Peaks project update: #13

[liamliam1234]

I also edited. But broadening on that point, most acknowledge that the one Minnesota year he had a decent teammate, he made the conference finals before losing after that teammate was injured.

Third guy is a joke of a second option, lol, and again factors into my issues with RAPM.

[freethedevil]

Edit: you messed up the quoting again; will respond to the other stuff as I edit.

My bad.

Anyway, I'll address your points without quoting, feel free to ding me if i miss anything.

YOu're right, from the graph, lebron's 11 circles align with the blurb saying lebron had 11+4 3 year stretches, so I'm not really sure what to make of the purple circle. However the explanation for the 2015-2017 stretch beinf rated low is explicit here:

Curry has played two of the last three postseasons on the heels of a knee injury, and that’s likely knocked his impact down a peg compared to his transcendent regular seasons.

This would be referring to 2018 and 2016.It would explain why his 2015-2017 stetch is rated lower than his 13-15 stretch. Regardless, his 13-15 stretch is still rated higher than any of kobe's. To nail single season impact, we can look at this:

This based nearl exclusively on postseason valuation and curry is the only one of the 4 players discussed with an all time season. Here's what the other seasons look like:
(wade)

(kobe)


The gap is huge. And it's largely based on imapct #'s. You can look at any impact stat and it will tell you the same thing. Additonally, the fact of the matter is for there to be a purple dot, that means curry posted a season higher than +6 as it would have to average out sub 6 seasons. Given that the author cites injuries from 2016 and 2018 and he states his 2013-2015 stretch as sub 6, there's really only one possible season here, and it's probably the going to be the one which the author has as way higher than curry' other seasons.


Interestingly enough, the author has david robinson with two all time seasons, however neither are quite at curry's 2017:



All this is to say, the link you used logically indicates curry's 2017 season had the highest impact. That stat aside lets address the other arguments.

So was Draymond’s three-year playoff peak better than Curry’s average from 2013-15?

Yes, per the stat. However it wasn't on par with the healthy postseason we've seen from peak curry. Regardless draymond's peak is listed way lower by the author, and he explains why here:

Watch on YouTube

Draymond green's impact #'s rate him as a top ten player in the postseason barring 2019, so him being portable would indicate those #"s underrate him. However, if you reject the notion that a defender who doesn't protect the rim as well as his tippity top contemporaries and his limited scoring makes him dependent on the warriors relatively unqiue switch everything and small ball scheme, it doesn't really say anything regarding curry, because again, Stats like rpm, which reduce players score depending on how good their teammates are and measure impact(how much better a team is) rather than effiency punish players on good teams and reward players on bad teams. It's why RPM had harden and paul george leading the league while luck adjusted metrics had giannis well ahead.

And yet Curry, Durant, and Draymond all rate highly together.

Right, not crazy when we consider currys' team had the goat postseason. So while logic, and the #'s would reflect, curry isn't the goat, they do, given that curry was the most impactful player on said team, still paint him as having a much better peak than kobe and wade's whose teams were nowhere near as good. If you're going to disregard, that, AND his portability, you shoudl have a better rebuttal than, curry's ppg and ts% dropped in seperate postseasons where he coming off injuries, or that curry having great teammates renders #'s which punish players on strong teams invalid.

I am saying to be consistent you should be arguing for Garnett like that.

And I do? I have garnett's peak higher than I have curry's.

Why is his impact so high? “Forwards need to score”? Why? Yeah, usually defensive anchors are traditional centres; why is Draymond’s impact so much higher? Why is his impact better despite limited scoring?

His passing isn't harmed much by his lack of scoring(which opens up reads) because of the warriors uniquely fast paced small ball and unrivaled abundance of player movement. His pnr coverage was nearly as valuable as rim protection(until 2019) because the warriors employed what was a rare defensive scheme with 5 switchable defenders, including curry. Regardless, draymond's impact isn't an argument against curry's so this really is a moot point.

Why was his impact higher? (Alright, Embiid was better this postseason; are you arguing him as the best in the game now because he led the postseason in RAPM?) Has Curry fallen off? Without Durant, and sometimes Klay, why did his impact metrics not explode the way you keep saying is true of solo stars with great metrics?

Uh, they did?:
https://www.bball-index.com/18-pipm/
After a slow start int he first two rounds, he nearly bridged a massive rs gap with giannis in the last two rounds only falling .4 short.

Per Ben's corp he was actually #1(slightly edging giannis) even with health concerns from his finger.

As for embid:
https://youtu.be/93zief7g2Jk?t=411

A. he played less minuites
B. HIs impact stats are nowhere close to #1.

Regardless, those stats have been saying the same thing about curry without durant as they did about curry with durant. There was no drop off.

Also, I brought up RPM, not rapm. RAPM gets useful with luck adjustment.




https://www.bball-index.com/18-pipm/

Not sure I agree but not especially relevant, disagree, disagree (he is more efficient but also has not held up in handling a heavy scoring load nearly as well as Wade/Kobe), agree but possibly offset by mediocre defence.

Curry has consistently across the board ranked as a slight postive on defense, just like kobe. Curry lowers players who play his position effiency more than most, fights screens, switches well, rebounds well and knows how to force bigs into help. His team's defenses have consistently been better without. That's kind of silly.
As for volume:

Significantly higher effiency on nearly as much volume as kobe. And scoreval has curry's 2017 scoring as worth more bpm than either wade.

Curry is the same, except without the passing, defence, and physical dominance (all of which Wade and Kobe have over him).

"Physical dominance" is already baked in to skillset, so it's silly to value that seperately. That aside, Passing is merely an aspect of playmaking and really it's only useful as so much as you're making your teammates better. Given that curry's effect on his offenses effiency is much higher than wade or kobe, this is not much more than a nitpick.

And fyi, we have proof this isn't teammate depedent in the nba finals. With kaly as his only option, curry's warriors posted a 110 rating against a defense that lowered it's opponents offence by 9 points(that's' second only to the 2004 pistons). Without klay and having basically bench players, they posted a rating of 107. Curry is +12 with only bench players.

Curry is great with the warriors but worse in a vacuum than other atg's is absolutely ludicrous. Kobe and Wade are more system dependent than curry, so if that's your argument here, it's not a very good one.

[liamliam1234]

However the explanation for the 2015-2017 stretch beinf rated low is explicit here:

Curry has played two of the last three postseasons on the heels of a knee injury, and that’s likely knocked his impact down a peg compared to his transcendent regular seasons.

This would be referring to 2018 and 2016.It would explain why his 2015-2017 stetch is rated lower than his 13-15 stretch. Regardless, his 13-15 stretch is still rated higher than any of Kobe’s.

And would you take 2013-15 Curry over 2008-10 Kobe?

For Curry, basically we can determine that 2013 > 2016 and 2013-14 > 2016-17, which to me suggests that either 2017 is overrated, 2014 is grossly underrated, or 2016 is downright gross (and while this is not your argument, it should bar 2016 from consideration for other voters).

To nail single season impact, we can look at this:

This based nearl exclusively on postseason valuation and curry is the only one of the 4 players discussed with an all time season. Here's what the other seasons look like:
(wade)

(kobe)


The gap is huge. And it's largely based on imapct #'s. You can look at any impact stat and it will tell you the same thing. Additonally, the fact of the matter is for there to be a purple dot, that means curry posted a season higher than +6 as it would have to average out sub 6 seasons. Given that the author cites injuries from 2016 and 2018 and he states his 2013-2015 stretch as sub 6, there's really only one possible season here, and it's probably the going to be the one which the author has as way higher than curry' other seasons.

Last part, fair enough, but remember that I was not arguing against 2017 being his best season. First part, are those Kobe and Wade graphs really from the postseason? Both graphs feature years where they missed the postseason...

Interestingly enough, the author has david robinson with two all time seasons, however neither are quite at curry's 2017:

My point was more to think about what this would suggest for prime Robinson’s impact.

This next bit about Draymond is kind-of a mess. Switchability is the modern playoff trend. The Jazz are the top defence resisting that trend, and plenty of people think their playoff “struggles” are because of that inflexibility (for me, I am waiting to see what happens now that they have an actual offence). But even then, players who can shift are easier to find than Rudy Gobert.

After that, you confuse the issue by criticising RPM and then saying it is better than RAPM. Maybe that was a typo. But I do not feel you have relied on PIPM much until right now, so it feels a bit slimy, or at least not especially consistent. If you think PIPM is far and away the best impact metric, just cite that rather than vaguely referring to impact metrics, because for a lot of people here that primarily means RAPM.

I am not done editing; going to look through PIPM a bit.

[Timmyyy]

Curry's season lost value for only making it to the finals but for some reason KG not even getting that far gets a pass because he played "great" whatever that means. And despite this ambiguous at best state of affairs in the post-season we should just ignore the regular season. So to sum up Curry advanced further in the post season and Curry had a better regular season despite it being KG's only MVP season but it was KG's year. I count at least two logical inconsistencies there. But great we have confidence in KG being superior anyway because he was able to have another MVP year right? Uh no. More titles? Nope. More finals appearances? No again. Well...at least KG has longevity.

Yeah exact, CURRY'S value is his TEAM'S result , we ignore that he played significantly worse in the PO's because he is Stephen Curry (you won't understand it I guess).
You should not ignore RS, you should evaluate both RS and PO, but you don't, you only look at RS for 16.
Curry advanced further because of his team not because of him alone.
Where do you get that Curry had the better RS? KG is better according to every RAPM number there is. KG literally had to anchor both defense and offense and still had the higher impact. I don't care about MVP's. KG was MVP level multiple times (2003 though Duncan deserved it and 2008).
What inconsistencies? KG was better in the RS and in the PO's for me resulting in a higher peak. Sorry that I do not look at the team success only to evaluate single players, but inconsistent? Nah.
MVP? Narrative. Titles and Finals? Team accomplishments.

Pffft. As if Magic and Bird didn't have the greatest teams you could ask for. You also seem to miss that Curry can claim not only a higher regular season peak but also a higher playoffs peak than Magic and Bird.

http://bkref.com/pi/shareit/EQ3Qy

Curry is currently in the top two for playoffs TS% among all top scorers in NBA history. He is also currently in the top ten all-time for playoffs ppg in a career as well as in finals ppg in a career. That's higher than Kobe, Nowitzki, Kareem, Malone, Shaq, Dr.J, etc. as well as Magic and Bird.

Yeah ok, since it is your favorite player we are allowed to mix 17 and 16 in a peaks project, alright.

Then Magic, Bird, Russell and maybe even Wilt shouldn't be ahead of him.

You are only black and white, aren't you? I want to extract the impact a guy has on his team. Curry had a great one, but it is not 'he had the best team = he is the best player' like you want it to be. The exact same standard I apply for the other guys.

Basically you're saying Curry was lucky that a team fell into his lap that was better than all the other teams in history including teams that were planned and constructed with the goal of being a superteam. Such luck!

Nope. I said he was fortunate with the situation he had and how certain things played out, but I have the biggest respect for him taking advantage of the situation. Opposed to you who just want to give all the credit to Curry (hey, Dray developed into a monster player! Let's credit Curry for it since Dray was only a 2nd round pick ).

In raw +/- Curry seems to be superior.
https://pbs.twimg.com/media/D7wepu0X4AAJfsz.jpg
https://www.wsj.com/articles/stephen-curry-changed-the-game-by-changing-every-game-11559140592

Ok so we rather take the raw data, since the superior adjusted data doesn't show your favorite player in front.

But as noted some of those other guys did have the luxury even longer than Curry and others eventually did get the luxury and still weren't able to produce more or better.

Teams aren't grouped in good or bad. It is a continuum and Curry had the luck to be on the absolute elite side of things. The others may have had good teams at some point but 1-12 the Warriors from 15-19 might be the best team ever. Curry was part of it, not all of it (for the 1000's time).

I am responding to clear logical inconsistencies. I'm providing statistics and facts in response to general feelings. Someone above referred to how a "proper appreciation of KG's defense" should quickly dismiss any Curry comparisons for example.
I would note a proper appreciation of KG's defense saw him defeated in short order while a proper appreciation of Curry's offense on the other hand has required the development of new defensive tactics that have seen big men like KG looking less and less important. I will respond to contempt in equal measure.

There are no logical inconsistencies no matter how much you would like to see them. I see things different, I don't judge players based on how their teams do, I apply context look at +/-, on/off, WOWY , on court ratings, team ratings, analyze with my eye.
Sometimes my results don't align with the PURE TEAM RESULTS, you are using. So it's no wonder it is inconsistent according to your criteria since I apply a completely different criteria.
You provided team statistics and facts. You would make a great argument for the Warriors being the best team of all time, not for Curry being the best.
I did mostly use my general feeling because I didn't argue my own points and I just wanted you to be a bit more open minded for the other side. But if you want numbers look how KG is way better in the RAPM category than Curry while having a clearly bigger load to handle being the best offensive and defensive player for his team. But hey he didn't even reach the finals right?

I want to let you know that I will drop this. I had such discussions multiple times. Guys with obvious screen names, not willing to give an inch even if it is sooo obvious that their player has flaw x or y, arguing with narratives, MVP's, rings and whatever.

I gave you a lot of thought processes why someone could see Curry lower than you see him, I gave you data that supports KG over Curry. I tried to give you a POV outside of team results and I tried to show you that I am not here to talk down Curry, that there isn't only black and white regarding the team results and who we are supposed to give credit.

All I took away from you is that you repeated the same points 3 times.

Such a discussion isn't giving me anything.

This is all fine. But how do any of these arguments hold against 2017 curry.

The starting point of the debate was specifically about 16. I wasn't discussing 17 Curry at all since I am very high on that season myself. The discussion just drifted to a general success debate sometimes, which I actually wasn't really interested in.

And on a general note, I hope at some point people do understand that...

...RAPM is not calculated with a formular

....and that this gitlab thing is the worst possible source for RAPM. We have data from the inventor JE himself, instead we use something derived from a paper of a guy from a university. Just because it is the easiest to find doesn't mean we should use it.

When you understand the first point it should also be easy to be skeptical about gitlab since it provides single year PO only RAPM, which is total nonsense when you know what RAPM really is.

[liamliam1234]

Yeah, it set off a few alarm bells, which is why I wanted to warn people about taking it as a definitive data set, but it also does not too radically deviate from better calculations.

This is a good write-up which goes into many of the issues. (https://squared2020.com/2018/12/24/regularized-adjusted-plus-minus-part-iii-what-had-really-happened-was/

I looked at PIPM, and it basically seems to be a glorified BPM, complete with a fair number of the same flaws. Again, I support using “advanced metrics”, but we need to be careful about building entire cases on them the way people are doing for Curry (sorry, cases beyond pointing to him being on three 67-win teams ). And I think it provides an interest contrast with the Garnett discussion where the reliance on “metrics” at least was always complemented by acknowledgements of what Garnett could do compared to Duncan, as well as an understanding that defensive impact is not especially well captured by the analytics which are available to us, and of course an appreciation that he was screwed in terms of public perception by his poor team quality (speaking of which, I love how with Curry we have flipped to script to say, “Actually, playing with the best possible team is a disadvantage for Curry compared to everyone else” ).

[freethedevil]

Yeah, I'm gonna need you to point out the issues specifically before I take your word for it. Calling it "glorified bpm" tells me you don't understand how luck adjustment works.

I love how with Curry we have flipped to script to say, actually, playing with the best possible team is a disadvantage for Curry compared to everyone else ).

I love how you literally just waved away paragraphs of explanation on a strawman. That's oddin tier.

No one said having better teammates makes it harder to win. Your garnett example falls apart because I never argued curry was better because he won.( i didn't even argue he was better). You're trying to claim, baselessly, that being on a good team inherently improves your impact. Reminder impact is x+ 30 = y. The twist is that the higher x gets, the more it takes it to lift it. So if you're getting +30 on a good team, and westbrook is getting +30 on a mediocre team, impact stats that don't use luck adjustment(like pipm) aren't going to factor that.

Luck adjustment looks and adjust for what an induvidual player can control, so, even if their 30 becomes a 22 because their team does 8 points better, they're not going to penalize you for your team doing 8 points better.



You and eballa repeating "teammates make thing easier" ad nauseum is literally the equivalent of odinn championing his "eff" stat. It being easier to win by 60 games means you would have to impact your team by less to win 60 games. So If you imapct your team by as much, that means you've won more than 60 games.

"It being harder to impact(not win)" is literally just common sense: it is harder to improve a 50 win team by 7 than a 30 win team by 7.

The stats i cited for garnett were luck adjusted. The stat i've been repeating for curry's 2017, rpm, is not. This is why your equivalency sucks.

TLDR: Imapct is derived from winning by, essentially, subtracting your teammates play from your team's play. It is not = to winning. So your was uncalled for.

[liamliam1234]

My phone browser crashed and lost my post.

Fine, I will do this slowly.

First, finishing up the previous post I said I would resume.

Right, not crazy when we consider currys' team had the goat postseason.

Again, was 2001 Kobe's peak?

If you're going to disregard, that, AND his portability, you shoudl have a better rebuttal than, curry's ppg and ts% dropped in seperate postseasons where he coming off injuries, or that curry having great teammates renders #'s which punish players on strong teams invalid.

It certainly is convenient how we can dismiss the bad playoffs for Curry such that we leave as a sample the single playoffs where his team was operating at its absolute peak and he faced zero adversity at all.

Regardless, draymond's impact isn't an argument against curry's so this really is a moot point.

It is when I am trying to call into question inconsistent reliance on "impact".

Uh, they did?:
https://www.bball-index.com/18-pipm/
After a slow start int he first two rounds, he nearly bridged a massive rs gap with giannis in the last two rounds only falling .4 short.

If that increases based on playoff value, how good was Nurkic if he ranks in the top ten without being in the playoffs? My guess is that the bigger reason was that Giannis dipped.

Per Ben's corp he was actually #1(slightly edging giannis) even with health concerns from his finger.

Cite the full CORP numbers if you are going to use them.

As for embid:
https://youtu.be/93zief7g2Jk?t=411

A. he played less minuites
B. HIs impact stats are nowhere close to #1.

What impact metrics are you checking? This is what I mean when I say you need to be clear, because I have seen plenty of impact metrics marking Embiid as possibly the league's most impactful force.

"Physical dominance" is already baked in to skillset, so it's silly to value that seperately.

Considering how we have watched Curry struggle with physical defence more than most, it obviously is not.

And fyi, we have proof this isn't teammate depedent in the nba finals. With kaly as his only option, curry's warriors posted a 110 rating against a defense that lowered it's opponents offence by 9 points(that's' second only to the 2004 pistons). Without klay and having basically bench players, they posted a rating of 107. Curry is +12 with only bench players.

Great sample size.

Curry is great with the warriors but worse in a vacuum than other atg's is absolutely ludicrous. Kobe and Wade are more system dependent than curry, so if that's your argument here, it's not a very good one.

Based on what? They were both very consistent in their playing style. They are less system flexible the way Curry hypothetically is (because of his off-ball value), but that is not the same as being system dependant (unless you mean the "system" is them having high usage, which is pedantic and can be applied to basically any top player).

Now, having covered that...

Here is PIPM:


Seem familiar? How is that meaningfully different from the exact complaints E-balla had of BPM?

If good players cannibalise each other, why do impact metrics paint the opposite picture? Why do the Warriors have three of the best six or eight guys? Why does Kevin Durant suddenly leap up the rankings once he joins; did he just happen to learn "winning" basketball in that 2016 offseason? Or if we use your PIPM metric, why do the 2018 Rockets have two of the league's three best players? Why is Danny Green a superstar impact player on the Raptors? Why is Robert Covington basically on par with Towns? Why is Otto Porter a perpetual metrics maestro? It is because they are all excelling in their roles. They are not being dinged because they are not being asked to do anything difficult for their skillset. They are not carrying the teams the way players like Garnett and Lebron and young Jordan and 2002-03 Duncan did. This is not the full story of basketball, and it should be obvious if you guys ever bothered to look at the players next to your specific points of focus.

[WarriorGM]

Again, I support using “advanced metrics”, but we need to be careful about building entire cases on them the way people are doing for Curry (sorry, cases beyond pointing to him being on three 67-win teams ). And I think it provides an interest contrast with the Garnett discussion where the reliance on “metrics” at least was always complemented by acknowledgements of what Garnett could do compared to Duncan, as well as an understanding that defensive impact is not especially well captured by the analytics which are available to us, and of course an appreciation that he was screwed in terms of public perception by his poor team quality

This is rich. "We need to be careful about building entire cases on 'advanced metrics' the way people are doing for Curry"? What!? First of all no one is building an entire case based only on advanced metrics for Curry. If anything that is a more apt description of the case made on behalf of Garnett aside from allusions to his defense which has been vaguely defined. I could do the same for Curry's offense and in contrast to Garnett's defense back it up with numbers. Second, what do people need to be careful about? Do they need to be careful because advanced metrics paint such a positive picture of Curry that he'd look too good?

As I've stated previously Curry has the strongest all-around case. He is strong on conventional metrics. He is strong on advanced metrics based on box score. He is strong on advanced metrics based on plus-minus. He has set records in various categories. He wins. He's done it by a large margin. He's done it repeatedly. One can even point to how the league has adjusted in response to him. Curry's peak was so high it changed the game. Simply put the case for Curry is very robust. The arguments for others like Garnett are thin and riddled with holes in comparison.

Let's face it even Curry 2015 is better than Garnett 2004.

http://bkref.com/pi/shareit/LxNHw

(speaking of which, I love how with Curry we have flipped to script to say, “Actually, playing with the best possible team is a disadvantage for Curry compared to everyone else” ).

Isn't this the same project that's chosen for its second best peak of all-time LeBron's 2013 year when he played with 3 other Hall of Fame guys in Wade, Bosh, and Allen? Then there's Wilt playing with three other Hall-of-Famers too, as well as Bird not to mention Russell. Basketball Reference doesn't even give Klay or Draymond a greater than 50% chance of making it into the Hall of Fame yet. Curry is being held to a higher standard and yet a fair evaluation would show he still manages to exceed them.

[liamliam1234]

Shocking perception from his biggest fan.

[E-Balla]

But if you want numbers look how KG is way better in the RAPM category than Curry while having a clearly bigger load to handle being the best offensive and defensive player for his team. But hey he didn't even reach the finals right?

Well will you look at that! The site referenced in an above post for RAPM (https://basketball-analytics.gitlab.io/rapm-data/) gives the following:

Stephen Curry 2016
Regular Season RAPM: 6.1032
Playoffs RAPM: 1.7193

Kevin Garnett 2004
Regular Season RAPM: 5.8716
Playoffs RAPM: 1.6113

As I suspected Garnett doesn't have an RAPM advantage either. Oh what else is this I see? Sam Cassell had a regular season RAPM of 4.2955 while Wally Szczerbiak had an RAPM in the playoffs of 1.5084 right behind Garnett. But I thought Garnett had trash teammates? So what's left?

Not getting into this one but that site has bad data and RAPM can't be compared year to year like that. Rankings/zscore is the best way to keep them on the same scale as RAPM doesn't keep a consistent... Umm... Scale I guess is the word I'm looking for?

[E-Balla]

Yeah, it set off a few alarm bells, which is why I wanted to warn people about taking it as a definitive data set, but it also does not too radically deviate from better calculations.

This is a good write-up which goes into many of the issues. (https://squared2020.com/2018/12/24/regularized-adjusted-plus-minus-part-iii-what-had-really-happened-was/

I looked at PIPM, and it basically seems to be a glorified BPM, complete with a fair number of the same flaws. Again, I support using “advanced metrics”, but we need to be careful about building entire cases on them the way people are doing for Curry (sorry, cases beyond pointing to him being on three 67-win teams ). And I think it provides an interest contrast with the Garnett discussion where the reliance on “metrics” at least was always complemented by acknowledgements of what Garnett could do compared to Duncan, as well as an understanding that defensive impact is not especially well captured by the analytics which are available to us, and of course an appreciation that he was screwed in terms of public perception by his poor team quality (speaking of which, I love how with Curry we have flipped to script to say, “Actually, playing with the best possible team is a disadvantage for Curry compared to everyone else” ).

Just want to point out RAPM is a valuable tool, but I agree. Too many people around here see a +4.5 RAPM and a +4.0 RAPM and see it as a definitive gap instead of an estimation of their impact. Personally I care more about tiers than the exact value because you do notice RAPM separates itself into tiers.

[Timmyyy]

Not getting into this one but that site has bad data and RAPM can't be compared year to year like that. Rankings/zscore is the best way to keep them on the same scale as RAPM doesn't keep a consistent... Umm... Scale I guess is the word I'm looking for?

I agree with you that the RAPM scores can't be accurately compared across different years. The score is a measure against the mean, since the mean can vary a little from year to year there can be differences. Still want to add that the mean of the league isn't in some sort jumping from A to Z, so at least a ballpark analysis should be reasonably possible. Looking at the ranks in the year in question and comparing the score compared to the rest of the field in the year, while comparing the ballpark in which the scores are between the years should actually give us a solid understanding.

The way bigger problem is comparing the RAPM scores between different methods and sample sizes. For example PI usually has a higher 'scale' than NPI. NPI RS+PS should have a higher scale than NPI RS only. The more data the better the real impact guys can be separated from the field with higher scores. At least I saw JE talk about that in some post, but it clearly shows in his data. So the above mentioned ballpark evaluation of the scores is only possible if we are looking at the same methodology.

But when you are aware of all of that, it is quite useful for comparisons.

[trex_8063]

1st ballot - '95 David Robinson
The near-reality as I see is that David Robinson was asked [by the Spurs] to be Bill Russell on defense and simultaneously be Shaquille O'Neal on offense.......and he kinda takes some flack for not being up to the task [primarily in the playoffs]. But realistically, if he'd been consistently capable of maintaining his rs standard of offensive performance and efficiency during the playoffs, we'd have been discussing him in the top 3 positions of this project. So I don't think it's off base to give him some consideration now around #10. This version of Robinson anchored a -2.9 rDRTG (5th/27) with a principle cast [in descending order of minutes played] of Avery Johnson (scrappy and energetic, but seriously undersized even for a PG; mediocre defender overall), Sean Elliott (mediocre defender), Vinny Del Negro (probably slightly weak defensively, iirc), Chuck Person (a pinch past prime, never a good defender anyway), Dennis Rodman (erratic defensively [awful in the Houston series, fwiw], and missed 33 games), and JR Reid, Terry Cummings, post-prime Doc Rivers (Rivers probably the only one of those three I'd say was passable good defensively [edit: +/- maybe Cummings]).
This version of Robinson simultaneously anchored a +3.4 rORTG (5th/27) with the aforementioned cast; they won 62 games (+5.90 SRS) overall. Made it to the WCF where Dennis Rodman had a total [and very public] meltdown, and the Spurs lost the series to Houston (with Hakeem in God-mode) in six games (outscored by a grand total of 10 pts in the entire series). Typically stated as Hakeem owning DRob and making him a helpless play-thing, though it's rarely acknowledged that Hakeem [because of how their offense and roster was structured] largely enjoyed single coverage (by Robinson), while Robinson was largely guarded by Olajuwon + 1-2 friends.
It's rarely acknowledged that DRob's cast [which had shot 37.5% from beyond the arc in the rs] somewhat crapped the bed shooting just 31.9% in this series (and did I mention they were only outscored by 10 points total in the entire series?); and again Rodman's meltdown and poor play is rarely given light of day in the construction of the usual narrative.
jsia, I think he deserves a look around now.

1.) You say that David Robinson was asked by the Spurs to be both Bill Russell on defense and Shaq on offense. I think that's a true statement. Do you also feel that is true of both Ewing and Olajuwon or do you think Robinson's responsibilities were somehow more grueling? You also say you think Robinson receives some undue flack for this (I can also agree with this). Do you think Ewing's flack for that same situation is equally undue? more? less? Or do you feel in his case it's flack that has merit?

To a degree, yes. Were they both [like Robinson] asked to be the anchor on BOTH sides of the court? Yes. Did they both shoulder similar usage? Yes (peak levels, trough levels [during primes], and avg prime levels are all extremely similar for all three players).

However, I would say in terms of standard or precedent set during the rs, Robinson set the bar higher than the other two (and thus had further to fall in the playoffs). This is why [as I'd noted previously] that even with his infamous playoff struggles, Robinson's avg offensive performance in the playoffs from '93-'96 was similar to [even marginally better than] what rs Patrick Ewing was averaging from '92-'96.

To further qualify that Robinson was "setting the bar higher": If we compare rs Robinson ('90-'98) to rs Ewing ('89-'97) and rs Olajuwon ('89-'97), Robinson [despite basically same USG%] has slightly higher scoring volume by way of better shooting efficiency (because he's got the same mpg avg as Ewing, which is actually 1 mpg LESS than Hakeem in this sample); and it can't be fully explained away with pace, either, as even in per 100 possession terms he comes out first in pts. He also has slightly higher volume of orpg, same or more apg, and the lowest topg.....

'89-'97 rs Ewing: 24.4 ppg @ 56.1% TS, 2.5 orpg, 2.3 apg, 3.2 topg, 108 ORtg
'89-'97 rs Hakeem: 24.9 ppg @ 55.8% TS, 3.2 orpg, 3.0 apg, 3.3 topg, 109 ORtg
'90-'98 rs Robinson: 25.1 ppg @ 59.0% TS, 3.4 orpg, 3.0 apg, 2.9 topg, 118 ORtg

It's certainly not a huge difference, but there's no denying Robinson appears as the most statistically dominant offensive player in their respective primes during the rs (and note that's even with including some of his post-injury years). And this is arguably reflected in the performance of their respective team offenses:

Ewing's worst team offense ('89-'97): -2.3 rORTG
Ewing's best team offense ('89-'97): +3.3 rORTG
Ewing's avg team offense ('89-'97): -0.4 rORTG

Hakeem's worst team offense ('89-'97): -3.3 rORTG
Hakeem's best team offense (""): +2.1 rORTG
Hakeem's avg team offense (""): -0.17 rORTG

Robinson's worst team offense ('90-'98): -3.4 rORTG* (*that's '97 when he played just 6 games; outside of that year the worst is a -1.2 rORTG)
Robinson's best team offense ('90-'98): +4.1 rORTG
Robinson's avg team offense ('90-'98): +0.66 rORTG* (*if we removed '97 from the sample, the avg was +1.03 rORTG).


From the above one might infer that the Spur's success was more closely tied to the offensive performance of Robinson and the team in general. So from that standpoint, one could suggest that slightly more was "expected" of Robinson on the offensive end than was "expected" of the other two.
He didn't live up to that expectation, I'm not denying that, which is why they almost never lived up to their rs standards.

For completeness, the playoff avg's of Ewing and Robinson [omitting Hakeem at this point, as he was the clear best post-season performer and I was never at any point arguing Robinson over him anyway] in the years stated above....

'89-'97 Ewing ps: 22.6 ppg @ 52.7% TS, 2.6 orpg, 2.4 apg, 2.8 topg, 106 ORtg
'90-'98 Robinson ps: 23.4 ppg @ 54.9% TS, 3.6 orpg, 2.9 apg, 2.9 topg, 112 ORtg

So Robinson has the larger drop-off from his rs averages noted above, but still [on average] appears like the superior offensive playoff performer. But it still just wasn't enough, based on the expectation he'd set for himself by his titan-level rs performance. It's true he had struggles performing against strong defenses and was never a true ideal offensive center-piece. He'd likely have done better even in a 1a/1b role offensively (particularly if paired with a perimeter-oriented 1a/1b option), or at least having one other truly reliable 2nd-option scorer; but he never had opportunity until Duncan came along (and just one year where Robinson was post-injury and 32 years old----arguably not even in his true prime anymore----after which Robinson was clearly past his prime). Sean Elliott or a post-prime Dale Ellis were the best he ever had prior to that.

[DatAsh]

1. 77 Walton - Best passing big man ever before Jokic. Top 5 defensive players ever, and one of the best offensive big men ever. Arguably better than 77 Kareem. Box score favors Kareem, impact metrics favor Walton.

2. 2017 Curry - Didn't get injured in the playoffs and still one of the best offensive seasons ever. I'm not entirely sure his offensive decline was a true decline, rather than a reduced primacy due to the arrival of Kevin Durant; I tend to split the difference here.

3. 64 Oscar - I'm picking this year as his peak. Best offensive player prior to Magic. Consistently lead the leagues best offenses while posting incredible scoring numbers on incredible efficiency, while leading the league in assists.

[liamliam1234]

I'm all but certain my vote here will go to Oscar Robertson, but I want to look a little deeper at some of the other candidates.

I feel like a have a pretty good handle on the case for Steph Curry, David Robinson and Charles Barkley, but I feel much like I might be missing out on the case for a few others.

I've put some questions I'd love responses to in bold:

1983 Moses Malone - I'd like to hear any arguments against him, because generally, I felt for a longtime like his 1983 season was among the top ten peaks in league history. Because of that, I feel like there is a compelling case for him, but just as he was once underrated historically, it feels like in the last decade or so he's become overrated in the eyes of many informed basketball fans. So let's hear what everyone here thinks...

1977 Bill Walton - I'm going to deep dive on this one because if I could guarantee health, he'd be a top ten for me for sure. The issue is of course availability being among the most important of abilities and being his Achilles heel, or perhaps more aptly his entire foot. Regardless, I'd like to know how others weigh how great he was on the court vs. how often he wasn't able to be on the court.

There is also a quartet of forwards who aren't getting a lot of traction that I think could if not should be mentioned at this point. Bob Pettit, Karl Malone, Dirk Nowitzki and Kevin Durant. Who stands out among this group to you and how soon will they be appearing on your ballot?

Beyond that, and maybe it's a bit too soon for this discussion, Jerry West vs. Kobe Bryant vs. Dwyane Wade - whose peak do we prefer and why?

Figured I would respond to this, because not many people did and because discussion is dying down.

Moses: Comparing him to other bigs, I do not think his individual offence, and especially not his team offence, was on the level of Dirk or Barkley. Skeptical of the suggestion that his defence makes up for that gap. Move to the 76ers was not the literal equivalent (in terms of basketball, not narrative) impact as Durant to the Warriors... but for a season it kind-of was, so even if his playoff numbers are a bit better than Barkley’s, I am not sure how much I would value that. Robinson is an interesting conversation; bad playoff dip on offence, but defensive impact may have meant he was a lot better than the box score would suggest (of course, for 1995, tough to argue considering his series against Hakeem). So after Walton is off the board, probably a tentative second on my list for bigs.

Side note: if Moses is on the board (heh), I would like to see some discussion of Gilmore popping up soon. Think his case is interesting, although picking a definitive year may torpedo that at the start.

Walton: Peaks project, obviously missed time did not cost them a title (plus, his regular season value was still high).

Forwards: Nowitzki, then tentatively Rick Barry before the rest. Willing to hear the case for McAdoo or Pettit. 2014 Durant dipped in the playoffs too much (and that is the year for which he deserves the most credit). Still probably over KMalone, though, as is Barkley. Will take Robinson somewhere in that range.

Guards: Total mess, as I expected they would be several rounds ago. Have five guys each with their own arguments. Made case for Wade’s and Kobe’s title seasons, but still nagged by the idea that no one thought 2006 Wade achieved something 2006 Kobe could have done in his place (you know, obvious Shaq element aside). Similar argument could be made swapping 2009-10 Wade with Kobe (and that one is more supported by “metrics”). Paul’s metrics are great, but occur without translating into postseason success. Tough to compare West because of era, but he is certainly the best passer after Paul (however, not at his passing peak until later in his career). West also has great playoff performances, but would not say he has an innate scoring advantage over any of the other three score-first guards, nor would I say he necessarily has the defensive edge over Wade. For all of them, lack of easily readable defensive data might overrate them by comparison with, say, a Ewing or Robinson type player. Think Rick Barry is interesting point of comparison. Offensive focus means Barkley and Nowitzki (and Durant and Pettit, and maybe McAdoo) can also be compared more easily.

[Bel]

1. 63 Oscar
2. 77 Walton
3. 90 Barkley

I explained my reasoning in the previous thread, though Dr.J was a fine choice for the 12th spot. The Curry arguments are unconvincing (sorry - do not want to wade into this morass due to time), as are Robinsons. David Robinson should not get any compensation for Rodman acting up. Thomas/Daly and Jordan/Jackson proved that Rodman could be properly harnessed with the right type of leadership, but Rodman did not respect Robinson, got into spats with Popovich and the Spurs allowed him to get away with his bull in a way that his prior and post-Spurs teams did not. If Rodman was a cancer on all of his teams, that would be justified to give Robinson compensation for Rodman's poor performance in the 95 playoffs, but he was only a cancer on the Spurs (I guess you could include the 99 Lakers who did not want him). As the leader of a team that consistently underperformed in the playoffs prior to Duncan, Robinson must bear responsibility. One aberration is fine, a consistent pattern is not.

I'm definitely interested in the Jerry West arguments being made: would love to see more detail. Sadly, as usual Barkley seems to be ignored as though he's a longevity monster and not a guy with a very high peak/prime and poor longevity. It's one thing if people who have watched sufficient footage reject him for good reasons, but...bah.

[liamliam1234]

What is the case for Barkley over Dirk? (Or vice versa, for those who want to implicitly argue against Barkley’s inclusion at this point.)

[Mavericksfan]

What is the case for Barkley over Dirk? (Or vice versa, for those who want to implicitly argue against Barkley’s inclusion at this point.)

I don’t think Barkley has much of an argument over Dirk. Dirk led his team to more regular season and playoff success. Dirk’s also one of the few who’s game translated against tougher defenses as well.

One thing I dont see mentioned enough is that Barkley routinely saw a pretty steep decrease in efficiency come playoff time. When he didnt he wasnt putting up the same eye popping regular season numbers or his teams werent very good. He also layed a couple of eggs against the Rockets in 94-95.

I think Dirk in 06 ,( incredible reg season and good overall playoffs) or Dirk in 2011 (good regular season, amazing playoffs) are just straight up better than anything Barkley has done.

My criteria is usually team results(whether a floor or ceiling raiser), individual production, portability, and playoff performance. Not necessarily in that order.

[70sFan]

What is the case for Barkley over Dirk? (Or vice versa, for those who want to implicitly argue against Barkley’s inclusion at this point.)

I don’t think Barkley has much of an argument over Dirk. Dirk led his team to more regular season and playoff success. Dirk’s also one of the few who’s game translated against tougher defenses as well.

One thing I dont see mentioned enough is that Barkley routinely saw a pretty steep decrease in efficiency come playoff time. When he didnt he wasnt putting up the same eye popping regular season numbers or his teams werent very good. He also layed a couple of eggs against the Rockets in 94-95.

I think Dirk in 06 ,( incredible reg season and good overall playoffs) or Dirk in 2011 (good regular season, amazing playoffs) are just straight up better than anything Barkley has done.

My criteria is usually team results(whether a floor or ceiling raiser), individual production, portability, and playoff performance. Not necessarily in that order.

It's not true that Barkley had problems against playoffs defense.

Both players against -2.0 or better rDRtg teams:

1987-96 Charles Barkley:(41.89% of playoffs games): 41.8 mpg, 12.8 rpg, 3.9 apg, 2.6 tov, 25.7 ppg on 52.0% FG, 23.1% 3FG, 77.6% FT and 59.9% TS (+6.30% rTS)
2001-11 Dirk Nowitzki: (47.58% of playoffs games): 40.9 mpg, 10.3 rpg, 2.6 apg, 2.2 tov, 24.4 ppg on 46.1% FG, 36.7% 3FG, 90.5% FT and 57.6% TS (+4.41% rTS)

Against -4.0 or better:

1987-96 Charles Barkley:(14.86% of playoffs games): 42.2 mpg, 13.1 rpg, 4.0 apg, 2.3 tov, 24.2 ppg on 45.3% FG, 29.6% 3FG, 79.8% FT and 54.3% TS (+1.03% rTS)
2001-11 Dirk Nowitzki (28.23% of playoffs games): 41.1 mpg, 10.6 rpg, 2.6 apg, 2.0 tov, 23.4 ppg on 45.5% FG, 29.3% 3FG, 88.4% FT and 57.4% TS (+4.41% rTS)

Barkley had some efficiency decrease against the best defensive teams, but he didn't face them often. Against good defenses they are comparable.

[penbeast0]

What is the case for Barkley over Dirk? (Or vice versa, for those who want to implicitly argue against Barkley’s inclusion at this point.)

I don’t think Barkley has much of an argument over Dirk. Dirk led his team to more regular season and playoff success. Dirk’s also one of the few who’s game translated against tougher defenses as well.

One thing I dont see mentioned enough is that Barkley routinely saw a pretty steep decrease in efficiency come playoff time. When he didnt he wasnt putting up the same eye popping regular season numbers or his teams werent very good. He also layed a couple of eggs against the Rockets in 94-95.

I think Dirk in 06 ,( incredible reg season and good overall playoffs) or Dirk in 2011 (good regular season, amazing playoffs) are just straight up better than anything Barkley has done.

My criteria is usually team results(whether a floor or ceiling raiser), individual production, portability, and playoff performance. Not necessarily in that order.

Taking a quick look at Barkley's stats on B-R.com, I don't see a dropoff, slight but less than a normal star in fact (which is more in line with my memory of Barkley being a good playoff performer. Even his defense improved from James Harden to mediocre).

Philly
RS 22pts 11reb 3.5ast .640 ts%
PO 22pts 13reb 4ast .618ts%

Phoenix
RS 23 pts 12 reb 3ast .580ts%
PO 27pts 13reb 4ast .563ts%

Houston 17pts 12reb 2ast .562ts%
RS PO 17pts 11pts 3ast .562ts%