RealGM 2023 Top 100 Project - #8 (Shaquille O'Neal)

[OldSchoolNoBull]

You're certainly not wrong about it being about money. Some quotes from Shaq:

link


link


link


Well and good to say that if the Magic would have just offered Shaq the kind of money other teams were spending on bigs he'd have stayed, but both he and the Magic were focused on the comparison with Hardaway, and Shaq would insult Hardaway years after the fact ( ""He thinks he's smart, but he's really not.") continuing to hurt Hardaway who looked up to Shaq as a big brother, and who went through absolute hell with injuries for years by that point.

If you want to take issue with my choice of words because they don't tell the whole story that's fine, but the emotion I attributed Shaq was absolutely accurate by his own admission.

Incidentally I think Orlando was silly for trying to pinch pennies, but they weren't actually wrong to think that Hardaway was the superior franchise player...except for their ignorance of his future injury issues.

I'm going to be honest, Doc, we all have biases, and you don't seem to like Shaq much, so I feel like continuing to argue in his favor to you may be futile, but nontheless...

I mean, we all have our biases, but I think you're being kind of selective with when RAPM matters and when it doesn't...

(And for the record, this is JE's Regular Season+Playoffs RAPM, so it takes everything into consideration.)

Okay, but JE's Regular Season+Playoffs RAPM very firmly supports Garnett(and Curry) over Shaq here with basically any frame and it doesn't include a guy with an arguably era-best impact portfolio(magic).

And the only other extended(sourced) set I know of has shaq scoring even worse:
https://public.tableau.com/app/profile/ahmed.cheema8618/viz/FiveYearRAPMPeaks1997-2021/FiveYearRAPMPeaks1997-2021


(5-year frame is Shaq's best look. He's the left-most circle, prime kobe(2006-2010) is the second one, wade's best is the third, lebron, kg, duncan, cp3, and curry make up dots above)

with short-stretch raw stuff(00-02) Shaq looks pretty good in the rs, but KG looks better and you've voiced strong disagreement even after the thread closed with Hakeem(someone who also looks pretty good with a raw approach).

To be clear, this is not meant as a dig. Doc recently(rightly) found me getting lost in bias with my approach regarding Lakers Kareem and Shaq. But I don't really understand why you toss rapm here as some vindication for Shaq when it really doesn't like him compared to the players he's up against here.

If JE's set is valued, Shaq should not be getting voted here. And that is not Shaq's worst look in this sort of metric.

I was citing that RAPM only in relation to Shaq vs Kobe, because that is the comparison Doc brought up.

[lessthanjake]

Not really. Cannabalization as the term is currently being used is somehow both

A. when a player becomes less impactful with a teammate in actual lineups
B. when a player's apm actually misrepresents how impactful a teammate is

The former isn't really a matter of collinearity. APM would not be misreading impact if it had player a above the cannibalized teammate. Especially if said player stayed very impactful without said teammate. With point 2 where we are now defining it as a matter of colinearity which is outright contradictory with the first use.

Collinearity becomes a problem when you play heavy minutes with someone whose very good, play very few minutes without that guy. In the original case Dray looks really good without Steph in similar minutes. So playing high minutes with a guy who is very good without you and low minutes without said guy benefits(and dray too benefits to be clear).

With Duncan and teammates who play way less than him(cough Robinson cough) we have the opposite problem where Duncan is playing with the good players and the not so good replacements while his teammates barely play without timmy. Aenigma expounds on this pretty well here:

You'll have to clarify for me, how are those two are "outright contradictory" to you. Since they aren't addressing the same statement, by any definition I know they can't be contradictory.

So wait... your position is that both Steph/Dray are 'benefitting' from collinearity with one another here (somewhat notably to bother mentioning it)?

statement a: -> player b actually isn't impactful with player a(lineup does not improve much from when it was just player a)
statement b: -> player's a rapm is inflated by player b's actual impact

both cannot simultaneously be true

And yes. That is my position. Both go off the team gets way worse because it has neither. Both come on, team gets way better because it has both.

You’re arguing against something I said, but “statement b” isn’t what I said. The point is that when “statement a” is true and there’s a lot of overlap in the two players’ minutes, then player B looks like an unimpactful player to the model, and therefore, when the model controls for player B’s presence on the court with player A, it is going to conclude that player A is responsible for almost all the impact of the two of them together because the model will see player B as an unimpactful player. And in a sense, that’s right—player B actually hasn’t had much impact. But the model cannot really take into account that player B is being artificially made into a less impactful player in a huge portion of his minutes. So it’s not really wrongly estimating player A’s impact in reality, but it is overestimating it if we take into account the negative effect player A is having on player B’s impact. A RAPM model is not trying to model what effect a player might have on another player’s impact. It cannot figure out if player A’s presence is causing player B’s impact to be lower, as opposed to player B’s impact simply randomly being lower when player A is on the court. It basically is just seeing what impact player B seems to have on average and controlling for that when estimating player A’s impact. Anyways, it’s a minor point and is actually yet another example of arguing about LeBron James when he is not directly relevant, so I don’t think I’m going to discuss it further.

[OhayoKD]

You'll have to clarify for me, how are those two are "outright contradictory" to you. Since they aren't addressing the same statement, by any definition I know they can't be contradictory.

So wait... your position is that both Steph/Dray are 'benefitting' from collinearity with one another here (somewhat notably to bother mentioning it)?

statement a: -> player b actually isn't impactful with player a(lineup does not improve much from when it was just player a)
statement b: -> player's a rapm is inflated by player b's actual impact

both cannot simultaneously be true

And yes. That is my position. Both go off the team gets way worse because it has neither. Both come on, team gets way better because it has both.

Makes more sense with those statements. I don't personally agree with statement A there.

Ahh, well, your position is... not statistically sound. Collinearity increases variance in the paired variables, it does not bias (inflate or deflate) estimated values.

I...do not see how deflation would be possible there. The upper/lower bound delta also doesn't look unusually high in JE's set:


That said, even if that's true, over the two extended sets we have for data-ball(where I'm guessing variance diminishes?), Steph does not look like an impact king or really even credible threat for the best looking guys. Draymond also does pretty well himself.

[lessthanjake]

statement a: -> player b actually isn't impactful with player a(lineup does not improve much from when it was just player a)
statement b: -> player's a rapm is inflated by player b's actual impact

both cannot simultaneously be true

And yes. That is my position. Both go off the team gets way worse because it has neither. Both come on, team gets way better because it has both.

Makes more sense with those statements. I don't personally agree with statement A there.

Ahh, well, your position is... not statistically sound. Collinearity increases variance in the paired variables, it does not bias (inflate or deflate) estimated values.

I...do not see how deflation would be possible there. The upper/lower bound delta also doesn't look unusually high in JE's set:


That said, even if that's true, over the two extended sets we have for data-ball(where I'm guessing variance diminishes?), Steph does not look like an impact king or really even credible threat for the best looking guys. Draymond also does pretty well himself.

Steph Curry is very clearly the impact-metric king of the last decade. It’s not even really a debatable premise. If you want to say that if you add in his first several seasons and compare to other players’ whole careers as well, then other people look more impactful on average than him, then that’s totally fine, and I’d tend to agree actually. But Steph Curry is clearly the impact king of the last decade. I’ve provided an absolutely enormous amount of data backing this up. The totality of the evidence is plainly in Steph’s favor, and you keep just making declarative statements that this isn’t true, without backing up your point in any meaningful way besides applying an unreasonable standard to say that Steph cannot be the impact king of the last decade because he isn’t 1st place in every metric all the time.

Anyways, I’ve provided an incredible amount of info that provides this, but here’s some more, using a metric that you yourself say you prefer:

Average EPM in the last decade (excluding injury seasons)

1. Steph Curry: 7.76
2. Nikola Jokic (last 5 years only): 7.48
3. Giannis Antetokounmpo (last 5 years only): 7.06
4. Kawhi Leonard: 6.61
5. Joel Embiid (last 5 years only): 6.54
6. LeBron James: 6.23
7. James Harden: 6.17
8. Chris Paul: 5.97
9. Kevin Durant: 5.68
…
Really high number: Draymond Green: 2.98

[For reference, while it’s a different time interval, so this is not necessarily directly comparable, in the decade from 2008-2009 to 2017-2018, LeBron had an average EPM of 7.59]

Average League Placement in EPM in the last decade (excluding injury seasons)

1. Steph Curry: 2.89
2. Nikola Jokic (last 5 years only): 4.80
3. Giannis Antetokounmpo (last 5 years only): 5.00
4. Kawhi Leonard: 5.50
5. Joel Embiid (last 5 years only): 6.00
6. LeBron James: 6.80
7. Kevin Durant: 9.38
8. James Harden: 9.70
9. Chris Paul: 12.2
…
Really high number: Draymond Green: 52.7

[For reference regarding Steph’s level of dominance, in the decade from 2008-2009 to 2017-2018, LeBron’s average league placement in EPM was 3.4]

[OhayoKD]

Makes more sense with those statements. I don't personally agree with statement A there.

Ahh, well, your position is... not statistically sound. Collinearity increases variance in the paired variables, it does not bias (inflate or deflate) estimated values.

I...do not see how deflation would be possible there. The upper/lower bound delta also doesn't look unusually high in JE's set:


That said, even if that's true, over the two extended sets we have for data-ball(where I'm guessing variance diminishes?), Steph does not look like an impact king or really even credible threat for the best looking guys. Draymond also does pretty well himself.

Steph Curry is very clearly the impact-metric king of the last decade.

And the two metrics you preferred had Jokic dominate by your original standard:
https://forums.realgm.com/boards/viewtopic.php?p=107743950#p107743950
Idk what you want me to say. "very clearly the impact king" doesn't really work when you have to aggressively filter to get him looking the best....

And ofc none of this stuff covers data-ball...

[lessthanjake]

I...do not see how deflation would be possible there. The upper/lower bound delta also doesn't look unusually high in JE's set:


That said, even if that's true, over the two extended sets we have for data-ball(where I'm guessing variance diminishes?), Steph does not look like an impact king or really even credible threat for the best looking guys. Draymond also does pretty well himself.

Steph Curry is very clearly the impact-metric king of the last decade.

And the two metrics you preferred had Jokic dominate by your original standard:
https://forums.realgm.com/boards/viewtopic.php?p=107743950#p107743950
Idk what you want me to say. "very clearly the impact king" doesn't really work when you have to aggressively filter to get him looking the best....

And ofc none of this stuff covers data-ball...

You are seemingly referring to criteria I mentioned when talking about ways that we might possibly compare different players’ performance in impact metrics across different time periods. I said that it’s not something we can really directly compare since the scaling of different time intervals are different, but that we might be able to draw inferences regarding players’ performance in different years or time periods if they look radically different in certain factors like that. But that’s not relevant to the question of who was the impact king overall over the specific time period of the last decade. To determine that, we don’t need any inferences. We can just directly compare the players’ performances in the metrics in the exact same years as each other. And there, Steph is clearly the top performer of the last decade—though it is true that Jokic has been the clear top performer if we just zeroed into the last few years.

As a sidenote, I don’t have metrics that I have said I prefer. My position is that they’re all almost certainly flawed, and that we should therefore just look at the totality of them in the hopes that the flaws roughly cancel out. You are the one that has preferred metrics, which end up being ones that are most charitable to your arguments. I look at all of them, and the totality in the last decade is plainly in Steph’s favor. It’s not a debatable premise.

[OhayoKD]

You'll have to clarify for me, how are those two are "outright contradictory" to you. Since they aren't addressing the same statement, by any definition I know they can't be contradictory.

So wait... your position is that both Steph/Dray are 'benefitting' from collinearity with one another here (somewhat notably to bother mentioning it)?

statement a: -> player b actually isn't impactful with player a(lineup does not improve much from when it was just player a)
statement b: -> player's a rapm is inflated by player b's actual impact

both cannot simultaneously be true

And yes. That is my position. Both go off the team gets way worse because it has neither. Both come on, team gets way better because it has both.

You’re arguing against something I said, but “statement b” isn’t what I said. The point is that when “statement a” is true and there’s a lot of overlap in the two players’ minutes, then player B looks like an unimpactful player to the model, and therefore, when the model controls for player B’s presence on the court with player A, it is going to conclude that player A is responsible for almost all the impact of the two of them together because the model will see player B as an unimpactful player. And in a sense, that’s right—player B actually hasn’t had much impact. But the model cannot really take into account that player B is being artificially made into a less impactful player in a huge portion of his minutes. So it’s not really wrongly estimating player A’s impact in reality, but it is overestimating it if we take into account the negative effect player A is having on player B’s impact. A RAPM model is not trying to model what effect a player might have on another player’s impact. It cannot figure out if player A’s presence is causing player B’s impact to be lower, as opposed to player B’s impact simply randomly being lower when player A is on the court. It basically is just seeing what impact player B seems to have on average and controlling for that when estimating player A’s impact. Anyways, it’s a minor point and is actually yet another example of arguing about LeBron James when he is not directly relevant, so I don’t think I’m going to discuss it further.

I mean..

In contrast, we can look to pairings that vultured impact from one another. For example, in their years together in Cleveland, Kyrie Irving had a +9.02 effect on the Cavaliers’ net rating in minutes that LeBron was off, but he had just a +0.38 effect on the Cavaliers’ net rating in minutes that LeBron was on. Kyrie was extremely impactful without LeBron, and basically had zero impact whatsoever if LeBron was on the court. This kind of vulturing of impact is obviously way more suboptimal. It may juice the superstar’s RAPM—because the secondary star’s lack of impact when the superstar is on the floor makes RAPM regard the secondary star as less good than other guys’ secondary stars, and therefore results in giving the superstar more of the credit—but it’s not actually reflective of a good thing.

How is a guy being "juiced" if his lineups look as good without the player whose supposedly juicing him?

Kyrie's rapm being "drained" is also dependent on what you are looking at:

Spoiler:

The above graph also jibes with the scouting report; as LeBron’s passing steadily improved and his shot selection grew more judicious, he synthesized with better talent, correlating with larger and larger scoreboard shifts after a nadir in 2012. This was a two-way street: As LeBron’s more efficient passing helped the talent around him — Kyrie Irving and Kevin Love posted career-best marks in scaled APM in 2017 — his improved 3-point shooting allowed him to finish more plays setup by his teammates. (Notice in the previous charts how LeBron’s efficiency improved alongside Irving.)

Luck adjusted

+1.75, 2017, +2.29 ORAPM, -.54 DRAPM
+1.73, 2015, +2.07 ORAPM, -.34 DRAPM
+1.69, 2019, +1.35 ORAPM, +.34 DRAPM
+1.49, 2021, +1.79 ORAPM, -.3 DRAPM
+1.43



Cheema’s set has it at:

2017-2021: +2.52
2015-2019: +2.41
2014-2018: +2.07
2016-2020: +1.97
2013-2017: +1.23

And there is always the possibility that Kyrie improved as a player over the course of his 20's?

Regardless, I'm fine with the definition of "cannabalizing" being when a players actual impact goes down. Though the idea this means Lebron winning 3 in 5 is floor-raising and Curry winning 3 in 5 is ceiling-raising seems dubious to me.

[OhayoKD]

Steph Curry is very clearly the impact-metric king of the last decade.

And the two metrics you preferred had Jokic dominate by your original standard:
https://forums.realgm.com/boards/viewtopic.php?p=107743950#p107743950
Idk what you want me to say. "very clearly the impact king" doesn't really work when you have to aggressively filter to get him looking the best....

And ofc none of this stuff covers data-ball...

You are seemingly referring to criteria I mentioned when talking about ways that we might possibly compare different players’ performance in impact metrics across different time periods. I said that it’s not something we can really directly compare since the scaling of different time intervals are different, but that we might be able to draw inferences regarding players’ performance in different years or time periods if they look radically different in certain factors like that. But that’s not relevant to the question of who was the impact king overall over the specific time period of the last decade. To determine that, we don’t need any inferences. We can just directly compare the players’ performances in the metrics in the exact same years as each other. And there, Steph is clearly the top performer of the last decade—though it is true that Jokic has been the clear top performer if we just zeroed into the last few years.

[/quote][/quote]
By the standard of "biggest gaps", "leads the most times", which you were using, Jokic is pretty clearly #1 in RAPTOR's overall and its on/off bit as he's #1 more times and he beats steph's biggest gaps over #2 again and again. That gets worse If we have to use specific data and change how we're interpreting it for Steph to be the impact king then Steph is not "clearly the impact king" and the premise is absolutely debatable.

If you want to turn it into something narrower like "between 2019 and 2021", that might work(though Harden and Giannis may have something to say about that), but "clearly impact king of the last decade" is just not viable as there are clear counter-cases even by means of interpretation you have deemed valid enough to use.

[One_and_Done]

KD clearly isn't getting any supoort, even though that's ridiculous, and I can't stomach nominating Mikan. Kobe isn't close for me yet. Anyone else want to make the Dirk/K.Malone/DrJ case? I'd be interested in D.Rob/Giannis too, but can't see them getting enough support due to longevity (even though I'd likely take both over Oscar/West/Kobe).

[lessthanjake]

Regardless, I'm fine with the definition of "cannabalizing" being when a players actual impact goes down. Though the idea this means Lebron winning 3 in 5 is floor-raising and Curry winning 3 in 5 is ceiling-raising seems dubious to me.

The manner in which Steph and LeBron won those titles was different though. By any measure you want to look at (regular season wins, SRS, playoff record, playoff SRS, number of game 7’s, etc.), the Warriors’ title teams were more dominant. One of those Warriors teams was probably the best team ever. Their ceiling was raised higher, even if in the end, as you point out, they got to the same place as LeBron’s teams did—three titles in five years, with finals appearances the other two years. That matters, and is a big deal when assessing the main players on those Warriors. It does actually matter *how* you win, and we can draw conclusions from it, instead of just reverting to a binary assessment of championship vs. no championship. Not all championships are the same, and leading the best team ever is a really impressive achievement, in a league where there’s been a lot of extremely talented teams put together over the years. (And, of course, it’s also a really impressive achievement to win a less dominant but still not particularly difficult title with a team that’s only as talented as the 2022 Warriors—a demonstration of extremely strong floor raising along the lines of Dirk’s 2011 Mavericks, Hakeem’s 1994 Rockets, Duncan’s 2003 Spurs, etc.).

[One_and_Done]

I've already had threads discussing Malone and D.Rob's case, but let's look at Dr J. Underrated due to injuries later in his career that slowed him a little, and forced to take less shots to help manage the egos on his early NBA teams. However there's really no doubt in my mind he peaked higher than Kobe and had longer longevity than people think at first. He also has size, length, hands and athleticism that let him do stuff on both ends that Kobe never could.

Peak Dr J absolutely kills Kobe's best year.

1976 RS Erving: 34.4 pp 100, 12.9 r, 5.9 a, 116 Ortg/97 Drtg, 569 TS%

1976 PS Erving: 37.4 pp 100, 13.6 r, 5.3a, 2.1, 2.2, 128 Ortg/103 Drtg, 610 TS%, and a title.

1976 ABA was as strong or stronger than 1976 NBA in terms of top teams.

[Bklynborn682]

I found this RAPM site saved on my phone do the numbers seem legitimate ?(as they don’t seem to match up with J.E. Or Cheemas though it does seem closer in line with ESPNs RAPM)
https://web.archive.org/web/20150408042813/http://stats-for-the-nba.appspot.com:80/

Edit this apparently is Englemans page though it does seem quite different than a different RAPM spreadsheet I have of his.

[lessthanjake]

By the standard of "biggest gaps", "leads the most times", which you were using, Jokic is pretty clearly #1 in RAPTOR's overall and its on/off bit as he's #1 more times and he beats steph's biggest gaps over #2 again and again. That gets worse If we have to use specific data and change how we're interpreting it for Steph to be the impact king then Steph is not "clearly the impact king" and the premise is absolutely debatable.

If you want to turn it into something narrower like "between 2019 and 2021", that might work(though Harden and Giannis may have something to say about that), but "clearly impact king of the last decade" is just not viable as there are clear counter-cases even by means of interpretation you have deemed valid enough to use.

No, to repeat myself, I mentioned those as examples of factors one could look at to potentially draw inferences *when comparing impact-metric performance of players across different time intervals*. It was an off-handed comment I made while actually making the point that we can’t directly compare across different time intervals in a particularly meaningful or concrete way. However, when looking at players within the same time intervals, we can just directly compare how they performed. There’s no need for inferences about how two players did compared to their different peers, because we’re simply comparing peers directly. Overall, if we look at the panoply of impact metrics we have, no one has performed better than Steph in the last decade. You can point to certain shorter time intervals within the last decade in which people have performed better in those metrics than Steph. Jokic in the last few years is an example. But no one has performed better than Steph overall in impact metrics over the course of the last decade.

This isn’t debatable. And your entire argument is just to go around and around saying that different players have finished ahead of Steph in certain years in the past decade and therefore he cannot be the “impact king” of the last decade as whole, ignoring the fact that these are noisy metrics and players’ form ebbs and flows year to year and so it’s essentially impossible to finish 1st in them all the time. As an example, as I posted earlier, Steph’s average league placement in EPM (a metric you’ve specifically said you like) in the last decade was 2.89. And he actually only finished 1st twice. Your argument would basically be to look at that and say that he clearly wasn’t the impact king because he wasn’t in 1st most of the time. But no one else in the last decade comes even close to that average league placement (not even Jokic, Giannis, or Embiid if we just charitably look at the last 5 years for them—and obviously the full decade is significantly less good for all three of them). And from 2008-2009 to 2017-2018, LeBron’s average league placement in EPM was 3.40, and he too only finished 1st in EPM twice in that timeframe. Steph’s average EPM in the last decade is also ahead of everyone else’s and FWIW is also slightly ahead of LeBron’s average EPM from 2008-2009 to 2017-2018. Steph’s performance in these sorts of metrics is what being the impact king of the decade looks like with fairly noisy metrics, just like it was for LeBron in that 2008-2009 to 2017-2018 period. Your argument just amounts to saying I’m wrong because Steph hasn’t met a completely unrealistic bar.

The bottom line is that if no individual player has outdone Steph in impact metrics overall over the course of the last decade (and no one has), then Steph is the impact metric king of the last decade.

[iggymcfrack]

KD clearly isn't getting any supoort, even though that's ridiculous, and I can't stomach nominating Mikan. Kobe isn't close for me yet. Anyone else want to make the Dirk/K.Malone/DrJ case? I'd be interested in D.Rob/Giannis too, but can't see them getting enough support due to longevity (even though I'd likely take both over Oscar/West/Kobe).

Robinson's longevity only gets dragged because people act like he ceased to exist as a meaningful player as soon as he ceded the lead scorer to Duncan which is ridiculous. Robinson led the league in BPM in 1998. He led the league in WS/48 in 1998, 1999, and 2001. Did those numbers dip in the playoffs? A little, but not as much as you'd think. From 1998-2001, Robinson had a postseason PER of 24.0 and a postseason BPM of 6.8 which is almost identical to Durant's career playoff averages of 24.1 and 6.7.

After Duncan joined the team in 1998, Robinson had an on/off of +6.0 in the regular season and +18.9 in the playoffs for the last 6 seasons of his career from age 32-37. He's actually played exactly one game more than Durant over the course of his career. And while, there was a little decline at the end of his career, he was still MUCH more impactful at age 36-37 than KD was at 19 and 20 his first 2 seasons in the league. I would actually give Robinson a small longevity edge over KD if anything although not enough to be anything other than a tiebreaker.

[ShaqAttac]

I'm going to be honest, Doc, we all have biases, and you don't seem to like Shaq much, so I feel like continuing to argue in his favor to you may be futile, but nontheless...

I mean, we all have our biases, but I think you're being kind of selective with when RAPM matters and when it doesn't...

(And for the record, this is JE's Regular Season+Playoffs RAPM, so it takes everything into consideration.)

Okay, but JE's Regular Season+Playoffs RAPM very firmly supports Garnett(and Curry) over Shaq here with basically any frame and it doesn't include a guy with an arguably era-best impact portfolio(magic).

And the only other extended(sourced) set I know of has shaq scoring even worse:
https://public.tableau.com/app/profile/ahmed.cheema8618/viz/FiveYearRAPMPeaks1997-2021/FiveYearRAPMPeaks1997-2021


(5-year frame is Shaq's best look. He's the left-most circle, prime kobe(2006-2010) is the second one, wade's best is the third, lebron, kg, duncan, cp3, and curry make up dots above)

with short-stretch raw stuff(00-02) Shaq looks pretty good in the rs, but KG looks better and you've voiced strong disagreement even after the thread closed with Hakeem(someone who also looks pretty good with a raw approach).

To be clear, this is not meant as a dig. Doc recently(rightly) found me getting lost in bias with my approach regarding Lakers Kareem and Shaq. But I don't really understand why you toss rapm here as some vindication for Shaq when it really doesn't like him compared to the players he's up against here.

If JE's set is valued, Shaq should not be getting voted here. And that is not Shaq's worst look in this sort of metric.

I was citing that RAPM only in relation to Shaq vs Kobe, because that is the comparison Doc brought up.

if it matter when it says shaq > kob. shouldnt it also matter when it says other players > shaq?

[eminence]

Voting post, third time in a row, so kinda brief.

Vote #1: Kevin Garnett
Vote #2: Shaquille O'Neal

Nomination: George Mikan

KG was great for a long time, any individual metric you can possibly like loves him, and he aces the eye test. Only way I can see not having him in your top 10 is if you thought for some reason his Minnesota version couldn't translate to better teams - which I don't really see any evidence for. They were pretty good the 1 time he got some half decent support (and it really was still just half decent - Cassell/Hoiberg/Sprewell/Hassell is not much), and he immediately won a title the first season he left.

Shaq, had a bit of a blessed NBA life, and brought himself down at times, but still had a very accomplished career that lasted for a surprisingly long time given his frame. Had more clear weaknesses than some of the guys around him (agreeing his PnR D was bad, and obviously FTs), but his strengths are comparably strong. Constantly piled up wins everywhere through his prime.

Considered Magic over Shaq, but not quite, wish he was. Bird/Curry I don't consider all that close to the top of my ballot currently, just terrible longevity compared to the top 2. I'm currently leaning Bird a bit over Curry, but there are a handful of guys not on the ballot yet that could beat them out for me (Mikan/Oscar/West/Kobe/Dirk all up for serious consideration).

Mikan keeping my nomination for now, I think Oscar my #2. Mikan didn't have that short of career (8 seasons by project rules), and he was probably the best player in the world each of those 8 seasons, sometimes by huge margins. The other potential nominees are fighting for more than a season or two, and never by even close to such a margin.

[LukaTheGOAT]

I found this RAPM site saved on my phone do the numbers seem legitimate ?(as they don’t seem to match up with J.E. Or Cheemas though it does seem closer in line with ESPNs RAPM)
https://web.archive.org/web/20150408042813/http://stats-for-the-nba.appspot.com:80/

Edit this apparently is Englemans page though it does seem quite different than a different RAPM spreadsheet I have of his.

https://www.apbr.org/metrics/viewtopic.php?t=8067

Not real RAPM, rather it's an estimate.

[Bklynborn682]

I found this RAPM site saved on my phone do the numbers seem legitimate ?(as they don’t seem to match up with J.E. Or Cheemas though it does seem closer in line with ESPNs RAPM)
https://web.archive.org/web/20150408042813/http://stats-for-the-nba.appspot.com:80/

Edit this apparently is Englemans page though it does seem quite different than a different RAPM spreadsheet I have of his.

https://www.apbr.org/metrics/viewtopic.php?t=8067

Not real RAPM, rather it's an estimate.

Appreciate it, I was wondering why it didn’t seem to match up closely/if at all with any of my other RAPM Files.

[trelos6]

Considering Steph is already nominated, and I've been suggesting Kobe for the last few rounds, I thought it was a good time to breakdown the Steph v Kobe comparison.

By my estimation, Steph has the higher peak, with 1 GOAT tier season compared to Kobe's 0.

They both have 8 seasons as an arguable top 3 player in the NBA.

Steph has 9 All NBA quality seasons, whereas Kobe has 15. Again, Steph has 9 All-Star quality seasons, and Kobe had 16.

Kobe also had 6 All-D worthy seasons, compared to 0 from Steph.

Onto what they're both known for, scoring. Curry's phenomenal '16 campaign was 31.9 pp75 at +12.8 rTS%. Simply incredible. He also had a brilliant 29.9 pp75 on +11.9% in '18. His 3 yr PS peak was 28.8 pp75 on +8.3%. so obviously a decline from the lofty heights of his RS, but still absolutely fantastic numbers. Also, the team rOrtg was fantastic those seasons, all above a 5, peaking at 8.1 in '16.

Kobe, had his volume peak in '06 with 34.2 pp75 on +2.4 rTS%, and his 3 yr PS from '08-'10 was 30.4 pp75 on +3.9%. He was certainly a very resilient scorer. The team rOrtg was generally quite good, though not as good as the Warriors, with ratings around +5, except the '10 season where it was +1.2 (that team hung its hat on D).

Looking at creation, Steph had an adjusted creation around 16 in his peak, with a passer rating of 8.2. Kobe on the other hand, had an adjusted creation of around 10 with a passer rating of 6.7.

So it's clear, Steph was the superior scorer, creator, but Kobe was the better defender. Kobe also has the longevity. The question remains, is Steph's peak enough to surpass the extra quality seasons from Kobe?

[lessthanjake]

Some data on top players’ average placement in the league in various impact metrics in the last decade. I’ve included Jokic, Giannis, and Embiid’s average placement in the last five years, just for reference as to how they’ve done in these metrics more recently—obviously if we looked at the whole decade they’d look worse.

Average League Placement in EPM in the last decade (excluding injury seasons)

1. Steph Curry: 2.89
2. Nikola Jokic (last 5 years): 4.80
3. Giannis Antetokounmpo (last 5 years): 5.00
4. Kawhi Leonard: 5.50
5. Joel Embiid (last 5 years): 6.00
6. LeBron James: 6.80
7. Kevin Durant: 9.38
8. James Harden: 9.70
9. Chris Paul: 12.2

Average League Placement in RPM in the last decade (excluding injury seasons)

1. Steph Curry: 4.56
2. Nikola Jokic (last 5 years): 5.80
3. Giannis Antetokounmpo (last 5 years): 6.00
4. LeBron James: 6.60
5. James Harden: 10.30
6. Joel Embiid (last 5 years): 13.20
7. Chris Paul: 13.60
8. Kevin Durant: 15.62
9. Kawhi Leonard: 17.75

Average League Placement in full season RAPTOR in the last decade (excluding injury seasons)

1. Steph Curry: 3.56
2. Nikola Jokic (last 5 years): 3.80
3. Kawhi Leonard: 3.88
4. Joel Embiid (last 5 years): 6.80
5. Giannis Antetokounmpo (last 5 years): 7.00
6. LeBron James: 10.9
7. Kevin Durant: 12.5
8. James Harden: 12.7
9. Chris Paul: 18.2

Average League Placement in NBAshotcharts RAPM in the last decade (excluding injury seasons)

1. Steph Curry: 8.11
2. Giannis Antetokounmpo (last 5 years): 12.8
3. Kawhi Leonard: 13.63
4. Kevin Durant: 13.63
5. Joel Embiid (last 5 years): 17.4
6. Chris Paul: 24.3
7. Nikola Jokic (last 5 years): 31.6
8. James Harden: 32.5
9. LeBron James: 34.1

Average League Placement in LEBRON in the last decade (excluding injury seasons)

1. Giannis Antetokounmpo (last 5 years): 2.00
2. Steph Curry: 5.11
3. Nikola Jokic (last 5 years): 5.20
4. Joel Embiid (last 5 years): 6.00
5. LeBron James: 6.90
6. Kawhi Leonard: 7.00
7. James Harden: 8.40
8. Chris Paul: 10.40
9. Kevin Durant: 10.75

Average League Placement in AuPM/g in the last decade (excluding injury seasons)

1. Steph Curry: 3.78
2. Nikola Jokic (last 5 years): 4.6
3. Giannis Antetokounmpo (last 5 years): 5.4
4. Kevin Durant: 6.13
5. Joel Embiid (last 5 years): 7.40
6. James Harden: 11.5
7. LeBron James: 13.70
8. Kawhi Leonard: 13.75
9. Chris Paul: 17.7

Steph has the best average placement in every single one of these impact metrics, except one. And that’s even above just the last 5 years of Jokic, Giannis, and Embiid—obviously it’s harder to maintain such league placement over a longer timespan. In the one measure where Steph’s average placement isn’t the best of anyone’s, he’s 2nd to Giannis, but of course I’m just listing Giannis’s last 5 years—if we extended it back even just a few years to start at Giannis’s 3rd season, he’d be below Steph in that too.