2021-22 NBA Season Discussion

[Colbinii]

Griffin finished 3rd in MVP because Paul missed time and he thrived once he was able to take on more of a playmaking role.

For comparison, here's how Griffin did in 2013-14 in the games with Paul compared to without.

With Paul: 24.3 points, 3.9 assists & 3.0 TO per 75 possessions shooting 57.6% TS with an On/48 of 9.9.
W/out Paul: 29.2 points, 4.5 assists & 2.7 TO per 75 possessions shooting 60.3% TS with an On/48 of 8.5.

To me it's long seemed pretty clear that adding Paul was awful for allowing Griffin to become the most valuable player he could possibly be...but since Paul was still the better player and the team was better off with him, it made sense to acquire him and let him run the offense how he seemed fit if you did acquire him.

Of course, it certainly would have been smarter, in retrospect, to try to acquire Curry, who could have allowed Griffin to develop on-ball while making life easier for Griffin with spacing.

I'm not really sure anything you are saying is counteracting what I am saying.

I already stated Draymond wouldn't be as useful next to CP3 as he is Curry on Offense but also disagreed with the notion that a player like CP3 would marginalize a basketball savant like Draymond Green or that Draymond Green would be booty.

The think is, Griffin was still an all-nba level player even with Chris Paul just as Draymond would be an all-nba level player next to Chris Paul.

Do you think of Wade in the same vein as CP3 then?

Both played next to stars [Griffin and LeBron] who seemingly had much better statistics without the co-star than with the co-star.

There is going to be give and take between every two teammates but something we have seen is when all-time great players team up, they have success. Some end up fitting better than others but that shouldn't mean we assume players of the caliber of Draymond all-of-a-sudden become completely marginalized next to a different superstar with a differing skill-set.

Well, I just pointed out how Griffin was marginalized by Paul, so I think we are in disagreement here.

And I think it would be considerably worse on offense for Green. Griffin off-ball still scares the hell out of defenses, Green does not.

Would Green be All-NBA next to Paul? He might based on his defense, but is that really your point? I would think it would be clear that neither Curry nor Paul would affect Green's defense. Offensively though, Green has a much bigger role to play when he's not playing with a ball dominant point guard.

Re: Wade similar to CP3? Well in the sense they are both on-ball, yes, but I'm not saying they are all that similar beyond that.

And listen, I think you and everyone else knows I'm a Nash guy and a Magic guy. The same holds true for all of them. There's only 1 ball, so if you're best suited to dominating the ball, then Green is not going to be a great offensive fit next to you.

Re: shouldn't assume marginalized next to a player with a different skillset. Disagree. Player fit is a real thing. It's one thing to argue that a particular player combo would work even if superficially it seems like it wouldn't, and quite another thing to not bother looking at fit at all when considering how to build a team.

Good post, but marginalized isn't the term to describe how Griffin was next to CP3 as Griffin was still a very significant offensive player next to CP3 and when he [Griffin] had the ball in his hands.

[falcolombardi]

Well, I just pointed out how Griffin was marginalized by Paul, so I think we are in disagreement here.

And I think it would be considerably worse on offense for Green. Griffin off-ball still scares the hell out of defenses, Green does not.

Would Green be All-NBA next to Paul? He might based on his defense, but is that really your point? I would think it would be clear that neither Curry nor Paul would affect Green's defense. Offensively though, Green has a much bigger role to play when he's not playing with a ball dominant point guard.

Re: Wade similar to CP3? Well in the sense they are both on-ball, yes, but I'm not saying they are all that similar beyond that.

And listen, I think you and everyone else knows I'm a Nash guy and a Magic guy. The same holds true for all of them. There's only 1 ball, so if you're best suited to dominating the ball, then Green is not going to be a great offensive fit next to you.

Re: shouldn't assume marginalized next to a player with a different skillset. Disagree. Player fit is a real thing. It's one thing to argue that a particular player combo would work even if superficially it seems like it wouldn't, and quite another thing to not bother looking at fit at all when considering how to build a team.

marginalizing sounds like griffin went from star to 3 and D specialist which is a bit hyperbolic for the second star of that clippers team

that would be like saying that jordan coming back to bulls marginalized pippen by reducing his offensive role

a guy going from his team only star to having a co-star is naturally gonna have some degree of prominence reduced, it happened to even curry when durant came

What word would you prefer me to use to say "Player A's presence resulted in Player B having lower primacy and lower effectiveness"?

dunno, i jusy think marginalizing gives people the idea of a player being wasted in a role way belpw their capabilities, the equivalent of using curry as a spot up shooter who never leaves the corner or handles the ball

a player having a bit leas touches and shots when he plays with talent rather than as the only focal point is not what i would call it tho

if prime klay went to be the franchise player of a loterry team and took enough shots to average 27 points or whatever would that mean curry was marginalizing him?

was pippen marginalized by jordan? did lebron get marginalized when he took less shots and was less heliocentric in miami? did lebron marginalize kyrie by causing him to take like 2 shots less per game?

in a odd sense, this point makes it sound like you advocate for extreme heliocentrism

[jalengreen]

draymond is such a great fit with curry IMO which, as a fan of lillard, makes me quite jealous because i think he's exactly the type of player the dame/cj blazers needed and makes more sense with dray than a cp3 or ja.

Honestly, the way Dame has actually played to me is more of an on-ball guy like CP3 or Ja. Are you saying that Dame doesn't play like that or that he wouldn't have played like that if he had Dray?

oh yeah, dame is definitely more of an on-ball guy

but i do think we've seen him play off-ball.. not as much as steph but more than other on-ball PGs imo..the passing of jusuf nurkic and evan turner was definitely utilized by stotts offenses, especially because dame didn't truly develop as a playmaker until the past few years (the dame that got swept by the pelicans was a far worse playmaker than the dame of the past three seasons IMO).

but i saw situations far too often where dame would get doubled in the backcourt, pass it to evan turner to run a mini 4 on 3 around the top free throw line, just for the opportunity to be wasted. a situation where we see the warriors offense thrives in

and that's ignoring the obvious defensive boon to portland teams that desperately needed it

i definitely do think dame could play in more of a designed off-ball role, though. not that he'd be steph though

[eminence]

'15, '16, '21, '22 - Full combined seasons

Curry/Dray on: 8707 minutes, 116.9 Ortg, 103.0 Drtg, +13.9 Net
Curry on/Dray off: 2893 minutes, 113.3 Ortg, 108.5 Drtg, +4.8 Net
Dray on/Curry off: 2015 minutes, 108.7 Ortg, 103.9 Drtg, +4.8 Net
Curry/Dray off: 4574 minutes, 104.8 Ortg, 109.6 Drtg, -4.8 Net

I appreciate they have the same solo Net rating over their period as the two clear leaders of the squad.

[Doctor MJ]

i think that doesnt really answer the question






the convo stars with a comment that curry is not in giannis or jokic level and you answer asking if they could play curry role which is completely unrelated to how good curry is now in relation to giannis and jokic

giannis cannot play like curry but cury obviouslt cannot play like giannis

giannis also cannot play like damian lillard either but that doesnt mean we cannot compare them, curry cannot replace jarret allen role in cleveland either but that doesnt mean we cannot say curry is better

Hmm, you're asking what my point is when I ask questions that are relevant to the post I was replying to. I don't think I need a "point" for that.

But I can speak to what you're bringing in here. I'm not saying we cannot compare guys in different roles - in fact I insist we should make such comparisons too because that's what coaches & GMs do - but the comparison is definitionally quite hard and in general I find that the challenge is not finding arguments for the two sides, but determining how to choose one over the other when neither can dominate the comparison (by being superior in all situations). So when someone says they struggle to see any argument for why the guy on the more successful team, with the more impressive +/-, in a completely different role, could conceivably be ranked ahead of the other guy, I wonder how they've come to the conclusions they've come to.

i'm not sure i agree with this

+/-
giannis +8.0
steph +10.1

on/off
giannis +11.0
steph +10.0

postseason +/-
giannis +7.7
steph +7.3

postseason on/off
giannis +28.8
steph +3.4

EPM
giannis +7.3
steph +7.2

LEBRON
giannis +6.54
steph +4.40

i'm not the biggest fan of +/- or single-season impact metrics due to the standard errors, but i must say i think giannis' profile here is more impressive *and* i think he's been more impressive in the postseason (wasn't a huge fan of the giannis DPOY arguments because i felt he coasted more on that end than in years past).

unless by more impressive +/- you just meant basic +/- itself and nothing else. in which case, yeah, but i'm not entirely sold on how meaningful that is.

So first, I was focusing on the regular season. If you'd like to make a playoff specific argument you certainly can.

In terms of what numbers I'm referring to, I'd tend to point to raw +/- in comparison to teammates and RAPM.

By raw, here's the top 3 on GS:

Curry +507
Poole +324
Wiggins +265

And on the Bucks:

Holiday +426
Giannis +397
Portis +314

So that's telling that as a first pass, Curry has a much bigger separation from his teammates this year than Giannis.

Go over to NBA Shot Charts' RAPM and looking at the same 3 guys from each team:

Curry +3.67
Poole +1.88
Wiggins +0.62

Holiday +3.41
Giannis +2.42
Portis +1.21

And the regressed metric tells us basically the same thing.

The stuff you're looking at adds more stuff into the mix including a thing that gets called "Luck-Adjustment" that in my experience assumes that a player has no effect on his teammates' ability to make 3's at a higher percentage...which just isn't true.

But let me emphasize again:

I'm not saying this data represents how the players will always be or how they are at their apex.

And I'll also emphasize that in our POY polls, I'm literally the only person who has given my #1 vote to Giannis for both POY and DPOY twice (in '18-19 & '20-21), so I'm not anti-Giannis.

What I'm getting at is that I've seen a lot of people over the course of this season dismiss Curry as not playing that great while talk about Giannis being dominant on both ends of the floor, and I get why their eyeballs tell them this...but impact is sneaky - easy to miss, and too easily dismissed imho.

[Doctor MJ]

marginalizing sounds like griffin went from star to 3 and D specialist which is a bit hyperbolic for the second star of that clippers team

that would be like saying that jordan coming back to bulls marginalized pippen by reducing his offensive role

a guy going from his team only star to having a co-star is naturally gonna have some degree of prominence reduced, it happened to even curry when durant came

What word would you prefer me to use to say "Player A's presence resulted in Player B having lower primacy and lower effectiveness"?

dunno, i jusy think marginalizing gives people the idea of a player being wasted in a role way belpw their capabilities, the equivalent of using curry as a spot up shooter who never leaves the corner or handles the ball

a player having a bit leas touches and shots when he plays with talent rather than as the only focal point is not what i would call it tho

if prime klay went to be the franchise player of a loterry team and took enough shots to average 27 points or whatever would that mean curry was marginalizing him?

was pippen marginalized by jordan? did lebron get marginalized when he took less shots and was less heliocentric in miami? did lebron marginalize kyrie by causing him to take like 2 shots less per game?

Not sure what to do with the feeling of hyperbole in the absence of another term to use in its place.

I'll tell you one reason I push back on Griffin: Because he chafed next to Paul. This wasn't just a situation where Griffin could have been more of an MVP candidate had Paul not been there. This was a situation where it was Griffin's team, then they brought Paul in and gave him control of the offense, and the relationship between Paul & Griffin didn't ever work with a lot of positivity.

Now, you might say this is less about on-court fit, and more about Griffin disliking the way Paul criticized him, but it's hard to know because both things are true.

By contrast, one could argue that "marginalized" isn't the best word for what Durant did to Curry because Curry chose to let that happen to make Durant happy.

On the other:

I think you can make arguments in most of these cases that someone was getting marginalized to some degree, and I'd just point back to the question of what word to use if we can't use "marginalized".

[falcolombardi]

What word would you prefer me to use to say "Player A's presence resulted in Player B having lower primacy and lower effectiveness"?

dunno, i jusy think marginalizing gives people the idea of a player being wasted in a role way belpw their capabilities, the equivalent of using curry as a spot up shooter who never leaves the corner or handles the ball

a player having a bit leas touches and shots when he plays with talent rather than as the only focal point is not what i would call it tho

if prime klay went to be the franchise player of a loterry team and took enough shots to average 27 points or whatever would that mean curry was marginalizing him?

was pippen marginalized by jordan? did lebron get marginalized when he took less shots and was less heliocentric in miami? did lebron marginalize kyrie by causing him to take like 2 shots less per game?

Not sure what to do with the feeling of hyperbole in the absence of another term to use in its place.

I'll tell you one reason I push back on Griffin: Because he chafed next to Paul. This wasn't just a situation where Griffin could have been more of an MVP candidate had Paul not been there. This was a situation where it was Griffin's team, then they brought Paul in and gave him control of the offense, and the relationship between Paul & Griffin didn't ever work with a lot of positivity.

Now, you might say this is less about on-court fit, and more about Griffin disliking the way Paul criticized him, but it's hard to know because both things are true.

By contrast, one could argue that "marginalized" isn't the best word for what Durant did to Curry because Curry chose to let that happen to make Durant happy.

On the other:

I think you can make arguments in most of these cases that someone was getting marginalized to some degree, and I'd just point back to the question of what word to use if we can't use "marginalized".

can we agree that a (offensive) player is always gonna be more prominent the most he is asked to do and the less talent he plays with but almost by definition will have a smaller role if more (offensive) talent joins his team?

there is a "normal"or even ideal amount of role reduction by playing with better talent, it happened to jordan from the 80's into the 90's, as his team improved he took less shots and let pippen handle the ball more and the bulls imrpoved with it, was he better off gunning for 37 a game in a offense deprived team?

griffing taking less shots or handling the ball less playing with another star compared to when playing with no other star is the expected thingh, i dont think it deserves the negative connotation

if you argue instead that paul caused griffin contributions to reduce more than the other league stars would have that is a valid discussion

[jalengreen]

So first, I was focusing on the regular season. If you'd like to make a playoff specific argument you certainly can.

In terms of what numbers I'm referring to, I'd tend to point to raw +/- in comparison to teammates and RAPM.

By raw, here's the top 3 on GS:

Curry +507
Poole +324
Wiggins +265

And on the Bucks:

Holiday +426
Giannis +397
Portis +314

So that's telling that as a first pass, Curry has a much bigger separation from his teammates this year than Giannis.

so this can certainly provide some insight onto a team/its players on its own and i appreciate you bringing it up

something interesting to consider, though

the problem with this, as you know, is that it doesn't account for something like rotations. of course you said you were talking about raw +/- so that's fine, i agree based on raw +/- that steph's numbers are better then giannis'

Go over to NBA Shot Charts' RAPM and looking at the same 3 guys from each team:

Curry +3.67
Poole +1.88
Wiggins +0.62

Holiday +3.41
Giannis +2.42
Portis +1.21

And the regressed metric tells us basically the same thing.

The stuff you're looking at adds more stuff into the mix including a thing that gets called "Luck-Adjustment" that in my experience assumes that a player has no effect on his teammates' ability to make 3's at a higher percentage...which just isn't true.

this is also interesting because it raises the question of how much value should we place onto single-season raw RAPM?

something i found from calculating RAPM myself is that the standard errors on single-season RAPM are very, very high. perhaps higher than some people would think.

here's an example that shows just how large the confidence intervals are for single-season RAPM from https://squared2020.com/2019/10/03/exercising-error-quantifying-statistical-tests-under-rapm-part-iv/

Code: Select all

"Comparing this to the rest of the league, we see that 39 other players fit within this confidence bound, indicating that despite being the “league leader,” Danny Green really identifies within the Top 39 players in the league. This is a “best case scenario” for identifiability. In fact, if we grab the 200th player in the league, Kyle Korver, we find that his confidence interval is [ -1.21, 4.16]. This indicates that Korver is equivalent to 460 other players in the league; ranging from Giannis Antetokounmpo (6th) to Damyean Dotson (464th)."

it's quite difficult to isolate the impact of individual players in smaller samples like a single regular season. which is why for single-season purposes, impact metrics like EPM and LEBRON are generally preferrable. IMO at least, i guess, although i'd add that there's generally a consensus on that:

Read on Twitter

the luck-adjustment point is a good one. but i do think it's reasonable to deal with the issues of small-sample size. and in my experience, they don't completely assume that a player has no effect on their teammates' ability to hit 3s, but they do adjust it.

for example, PIPM used a bayesian approach to adjusting teammate shooting:

Code: Select all

Leonard’s teammates shot 780 3-pointers while he was on the floor and they made them at a 41.8 percent clip. Overall during the season, Spurs players not named Kawhi shot 1533 3s and made them at a 39.5 percent clip. The following regression is used to reach the luck-adjusted 3-point total for Kawhi: [ (780 x 41.8%) + (1533 x 39.5%) ] / (780 + 1533) = 40.3%. This regressed total allows for the individual’s impact to be felt while limiting noise and small sample size issues. In reality, Kawhi’s teammates made 326 3s when he was on-court, but the total drops slightly to 314.3 made 3s after adjusting for luck. Thus, his teammate 3-point adjustment would be -35.1 (3 time 314.3 minus 326).

in larger samples the adjustment would be smaller because there can be more confidence in the player actually having that impact.. in smaller samples there is less such confidence so the adjustment would be greater.

there are other forms of luck-adjustment too, like free throw shooting. which i think is also reasonable. if player X's teammates shoot 5% better from the free throw line in the 2020 season with him on the floor, are we really sure that has anything to do with player X? or would we rather adjust for that and decrease the variance in the regression estimates?

it should also be stated that these adjusted metrics perform better than RAPM in tests.

And I'll also emphasize that in our POY polls, I'm literally the only person who has given my #1 vote to Giannis for both POY and DPOY twice (in '18-19 & '20-21), so I'm not anti-Giannis.

What I'm getting at is that I've seen a lot of people over the course of this season dismiss Curry as not playing that great while talk about Giannis being dominant on both ends of the floor, and I get why their eyeballs tell them this...but impact is sneaky - easy to miss, and too easily dismissed imho.

i think this discussion has been very reasonable, i would certainly not call you anti-giannis for asking me to defend my stance

[eminence]

LAC '12-'17 - 6 full combined seasons, 25700 minutes

CP3/Griffin on: 11944 minutes (46.5%), 115.9 Ortg, 106.1 Drtg, +9.8 Net
CP3 on/Griffin off: 3891 minutes (15.1%), 112.9 Ortg, 103.6 Drtg, +9.3 Net
Griffin on/CP3 off: 3326 minutes (12.9%), 108.7 Ortg, 107.3 Drtg, +1.4 Net
CP3/Griffin off: 6539 minutes (25.4%), 101.3 Ortg, 107.0 Drtg, -5.7 Net

GSW '15, '16, '21, '22 - 4 full combined seasons, 18189 minutes

Curry/Dray on: 8707 minutes (47.9%), 116.9 Ortg, 103.0 Drtg, +13.9 Net
Curry on/Dray off: 2893 minutes (15.9%), 113.3 Ortg, 108.5 Drtg, +4.8 Net
Dray on/Curry off: 2015 minutes (11.1%), 108.7 Ortg, 103.9 Drtg, +4.8 Net
Curry/Dray off: 4574 minutes (25.1%), 104.8 Ortg, 109.6 Drtg, -4.8 Net

I think there's something to CP3/Griffin being notably less synergistic than Steph/Dray. Not having looked at enough large sample star pairings I'm unsure we should lean towards this saying that Steph/Dray have especially notable synergy, or CP3/Griffin have more negative synergy than most. Probably a bit of both.

[MartinToVaught]

Unsurprisingly, I have to agree that CP3 marginalized Blake. Not going to lie, it stings a little bit to look at Blake's 2018-19 season in Detroit and realize that he could have been doing that earlier in his career, before all the injuries finally caught up to him, if he hadn't been stuck watching CP3 spend 90% of the game dribbling and the other 10% deferring to Jamal and JJ. (Before the CP3 fans descend on me again, I'm of course being hyperbolic here, but the point remains.)

It's noteworthy to me that Ayton has been having similar problems with CP3 on the Suns all these years later - worse in his case because he's had the contract situation and the stigma of being drafted over Luka hanging over his head, which Blake didn't have to deal with. Granted, there's legitimate questions about Ayton's maturity and toughness. But it also could be that CP3 simply needs Chandler/DJ/Capela type big men who are just there on offense to get spoonfed, and he struggles to work with big men who have a different skillset and need the ball more. David West would be the exception here, but he was a late first-round pick who was happy to take a backseat to CP3, while Blake and Ayton were both #1 picks and hungry to prove they were worth it. I honestly can't say I blame Ayton for getting frustrated with being a background character on the Chris Paul Show all season when Sarver was already cheaping out on his contract.

[falcolombardi]

LAC '12-'17 - 6 full combined seasons, 25700 minutes

CP3/Griffin on: 11944 minutes (46.5%), 115.9 Ortg, 106.1 Drtg, +9.8 Net
CP3 on/Griffin off: 3891 minutes (15.1%), 112.9 Ortg, 103.6 Drtg, +9.3 Net
Griffin on/CP3 off: 3326 minutes (12.9%), 108.7 Ortg, 107.3 Drtg, +1.4 Net
CP3/Griffin off: 6539 minutes (25.4%), 101.3 Ortg, 107.0 Drtg, -5.7 Net

GSW '15, '16, '21, '22 - 4 full combined seasons, 18189 minutes

Curry/Dray on: 8707 minutes (47.9%), 116.9 Ortg, 103.0 Drtg, +13.9 Net
Curry on/Dray off: 2893 minutes (15.9%), 113.3 Ortg, 108.5 Drtg, +4.8 Net
Dray on/Curry off: 2015 minutes (11.1%), 108.7 Ortg, 103.9 Drtg, +4.8 Net
Curry/Dray off: 4574 minutes (25.1%), 104.8 Ortg, 109.6 Drtg, -4.8 Net

I think there's something to CP3/Griffin being notably less synergistic than Steph/Dray. Not having looked at enough large sample star pairings I'm unsure we should lean towards this saying that Steph/Dray have especially notable synergy, or CP3/Griffin have more negative synergy than most. Probably a bit of both.

draymond having most of his impact come from defense helps with this as defense doesnr have the same diminishing returns offense does

but even then dray alone has better results than griffin alone suggesting griffing as a first option in the most prominent role possible may be the ideal to maximize blake stats but not a team offense

draymond without curry still had the same offense than griffin without paul

notice however how the offense still inproves 3 points adding griffinh to paul, while getting 2.5 points worse on defense (maybe by playing other weaker defenders in griffin/paul lineups, not nevesarrily blaming griffin there)

improving a 112 (+4 for the era) offense by 3 points (+7) is actually very damn good, and a clear case of offense ceiling raising, comparable to what happens with draymomd and curry

[falcolombardi]

Unsurprisingly, I have to agree that CP3 marginalized Blake. Not going to lie, it stings a little bit to look at Blake's 2018-19 season in Detroit and realize that he could have been doing that earlier in his career, before all the injuries finally caught up to him, if he hadn't been stuck watching CP3 spend 90% of the game dribbling and the other 10% deferring to Jamal and JJ. (Before the CP3 fans descend on me again, I'm of course being hyperbolic here, but the point remains.)

It's noteworthy to me that Ayton has been having similar problems with CP3 on the Suns all these years later - worse in his case because he's had the contract situation and the stigma of being drafted over Luka hanging over his head, which Blake didn't have to deal with. Granted, there's legitimate questions about Ayton's maturity and toughness. But it also could be that CP3 simply needs Chandler/DJ/Capela type big men who are just there on offense to get spoonfed, and he struggles to work with big men who have a different skillset and need the ball more. David West would be the exception here, but he was a late first-round pick who was happy to take a backseat to CP3, while Blake and Ayton were both #1 picks and hungry to prove they were worth it. I honestly can't say I blame Ayton for getting frustrated with being a background character on the Chris Paul Show all season when Sarver was already cheaping out on his contract.

i dont think griffin was good enough offensively to have that kind of role

eminence posted stats for it and the offense with griffin but not cp3 was fairly mediocre over the years

LAC '12-'17 - 6 full combined seasons, 25700 minutes

CP3/Griffin on: 11944 minutes (46.5%), 115.9 Ortg, 106.1 Drtg, +9.8 Net
CP3 on/Griffin off: 3891 minutes (15.1%), 112.9 Ortg, 103.6 Drtg, +9.3 Net
Griffin on/CP3 off: 3326 minutes (12.9%), 108.7 Ortg, 107.3 Drtg, +1.4 Net
CP3/Griffin off: 6539 minutes (25.4%), 101.3 Ortg, 107.0 Drtg, -5.7 Net

griffin "alone"= average offense (+1~)
paul "alone" = really good offense (+5~)
griffin + paul = really fantastic offense (+8~)

which to be fair may partially be the fault of doc running so much offense through crawford

[eminence]

.

it should also be stated that these adjusted metrics perform better than RAPM in tests.

A bit of an aside, but something I've been stewing on awhile and trying to think of a decent way to analyze.

I don't believe the above statement to hold much usefulness for questions we are usually asking here on the PC board (it would hold true for something like a fantasy draft question, or maybe if one leaned very heavily hypothetical in their CORP style lists).

xRAPM variants (EPM and the like) outperform RAPM in predictive tests (useful for FO people and gamblers), personally when I (and I think Doc, maybe others) am doing something like a POY vote I'm trying to describe the past season/seasons. I haven't seen any evidence that xRAPM variants are better at that than RAPM (or even just APM).

[MartinToVaught]

which to be fair may partially be the fault of doc running so much offense through crawford

Bingo. Any analysis that doesn't factor in the worst coach in sports history is flawed.

I do think it's worth pointing out that the season Blake finished top 3 in MVP voting was the one where CP3 missed over a month of the season with an injury and Blake kept us afloat as the first option without him.

[jalengreen]

.

it should also be stated that these adjusted metrics perform better than RAPM in tests.

A bit of an aside, but something I've been stewing on awhile and trying to think of a decent way to analyze.

I don't believe the above statement to hold much usefulness for questions we are usually asking here on the PC board (it would hold true for something like a fantasy draft question, or maybe if one leaned very heavily hypothetical in their CORP style lists).

xRAPM variants (EPM and the like) outperform RAPM in predictive tests (useful for FO people and gamblers), personally when I (and I think Doc, maybe others) am doing something like a POY vote I'm trying to describe the past season/seasons. I haven't seen any evidence that xRAPM variants are better at that than RAPM (or even just APM).

that's a fair point

however i do believe that the lower variance of xRAPM metrics is more representative of player ability & impact than RAPM itself (in terms of how much player ability changes season to season)

i also don't think descriptive vs predictive are two unlinked things. you need accurate information from the past to predict the future well and i do think you would expect more descriptive metrics from the past to also be more predictive for the future

[falcolombardi]

which to be fair may partially be the fault of doc running so much offense through crawford

Bingo. Any analysis that doesn't factor in the worst coach in sports history is flawed.

I do think it's worth pointing out that the season Blake finished top 3 in MVP voting was the one where CP3 missed over a month of the season with an injury and Blake kept us afloat as the first option without him.

keeping afloat a team for a month =/= doing a good enough job to merit being the first option, pippen could rise his offense as the lone star and keep the bulls offense afloat (in fact he won mvp consideration foing exactly that) but that was obviously not ideal compared to runninh it through jordan instead and have pippen be more of a facilitator

i dont have anythingh against griffin i think is impressive he kept the offense afloat with a team composed of crawford (who i love aesthetically and seems a great person but was not a guy i wanted on my team even on offense alone) deandre (solid defender at his best but not a strong offensive player), etc

but paul (+5 offense without blake) could do the same thingh a lot better than griffin ( +1 offensen without paul) and playing them both (+8 offense) clearly was a lot more effective than running the offense through blake

the clippers offense, again, was -fantastic- with griffin and paul, their problems were not there, but in their weak defense amd lack of useful depth

did doc run more offense through jamal with paul off the court and griffing on than otherwise? that would actuslly be really useful to know as i think crawford may have been a net negative in offense

[Texas Chuck]

none of them isolate impact nearly as well as they purport. Not sure why they continue to be relied upon so heavily especially in small samples. Especially when sometimes we don't rely on them at all if they give us numbers that don't fit what we've already decided.

I don't understand why we can't accept there is no single-metric available to tell us how good a basketball player is or isn't. That we have to work harder, and look beyond just someone's attempt at an all-in-one metric. Nothing against the creators of those who I'm sure all all brilliant stats guys doing their best, but its too dynamic and the samples sizes with the same people on the court are just way too small for us to say yep, this guy is the best or even yep this guy is the best in his role.

And if its really is as simple as just who has the best +/- then we should just shut this subforum down, because you can just look at the numbers and the job is done for you.

[parsnips33]

none of them isolate impact nearly as well as they purport. Not sure why they continue to be relied upon so heavily especially in small samples. Especially when sometimes we don't rely on them at all if they give us numbers that don't fit what we've already decided.

I don't understand why we can't accept there is no single-metric available to tell us how good a basketball player is or isn't. That we have to work harder, and look beyond just someone's attempt at an all-in-one metric. Nothing against the creators of those who I'm sure all all brilliant stats guys doing their best, but its too dynamic and the samples sizes with the same people on the court are just way too small for us to say yep, this guy is the best or even yep this guy is the best in his role.

And if its really is as simple as just who has the best +/- then we should just shut this subforum down, because you can just look at the numbers and the job is done for you.

Even if there was a single stat that totally captured the value of any given player (like you said, there's not and doubt there ever will be one) it would be so much less interesting from a basketball fan discussion perspective than just talking about what we see in the game, even if we all acknowledge that our perspective is inherently limited and flawed.

[ardee]

'15, '16, '21, '22 - Full combined seasons

Curry/Dray on: 8707 minutes, 116.9 Ortg, 103.0 Drtg, +13.9 Net
Curry on/Dray off: 2893 minutes, 113.3 Ortg, 108.5 Drtg, +4.8 Net
Dray on/Curry off: 2015 minutes, 108.7 Ortg, 103.9 Drtg, +4.8 Net
Curry/Dray off: 4574 minutes, 104.8 Ortg, 109.6 Drtg, -4.8 Net

I appreciate they have the same solo Net rating over their period as the two clear leaders of the squad.

Where do you find these stats since BBR put their stat finder behind a paywall?

[MyUniBroDavis]

none of them isolate impact nearly as well as they purport. Not sure why they continue to be relied upon so heavily especially in small samples. Especially when sometimes we don't rely on them at all if they give us numbers that don't fit what we've already decided.

I don't understand why we can't accept there is no single-metric available to tell us how good a basketball player is or isn't. That we have to work harder, and look beyond just someone's attempt at an all-in-one metric. Nothing against the creators of those who I'm sure all all brilliant stats guys doing their best, but its too dynamic and the samples sizes with the same people on the court are just way too small for us to say yep, this guy is the best or even yep this guy is the best in his role.

And if its really is as simple as just who has the best +/- then we should just shut this subforum down, because you can just look at the numbers and the job is done for you.

I’m actually curious, what’s the justification for RAPM being more descriptive than other ones and for it being super descriptive in general?

I don’t think using all in one metrics alone is ever a full enough argument, you have to justify a why it’s like that and if you can’t or the reasoning is absurd then it isn’t a good argument

On predictive vs descriptive, if anything I’d think something that can predict impact for the next year would probably be more accurate at measuring what’s going on this year, or eliminating “noise”

But i agree that some takes with only impact data backing it up can be kinda crazy. Even if it does measure “impact”there are so many variables that can go into that as well that players don’t necessarily effect

Esp in the playoffs people obsess way too much over minute differences in teams SRS accross years because of a correlation between it and winning which should be obvious