year and less and less weight given to previous years, reference matrix of 0s.

[#3]Prior-informed RAPM is essentially the best (ITO out-of-sample prediction) version of this family without introducing the box-score. It's built the same way as RAPM, but the reference matrix uses RAPM values from the previous year, instead of all players being set at 0. Again, as the sample size of the season grows, the reference value holds less and less weight. Obviously this only works once we have multiple years of data, in the 1st year, there is no prior. In the 2nd, we have a prior but it is vanilla RAPM [#2A], but by the 3rd year we can use PI RAPM of the previous year to inform the current year.

[#4A]RPM is RAPM, but the reference matrix is made up of SPM values (SPM is again, regression of box-score metrics on a multi-year non-box-score model such as RAPM). There is also [#4B]multi-year RPM, which is the same as multi-year RAPM, except it presumably uses a reference matrix of multi-year SPM values.

There is also [#5]prior informed RPM. Again, same idea as PI RAPM [#3], single-year, reference matrix of prior-year's RPM values.

[kayess]

(3) Last point: The converse of someone like KG allowing other stars to shine because he shores up the D, is that, like Love and Bosh, they can focus on other things that are additive (shooting and defense) because LeBron can take up so much of the primary scoring/playmaking burden, without taking away from your ability to make plays as well (IF you can). We've seen this with other stars and situations too - Jordan on the Bulls (great D, and the guys had tons of offensive value just by their great OREB%), LeBron himself on the 2007 team (side note: did not realize his RAPM was +7 there. Can someone shed some light on this?? I can't believe this is accurate, tbh), whose calling card was defense and a not-terrible offense courtesy of LeBron, also some Dirk seasons/teams have this. I don't think one is more valuable than the other: it depends on the available talent pool, which are based on factors beyond a player's control, so I don't think punishing/rewarding players for this is fair.

Few things - this is his first truly good defensive year (> +1 by my personal evaluation), this is his first year with a strong prior in 2006 (assuming PI RAPM is what you're talking about), and this is the year in which he debuted the 07-10 athleticism/body, so his finishing at the rim improved (meaning his offensive gravity improved).

Judging him in a vacuum - 07 regressed from 06 slightly on offense due to poorer shooting (likely due to the adjustment curve for the added mass), but got substantially better on defense (great defensive playoff run by him and the Cavs as well - Brown is hugely responsible for his defensive development, and I believe the 07-09 Cavs put up the best postseason DRTG in that 3-year stretch). Overall, 06 and 07 are the same level of player, but given what I described in the paragraph above, the situation was more favorable in 07 for him to have a higher RAPM score.

I don't remember much and I think I've been swept away by too many "that year is only remembered because of 1 great game!" comments in an attempt to be objective, but damn - I thought he started being good on D in 08 or something.

edit: 07-09 Cavs having the best DRTG is certainly proof that he can make it work with defensive minded players - I remember discussions previously where people were saying he could only make it work with shooters/good spacing, etc. +7 O on that '07 team... I've probably underrated that year a ton. Any good material on '07, SSB?

[kayess]

Obligatory RAPM Guide:

[#1]APM is Simple OLS, exactly as Spaceman described it. Set up every 5on5 matchup, set equal to the scoring margin and solve for each player across the league (I've run it for a few years, its like ~65000 lines of 5on5 matchups). The resulting coefficients (on each player) are the APM values. This needs a very large sample size to say anything of considerable meaning; a single-year APM has large error terms on each coefficient, multi-year (usually 2-year) studies are preferred.

RAPM is essentially the same thing (OLS) with one exception. It introduces what we'll call a "reference matrix", basically each player is given a baseline value, towards which their coefficient will be pulled. I believe this tries to reduce the multicollinearity problem.

In [#2A]vanilla/basic RAPM, every value in the reference matrix is set to 0. The greater the amount of games played, the less weight that reference of 0 has. It is almost the same as APM [#1], but the regression towards 0 in theory reduces the error within a single-season set. It's still fairly volatile, but it's better than APM [#1] is within a single year. There is also [#2B]multi-year RAPM, which just uses a larger number of seasons, with most weight given to the current year and less and less weight given to previous years, reference matrix of 0s.

[#3]Prior-informed RAPM is essentially the best (ITO out-of-sample prediction) version of this family without introducing the box-score. It's built the same way as RAPM, but the reference matrix uses RAPM values from the previous year, instead of all players being set at 0. Again, as the sample size of the season grows, the reference value holds less and less weight. Obviously this only works once we have multiple years of data, in the 1st year, there is no prior. In the 2nd, we have a prior but it is vanilla RAPM [#2A], but by the 3rd year we can use PI RAPM of the previous year to inform the current year.

[#4A]RPM is RAPM, but the reference matrix is made up of SPM values (SPM is again, regression of box-score metrics on a multi-year non-box-score model such as RAPM). There is also [#4B]multi-year RPM, which is the same as multi-year RAPM, except it presumably uses a reference matrix of multi-year SPM values.

There is also [#5]prior informed RPM. Again, same idea as PI RAPM [#3], single-year, reference matrix of prior-year's RPM values.

Thanks - I was just about to ask again because I'd forgotten how priors were set (I knew it was the previous year's RAPM, but forgot how many years, how far back before that prior is 0, what's the prior assumption for their rookie years, etc.)

Brought this up because I think at this level, while portability and all this matters, it essentially boils down to: do we think the impact scores for these people are valid? Does this reflect what our stats-blind (i.e., informed, unbiased eye-test) tells us about these players?

I feel like the only 2 questions against James now are: Is his longevity enough (for ~top 5: clearly yes, for GOAT? Not as clear cut, because it's somewhat dependent on question 2), and are his impact stats, for a lack of a better word, legit? His high stats are easier to accept than Garnett's because his impact is more evident in the box-score, but (trying to find nits to pick, playing devil's advocate here).

(1) Is it possible the priors might have overrated him in some way? s.t., when he finally got a good prior his impact numbers just snowballed from there?
(2) Are his tail-end prime numbers somewhat inflated by this effect as well?

Sure looking at NPI RAPM solves these problems somewhat - but it's too noisy over 1-year samples, etc. etc. Do we have multi-year no prior RAPM up until the present? I'd love to see how his scores evolved there, it's not on any of the sheets that people usually link to as repositories of RAPM.

[SideshowBob]

I don't remember much and I think I've been swept away by too many "that year is only remembered because of 1 great game!" comments in an attempt to be objective, but damn - I thought he started being good on D in 08 or something.

edit: 07-09 Cavs having the best DRTG is certainly proof that he can make it work with defensive minded players - I remember discussions previously where people were saying he could only make it work with shooters/good spacing, etc. +7 O on that '07 team... I've probably underrated that year a ton. Any good material on '07, SSB?

I've got limited writing on 07 so I'd have to dig a little. Problem is RealGM did a server transition at the turn of 07/08 so all content from prior to that is completely absent save for what users have personally archived so it's very difficult to pull older thoughtful write-ups that occurred in-season.

Elgee's the one who got me thinking on Lebron's 07 season so maybe he can chime-in with his thoughts.

[kayess]

I don't remember much and I think I've been swept away by too many "that year is only remembered because of 1 great game!" comments in an attempt to be objective, but damn - I thought he started being good on D in 08 or something.

edit: 07-09 Cavs having the best DRTG is certainly proof that he can make it work with defensive minded players - I remember discussions previously where people were saying he could only make it work with shooters/good spacing, etc. +7 O on that '07 team... I've probably underrated that year a ton. Any good material on '07, SSB?

I've got limited writing on 07 so I'd have to dig a little. Problem is RealGM did a server transition at the turn of 07/08 so all content from prior to that is completely absent save for what users have personally archived so it's very difficult to pull older thoughtful write-ups that occurred in-season.

Elgee's the one who got me thinking on Lebron's 07 season so maybe he can chime-in with his thoughts.

Link to this ElGee post? Was it in the top 100 project? I must've missed it.

Btw, I know you're not a voter, but thoughts on the discussion so far? How is your list looking like?

[trex_8063]

So.......it seems my question regarding Lebron's level of responsibility in how his teams perform without him has damn near derailed this thread. It was a question I've been pondering, particularly because I think it relates directly to how scaleable his impact is, and potential ceiling for his teams. I wanted some thoughts on it since he and Duncan are my top two candidates here.
We have quite a few thoughtful answers on all sides of the table.

So Lebron has perhaps had the FO backing to essentially choose his own roster (within certain reality constraints, of course), and has constructed a cast and style of play that is reliant on his talents (thus maximizing his impact). So from that standpoint, it's "his fault" that the Cavs are what they are.

RCM88x made the point that it's sort of disappointing that Kyrie and Love (two star-level players in their own right) haven't developed some more synergy between the two of them to keep the boat afloat. Part of that may be the frequent intermittent injuries, but he appears to be suggesting their failure to do so isn't Lebron's fault.

otoh, Doctor MJ and Outside have [probably correctly, to some degree] speculated that they're trained (in practice, and game-to-game) how to play alongside Lebron, and thus haven't sufficient time/practice in learning how to co-exist without him.
But then someone else (forget who) implied maybe a pinch of blame should be laid at Lue's doorstep, in that he's not a brilliant enough coach to come up with some manner of offensive scheme that can work for the [NOT insubstantial] offensive talent that they have OUTSIDE OF Lebron. Instead, the mostly just let Kyrie do his iso thing (which the circa-2000 NBA has taught us that you basically CANNOT run a "good" offense----by today's standards----playing iso ball). It's very unsophisticated, and rather like watching a pick-up game at the gym (except the players are REALLY good).


So what have I concluded wrt how much of the blame is Lebron's? idk; to put not too fine a point on it, I would say "some" of it rests with him, but not all.


However, if we are to say that it's "Lebron's fault" that the Cavs dynamic is playing out in this way, Joao Saraiva made [albeit somewhat sarcastically] a really good point that all the success this cast has had is likewise "Lebron's fault".
I mean, his recipe worked, no? Three consecutive finals appearances----(we can talk about the relatively weak strength of the EC, but let's be honest: the EC could have been substantially stronger, and the Cavs were still the very likely team to come out on top; they've pretty well steam-rolled the rest of the conference in each of the last three seasons, and were pretty much a foregone conclusion to be in the finals before each season even started)----winning one, which had to ripped from the grasp of a 73-win team. And who knows?: Maybe could have won TWO of them if both Love and Kyrie hadn't been injured in '15 (somewhat remarkable to take two games of the Warriors---losing game 1 by inches in overtime, too---given the circumstance).


And wrt his apparent impact on these casts.....
It's been implied that his impact is shall we say accentuated again, because his team is constructed to maximize what he brings to the table, and also rendering the cast impotent when he sits.
To be fair, though, we've seen him do similar (impact-wise) with more impromptu (and frequently poor) casts (see '08-'10 in first Cleveland stint), as well as decent casts that he didn't have as much of a hand in constructing (see '12-'13).


So overall, I've not been swayed from him as my top pick, though I don't feel as strong about it as I did at the start of the project (mostly because people have been making some excellent cases for Duncan). I'm gonna stick with Lebron for now, though, as he appears to have more traction anyway and it still feels like my "gut pick".

1st pick: Lebron James
2nd pick: Tim Duncan

[Outside]

Question to any/all:

Does Lebron, in your opinion, bear any of the blame for the fact that his team (even a presumably "good" cast) falls off a cliff any time he's not on the court? And if so, why?

I mean some of those old casts in Cleveland 1.0 it's obvious: those were just crummy casts.

But in Cleveland 2.0, where he has K.Love and Kyrie, and reasonable depth (though lacking in interior presence, especially defensively), they still seem to utterly drown without him.
Is it somehow [even partially, perhaps?] his fault? Or is this squarely on the the supporting cast? I mean, these are grown men, they're professionals (and almost exclusively veterans, too), should they be considered responsible for themselves?

Thoughts on this?

Sorry for derailing the thread. The LeBron discussion detracts from the main point of the thread, which is debating who should be number 3 on the list. Lesson learned on my part.

[trex_8063]

LeBron himself on the 2007 team (side note: did not realize his RAPM was +7 there. Can someone shed some light on this?? I can't believe this is accurate, tbh)

It isn't, he has a +5.49 that year (by Engelmann's PI RAPM).

Can you provide link?

I thought this was J.E.'s data (it shows +7.3 RAPM for Lebron that year).

[trex_8063]

Question to any/all:

Does Lebron, in your opinion, bear any of the blame for the fact that his team (even a presumably "good" cast) falls off a cliff any time he's not on the court? And if so, why?

I mean some of those old casts in Cleveland 1.0 it's obvious: those were just crummy casts.

But in Cleveland 2.0, where he has K.Love and Kyrie, and reasonable depth (though lacking in interior presence, especially defensively), they still seem to utterly drown without him.
Is it somehow [even partially, perhaps?] his fault? Or is this squarely on the the supporting cast? I mean, these are grown men, they're professionals (and almost exclusively veterans, too), should they be considered responsible for themselves?

Thoughts on this?

Sorry for derailing the thread. The LeBron discussion detracts from the main point of the thread, which is debating who should be number 3 on the list. Lesson learned on my part.

Don't be sorry, you didn't do anything wrong. You answered the question I put forth because it was something I was having difficulty deciding how relevant it was. I just didn't expect it to take over the thread as it did (though the past few months on this forum should have taught me that it would); but the discussion was still useful in exploring this aspect, and it also sprouted some tangent topics (wrt Garnett and such). So it's all good.

[Dr Positivity]

Why I don't trust injury in/out numbers

- A lot of variables beyond the star. Teams are pacing themselves during the regular season (which is why they have "another gear" in the playoffs), when a star goes out they can start playing harder to make up for it. Teams have different make-ups emotionally that will make them respond to adversity differently. Strategically there is an adjustment period where the other team's scouting plan was against the team's star, and they don't have as much time to prepare for the team without the star. In the middle of the season there's barely any practice time to get used to playing without the star.

eg. 2012 Rose gets injured and the Bulls play at a mid 60s Ws pace without him... I don't think there's a world where that Bulls team with Deng and Noah as its best players is mid 60s W caliber. They didn't bring everyone back but the next 2 seasons were 45 and 48 W. I prefer to chalk it up to weird variables such as the above and not "Rose has no impact"

- General SSS problem. Adjusted plus minus exists cause it's accepted that raw +/- isn't that trustworthy. So while it's not apples to apples, I have a hard time believing that 15-20 games or whatever of raw +/- data when a player gets injured, can be trusted as key information for a player.

[trex_8063]

Thru post #95 (if you've voted and don't see your name, please let me know; it's sometimes hard to see them all within large blocks of text, as I don't always have time to read EVERY post in it's entirety--->this is where bolding you picks helps me out):

Bill Russell - 8 (Texas Chuck, TrueLAfan, scabbarista, penbeast0, Outside, JordansBulls, Doctor MJ, BasketballFan7)
Lebron James - 7 (Bad Gatorade, Jaivl, Joao Saraiva, LABird, Narigo, Tesla, trex_8063)
Tim Duncan - 2 (kayess, Cyrusman12200)

Thread will be open a little less than 24 more hours. If you plan on voting/contributing, please do so within that time-frame. Few guys have been active but not yet cast ballots that I have seen (micahclay, drza).

@ Narigo - Please get at least some brief arguments next to your top pick before the deadline, or I may not be able to count it.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

[Jaivl]

It isn't, he has a +5.49 that year (by Engelmann's PI RAPM).

Can you provide link?

I thought this was J.E.'s data (it shows +7.3 RAPM for Lebron that year).

Yeah, you're right, it's not PI RAPM, it's multi-year NPI RAPM (https://drive.google.com/drive/u/0/folders/0B_AdaCB40YpgdU00TUYtYk42dXM).

I recall Engelmann saying that it was about the same, though.

EDIT: Multiyear RAPM with lower weight to older seasons (those who requested PI RAPM should use this instead)

[Dr Positivity]

Not voting for Shaq, Wilt cause of mental reasons, Duncan or KG cause Hakeem had more dynamic/clutch peak, Bird or Magic cause Lebron is rated over them (Defense in Magic's case, not being 84-88 level long enough in Bird's case)

Leaves LBJ, Hakeem and Russell

Russell: Incredible defensive impact, I see his impact warping the court as the defense version of Curry. So you have that court changing impact along with his literal blocks, all time rebounding and helping offensively by passing and what he can do by scoring. Still the question here is how big is Russell's defensive impact over Hakeem's? Is it enough to make up for significant offensive advantage? Put it this way Curry's ORPM (rough estimate for Russell defense) + Draymond's ORPM (rough estimate for Russell offense) is easily below say the combination of Gobert DRPM + Cousins or Towns ORPM, and Hakeem defense and Hakeem offense is arguably better than either of those players. I still think having elite impact on both ends gives players the highest ceiling.

That leaves Lebron vs Hakeem. Offensively, Lebron is better in both numbers and style of play, while Hakeem clearly by nature of being an all time great defensive C gets the edge on both ends. Both have sky high ceilings in terms of impact of big game playoff performances, they straight up stole at least one title (Lebron 17, Hakeem 95) by being clutch. I have a hard time deciding so I'm going to go with Lebron because his personality the first half of his career was more stable

Vote: Lebron James

2nd: Hakeem Olajuwon

[SideshowBob]

Thanks - I was just about to ask again because I'd forgotten how priors were set (I knew it was the previous year's RAPM, but forgot how many years, how far back before that prior is 0, what's the prior assumption for their rookie years, etc.)

JE's set goes back to 01, so that's the first year with 0 priors. After that, every year n is informed by the score from n-1.

Brought this up because I think at this level, while portability and all this matters, it essentially boils down to: do we think the impact scores for these people are valid? Does this reflect what our stats-blind (i.e., informed, unbiased eye-test) tells us about these players?

I feel like the only 2 questions against James now are: Is his longevity enough (for ~top 5: clearly yes, for GOAT? Not as clear cut, because it's somewhat dependent on question 2), and are his impact stats, for a lack of a better word, legit? His high stats are easier to accept than Garnett's because his impact is more evident in the box-score, but (trying to find nits to pick, playing devil's advocate here).

(1) Is it possible the priors might have overrated him in some way? s.t., when he finally got a good prior his impact numbers just snowballed from there?
(2) Are his tail-end prime numbers somewhat inflated by this effect as well?

I'm mobile so can't get into too much detail know, but healthy skepticism is always a good thing, as long as it is consistently and legitimately applied. It's possible that RAPM is overrating him in some way, sure - but we need to either consider the same for others or dive into why this would more likely be the case for him and not others.

Personally I don't put much more conclusive stock in the regression based data than the box-score data - they're all supplementary tools at this point. RAPM is most useful in giving an impact footprint IMO.

Sure looking at NPI RAPM solves these problems somewhat - but it's too noisy over 1-year samples, etc. etc. Do we have multi-year no prior RAPM up until the present? I'd love to see how his scores evolved there, it's not on any of the sheets that people usually link to as repositories of RAPM.

We have consistent single and multi-year RAPM from JE for 01-16 (will also post these later when I'm in front of a computer). Just tweeted him today requesting both for 2017.

Link to this ElGee post? Was it in the top 100 project? I must've missed it.

Btw, I know you're not a voter, but thoughts on the discussion so far? How is your list looking like?

2012 Peaks Project I think, I'll find it.

Haven't honestly messed with my list since 2014 - its mostly just been adding on the current year seasons. Project's been fun (I've followed the whole thing) but guys are just slipping in IMO - not much debate yet (that'll change when Kobe and KG discussions gain more presence).