Top 25 peaks of the 2001-25: #15-#16 Spots

[Owly]

I think when we’ve seen Nash in different systems and he’s SOOOOO much more impactful in one specific system that was tailored to him with an ahead of his time coach and a roster where he’s the only real playmaker than he was in other system, it makes sense to be a little skeptical. I think his poor defense and need to dominate the ball to show impact make it hard for him to scale up next to other elite talent. His actual value has to be somewhere between what he showed in Phoenix and what she showed in Dallas and LA.

Do you really think it’s just a super weird aging curve that made him elite from age 30-37 in Phoenix and not show any comparable impact before or after? We’ve seen Harden perform in a variety of roles and systems and be very effective. Nash probably has the biggest difference in performance I’ve seen for how a player did in one system vs. how he did in any other system, and he also has some of the biggest difference between his peak impact stats and his box stats. If there’s any player in NBA history where the context would make me question his peak impact stats, it’s Nash.

I don't have a dog in the fight, I was just wondering if there was a good reason for this ... I'm now even more doubtful.

Maybe I'm wrong in being persuaded that Nash was the core and the mind of that Phoenix system.

So my first instinct questions from this would be

Are you curving towards playoff success, deep runs as many seem to be? I'm not tracking every voters preferences, so maybe not. A lot of the early voting trended towards title-team players.
If yes, then how does this fit with penalizing a player for doing well in positive circumstances?

On the lines of reasoning offered though

I think when we’ve seen Nash in different systems and he’s SOOOOO much more impactful in one specific system that was tailored to him with an ahead of his time coach and a roster where he’s the only real playmaker than he was in other system

How many systems are we counting here, just to understand the framing? A personal taste aside but I trend a little ... skeptical ... on CAPS to make a point in general. My understanding is the core of the "one specific system is ..." often argued as get the ball up quickly, let Nash make decisions. Arguably a positive that a player does well when trusted to make decisions. "Only real playmaker" seems a stretch with Johnson and then Diaw. On D'Antoni (beyond above regarding system) ... it was replicated with Gentry once they realized Shaq wasn't the route to go down. How high one wants to go on D'Antoni is up for grabs (and some of his doesn't have to be zero-sum ... trusting a smart player to do smart things is ... smart even if perhaps you may only be enabling by not impeding) ... but it may potential ring a little hollow from anyone using Harden and any of his box-side numbers under that coach.

I think his poor defense and need to dominate the ball to show impact make it hard for him to scale up next to other elite talent.

My recollection of the older impact stuff suggested defensive weaknesses were somewhat overstated (and are baked into IRL impact). And given his smarts I'd argue they're the sort of relative weaknesses that could be covered, blended well with good defenders, good coaching, perhaps a rim-protector. And whilst at the margin there may be some scaling stuff ... the idea that your point guard who's having terrific impact playing on the ball and making decisions is a bad thing ... sits a touch uneasy. It's also possible some might lodge these criticisms at Harden.

His actual value has to be somewhere between what he showed in Phoenix and what she showed in Dallas and LA.

Wait ... LA is a significant part of your assessment of Nash. And specifically Nash at his peak?

Do you really think it’s just a super weird aging curve that made him elite from age 30-37 in Phoenix and not show any comparable impact before or after?

You want it after 37?

On before ... I'd argue it's largely moot for peaks unless everyone's getting this systematically, but one can debate rules changes (and how one accounts for this across eras), versus being given the keys to the offense, versus commitment to conditioning (and some may say pride, "I'll show you" as a motivator), versus did Dirk's primacy mute his offensive impact to why his impact signal was so much weaker in Dallas. Still I am inclined towards he largely was the system in Phoenix and that succeeded so whilst it probably has to take away from potential career accumulated value ... I'm not sure how much Dallas matters to his peak.

We’ve seen Harden perform in a variety of roles and systems and be very effective.

Fairly fuzzy so it somewhat depends what you mean. I really liked how final year OKC Harden was effective and impactful whilst he wasn't being given a lot of opportunity in terms of primacy. Still beyond that ... and whilst defensive concerns for him too were sometimes overblown ... it's not like he didn't like holding on to the ball. And on his best in-strong-prime chance, whilst injuries were a major blow and he faced elite opponents ... I think it's generally perceived that he and Paul didn't mix well and then Harden sought to move Paul on for a worse fit and worse player. He's a great player and absolutely worthy of contending here. Fwiw, to my knowledge barring some mental gymnastics I don't think he does have Nash's pure impact-side profile at apex but I'm not an expert.


IDK, I just kind of hoped there was something ... deeper to why we'd be using clearly lower level years versus another player's actual peak.

At age 37 in Phoenix, Nash has much better impact numbers than he did in any of his 24-29 seasons in Dallas. At age 38 in Los Angeles, he’s completely washed and a shell of himself. Yes, I absolutely think that’s relevant!

Look at these 2 year RAPMs:

Last 2 years in Dallas (age 28-29): -0.6 (177th in the league)

First 2 years in Phoenix: (age 30-31): 6.1 (3rd in the league)

Last 2 years in Phoenix (age 36-37): 4.0 (15th in the league)

2 years in Los Angeles (age 38-39): -1.5 (357th in the league)

I don’t see how that doesn’t give you any pause. If his RAPM was negative immediately surrounding these Phoenix years where he’s always top 20, AND the box numbers and hybrid systems all suggest he’s not as good as the impact stats would suggest, I don’t see how you can just take them as gospel.

On 38 relevant to peak because age 37 is really good? No ones arguing 2012 as his peak. Nor a year close to it. It's not D'Antoni if the idea is he was magically making Nash better. Nash was already down at 31.6mpg in '12. The idea that his age 38 season in which he's hurt, on a new team, that's had significant other turnover, turns out to be somewhat of a dumpster fire and he isn't as entrusted as a decision maker where that's a (the?) primary asset he gives you ... I just don't see that.

On

I don’t see how that doesn’t give you any pause ...

I'm not sure about the use of tense or assumption that everyone is familiar with and using the exact dataset as you.

But to that data ... per above Lakers to assess him at peak makes no sense to me. What it comes down to is ... is Dallas relevant, and was the way it was used earlier transparent and fair?

And I do worry on the latter, the answer on the latter questions is "no."
Because this now really does look make it look like it was "if I deliberately spike down a four or five year spell where I know there are multiple better ones and compare it to my candidates best one, my candidate comes out on top", because it seems you aware that in these already large samples there were multiple better ones and are intentionally dragging it down.

Now is Dallas relevant ... that's a reasonable question and I think the answer is yes. I don't think the use applied was sensible but having a discussion about why his Dallas impact was relatively low if you're picking a nearby year seems fair enough. There have been discussions of that in terms of why his Dallas impact was so much lower and how much one wants to buy into those is up for grabs. So whilst it has been suggested

... the impact numbers are so bad that we can't even look at the numbers to evaluate Nash because it's somehow unfair.

This doesn't seem to reflect any actually given position that I've seen on Dallas. It's just that "look[ing] at" and "using against someone else's best" are very different things.

The one other thing I'd note regarding context of '04 specifically is that roster is considered somewhat of a mess. I think I'm higher on it than some as a team, but it's not exactly a sensibly balanced construction.

And briefly touching on a couple of downstream points.

I don’t think his impact was weaker in Dallas because he was used wrong. Don Nelson wasn’t exactly a fuddy duddy who didn’t want to let an offensive maestro cook. If anything I’d say the team’s ORtg was evidence that he was used pretty optimally.

...

Nash has some serious weaknesses especially defensively that make it hard for him to have elite impact unless the offense is completely and utterly dependent on him

I'm struggling with internal coherence here. He was used optimally but needs the offense dependent on him ... well then surely a substantial stagger would be optimal? And I'm not sure why a good offense would necessarily mean optimal use given it now seems, though elite, perhaps less than the sum of its parts.
Defensive weaknesses make it so that offenses need to be centered around him to have elite impact? I'm not sure how that would follow. I guess this is supposed to read that such is the cost of his defense, his offensive value only dwarfs it if the offense is "completely and utterly dependent on him". Yet actually this seems to make a construction around him really easy. Defensive pieces tend to be cheaper than offensive ones. Nash can have elite net impact when given the keys to the offense. And Nash can have an elite offense
'05: 120.3 Ortg on in a 106.1 league
06: 114.2 Ortg on in a 106.2 league
07: 118.6 Ortg on in a 106.5 league
08: 118.2 Ortg on in a 107.5 league
with an offense that is "completely and utterly dependent on him". If one believes this entirely, then one merely needs to put together good defenders strong enough to cover for Nash enough to not bleed too much of the enormous offensive advantage he generates and one would seem to have a contender. If not then there's a little more room for maneuver ... but those weren't my words (and he does seem to be driving very substantial impact).

Both of them can show superstar impact if they have a team that perfectly fits their strengths.

I'm open to being wrong here as I've said before but my impression and that of others here has been perfectly fitting his strengths seems primarily to be giving him (a point guard) the ball and trusting him (and fwiw Nash is in different tier of impact to Iverson).

but it’s hard to win a title with them because their value is muted when they play with enough talent for their team to actually contend.

Whilst some seem to assume the hip-check necessarily turned the outcome and then perhaps go further in terms of assumptions ... implicit in this statement is the idea that the Suns didn't "actually contend" (which is hardly a given) and "it's hard to win a title" with Nash ... presumably because it's hard to put together better defensive pieces around Nash than the Suns did. And, in fairness, those teams get too much heat for "bad" defense where the reality is more around league average and there are some pieces there: Bell and Marion first coming to mind. Still it doesn't seem that much of a stretch to imagine a better defensive build, especially where a team can depend on him for offense.

[f4p]

Whoa iggy -

Do you think it was Nash's goal out there to maximize his personal BPM? I assume the answer is No, but if it wasn't his goal, what exactly are you criticizing him for?

Do you think the Suns would have been better if he had prioritized optimizing his BPM instead of what he did? How?

No, of course I don't think he was trying to maximize a particular stat that was incredibly obscure at the time if it was even invented which I'm pretty sure it wasn't. I'm saying that if the impact stats in Phoenix said he had a top 15 peak this century, and the impact stats outside of Phoenix said he was pretty much an average starter, and every all-in-one measure with a box score component said he was never a top 5 player for any individual season, maybe his peak wasn't actually QUITE as good as all-time seasons from Westbrook and Harden once you account for context.

I'm going to direct this back to the second question because I think it's really important:

The implication of knocking a guy for not placing higher with a stat like BPM is that he was doing something wrong, and thus cries out for us to ask what that is.

For contrast here:

If you were knocking a player because he had too low of a PPG the implication is that he should have shot & scored more.
If you were knocking a player because he had too many turnovers is that he should have turned it over less.
If you were knocking a player because his RAPM was too low, the implication is that he should have been more impactful at helping his team do better on the scoreboard.

But when you do this with a box score all-in-one, it's unclear what exactly the criticism is, only a vague faith that if the player were better, he'd look better in this stat because...

Now to be clear, we can look at, say, a Nash vs Paul comparison and say that if Nash had less turnovers he'd have had all-in-one's more like Paul. But of course Nash could have had the same improvement in all-in-one simply by increasing scoring or rebounding or whatever sufficiently.

I would say that the traditional Hollinger-style analysis would thus respond by saying "Yeah, any of that. Anything more of the good and less of the bad would suffice."

But the hitch with this thinking is that it would alter Nash's process. Whatever's Nash's instincts were in any given moment, they'd need to be changed to favor his own personal box score. Call his own number more, insist that that his teammates let him get the defensive boards, gamble for steals, not try for risky passes because turnovers tend to be so heavily punished in all-in-ones.

And so I'm asking you: If you had the chance to talk to Nash in 2004-05 and tell him what was wrong with his instincts, what would you say?

Because if you can't actually point to anything in his habits that was holding him and his team back, then I would suggest you don't any basis for saying he should have had a higher BPM/PER/whatever, or that he should be judged inferior to those with higher such metrics.

And big picture: This is the problem with using box score all-in-ones as a final step in our analysis. They abstract away the actual basketball decision making into a single number that implies "better or worse", while lacking the validity to justify being used in that way.

i know you have an almost allergic disregard for the box score, and tend to say something similar to the above and reduce any mention of the box score to it always being the "final step" in someone's analysis instead of just one more piece of information, but it doesn't seem any more problematic than using impact metrics that can apparently say the same player was barely doing anything one year (2004 nash) and was practically the best in the league the very next year in a different situation (2005 nash). seems they are highly dependent on the role a player is playing, often through no particular active decision on the player's part to change their role. if impact is so situation-dependent, then to me it certainly loses value as a catch all as many of you treat it. especially when impact numbers can say even crazier things than the box score about well known players we all watched and saw succeed/fail and also apparently are so unreliable in even 150+ game samples that it can't even be used for the postseason and so we acknowledge that players can rise/fall in the playoffs but have to treat impact as constant. because if it's situation-dependent, then we are getting into evaluating the coach and the GM and the teammates. players can look way better at seemingly worse portions of their career.

i believe your 4 year RAPM thread mentions that andre iguodala looks better on the warriors than before. do i think AI got better as he got older? pretty much a big no on that. basically nobody does. do i think he went from being asked to do things he was bad at (run an offense) and went on to focus only on things he was good at? pretty much yes. maybe nash just found the one situation where no one would care about his defense and basically everybody on the team was an elite finisher (dunker or 3 point shooter) but was almost entirely dependent on his creation, basically a perfect storm. a perfect storm where he could produce +14 and +15 offenses in playoff series and still somehow lose thanks to defense. i know you think it was seemingly unfair to look at nash's impact when he wasn't in a perfect situation for his whole career but most players aren't.

[lessthanjake]

Whilst some seem to assume the hip-check necessarily turned the outcome and then perhaps go further in terms of assumptions ... implicit in this statement is the idea that the Suns didn't "actually contend" (which is hardly a given) and "it's hard to win a title" with Nash ... presumably because it's hard to put together better defensive pieces around Nash than the Suns did. And, in fairness, those teams get too much heat for "bad" defense where the reality is more around league average and there are some pieces there: Bell and Marion first coming to mind. Still it doesn't seem that much of a stretch to imagine a better defensive build, especially where a team can depend on him for offense.

Great post!

Regarding the bolded, I just want to note that I think it’s super easy to imagine a better defensive build. All it would take is to have an athletic big (which is good for Nash offensively) who actually leveraged his athleticism to be a good defender. What they had instead was an athletic big who was somehow still one of the worst big-man defenders in NBA history (i.e. Amare). Of course, if you take Amare’s offense (or even something approaching it) and make his defense actually good, then you’re ending up with a better player. So in a sense I’m just saying the Suns would’ve been better if Nash’s teammates had been better. But I’d say that most great teams have a second-best player that’s significantly better than Amare. Seriously, the guy was a negative RAPM player outside of his years with Nash. Pretty catastrophically so, actually. This a guy who had a -3.6 two-year RAPM in his first two years on the Knicks, a -4.7 RAPM in his first three years there, a -2.7 RAPM in his initial two years before Nash, etc. And even with Nash he was not all that impactful. It’s *certainly* the case that the team that Nash kept losing to (i.e. the Spurs) had a much better second-best player in Manu Ginobili (or Duncan in 2005 ). So yeah, taking Amare Stoudemire and making him a good defender would not really break the bank in terms of the team’s overall supporting-cast-quality budget, and I think it is hard to imagine that it wouldn’t have led to the Suns being a really dominant team, given how good they were even in reality. The player I always think is most ideal for Nash is David Robinson—who has pretty similar strengths offensively as Amare but is an elite defender. I think the two of them would’ve just completely dumpstered the league. But you wouldn’t actually need a player nearly as good as David Robinson for Nash’s teams to have been quite dominant IMO. Just maybe not a big man who was a catastrophic negative defensively.

i believe your 4 year RAPM thread mentions that andre iguodala looks better on the warriors than before. do i think AI got better as he got older? pretty much a big no on that. basically nobody does. do i think he went from being asked to do things he was bad at (run an offense) and went on to focus only on things he was good at? pretty much yes. maybe nash just found the one situation where no one would care about his defense and basically everybody on the team was an elite finisher (dunker or 3 point shooter) but was almost entirely dependent on his creation, basically a perfect storm. a perfect storm where he could produce +14 and +15 offenses in playoff series and still somehow lose thanks to defense. i know you think it was seemingly unfair to look at nash's impact when he wasn't in a perfect situation for his whole career but most players aren't.

I really think the idea that the Suns were a “perfect storm” for Nash is definitely wrong. Like, just on its face, if you have an incredible offensive player whose weakness is his defense, then I think you’d want your big man to not be absolutely awful defensively. Surely having a good defensive big man would *actually* be a situation where Nash could “focus only on things he was good at” and where we could get to a situation where “no one would care about his defense.” That’s certainly not the case when you do not have a big man who can cover up for anyone else’s defensive weakness! I can buy that he had a good fit offensively, but having a big man as bad on defense as Amare should never really be called the “perfect storm.” And even on the good offensive fit thing, the roster of the team changed a lot over the years. There’s basically no one else who was there the whole time that they had their run of incredible offenses (just Nash and Barbosa were there the whole 2005-2010 time period, I believe). So if he had the perfect storm offensively, then it seems like it wasn’t that difficult to get what he needed, since the Suns were able to repeatedly get it (and that’s despite having a famously cheap owner not willing to shell out for players).

I think people sometimes just like to decide that players they’re looking to argue against had great fit around them, as reason to handwave away their impact. But it’s such a squishy concept that someone can just convince themselves it’s true no matter what. Like, I’m sure you’ve argued that the fact that Draymond was great defensively and defense is Steph’s weakest area made Draymond a great fit for Steph. But now you’re saying Nash has the “perfect storm” in terms of fit, while having the absolute opposite of that.

[Cavsfansince84]

I really think the idea that the Suns were a “perfect storm” for Nash is definitely wrong. Like, just on its face, if you have an incredible offensive player whose weakness is his defense, then I think you’d want your big man to not be absolutely awful defensively. Surely having a good defensive big man would *actually* be a situation where Nash could “focus only on things he was good at” and where we could get to a situation where “no one would care about his defense.” That’s certainly not the case when you do not have a big man who can cover up for anyone else’s defensive weakness! I can buy that he had a good fit offensively, but having a big man as bad on defense as Amare should never really be called the “perfect storm.” And even on the good offensive fit thing, the roster of the team changed a lot over the years. There’s basically no one else who was there the whole time that they had their run of incredible offenses (just Nash and Barbosa were there the whole 2005-2010 time period, I believe). So if he had the perfect storm offensively, then it seems like it wasn’t that difficult to get what he needed, since the Suns were able to repeatedly get it (and that’s despite having a famously cheap owner not willing to shell out for players).

I think people sometimes just like to decide that players they’re looking to argue against had great fit around them, as reason to handwave away their impact. But it’s such a squishy concept that someone can just convince themselves it’s true no matter what. Like, I’m sure you’ve argued that the fact that Draymond was great defensively and defense is Steph’s weakest area made Draymond a great fit for Steph. But now you’re saying Nash has the “perfect storm” in terms of fit, while having the absolute opposite of that.

Absolute opposite of perfect fit seems way off to me re Steve in Phx. I think for the most part it was a very good roster for him. I mean just look at the 3pt%'s of those guys(which Nash gets some credit for) plus A'mare to run pnr or finish with. What they lacked was defense and rebounding but were still able to be top 16 in DRtg from 06-08. So it was a team good enough to contend and destroy offensive records. Lots of guys who could shoot 3's and defend pretty well on the perimeter. I would rate his rosters in Phx like an 8/10 tbh. Same as I would rate Duncan's rosters in the 05-14 years about an 8/10 most years.

[f4p]

i believe your 4 year RAPM thread mentions that andre iguodala looks better on the warriors than before. do i think AI got better as he got older? pretty much a big no on that. basically nobody does. do i think he went from being asked to do things he was bad at (run an offense) and went on to focus only on things he was good at? pretty much yes. maybe nash just found the one situation where no one would care about his defense and basically everybody on the team was an elite finisher (dunker or 3 point shooter) but was almost entirely dependent on his creation, basically a perfect storm. a perfect storm where he could produce +14 and +15 offenses in playoff series and still somehow lose thanks to defense. i know you think it was seemingly unfair to look at nash's impact when he wasn't in a perfect situation for his whole career but most players aren't.

I really think the idea that the Suns were a “perfect storm” for Nash is definitely wrong. Like, just on its face, if you have an incredible offensive player whose weakness is his defense, then I think you’d want your big man to not be absolutely awful defensively. Surely having a good defensive big man would *actually* be a situation where Nash could “focus only on things he was good at” and where we could get to a situation where “no one would care about his defense.” That’s certainly not the case when you do not have a big man who can cover up for anyone else’s defensive weakness! I can buy that he had a good fit offensively, but having a big man as bad on defense as Amare should never really be called the “perfect storm.” And even on the good offensive fit thing, the roster of the team changed a lot over the years. There’s basically no one else who was there the whole time that they had their run of incredible offenses (just Nash and Barbosa were there the whole 2005-2010 time period, I believe). So if he had the perfect storm offensively, then it seems like it wasn’t that difficult to get what he needed, since the Suns were able to repeatedly get it (and that’s despite having a famously cheap owner not willing to shell out for players).

well i was talking more that they were a perfect storm for nash's impact. i agree with your general concept that a "less offense, more defense" amare probably gets us a still very elite offense and a better defense and would have resulted in better team results. but i don't know if a suns team with better defense but at least a few eric gordon's who can create for themselves or a backup PG who isn't basically an SG who has no idea how to run an offense like marcus banks (i think that was his name) doesn't suddenly make nash look less impactful.

as for draymond, i think it is essentially impossible to replicate his fit with steph or argue it wasn't the best thing steph could ever hope for. like we may never see as elite a defender who somehow only provides offensive value (after he forgot how to shoot) in the one place his team's best player is most suited to and his team's offensive system is most suited to. i mean it would be like having rudy gobert next to lebron/harden/luka except his one offensive skill is hitting 40% of his 3's on volume. it would be game breaking for those teams.

I think people sometimes just like to decide that players they’re looking to argue against had great fit around them, as reason to handwave away their impact. But it’s such a squishy concept that someone can just convince themselves it’s true no matter what. Like, I’m sure you’ve argued that the fact that Draymond was great defensively and defense is Steph’s weakest area made Draymond a great fit for Steph. But now you’re saying Nash has the “perfect storm” in terms of fit, while having the absolute opposite of that.

i mean impact in general is incredibly squishy. steve nash went from average to basically MVP in one season at age 30. james harden apparently got worse as he got better and then better again as he got worse. kevin garnett couldn't win a first found series but was so impactful you could seemingly put 4 guys from the gym around him and get like a +5 team. kobe doesn't stand out but tripped and fell into 7 finals and 5 championships. you've used EPM a lot but peak harden fairly well crushes peak nash but it's not a conclusive result because nash is the one guy who simply has to be highly impactful no matter what. i mean to some degree we probably all gravitate toward stats that most match our eye test because we probably instinctively trust that the most.

arguably no one has had more handwaving than nash on this board. i remember someone in the last peaks project (i won't say who, not you) basically making kawhi and nash equal defensively because kawhi wasn't DPOY level any more and so it should just become an offense only argument. and offense only has been a very consistent steve nash argument for the last 3 years here, no matter how devastating the defensive results. and again, the next Top 100 will be interesting. all the nash supporters have now said his impact in the first half of his career wasn't good and this is just a peaks project so we'll only focus on phoenix. i hope we can focus on the full career numbers when the Top 100 is done, which combined with also "meh" box numbers would not indicate a high ranking for a player.

[f4p]

I really think the idea that the Suns were a “perfect storm” for Nash is definitely wrong. Like, just on its face, if you have an incredible offensive player whose weakness is his defense, then I think you’d want your big man to not be absolutely awful defensively. Surely having a good defensive big man would *actually* be a situation where Nash could “focus only on things he was good at” and where we could get to a situation where “no one would care about his defense.” That’s certainly not the case when you do not have a big man who can cover up for anyone else’s defensive weakness! I can buy that he had a good fit offensively, but having a big man as bad on defense as Amare should never really be called the “perfect storm.” And even on the good offensive fit thing, the roster of the team changed a lot over the years. There’s basically no one else who was there the whole time that they had their run of incredible offenses (just Nash and Barbosa were there the whole 2005-2010 time period, I believe). So if he had the perfect storm offensively, then it seems like it wasn’t that difficult to get what he needed, since the Suns were able to repeatedly get it (and that’s despite having a famously cheap owner not willing to shell out for players).

I think people sometimes just like to decide that players they’re looking to argue against had great fit around them, as reason to handwave away their impact. But it’s such a squishy concept that someone can just convince themselves it’s true no matter what. Like, I’m sure you’ve argued that the fact that Draymond was great defensively and defense is Steph’s weakest area made Draymond a great fit for Steph. But now you’re saying Nash has the “perfect storm” in terms of fit, while having the absolute opposite of that.

Absolute opposite of perfect fit seems way off to me re Steve in Phx. I think for the most part it was a very good roster for him. I mean just look at the 3pt%'s of those guys(which Nash gets some credit for)

on the 3 point shooting, which again Nash gets some credit for, you could remove nash's own remarkable shooting and his 3 point makes and the suns would have still led the league each year from 2005-2007 in 3 point percentage and would have led in made 3's in 2005 and 2006 and finished 3rd in 2007.

compare that to something like the rockets who are usually thought of as being as ahead of the league as the suns but when you remove harden you get somewhat good rankings in made 3's with 2nd in 2018, 11th in 2019, but down to 19th in 2020, but in 3 point percentage you get 14th in 2018, 20th in 2019, and down near the bottom at 26th in 2020. so basically not even league average 3 point shooting percentage around harden.

[Cavsfansince84]

on the 3 point shooting, which again Nash gets some credit for, you could remove nash's own remarkable shooting and his 3 point makes and the suns would have still led the league each year from 2005-2007 in 3 point percentage and would have led in made 3's in 2005 and 2006 and finished 3rd in 2007.

compare that to something like the rockets who are usually thought of as being as ahead of the league as the suns but when you remove harden you get somewhat good rankings in made 3's with 2nd in 2018, 11th in 2019, but down to 19th in 2020, but in 3 point percentage you get 14th in 2018, 20th in 2019, and down near the bottom at 26th in 2020. so basically not even league average 3 point shooting percentage around harden.

Ya, I mean they had so many guys who were weapons on the court in those years. Joe Johnson went from 3.3 3's at 30.5%in 04 to 4.5 at 47.8%(by far his career high) in 05, Jim Jackson shot 3.7 at 45.9%(career 36.5), so there were some near outliers that occurred in 05 specifically which not only vaulted Phx to the top of the league but I think by the largest margin of all time. You can credit Nash for some of it of course but it definitely was a roster that complimented him very well. It just had its weaknesses like most any roster does short of a few.

[Doctor MJ]

No, of course I don't think he was trying to maximize a particular stat that was incredibly obscure at the time if it was even invented which I'm pretty sure it wasn't. I'm saying that if the impact stats in Phoenix said he had a top 15 peak this century, and the impact stats outside of Phoenix said he was pretty much an average starter, and every all-in-one measure with a box score component said he was never a top 5 player for any individual season, maybe his peak wasn't actually QUITE as good as all-time seasons from Westbrook and Harden once you account for context.

I'm going to direct this back to the second question because I think it's really important:

The implication of knocking a guy for not placing higher with a stat like BPM is that he was doing something wrong, and thus cries out for us to ask what that is.

For contrast here:

If you were knocking a player because he had too low of a PPG the implication is that he should have shot & scored more.
If you were knocking a player because he had too many turnovers is that he should have turned it over less.
If you were knocking a player because his RAPM was too low, the implication is that he should have been more impactful at helping his team do better on the scoreboard.

But when you do this with a box score all-in-one, it's unclear what exactly the criticism is, only a vague faith that if the player were better, he'd look better in this stat because...

Now to be clear, we can look at, say, a Nash vs Paul comparison and say that if Nash had less turnovers he'd have had all-in-one's more like Paul. But of course Nash could have had the same improvement in all-in-one simply by increasing scoring or rebounding or whatever sufficiently.

I would say that the traditional Hollinger-style analysis would thus respond by saying "Yeah, any of that. Anything more of the good and less of the bad would suffice."

But the hitch with this thinking is that it would alter Nash's process. Whatever's Nash's instincts were in any given moment, they'd need to be changed to favor his own personal box score. Call his own number more, insist that that his teammates let him get the defensive boards, gamble for steals, not try for risky passes because turnovers tend to be so heavily punished in all-in-ones.

And so I'm asking you: If you had the chance to talk to Nash in 2004-05 and tell him what was wrong with his instincts, what would you say?

Because if you can't actually point to anything in his habits that was holding him and his team back, then I would suggest you don't any basis for saying he should have had a higher BPM/PER/whatever, or that he should be judged inferior to those with higher such metrics.

And big picture: This is the problem with using box score all-in-ones as a final step in our analysis. They abstract away the actual basketball decision making into a single number that implies "better or worse", while lacking the validity to justify being used in that way.

i know you have an almost allergic disregard for the box score, and tend to say something similar to the above and reduce any mention of the box score to it always being the "final step" in someone's analysis instead of just one more piece of information, but it doesn't seem any more problematic than using impact metrics that can apparently say the same player was barely doing anything one year (2004 nash) and was practically the best in the league the very next year in a different situation (2005 nash). seems they are highly dependent on the role a player is playing, often through no particular active decision on the player's part to change their role. if impact is so situation-dependent, then to me it certainly loses value as a catch all as many of you treat it. especially when impact numbers can say even crazier things than the box score about well known players we all watched and saw succeed/fail and also apparently are so unreliable in even 150+ game samples that it can't even be used for the postseason and so we acknowledge that players can rise/fall in the playoffs but have to treat impact as constant. because if it's situation-dependent, then we are getting into evaluating the coach and the GM and the teammates. players can look way better at seemingly worse portions of their career.

i believe your 4 year RAPM thread mentions that andre iguodala looks better on the warriors than before. do i think AI got better as he got older? pretty much a big no on that. basically nobody does. do i think he went from being asked to do things he was bad at (run an offense) and went on to focus only on things he was good at? pretty much yes. maybe nash just found the one situation where no one would care about his defense and basically everybody on the team was an elite finisher (dunker or 3 point shooter) but was almost entirely dependent on his creation, basically a perfect storm. a perfect storm where he could produce +14 and +15 offenses in playoff series and still somehow lose thanks to defense. i know you think it was seemingly unfair to look at nash's impact when he wasn't in a perfect situation for his whole career but most players aren't.

So first, I don't know if you interpreted my words as being insulting toward iggy, but from my perspective you're inserting yourself into a dialogue between two people and frontloading as much belittling as possible. I can't even imagine you'd walk up to me in person and start talking to me like this. Despite this, I think this is an important topic so I'm going to respond.

To me the key thing is this idea that we would look to relegate player impact when evaluating players because it is contextually dependent. Of course it's contextually dependent! How could it not be in basketball? It's a sport full of complex interactions, to say nothing of differences in strategy.

Further, the implication here is that we should care more about stats that are less contextually dependent, but as I've noted, the only reason why any of the production stats aren't contextually dependent is because some superstars are allowed to just play the same way all the time. Take Shaq and turn him into a 3 & D guy and all of his stats - production or impact - will be terrible. There's absolutely nothing protecting Shaq from this other than coaches all looking to use him in pretty much the same way.

Re: "if impact is so situation-dependent, then to me it certainly loses value". I'll just point out that impact & value are largely synonyms. They don't quite mean that here because the former refers to basketball play and the latter refers to your personal process for evaluating basketball play, but I'd still say it's telling.

You can't expect to be able to get a better picture of player value by refusing to use the stats that have validity pertaining to player value - and that's what the +/- stats give us.

Listen, we'd all love to have access to a stat that is both optimally valid and perfectly reliable, but they don't exist, so we work with many tools to try to get as clear of picture. But it's hard, we all do it imperfectly, and we could all be more patient with each other - myself included.

But I worry that you're like the proverbial man looking for his classes under a streetlight even though he lost them somewhere else because he's more comfortable looking in the light.

If you care about whether a player is actually helping his team win, you should want to be using +/- stats. Period. Not exclusively of course, but you're just fooling yourself if you think you can see everything there is to see by ignoring stats that are contextually dependent in a contextually dependent sport.

Re: "i believe your 4 year RAPM thread mentions that andre iguodala looks better on the warriors than before. do i think AI got better as he got older? pretty much a big no on that. basically nobody does. do i think he went from being asked to do things he was bad at (run an offense) and went on to focus only on things he was good at? pretty much yes. "

Seems like what you're saying is that the RAPM data suggests Iguodala was more valuable playing in his role on the Warriors than he was playing for his previous teams, and you're then theorizing that this is because they were using him at the things he was best at and not using him for things he wasn't as good at.

That all sounds reasonable to me. I'm just confused why you don't see the value in a stat that gives you that information.

Re: "maybe nash just found the one situation where no one would care about his defense and basically everybody on the team was an elite finisher (dunker or 3 point shooter) but was almost entirely dependent on his creation, basically a perfect storm. a perfect storm where he could produce +14 and +15 offenses in playoff series and still somehow lose thanks to defense. i know you think it was seemingly unfair to look at nash's impact when he wasn't in a perfect situation for his whole career but most players aren't."

So there's a lot to hit here.

That "perfect storm" thought seemed a lot more plausible to me 20 years ago than it does right now, and it's because the game had shifted not just in the direction of the Suns, but passed them.

The '04-05 Suns shot 24.7 3PA/g.
In '24-25, the least 3-prone team in the league shoots 31.9.

Why would it make sense to treat a point guard as being in an unrealistically perfect/lucky/fluky/extreme situation when now every single team would yield that situation and then some?

On the Spurs series, let's first keep in mind that this isn't the Suns being an awful team, it's them losing to the eventual champs, who with that core won 4 titles and was consistently elite for a very long time. It's not crazy to favor players who had more team success than Nash, but let's not talk as if losing to the champ means that the things you did well don't matter.

Second, if we're going to bring up the Spurs having a high ORtg in that series, we should probably talk about what drove that. In a nutshell, if was offensive rebounding.

So then, it's a reasonable thing to point out that taking advantage of small ball lineups by crashing the offensive boards makes sense.

How does that relate to Nash? Not all that closely. It's bigs that determine a team's defensive rebounding prowess.

Now 20 years ago it would have made sense to associate what Nash was doing with "small ball", to blame the small ball for bad defensive rebounding, and argue that the whole pace & space approach was therefore a gimmick that you couldn't win a title with.

But now every team places pace & space, so clearly the concerns about this approach being doomed were just plain wrong. Not only can you win playing this way, but once the league adopts playing this way, you can't win playing any other way that we've found at present.

By this same token, back then many would have argued that Amar'e was the perfect big for this style of play - and that related to why the team simply had to be bad at rebounding - but Amar'e was nothing like perfect. He was bad on D, bad on the boards, and he didn't shoot the 3. He had his strengths don't get me wrong - he deserved his accolades for the most part - but there was really nothing about his weaknesses that simply had to be there to allow Nash to do his thing.

On your last sentence:

Ideally I'd like to judge everyone based on what they can do when properly used, whether in basketball or in any other walk of life. That doesn't mean I look to ignore contextual advantages they might have in a given situation and act like they weren't there, but I'm not looking to fixate on scenarios where things didn't work as well and say "That's all that person really was and it was just lucky context that resulted in all the good when their was good."

Shaq's outlier gifts are about his size, agility for his size, and interior scoring touch, so I'm looking to evaluate him in a situation where a team is trying to use those gifts. Fortunately in that case coaches are not so stupid as to try to use him as a 3-point shooter so I don't have worry about things all that much there, but it's not that it's even slightly difficult for me to prescribe a scenario where Shaq would look like an incompetent boob.

I think it's important that we don't end up underrating players simply because they weren't always used to their potential while other players were, let alone convince ourselves that it's the low draft picks who aren't given primacy for years that are the ones gaining the advantage of "perfect scenarios" while we watch all the #1 draft picks generally get far more perfect opportunities.

[lessthanjake]

I really think the idea that the Suns were a “perfect storm” for Nash is definitely wrong. Like, just on its face, if you have an incredible offensive player whose weakness is his defense, then I think you’d want your big man to not be absolutely awful defensively. Surely having a good defensive big man would *actually* be a situation where Nash could “focus only on things he was good at” and where we could get to a situation where “no one would care about his defense.” That’s certainly not the case when you do not have a big man who can cover up for anyone else’s defensive weakness! I can buy that he had a good fit offensively, but having a big man as bad on defense as Amare should never really be called the “perfect storm.” And even on the good offensive fit thing, the roster of the team changed a lot over the years. There’s basically no one else who was there the whole time that they had their run of incredible offenses (just Nash and Barbosa were there the whole 2005-2010 time period, I believe). So if he had the perfect storm offensively, then it seems like it wasn’t that difficult to get what he needed, since the Suns were able to repeatedly get it (and that’s despite having a famously cheap owner not willing to shell out for players).

I think people sometimes just like to decide that players they’re looking to argue against had great fit around them, as reason to handwave away their impact. But it’s such a squishy concept that someone can just convince themselves it’s true no matter what. Like, I’m sure you’ve argued that the fact that Draymond was great defensively and defense is Steph’s weakest area made Draymond a great fit for Steph. But now you’re saying Nash has the “perfect storm” in terms of fit, while having the absolute opposite of that.

Absolute opposite of perfect fit seems way off to me re Steve in Phx.

To be clear, I wasn’t saying it was the “absolute opposite” of a perfect fit in general. I was referring specifically to what I’d been talking about in the sentence just before that—which was Steph having his relative weakness on defense balanced out with Draymond’s great defense. Nash had his relative weakness on defense exacerbated by Amare’s all-time-level bad defense. If we’re talking in general, then I think the offensive fit was pretty good. But the overall picture here really was not a “perfect storm.”

I think for the most part it was a very good roster for him. I mean just look at the 3pt%'s of those guys(which Nash gets some credit for) plus A'mare to run pnr or finish with. What they lacked was defense and rebounding but were still able to be top 16 in DRtg from 06-08. So it was a team good enough to contend and destroy offensive records. Lots of guys who could shoot 3's and defend pretty well on the perimeter. I would rate his rosters in Phx like an 8/10 tbh. Same as I would rate Duncan's rosters in the 05-14 years about an 8/10 most years.

So I’d note that the Suns did have a good number of three-point shooters, but it gets a bit overblown. Guys like Amare, Diaw, Kurt Thomas, Shaq, Grant Hill, Robin Lopez, etc. all weren’t really three-point shooters. And Marion shot some of them, but was bad enough at it that I’d groan every time I saw him launch one.

Interestingly, it’s actually not so clear to me that Nash really needed a lot of shooters to produce incredible offense with that team. I just quickly ran some numbers to get at the question of how much Nash was helped by having good three-point shooters on the court. I defined the good shooters he had from 2005-2010 (the years the Suns had absolutely elite offense) to include: Jason Richardson, Quentin Richardson, Channing Frye, Jared Dudley, Leandro Barbosa, Goran Dragic, Raja Bell, Gordon Giricek, James Jones, Jalen Rose, Tim Thomas, Eddie House, Joe Johnson, Jim Jackson, and Walter McCarty. In the playoffs, when Nash played with only 2 or fewer of those guys on the court, the Suns overall offensive rating from 2005-2010 was 116.24. This is basically exactly the same as the Suns’s 116.31 overall offensive rating with Nash on the court in the playoffs those years. If we add Marion to the list as well, then it is a virtually identical 116.07. So the Suns’s offensive rating with Nash on the court in the playoffs really wasn’t dependent at all on there being a lot of spacers on the court. And the same thing is true of the regular season. The Suns’s overall ORTG with Nash on the court from 2005-2010 was 117.50. And it was a nearly identical 117.37 with only 2 or fewer of those listed guys on the court. It goes down slightly to 116.11 in the regular season if you add Marion to the list.

So yeah, we have very good reason to think that Nash was producing ridiculous offenses on the Suns even without lots of three-point shooters on the court with him. The offense was basically the same even when he only had 2 or fewer other shooters on the court with him.

Also, as for the Spurs comparison, I just think you could only get to that conclusion by way overrating Amare. Consider that Amare consistently has a negative RAPM outside of the years when he played with Nash. And it wasn’t all that great even with Nash. The difference between having Manu Ginobili and Amare Stoudemire is like…really really big. The Suns make up some of that gap by the fact that Marion was a better player than Tony Parker, but it doesn’t make up for that huge Ginobili/Stoudemire gap. And then we get to the fact that those Suns often had cripplingly little depth, especially in the early years Nash was there. They were a very flawed team.

I really think the idea that the Suns were a “perfect storm” for Nash is definitely wrong. Like, just on its face, if you have an incredible offensive player whose weakness is his defense, then I think you’d want your big man to not be absolutely awful defensively. Surely having a good defensive big man would *actually* be a situation where Nash could “focus only on things he was good at” and where we could get to a situation where “no one would care about his defense.” That’s certainly not the case when you do not have a big man who can cover up for anyone else’s defensive weakness! I can buy that he had a good fit offensively, but having a big man as bad on defense as Amare should never really be called the “perfect storm.” And even on the good offensive fit thing, the roster of the team changed a lot over the years. There’s basically no one else who was there the whole time that they had their run of incredible offenses (just Nash and Barbosa were there the whole 2005-2010 time period, I believe). So if he had the perfect storm offensively, then it seems like it wasn’t that difficult to get what he needed, since the Suns were able to repeatedly get it (and that’s despite having a famously cheap owner not willing to shell out for players).

I think people sometimes just like to decide that players they’re looking to argue against had great fit around them, as reason to handwave away their impact. But it’s such a squishy concept that someone can just convince themselves it’s true no matter what. Like, I’m sure you’ve argued that the fact that Draymond was great defensively and defense is Steph’s weakest area made Draymond a great fit for Steph. But now you’re saying Nash has the “perfect storm” in terms of fit, while having the absolute opposite of that.

Absolute opposite of perfect fit seems way off to me re Steve in Phx. I think for the most part it was a very good roster for him. I mean just look at the 3pt%'s of those guys(which Nash gets some credit for)

on the 3 point shooting, which again Nash gets some credit for, you could remove nash's own remarkable shooting and his 3 point makes and the suns would have still led the league each year from 2005-2007 in 3 point percentage and would have led in made 3's in 2005 and 2006 and finished 3rd in 2007.

compare that to something like the rockets who are usually thought of as being as ahead of the league as the suns but when you remove harden you get somewhat good rankings in made 3's with 2nd in 2018, 11th in 2019, but down to 19th in 2020, but in 3 point percentage you get 14th in 2018, 20th in 2019, and down near the bottom at 26th in 2020. so basically not even league average 3 point shooting percentage around harden.

So you might want to look up how the Suns players shot the three in their years with Nash versus the years of their career without Nash. It’s pretty obvious that Nash was having a huge impact there—which is not surprising for me, having watched him systematically creating great looks, passing into guys’ shooting pockets, etc. It’d be a pretty perverse thing to use Nash’s teammates’ 3P% against him, when improving his teammates’ shooting percentages is like genuinely the main mechanism through which he got so much positive impact.

[Cavsfansince84]

I think for the most part it was a very good roster for him. I mean just look at the 3pt%'s of those guys(which Nash gets some credit for) plus A'mare to run pnr or finish with. What they lacked was defense and rebounding but were still able to be top 16 in DRtg from 06-08. So it was a team good enough to contend and destroy offensive records. Lots of guys who could shoot 3's and defend pretty well on the perimeter. I would rate his rosters in Phx like an 8/10 tbh. Same as I would rate Duncan's rosters in the 05-14 years about an 8/10 most years.

So I’d note that the Suns did have a good number of three-point shooters, but it gets a bit overblown. Guys like Amare, Diaw, Kurt Thomas, Shaq, Grant Hill, Robin Lopez, etc. all weren’t really three-point shooters. And Marion shot some of them, but was bad enough at it that I’d groan every time I saw him launch one.

Interestingly, it’s actually not so clear to me that Nash really needed a lot of shooters to produce incredible offense with that team. I just quickly ran some numbers to get at the question of how much Nash was helped by having good three-point shooters on the court. I defined the good shooters he had from 2005-2010 (the years the Suns had absolutely elite offense) to include: Jason Richardson, Quentin Richardson, Channing Frye, Jared Dudley, Leandro Barbosa, Goran Dragic, Raja Bell, Gordon Giricek, James Jones, Jalen Rose, Tim Thomas, Eddie House, Joe Johnson, Jim Jackson, and Walter McCarty. In the playoffs, when Nash played with only 2 or fewer of those guys on the court, the Suns overall offensive rating from 2005-2010 was 116.24. This is basically exactly the same as the Suns’s 116.31 overall offensive rating with Nash on the court in the playoffs those years. If we add Marion to the list as well, then it is a virtually identical 116.07. So the Suns’s offensive rating with Nash on the court in the playoffs really wasn’t dependent at all on there being a lot of spacers on the court. And the same thing is true of the regular season. The Suns’s overall ORTG with Nash on the court from 2005-2010 was 117.50. And it was a nearly identical 117.37 with only 2 or fewer of those listed guys on the court. It goes down slightly to 116.11 in the regular season if you add Marion to the list.

So yeah, we have very good reason to think that Nash was producing ridiculous offenses on the Suns even without lots of three-point shooters on the court with him. The offense was basically the same even when he only had 2 or fewer other shooters on the court with him.

Keep in mind that I specifically said supporting cast for Nash. Taking A'mare out of there where he really only had one prime year isn't really relevant to how good of a complimentary piece he was for Steve in Phx. I'm not saying Phx's cast was = to what Duncan had every year but it was close a lot of those years imo. It just helps that Duncan anchored the defense and had some good offensive pieces around him. Also, let's be honest, Duncan>Nash.

[lessthanjake]

I think for the most part it was a very good roster for him. I mean just look at the 3pt%'s of those guys(which Nash gets some credit for) plus A'mare to run pnr or finish with. What they lacked was defense and rebounding but were still able to be top 16 in DRtg from 06-08. So it was a team good enough to contend and destroy offensive records. Lots of guys who could shoot 3's and defend pretty well on the perimeter. I would rate his rosters in Phx like an 8/10 tbh. Same as I would rate Duncan's rosters in the 05-14 years about an 8/10 most years.

So I’d note that the Suns did have a good number of three-point shooters, but it gets a bit overblown. Guys like Amare, Diaw, Kurt Thomas, Shaq, Grant Hill, Robin Lopez, etc. all weren’t really three-point shooters. And Marion shot some of them, but was bad enough at it that I’d groan every time I saw him launch one.

Interestingly, it’s actually not so clear to me that Nash really needed a lot of shooters to produce incredible offense with that team. I just quickly ran some numbers to get at the question of how much Nash was helped by having good three-point shooters on the court. I defined the good shooters he had from 2005-2010 (the years the Suns had absolutely elite offense) to include: Jason Richardson, Quentin Richardson, Channing Frye, Jared Dudley, Leandro Barbosa, Goran Dragic, Raja Bell, Gordon Giricek, James Jones, Jalen Rose, Tim Thomas, Eddie House, Joe Johnson, Jim Jackson, and Walter McCarty. In the playoffs, when Nash played with only 2 or fewer of those guys on the court, the Suns overall offensive rating from 2005-2010 was 116.24. This is basically exactly the same as the Suns’s 116.31 overall offensive rating with Nash on the court in the playoffs those years. If we add Marion to the list as well, then it is a virtually identical 116.07. So the Suns’s offensive rating with Nash on the court in the playoffs really wasn’t dependent at all on there being a lot of spacers on the court. And the same thing is true of the regular season. The Suns’s overall ORTG with Nash on the court from 2005-2010 was 117.50. And it was a nearly identical 117.37 with only 2 or fewer of those listed guys on the court. It goes down slightly to 116.11 in the regular season if you add Marion to the list.

So yeah, we have very good reason to think that Nash was producing ridiculous offenses on the Suns even without lots of three-point shooters on the court with him. The offense was basically the same even when he only had 2 or fewer other shooters on the court with him.

Keep in mind that I specifically said supporting cast for Nash. Taking A'mare out of there where he really only had one prime year isn't really relevant to how good of a complimentary piece he was for Steve in Phx. I'm not saying Phx's cast was = to what Duncan had every year but it was close a lot of those years imo. It just helps that Duncan anchored the defense and had some good offensive pieces around him. Also, let's be honest, Duncan>Nash.

So Amare seems to have probably been a negative-impact guy even that one prime year away from Nash (negative on-off, negative one-year RAPM, etc.). And leaving that aside, he wasn’t a very high-impact guy even in his years with Nash. His best 5-year RAPM was ranked 96th (note: he rocked an astoundingly bad league rank of 749th in his first five years after leaving Phoenix). His best 4-year RAPM was ranked 90th (note: he was ranked 699th in his first four years after Phoenix). The best RAPM signal we have is him being ranked 50th in two-year RAPM in 2005/2006, and he didn’t even really play in 2006. This is just not a player that should be placed in the same zip code as Manu. Amare not actually being a high-impact player is kind of the big reason the 2006 Suns unexpectedly did very well still (which helped Nash win MVP again). People overestimated how good Amare was, so they were shocked when the Suns were still a really good team despite losing Amare (and Joe Johnson). But it probably shouldn’t have really been a surprise, since Amare was largely an empty-calories guy.

As for Duncan>Nash, I definitely agree with that if comparing peaks or full career, but I wouldn’t say 2005-2010 Duncan was as good as 2005-2010 Nash. The Spurs mostly got the better of the Suns anyways in that era because Duncan’s supporting cast was significantly better than Nash’s.

[lessthanjake]

So just thought I’d leave this one here:

Watch on YouTube

That is a full-game video of the Suns’s close-out game against the Mavericks in the 2005 playoffs. Obviously it’s a very long video, but I’d encourage anyone to watch in particular the last few minutes of the 4th quarter and the OT period—so like, starting from maybe 1:38:28 in the video. Just incredible stuff from Nash.

And, even beyond just this game, that series really was one of the best series I’ve ever seen a player play. I’ll note that, in the last 3 games of the series, Nash averaged 40/10/9 on 71.1% TS%. It was 37/12/8 on 68.9% TS% in the last 4 games. And, of course, just amazing heroics from him at the end of Game 6, as one can see in the above video. This is against a Mavericks team that was a major title contender.

[Doctor MJ]

Vote:

1. Steve Nash 05 > 07 > 06
2. Manu Ginobili 05 > 06 > 07
3. Draymond Green 16 > 15 > 17
4. James Harden 18 > 19 > 12

Harden's the lone new vote for me this round. I spent my time mostly thinking about Harden vs Luka, and I already expressed the essence of the dilemma:

I think Luka at his best is better than Harden ever was, but he's never had a season consistently at his best. Frankly I'd say the same about Embiid, but Embiid's got bigger issues there.

I've tried to be pretty flexible in how I'm using the idea of a season-based peak and not get overly hung up on things that tempt me toward winning bias, but I do think, for example, that Harden deserved that MVP and that Luka hasn't had a season I'd give him the MVP over Harden in '17-18.

The playoffs could make up for that difference - and that's why I honestly went back & forth here - but I suppose I feel like Harden is one of the many rungs on the grand ladder that Luka hasn't definitively climbed passed yet.

I've mentioned that AD is on my mind, and he is, but I'll say that Tatum's starting to linger in my thoughts as a harder guy to not-pick-yet.

[lessthanjake]

I’ve noticed people saying that Luka is basically just a better version of Harden, and I do want to push back on that a bit. Peak Harden really is a huge outlier in terms of his ability to produce really efficient offense from isolations at a very high volume. Luka is good at this, but I think peak Harden was much more impressive in this regard. The volume thing is in large part a coaching decision, but to maintain better efficiency on those plays at such high volume is definitely more impressive IMO.

Just to illustrate this, I’ll show Luka’s PPP and volume of isolations after his rookie year, compared to what peak Harden was doing:

Luka on Isolations

2025: 1.06 PPP on 6.2 possessions per game
2024: 1.09 PPP on 7.3 possessions per game
2023: 1.11 PPP on 7.1 possessions per game
2022: 1.11 PPP on 6.3 possessions per game
2021: 1.05 PPP on 4.6 possessions per game
2020: 1.02 PPP on 4.3 possessions per game

Peak Harden on Isolations

2020: 1.12 PPP on 14.1 possessions per game
2019: 1.10 PPP on 16.4 possessions per game
2018: 1.22 PPP on 10.0 possessions per game

These numbers should also be seen in the context of the league changing, with isolations (and pretty much everything else) being more efficient in more recent years.

Again, I know that the volume is largely a coaching decision, but I just think peak Harden was a noticeably better player in isolation—and the main explanation I have for that is that peak Harden was *a lot* quicker.

Of course, there’s things that Luka does better than Harden, so it’s not all one-way traffic, but just wanted to note that I definitely don’t think Luka is just an upgraded Harden.

[f4p]

I'm going to direct this back to the second question because I think it's really important:

The implication of knocking a guy for not placing higher with a stat like BPM is that he was doing something wrong, and thus cries out for us to ask what that is.

For contrast here:

If you were knocking a player because he had too low of a PPG the implication is that he should have shot & scored more.
If you were knocking a player because he had too many turnovers is that he should have turned it over less.
If you were knocking a player because his RAPM was too low, the implication is that he should have been more impactful at helping his team do better on the scoreboard.

But when you do this with a box score all-in-one, it's unclear what exactly the criticism is, only a vague faith that if the player were better, he'd look better in this stat because...

Now to be clear, we can look at, say, a Nash vs Paul comparison and say that if Nash had less turnovers he'd have had all-in-one's more like Paul. But of course Nash could have had the same improvement in all-in-one simply by increasing scoring or rebounding or whatever sufficiently.

I would say that the traditional Hollinger-style analysis would thus respond by saying "Yeah, any of that. Anything more of the good and less of the bad would suffice."

But the hitch with this thinking is that it would alter Nash's process. Whatever's Nash's instincts were in any given moment, they'd need to be changed to favor his own personal box score. Call his own number more, insist that that his teammates let him get the defensive boards, gamble for steals, not try for risky passes because turnovers tend to be so heavily punished in all-in-ones.

And so I'm asking you: If you had the chance to talk to Nash in 2004-05 and tell him what was wrong with his instincts, what would you say?

Because if you can't actually point to anything in his habits that was holding him and his team back, then I would suggest you don't any basis for saying he should have had a higher BPM/PER/whatever, or that he should be judged inferior to those with higher such metrics.

And big picture: This is the problem with using box score all-in-ones as a final step in our analysis. They abstract away the actual basketball decision making into a single number that implies "better or worse", while lacking the validity to justify being used in that way.

i know you have an almost allergic disregard for the box score, and tend to say something similar to the above and reduce any mention of the box score to it always being the "final step" in someone's analysis instead of just one more piece of information, but it doesn't seem any more problematic than using impact metrics that can apparently say the same player was barely doing anything one year (2004 nash) and was practically the best in the league the very next year in a different situation (2005 nash). seems they are highly dependent on the role a player is playing, often through no particular active decision on the player's part to change their role. if impact is so situation-dependent, then to me it certainly loses value as a catch all as many of you treat it. especially when impact numbers can say even crazier things than the box score about well known players we all watched and saw succeed/fail and also apparently are so unreliable in even 150+ game samples that it can't even be used for the postseason and so we acknowledge that players can rise/fall in the playoffs but have to treat impact as constant. because if it's situation-dependent, then we are getting into evaluating the coach and the GM and the teammates. players can look way better at seemingly worse portions of their career.

i believe your 4 year RAPM thread mentions that andre iguodala looks better on the warriors than before. do i think AI got better as he got older? pretty much a big no on that. basically nobody does. do i think he went from being asked to do things he was bad at (run an offense) and went on to focus only on things he was good at? pretty much yes. maybe nash just found the one situation where no one would care about his defense and basically everybody on the team was an elite finisher (dunker or 3 point shooter) but was almost entirely dependent on his creation, basically a perfect storm. a perfect storm where he could produce +14 and +15 offenses in playoff series and still somehow lose thanks to defense. i know you think it was seemingly unfair to look at nash's impact when he wasn't in a perfect situation for his whole career but most players aren't.

So first, I don't know if you interpreted my words as being insulting toward iggy, but from my perspective you're inserting yourself into a dialogue between two people

hmm, y''all were having the same nash/impact/box score data discussion that like 7 of us are having right now.

and frontloading as much belittling as possible.

other than maybe "allergic", i'm not sure what was considered belittling. and frankly, i was just looking for the correct word. you almost always recoil at box score usage more than basically anyone i can think of on this board (if you recall the drummond discussion we had maybe a year ago). iggy had even gone out of his way to say he was an impact guy and you described it as the final step in his analysis. if even this limited usage isn't appropriate, then you are basically abandoning it as meaning anything. which you are free to do, but i would then say "allergic" feels like a good word, whether it came off as belittling or not.

I can't even imagine you'd walk up to me in person and start talking to me like this.

i mean i don't walk up to people in person and just start talking so no, i probably wouldn't. but i didn't see much that i wouldn't say to someone in real life if they were someone i talked to about basketball all the time. the rest of my first paragraph after allergic just seemed like a recap of the impact discussion. it's fluctuates, we can't use it in the playoffs, it's situation dependent. maybe the tone in my head (frustration) is not how you read it (belittling). either way, this discussion has probably run it's course if it's gotten to this point. i'll just say you (and arguably jake) essentially want to credit nash for a career he could have had if the league had moved faster in his direction. i get it, but then it feels like that would work for many players. i'd like james harden and the 2018/2019 rockets to have faced the same WCF and finals opponents the 2020 lakers, 2021 bucks, 2022 warriors, 2023 nuggets, 2024 celtics, and 2025 thunder faced (i.e. mediocre teams). so they could crush them and harden/cp3/morey/d'antoni would all get their flowers. but then we would have to rank harden like 15 spots higher in the top 100 and like 5 to 8 spots higher here and no one is doing that.

[Djoker]

I’ve noticed people saying that Luka is basically just a better version of Harden, and I do want to push back on that a bit. Peak Harden really is a huge outlier in terms of his ability to produce really efficient offense from isolations at a very high volume. Luka is good at this, but I think peak Harden was much more impressive in this regard. The volume thing is in large part a coaching decision, but to maintain better efficiency on those plays at such high volume is definitely more impressive IMO.

Just to illustrate this, I’ll show Luka’s PPP and volume of isolations after his rookie year, compared to what peak Harden was doing:

Luka on Isolations

2025: 1.06 PPP on 6.2 possessions per game
2024: 1.09 PPP on 7.3 possessions per game
2023: 1.11 PPP on 7.1 possessions per game
2022: 1.11 PPP on 6.3 possessions per game
2021: 1.05 PPP on 4.6 possessions per game
2020: 1.02 PPP on 4.3 possessions per game

Peak Harden on Isolations

2020: 1.12 PPP on 14.1 possessions per game
2019: 1.10 PPP on 16.4 possessions per game
2018: 1.22 PPP on 10.0 possessions per game

These numbers should also be seen in the context of the league changing, with isolations (and pretty much everything else) being more efficient in more recent years.

Again, I know that the volume is largely a coaching decision, but I just think Harden was a noticeably better player in isolation—and the main explanation I have for that is that peak Harden was *a lot* quicker.

Of course, there’s things that Luka does better than Harden, so it’s not all one-way traffic, but just wanted to note that I definitely don’t think Luka is just an upgraded Harden.

Good data.

Do you happen to have the numbers for the playoffs?

[lessthanjake]

I’ve noticed people saying that Luka is basically just a better version of Harden, and I do want to push back on that a bit. Peak Harden really is a huge outlier in terms of his ability to produce really efficient offense from isolations at a very high volume. Luka is good at this, but I think peak Harden was much more impressive in this regard. The volume thing is in large part a coaching decision, but to maintain better efficiency on those plays at such high volume is definitely more impressive IMO.

Just to illustrate this, I’ll show Luka’s PPP and volume of isolations after his rookie year, compared to what peak Harden was doing:

Luka on Isolations

2025: 1.06 PPP on 6.2 possessions per game
2024: 1.09 PPP on 7.3 possessions per game
2023: 1.11 PPP on 7.1 possessions per game
2022: 1.11 PPP on 6.3 possessions per game
2021: 1.05 PPP on 4.6 possessions per game
2020: 1.02 PPP on 4.3 possessions per game

Peak Harden on Isolations

2020: 1.12 PPP on 14.1 possessions per game
2019: 1.10 PPP on 16.4 possessions per game
2018: 1.22 PPP on 10.0 possessions per game

These numbers should also be seen in the context of the league changing, with isolations (and pretty much everything else) being more efficient in more recent years.

Again, I know that the volume is largely a coaching decision, but I just think Harden was a noticeably better player in isolation—and the main explanation I have for that is that peak Harden was *a lot* quicker.

Of course, there’s things that Luka does better than Harden, so it’s not all one-way traffic, but just wanted to note that I definitely don’t think Luka is just an upgraded Harden.

Good data.

Do you happen to have the numbers for the playoffs?

Yeah here’s the numbers for Luka:

2025: 1.05 PPP on 7.8 possessions per game
2024: 1.02 PPP on 6.0 possessions per game
2023: N/A
2022: 0.90 PPP on 8.5 possessions per game
2021: 0.93 PPP on 8.3 possessions per game
2020: 0.92 PPP on 6.3 possessions per game


And here’s the numbers for peak Harden:

2020: 1.08 PPP on 10.3 possessions per game
2019: 1.04 PPP on 15.0 possessions per game
2018: 1.00 PPP on 12.9 possessions per game

Each individual year is not a large sample, of course. But yeah, I’d say Harden looks better by this too, especially when we keep in mind that offense is significantly more efficient in general in the last few years than it was in those years with the Harden numbers.

[Djoker]

I’ve noticed people saying that Luka is basically just a better version of Harden, and I do want to push back on that a bit. Peak Harden really is a huge outlier in terms of his ability to produce really efficient offense from isolations at a very high volume. Luka is good at this, but I think peak Harden was much more impressive in this regard. The volume thing is in large part a coaching decision, but to maintain better efficiency on those plays at such high volume is definitely more impressive IMO.

Just to illustrate this, I’ll show Luka’s PPP and volume of isolations after his rookie year, compared to what peak Harden was doing:

Luka on Isolations

2025: 1.06 PPP on 6.2 possessions per game
2024: 1.09 PPP on 7.3 possessions per game
2023: 1.11 PPP on 7.1 possessions per game
2022: 1.11 PPP on 6.3 possessions per game
2021: 1.05 PPP on 4.6 possessions per game
2020: 1.02 PPP on 4.3 possessions per game

Peak Harden on Isolations

2020: 1.12 PPP on 14.1 possessions per game
2019: 1.10 PPP on 16.4 possessions per game
2018: 1.22 PPP on 10.0 possessions per game

These numbers should also be seen in the context of the league changing, with isolations (and pretty much everything else) being more efficient in more recent years.

Again, I know that the volume is largely a coaching decision, but I just think Harden was a noticeably better player in isolation—and the main explanation I have for that is that peak Harden was *a lot* quicker.

Of course, there’s things that Luka does better than Harden, so it’s not all one-way traffic, but just wanted to note that I definitely don’t think Luka is just an upgraded Harden.

Good data.

Do you happen to have the numbers for the playoffs?

Yeah here’s the numbers for Luka:

2025: 1.05 PPP on 7.8 possessions per game
2024: 1.02 PPP on 6.0 possessions per game
2023: N/A
2022: 0.90 PPP on 8.5 possessions per game
2021: 0.93 PPP on 8.3 possessions per game
2020: 0.92 PPP on 6.3 possessions per game


And here’s the numbers for peak Harden:

2020: 1.08 PPP on 10.3 possessions per game
2019: 1.04 PPP on 15.0 possessions per game
2018: 1.00 PPP on 12.9 possessions per game

Each individual year is not a large sample, of course. But yeah, I’d say Harden looks better by this too, especially when we keep in mind that offense is significantly more efficient in general in the last few years than it was in those years with the Harden numbers.

Yea I always knew Harden was a beast in iso situations but this is great to see.

It's notable that both guys increase their iso frequency in the PS while their PPP predictably drops. Not sure it's a recipe for success. Especially in Harden's case where the volume is absurd. Maybe his teammates can get better shots than what he creates from these plays. 1.04 PPP works out to 52 %TS and isn't exactly efficient offense.

[lessthanjake]

Good data.

Do you happen to have the numbers for the playoffs?

Yeah here’s the numbers for Luka:

2025: 1.05 PPP on 7.8 possessions per game
2024: 1.02 PPP on 6.0 possessions per game
2023: N/A
2022: 0.90 PPP on 8.5 possessions per game
2021: 0.93 PPP on 8.3 possessions per game
2020: 0.92 PPP on 6.3 possessions per game


And here’s the numbers for peak Harden:

2020: 1.08 PPP on 10.3 possessions per game
2019: 1.04 PPP on 15.0 possessions per game
2018: 1.00 PPP on 12.9 possessions per game

Each individual year is not a large sample, of course. But yeah, I’d say Harden looks better by this too, especially when we keep in mind that offense is significantly more efficient in general in the last few years than it was in those years with the Harden numbers.

Yea I always knew Harden was a beast in iso situations but this is great to see.

It's notable that both guys increase their iso frequency in the PS while their PPP predictably drops. Not sure it's a recipe for success. Especially in Harden's case where the volume is absurd. Maybe his teammates can get better shots than what he creates from these plays. 1.04 PPP works out to 52 %TS and isn't exactly efficient offense.

I think 1.04 PPP is a lot better than that. The PPP calculation is more comparable to offensive efficiency, rather than TS%, since it is taking turnovers into account too and is basically subtracting out offensive rebounds that occurred on those possessions. So you might say, okay but a 104 ORTG isn’t good either. However, we can’t really compare it to general offensive ratings, because those are being dragged up a lot by situations that aren’t analogous to the situations in which isolations occur. In particular, offensive ratings are dragged up a lot by the fact that transition and second-chance shot attempts (particularly putbacks) are really efficient. But those aren’t the situations that isolations happen in.

What efficiency on isolations should be compared to is points per play in half-court offense. We have that from cleaningtheglass. See for instance, here: https://cleaningtheglass.com/stats/league/context?season=2017&seasontype=regseason&start=10/1/2017&end=10/15/2018#tab-offense_halfcourt_putbacks. At that link, you can see that the league average PPP in the half-court was 93.4 in the 2018 season. It was 94.2 in the 2018 playoffs. In 2019, it was 94.7 in the regular season and 93.9 in the playoffs. And, in 2020, it was 95.2 in the regular season and 98.3 in the playoffs. So yeah, peak Harden isolations were actually quite a lot more effective than average offense in half-court situations, including in the playoffs.

And, for reference, just perusing the regular-season numbers on this in the last several years, the league average is around 3-5 points higher in the last three seasons than it was before that, so that puts some context on Luka’s numbers that make it look less good.

[lessthanjake]

Just to pull my last two posts together a bit, here’s these guys’ PPP relative to league average half-court offense, along with the volume in parentheses. Note that, for the playoffs, I used the playoff average, rather than using what the specific opponents gave up in the RS, since I’d need to subscribe to CleaningTheGlass in order to get that info.

_____

Luka rPPP on Isolations

2025 RS: +8.2 (6.2 possessions per game)
2024 RS: +10.3 (7.3 possessions per game)
2023 RS: +12.6 (7.1 possessions per game)
2022 RS: +15.4 (6.3 possessions per game)
2021 RS: +8.1 (4.6 possessions per game)
2020 RS: +6.8 (4.3 possessions per game)

2025 PS: +9.3 (7.8 possessions per game)
2024 PS: +4.4 (6.0 possessions per game)
2023 PS: N/A
2022 PS: -5.4 (8.5 possessions per game)
2021 PS: +5.8 (8.3 possessions per game)
2020 PS: -6.3 (6.3 possessions per game)

Peak Harden rPPP on Isolations

2020 RS: +16.8 (14.1 possessions per game)
2019 RS: +15.3 (16.4 possessions per game)
2018 RS: +28.6 (10.0 possessions per game)

2020 PS: +9.7 (10.3 possessions per game)
2019 PS: +10.1 (15.0 possessions per game)
2018 PS: +5.8 (12.9 possessions per game)

_____

So yeah, Harden definitely looks a lot better in this regard. He does go down in his relative efficiency on isolations in the playoffs, but he still looks great (not to mention that some of him going down is surely that he had harder-than-average playoff opponents, and I’m only doing this relative to league average in the playoffs). Luka looks worse in both regular season and playoffs (with both noticeably lower efficiency and lower volume), and his playoff relative efficiency numbers actually are sometimes negative.

With peak Harden, I think his isolation plays were genuinely extremely effective half-court offense. Luka’s isolation plays are still effective in the regular season, but not to the same degree, and they’re a mixed bag in the playoffs. This is probably a significant reason why coaches had Harden isolate so much more. And, considering Harden’s higher effectiveness and significantly higher volume on these plays, I think this is just a way bigger weapon for peak Harden in a way that Luka just can’t compare to at all (and honestly, not sure anyone can—maybe a guy like Jordan if we go back far enough, but Harden is a massive outlier for guys we have this data for).

NOTE: One caveat I’d give on this is that I believe the PPP calculations for isolation plays aren’t taking into account points scored off of assists and whatnot. So the actual overall effectiveness of these plays is a bit a bit more complicated a question, since I believe this is just talking about how effective it is when they isolated and attempted to score (or turned the ball over).