John Stockton is underrated here

[lessthanjake]

So this is all good stuff to point out! Let me note though:

1. While Stockton looks amazing by nbarapm given his age, I wouldn't say that the indicators are giving him a clearly better overall RAPM than Nash. By the 5 year models you're looking at, Nash still has the higher absolute value, and while I'd urge people to avoid using absolute values comparing different epochs as if they are apples-to-apples comparisons, just in general, Nash's prime numbers are looking about as good as the numbers we have for Stockton.

(Incidentally, for me nbarapm defaults to the 4-year model which to me raises the question of what time frame we should be most interested in. Obviously it depends on the player which time frame will best measure him, but among those who are serious about using many years to reduce noise, I've seen 3, 4, and 5 year models being preferred relative to each other depending on the course.)

To be clear, there's definitely a good Stockton career argument to make that if he's as good in those later years as Nash's best, then his longevity should give him the career edge even if he didn't actually impact the game more during his prime, it's just that's not actually the same thing as having evidence that Stockton's defensive edge was making him more valuable overall than Nash.

2. But the scale of how good old man Stockton's DRAPM numbers look, concern me, because it's undeniable that he had some physical limitations compared to when he was younger, and yet looks basically right up there with the best perimeter defenders by these metrics with those limitations. What's up with that?

So, if we zoom in over at nbarapm, what we see is that Stockton's best DRAPM runs are focused on the years '99-00 through '01-02, when Stockton was age 37-39. As in, we have data beginning when Stockton was 34, but he appears to be peaking later than that. That's weird.

Now if we look at other defensive guards, how did they rank by DRAPM in the age 37-39 range in comparison?

Stockton 3.8, 3rd in NBA (for 3-year RAPM '99-00 - '01-02).
Kidd 0.9, 85th in NBA.
Lowry 0.5, 133rd in NBA (technically only age 36-38, because didn't turn 39 until late March this year)
Paul 0.5, 136th in NBA

This then is what's so weird. It's not strange to make the argument that Stockton was a defender on the level of these other dudes, but him appearing to be a defensive superstar by DRAPM this late in his career, doesn't fit with how anyone else looked.

And so that's the thing I think we have to focus on when just interpreting the data we have:

How real was Stockton's defensive data in this time period when he's super-old, he's not playing star minutes, his team doesn't have a Top 10 defense, and his team isn't a contender going deep into the playoffs each year?

If the question is: How could that data look good if he isn't actually that good? There I'd point to the stodgy nature of offensive strategy in the era. I think we're largely looking at data where teams weren't trying to target Stockton at a time when targeting Stockton probably would have worked pretty well even then, and would have worked even better today.

3. As I say all of that let me acknowledge:

The distinction between ORAPM and DRAPM isn't as clean as we'd like it and that's why I always emphasize that it's the overall RAPM we should focus on most. It's possible that lineup weirdness is pushing Stockton's offensive impact into the defensive measure - which could happen if he's disproportionately playing with defense-oriented lineups.

So perhaps his overall RAPM is fundamentally legitimate even though his DRAPM overrates his defense because of this weirdness.

I think it’s reasonable to be a little skeptical of just how good Stockton’s defensive impact numbers are. A few points I’d make on this though:

1. My point about this is more a directional one than one that relies on the precise numbers. Even if we did think old Stockton wasn’t quite as good defensively as the DRAPM data indicates, he’d still be quite a lot better defensively than Nash. And even discounting Stockton’s DRAPM a good bit, that gap would be quite significant.

2. Your point about the distinction between ORAPM and DRAPM not being super clean is a good one and might be at play here to at least some degree. I think one aspect that may be relevant here is that Stockton being on the court likely made the Jazz more slow and methodical on the offensive end, particularly when the other team was making runs. Basically, a calming influence from a veteran PG—which can have a significant effect on the defense, since panicking teams often end up being killed in transition and whatnot. Just a theory, but this is the type of thing that you’re talking about where the distinction between ORAPM and DRAPM isn’t clean. I think it’s perfectly plausible that things Stockton was doing on the offensive end were bleeding into DRAPM instead of ORAPM, contributing to DRAPM that might feel a little too high.

3. While Nash’s numbers in the NBArapm data set do look about as good as old Stockton, I think it’s reasonable to infer that old Stockton was probably not Stockton’s RAPM peak though, so that makes me pretty bullish on the idea that prime Stockton was more impactful than prime Nash.

4. I also note that there’s other RAPM sets that are even better for old Stockton. For instance, in TheBasketballDatabase, old Stockton has higher five-year RAPM than Nash ever did. As another example, the NBArapm website actually has its separate “6Factor” RAPM. That starts at 2000, so it only has four years for Stockton, but Stockton’s four-year RAPM in that measure is 7.7, which is higher than Nash’s best four-year RAPM of 6.6. Similarly, in Cheema’s five-year RAPM, old Stockton has two five-year spans that are above any five-year span for Nash, and another span that is above all but one of Nash’s spans. Overall, I think the RAPM picture is better for old Stockton than it is for prime Nash. And it’s hard for me to look at that and conclude that prime Stockton wouldn’t be ahead of prime Nash. We don’t have much data for prime Stockton, so it’s possible, but the RAPM snippets we have from those earlier years look good for Stockton, so it doesn’t feel likely.

[tsherkin]

Oh, the episode was so good! I knew about Nash thriving in basically all open field sports as a passer, but I love him specifically talking about the influence of Gretzky's Office - as in Gretzky's tendency to take the puck behind the net, and how what we now call "Nashing" is essentially the basketball version of that.

Yes, that was a fun little section. And you could immediately see the parallels, right? It was great stuff.

We should note: In basketball you don't get to patiently wait while you do this. Nash's success while Nashing was entirely dependent on him being able to read the entire half court as soon as he reached the requisite height, and make the right pass instantly. While Gretzky may well have had just as fast of a visual efficiency in his brain, the layout of the hockey rink gave him a bigger advantage in doing this than Nash got on the basketball court.

I particularly enjoyed his discussion in that episode of his approach. He'd curl around with the continuity dribble, and at least one defender was going to be out of sorts from looking man or ball, and then one of Nash's guys would cut somewhere and he could hit the on the way to the basket... or he'd get a seam to attack for a layup or something. And then of course, his whole discussion of how he approached the perimeter game and hesitations, etc, etc. How that sprained ankle helped him get comfortable pushing off his right and going left, etc.

Oh to be clear, I wasn't seeking to rebut you specifically, I was just looking to clarify something that confuses people in my experience.

Fair!

The statement "Nash didn't have his big impact prior to D'Antoni" is what led people to start trying to neg Nash as a "system player" - a term borrowed from football that never made sense in basketball, and frankly we now know it didn't actually make that much sense in football either as a guy like Patrick Mahomes would have had that label in an earlier era. (The lesson the NFL eventually learned is that they should be learning from innovative coaches in college, rather than bashing the players who thrived in college but failed to thrive in the NFL's then old-fashioned offensive norms.)

Yeah, that really only applies to guys for whom offense is created, not guys who ARE the offense. And as you say later, it's important to grok that Nash's impact in Dallas vs. Phoenix is more about usage.

It's fascinating to me looking back now 20 years after I was first heavily engaged in NBA analytics and see how some of these guys from back then skated by without it severely damaging their demand. Walker was someone everyone had issues with...and yet he still kept getting seen as someone who could be a core part of a contender up through the Heat acquiring him at age 29 and winning a chip with him playing major minutes. The fact they could do that means he wasn't super-unsuccessful on the court back then despite his bad habits...on the other hand the fact he was out of the league not long afterward without their being any earth shattering injuries is telling.

The worst is that we all knew Walker was a useless moron AT THE TIME! He had horrible shot selection, he had low FG%, he wasn't a good 3pt shooter a lot of the time, he wans't a good defender... he was just... a disaster.

Definitely. Whereas in a sport like baseball, anyone who gets a chance at bat gets basically the same chance to hit a homer, in basketball you're either the one the scheme is built around, or your job is to fit in where there's a need.

Yeah, it's just a far less discrete environment.

I remember one story of someone who played in a pick up game against Ben Wallace, and was shocked to realize that Ben's handle and shot were pretty good. They had assumed that Ben was a guy who couldn't do the basics of what make you a star basketball player and had simply lucked into having skills that an NBA team needed, only to find out that Ben was actually a pretty good all around player, he just wasn't NBA level at the stuff that makes you an offensive NBA star.

Witness Shaq playing with the ball at the All-Star game as a member of the Heat, breaking Brad Miller's ankles and such, you know?

When a guy like Jahlil Okafor falls out of the league, it's literally because when you try to let him do the only stuff he was ever good at at lower levels of ball, it just doesn't work in the NBA, and so he literally has no place in the NBA unless he completely forgets what he thought he was and works hard at developing new skills.

To be fair, Okafor was a crap defender, a bit tepid as a defensive rebounder and had persistent issues with his meniscus, so it was more than just not letting him do what he was good at. He was efficient when he touched the ball, he just never played 60 games and didn't do any of the other stuff which made it worthwhile to have him on the court (except offensive rebounding).

Another guy on my mind recently: Trae Young. The talk in podcasts right now is that it might actually be for the best if the Hawks keep building around him, and not because the original decision to build around him was a good one, but because everything they've done since making that original decision has been about building the best team they could around him. Going hand and hand in this is the realization that there may be no other NBA team that wants Young on their roster at all given his contract and expectations. This then to say, we may end up in a situation where Young is the Hawk franchise player for 15 years, not because he should have ever been any team's franchise player, but because once you've made moves in a certain direction, it may not make sense to blow it up for a long time if what you're trying to do is keep relevance in your town (which the Hawks have lone struggled with).

I think the talk should be more about putting a legit team around him instead of airing him out to dry with play-finishers and perimeter guys who aren't good enough to have the ball instead of him. He's a high-end playmaker who hasn't had any kind of legit perimeter partner, ever.

They SHOULD focus on continuing to build around him, they just need to suck less at it. Trae has his own issues trying to rediscover his shot and learning how not to be an awful, awful finisher in close, but he's still an impact player on offense because of his playmaking... which is the more impressive for what he does and doesn't have on that team.

Before moving on from discussing Young, I should acknowledge that Young is arguably the player most similar to Nash in the modern game, so why isn't it working?

Consider difference in shooting proficiency would be a starting point, but roster disposition is also worth significant mention. And as you note, Nash was a little taller.

This is one of the big question marks about Stockton so I'm glad you brought it up. The reality is that even in Nash's time, opposing offenses weren't in full "bum hunting" mode, and this was all the more so in Stockton's era. People might think Stockton would be immune to this because he had a good defensive rep, but of course part of the strategy here isn't about the defender being bad, but about attacking with a man he's not suited to guard, and doing so in part to wear him down and lessen his offensive advantage.

Like, for the thread's sake, I want to clarify I'm not bagging on Stockton. I figure he'd be a top-15 ish type of player in today's game. Perhaps higher. He was an extremely good player. But he was a small dude, and he had small dude problems. And because he has All-D selections, people automatically assume his D was amazing, all the time, in all facets. Like with MJ, who was totally not a gambler in the passing lanes, and whose teammates and opponents totally didn't consider a guy who was more of that than a man-on guy (by focus), right? Right? Aherm... Different, because Jordan COULD do it when he wanted to, of course, but you take my point, I'm sure. Stockton had all the same issues Nash did with man defense, but he was an exceptional team defender, whereas of course Nash wasn't at that level at all.

When I've seen impact indicators that say that old man Stockton is up there with the best non-DPOY level guys, to me that's raising a red flag about primitive strategy.

And the particulars of what RAPM and such think they are measuring, IMHO. Lots of stuff around Stockton specifically makes me wonder.

I do think Stockton's defense would be better in any era than Nash's, but I also think it's problematic to look the Stockton I watched late in his career, and think he was something like the most capable perimeter defender in the world.

Agreed on both counts.

Re: team defense. I'll say on Nash's behalf, I actually think his team defense in the sense of him doing what the scheme asked of him, was quite good.

Yes, he just wasn't Stockton in that regard.

Nash was vulnerable to direct attacks, and that would be a bigger concern today than it was back then, but this didn't lead to the Suns having a horribly worked-over defense generally. With a wise scheme, and good defenders around him, like what we saw in '05-06 with Kurt Thomas in place of Amar'e Stoudemire, it was actually pretty solid.

Yep, it wasn't until later on that it started to drop off from around league average. There were weaknesses, but they were able to cover in other ways, to at least some extent. But like, I think people forget that there are many paths to the same destination. You do not need to be the best defense in the league to win the title. You do not need to be the best offense, either. You can find a comfortable balance of both and get it done, depending on what talent you have, as long as you're good enough to get past what's in front of you. It's why I'm not thrilled with Mazzula's approach to Boston's offense, you know? SMASHING 3PAr like beasts and having nothing in the middle is a great way to flame out in the playoffs, and it would have burned them again in 2024 if Dallas hadn't collapsed as horribly as they did (because that wasn't all on Boston's defense). Nash had some words on Mind the Game about that too, which were nice to hear.

1. "Science moves forward one funeral at a time". Meaning, it's a mistake to think that when the paradigm shift occurs, everyone sees the light. What actually happens is a much slower process wherein a new paradigm gains new followers while the old paradigm does not, and the stodgy followers of the old paradigm tend to cling to their beliefs until the end.

Let's note that when we talk like this, the new paradigm is not necessarily more right than the old one, but in fields where there is sufficient objective data, this will typically be so. (By contrast, public policy doesn't have that, and while one might argue that public policy doesn't actually ever achieve the type of consensus to earn the word "paradigm", there is no doubt that public policy ideas spread like wildfire.)

Yeah, I mean, there are hangers-on all the time. We see that especially in basketball discourse, hearing old philosophies long-since disavowed/disproven/etc still being espoused by some folk who just don't want to accept things, or who haven't put in the work to keep up, or who have just too much emotional investment in older ideas with new twists.

2. "Gimmick to Gospel". Meaning, you know when a paradigm shift takes place when the same sort of think that was once dismissed as a gimmick, now becomes embraced as the standard. This is absolutely what has happened in the NBA with pace & space.

Sure... but pace and space works in some ways because we don't have the same kind of talent in the league. You CAN still punish small teams with size, you just need the talent, and there's only so much interior scoring talent in the league right now. There's more now than there was a decade ago, it's been filtering back in... and that brings us to the talent cycle and how everything old is new again, haha!

Now, that said, we are starting to see larger guys with shooting ability, which is sort of both worlds joining together. You have yourself someone like Embiid (when he's healthy), for example, and life starts looking a lot different.

And I also think the world of basketball would have a very different narrative over Morey's idea of shot distribution and the mid-range if Kawhi had been healthy, because it's a copycat league and Kawhi is a guy who helps prove that the mid-range is still QUITE a viable scoring space (as does KD) if you're good enough at it... the way Joker and Embiid prove the same about the post.

You always give up SOMETHING. You go small, it's easier to attack you on the boards and with interior offense. You go large, it's generally harder to keep up the tempo, and historically to space the floor, though obviously that's starting to change quite dramatically.

But like, in retrospect, you can see this spacing aspect changing basketball from ages and ages back. I could go to a comparatively recent event and look at Horry's impact on Olajuwon, Shaq and Duncan... but we could look at Sikma. Bob McAdoo. How Bill Walton was pulled from the interior to be a high post passing hub.

It's just it happened SO FAST, ironically. Right? People need time to adjust. They've been bitching about the 3pt line for 45 years already, calling it a gimmick, condescending towards those who use it. Reggie, Ray Ray, looked as lesser because they weren't slashers, for example.

It took Steph winning a title for Barkley to get over his "jump shooting teams don't win titles" mantra, and even that happened with Golden State being the best D in the league... a lot like Boston, 8 years later (okay, they were 3rd, but still), and he still leans a little heavy on the older generation mantras, right?

[Doctor MJ]

So this is all good stuff to point out! Let me note though:

1. While Stockton looks amazing by nbarapm given his age, I wouldn't say that the indicators are giving him a clearly better overall RAPM than Nash. By the 5 year models you're looking at, Nash still has the higher absolute value, and while I'd urge people to avoid using absolute values comparing different epochs as if they are apples-to-apples comparisons, just in general, Nash's prime numbers are looking about as good as the numbers we have for Stockton.

(Incidentally, for me nbarapm defaults to the 4-year model which to me raises the question of what time frame we should be most interested in. Obviously it depends on the player which time frame will best measure him, but among those who are serious about using many years to reduce noise, I've seen 3, 4, and 5 year models being preferred relative to each other depending on the course.)

To be clear, there's definitely a good Stockton career argument to make that if he's as good in those later years as Nash's best, then his longevity should give him the career edge even if he didn't actually impact the game more during his prime, it's just that's not actually the same thing as having evidence that Stockton's defensive edge was making him more valuable overall than Nash.

2. But the scale of how good old man Stockton's DRAPM numbers look, concern me, because it's undeniable that he had some physical limitations compared to when he was younger, and yet looks basically right up there with the best perimeter defenders by these metrics with those limitations. What's up with that?

So, if we zoom in over at nbarapm, what we see is that Stockton's best DRAPM runs are focused on the years '99-00 through '01-02, when Stockton was age 37-39. As in, we have data beginning when Stockton was 34, but he appears to be peaking later than that. That's weird.

Now if we look at other defensive guards, how did they rank by DRAPM in the age 37-39 range in comparison?

Stockton 3.8, 3rd in NBA (for 3-year RAPM '99-00 - '01-02).
Kidd 0.9, 85th in NBA.
Lowry 0.5, 133rd in NBA (technically only age 36-38, because didn't turn 39 until late March this year)
Paul 0.5, 136th in NBA

This then is what's so weird. It's not strange to make the argument that Stockton was a defender on the level of these other dudes, but him appearing to be a defensive superstar by DRAPM this late in his career, doesn't fit with how anyone else looked.

And so that's the thing I think we have to focus on when just interpreting the data we have:

How real was Stockton's defensive data in this time period when he's super-old, he's not playing star minutes, his team doesn't have a Top 10 defense, and his team isn't a contender going deep into the playoffs each year?

If the question is: How could that data look good if he isn't actually that good? There I'd point to the stodgy nature of offensive strategy in the era. I think we're largely looking at data where teams weren't trying to target Stockton at a time when targeting Stockton probably would have worked pretty well even then, and would have worked even better today.

3. As I say all of that let me acknowledge:

The distinction between ORAPM and DRAPM isn't as clean as we'd like it and that's why I always emphasize that it's the overall RAPM we should focus on most. It's possible that lineup weirdness is pushing Stockton's offensive impact into the defensive measure - which could happen if he's disproportionately playing with defense-oriented lineups.

So perhaps his overall RAPM is fundamentally legitimate even though his DRAPM overrates his defense because of this weirdness.

I think it’s reasonable to be a little skeptical of just how good Stockton’s defensive impact numbers are. A few points I’d make on this though:

1. My point about this is more a directional one than one that relies on the precise numbers. Even if we did think old Stockton wasn’t quite as good defensively as the DRAPM data indicates, he’d still be quite a lot better defensively than Nash. And even discounting Stockton’s DRAPM a good bit, that gap would be quite significant.

2. Your point about the distinction between ORAPM and DRAPM not being super clean is a good one and might be at play here to at least some degree. I think one aspect that may be relevant here is that Stockton being on the court likely made the Jazz more slow and methodical on the offensive end, particularly when the other team was making runs. Basically, a calming influence from a veteran PG—which can have a significant effect on the defense, since panicking teams often end up being killed in transition and whatnot. Just a theory, but this is the type of thing that you’re talking about where the distinction between ORAPM and DRAPM isn’t clean. I think it’s perfectly plausible that things Stockton was doing on the offensive end were bleeding into DRAPM instead of ORAPM, contributing to DRAPM that might feel a little too high.

3. While Nash’s numbers in the NBArapm data set do look about as good as old Stockton, I think it’s reasonable to infer that old Stockton was probably not Stockton’s RAPM peak though, so that makes me pretty bullish on the idea that prime Stockton was more impactful than prime Nash.

4. I also note that there’s other RAPM sets that are even better for old Stockton. For instance, in TheBasketballDatabase, old Stockton has higher five-year RAPM than Nash ever did. As another example, the NBArapm website actually has its separate “6Factor” RAPM. That starts at 2000, so it only has four years for Stockton, but Stockton’s four-year RAPM in that measure is 7.7, which is higher than Nash’s best four-year RAPM of 6.6. Similarly, in Cheema’s five-year RAPM, old Stockton has two five-year spans that are above any five-year span for Nash, and another span that is above all but one of Nash’s spans. Overall, I think the RAPM picture is better for old Stockton than it is for prime Nash. And it’s hard for me to look at that and conclude that prime Stockton wouldn’t be ahead of prime Nash. We don’t have much data for prime Stockton, so it’s possible, but the RAPM snippets we have from those earlier years look good for Stockton, so it doesn’t feel likely.

Okay by point:

1. Even if Stockton's defense is a "quite a lot better" than Nash, if it's less than the DRAPM indicates, then it's less than the overall indicators we have for Nash, which renders a prime superiority argument supporting Stockton based on the assumption that whatever Stockton's numbers were late in his career, they must have been better in his prime (point 3).

2. It would absolutely make sense if old man Stockton slowed the Jazz down, but the Pace data if anything suggests the opposite.

3. Okay so while the logic of expecting a player's impact numbers to be better nearer in his prime absolutely makes sense, I'm not suggesting this might not be the case simply because we don't know it. Here's a simple stat I track:

Raw +/- leaders for a team each year, here's what we have for Utah during Stockton's career:

2003 - Ostertag
2002 - Kirilenko
2001 - Stockton
2000 - Stockton
1999 - Malone
1998 - Malone
1997 - Malone
1996 - Stockton (RS only)
1995 - Malone (RS only)
1994 - Malone (RS only)

So we note that that '00-02 samples where Stockton's DRtg really spikes coincide with the only time frame we know of where Stockton leads the team in +/- twice in a row.

My perspective here is shaped because we got access to the later years first. I remember us looking back on data that went back I believe to 2001 wondering if Stockton would continue to hold this lead when we had more data - back to the time period when the Jazz were contenders, with Malone absolutely at the fulcrum of their offense (compared to earlier in their run together) and winning MVPs that nobody was taking Stockton seriously for.

The thought wasn't so much "X couldn't possibly be true" so much as that if it turned out Stockton just kept being the guy leading this stat all the way back, then Stockton really should have been the MVP candidate from the team.

But as soon as the '90s data came for those critical years, they largely favored Malone, which made it all the harder to argue against conventional wisdom.

We still don't have access to the earlier years where Stockton put up his biggest box scores and so time will tell on that, but I do think the argument of the MVP of those Jazz finals teams is settled in favor of Malone.

And that makes it all the more interesting to consider that while Stockton was ranking #1 in DRAPM circa '00-02, Malone was ranking 576th and with an overall of 138th. Did Malone seriously get that much worse that quickly?

Well, it's definitely possible. These things can absolutely happen even without glaring injuries when you're talking about high primacy guys. It's like you're playing the same poker strategy you always were, but all of a sudden it's effectiveness goes from net gain to net loss...multiplied by however long you keep at it.

The thing here in relation to Malone & Stockton is that we're largely talking about - from my eyeballing of the lineup data - is a team that wasn't staggering the stars so much as they were just playing Stockton as much as they felt they could handle, and then Malone & company would play the rest.

I would say when they started doing this - cutting Stockton's minutes - the lineups with Malone without Stockton still mostly gave the Jazz a net gain, but eventually Malone reached a point that without a partner as good as Stockton, it was a net loss.

I would also note then the aspect of it being possible for Player A to be the Raw +/- leader for the team while Player B is the higher RAPM player - all it takes is a situation where the team really functions best with Player B out there next to Player A, but then Player B needs to rest and so Player A has to slog through without his partner. In such a case, who should be the MVP? Context matters of course, but in general I'd side with Player A.

For the Malone vs Stockton comparison specifically though, we need to remember that this isn't about Stockton holding steady in RAPM while Malone dropped, but about Stockton actually seeing his RAPM numbers here rise as Malone and the Jazz fell from contention, which is precisely the opposite of the prediction of those numbers getting better as we go back in time.

Projecting superior performance in a metric based on the continuation of a trend makes sense...but the thing about Stockton is that continuing the trend we see won't actually lead to better performance in the deeper past. That's why it's so weird, and why we should think really hard about what to make of it.

[doogie_hauser]

Thug of a player and a toxic human being..

Rodman's autobiography was very candid and interested in how much dirty plays Stockton and Malone were allowed to get away with in the 80s and 90s.

[Djoker]

Fantastic thread that I somehow missed till now. Read through most of it.

I think my ranking of Stockton is somewhere around #30 all-time. Let's say #25-35 range more likely. I feel like he's one of those guys that dropped a lot on all-time lists in the last 10-15 years. With no championships or major accolades like MVP's, I guess it was only a matter of time before his stature falls.

A couple of factors in this debate that I haven't seen mentioned.

Stockton and the Jazz were really fortunate to have two superstars as healthy as Malone and Stockton. I don't think people appreciate just how few games these two guys missed. Stockton missed 22 games in his entire career, 4 in 1990 and 18 in 1998. Malone missed 50 games but 40 of them came in his final LA season in 2004 and in his 18 Utah seasons, he missed a total of just 10 games and never missed more than 2 games in any single season. When your two superstars are THAT healthy, you should be expected to win quite a few games. These two guys are outliers of outliers when it comes to durability. It also may partly explain the Jazz' struggles in the playoffs. Other teams may have stars missing more regular season games which affects their win totals (and thus seeding) as well as SRS but if healthy in the playoffs, those teams can suddenly find themselves on relatively stronger footing relative to the Jazz.

I'm not sure about this but Stockton may also genuinely be a player who loses value in the playoffs. Teams may have hunted him defensively putting him in tougher matchups where he had to guard bigger, more athletic players. Teams also play him as a scorer and actively looking for his own shot isn't something he was always comfortable with. He was a conservative player and as such defenses didn't need to worry about him scoring a lot of points and could focus their defensive attention on others, namely his teammate Malone.

[lessthanjake]

So this is all good stuff to point out! Let me note though:

1. While Stockton looks amazing by nbarapm given his age, I wouldn't say that the indicators are giving him a clearly better overall RAPM than Nash. By the 5 year models you're looking at, Nash still has the higher absolute value, and while I'd urge people to avoid using absolute values comparing different epochs as if they are apples-to-apples comparisons, just in general, Nash's prime numbers are looking about as good as the numbers we have for Stockton.

(Incidentally, for me nbarapm defaults to the 4-year model which to me raises the question of what time frame we should be most interested in. Obviously it depends on the player which time frame will best measure him, but among those who are serious about using many years to reduce noise, I've seen 3, 4, and 5 year models being preferred relative to each other depending on the course.)

To be clear, there's definitely a good Stockton career argument to make that if he's as good in those later years as Nash's best, then his longevity should give him the career edge even if he didn't actually impact the game more during his prime, it's just that's not actually the same thing as having evidence that Stockton's defensive edge was making him more valuable overall than Nash.

2. But the scale of how good old man Stockton's DRAPM numbers look, concern me, because it's undeniable that he had some physical limitations compared to when he was younger, and yet looks basically right up there with the best perimeter defenders by these metrics with those limitations. What's up with that?

So, if we zoom in over at nbarapm, what we see is that Stockton's best DRAPM runs are focused on the years '99-00 through '01-02, when Stockton was age 37-39. As in, we have data beginning when Stockton was 34, but he appears to be peaking later than that. That's weird.

Now if we look at other defensive guards, how did they rank by DRAPM in the age 37-39 range in comparison?

Stockton 3.8, 3rd in NBA (for 3-year RAPM '99-00 - '01-02).
Kidd 0.9, 85th in NBA.
Lowry 0.5, 133rd in NBA (technically only age 36-38, because didn't turn 39 until late March this year)
Paul 0.5, 136th in NBA

This then is what's so weird. It's not strange to make the argument that Stockton was a defender on the level of these other dudes, but him appearing to be a defensive superstar by DRAPM this late in his career, doesn't fit with how anyone else looked.

And so that's the thing I think we have to focus on when just interpreting the data we have:

How real was Stockton's defensive data in this time period when he's super-old, he's not playing star minutes, his team doesn't have a Top 10 defense, and his team isn't a contender going deep into the playoffs each year?

If the question is: How could that data look good if he isn't actually that good? There I'd point to the stodgy nature of offensive strategy in the era. I think we're largely looking at data where teams weren't trying to target Stockton at a time when targeting Stockton probably would have worked pretty well even then, and would have worked even better today.

3. As I say all of that let me acknowledge:

The distinction between ORAPM and DRAPM isn't as clean as we'd like it and that's why I always emphasize that it's the overall RAPM we should focus on most. It's possible that lineup weirdness is pushing Stockton's offensive impact into the defensive measure - which could happen if he's disproportionately playing with defense-oriented lineups.

So perhaps his overall RAPM is fundamentally legitimate even though his DRAPM overrates his defense because of this weirdness.

I think it’s reasonable to be a little skeptical of just how good Stockton’s defensive impact numbers are. A few points I’d make on this though:

1. My point about this is more a directional one than one that relies on the precise numbers. Even if we did think old Stockton wasn’t quite as good defensively as the DRAPM data indicates, he’d still be quite a lot better defensively than Nash. And even discounting Stockton’s DRAPM a good bit, that gap would be quite significant.

2. Your point about the distinction between ORAPM and DRAPM not being super clean is a good one and might be at play here to at least some degree. I think one aspect that may be relevant here is that Stockton being on the court likely made the Jazz more slow and methodical on the offensive end, particularly when the other team was making runs. Basically, a calming influence from a veteran PG—which can have a significant effect on the defense, since panicking teams often end up being killed in transition and whatnot. Just a theory, but this is the type of thing that you’re talking about where the distinction between ORAPM and DRAPM isn’t clean. I think it’s perfectly plausible that things Stockton was doing on the offensive end were bleeding into DRAPM instead of ORAPM, contributing to DRAPM that might feel a little too high.

3. While Nash’s numbers in the NBArapm data set do look about as good as old Stockton, I think it’s reasonable to infer that old Stockton was probably not Stockton’s RAPM peak though, so that makes me pretty bullish on the idea that prime Stockton was more impactful than prime Nash.

4. I also note that there’s other RAPM sets that are even better for old Stockton. For instance, in TheBasketballDatabase, old Stockton has higher five-year RAPM than Nash ever did. As another example, the NBArapm website actually has its separate “6Factor” RAPM. That starts at 2000, so it only has four years for Stockton, but Stockton’s four-year RAPM in that measure is 7.7, which is higher than Nash’s best four-year RAPM of 6.6. Similarly, in Cheema’s five-year RAPM, old Stockton has two five-year spans that are above any five-year span for Nash, and another span that is above all but one of Nash’s spans. Overall, I think the RAPM picture is better for old Stockton than it is for prime Nash. And it’s hard for me to look at that and conclude that prime Stockton wouldn’t be ahead of prime Nash. We don’t have much data for prime Stockton, so it’s possible, but the RAPM snippets we have from those earlier years look good for Stockton, so it doesn’t feel likely.

Okay by point:

1. Even if Stockton's defense is a "quite a lot better" than Nash, if it's less than the DRAPM indicates, then it's less than the overall indicators we have for Nash, which renders a prime superiority argument supporting Stockton based on the assumption that whatever Stockton's numbers were late in his career, they must have been better in his prime (point 3).

2. It would absolutely make sense if old man Stockton slowed the Jazz down, but the Pace data if anything suggests the opposite.

3. Okay so while the logic of expecting a player's impact numbers to be better nearer in his prime absolutely makes sense, I'm not suggesting this might not be the case simply because we don't know it. Here's a simple stat I track:

Raw +/- leaders for a team each year, here's what we have for Utah during Stockton's career:

2003 - Ostertag
2002 - Kirilenko
2001 - Stockton
2000 - Stockton
1999 - Malone
1998 - Malone
1997 - Malone
1996 - Stockton (RS only)
1995 - Malone (RS only)
1994 - Malone (RS only)

So we note that that '00-02 samples where Stockton's DRtg really spikes coincide with the only time frame we know of where Stockton leads the team in +/- twice in a row.

My perspective here is shaped because we got access to the later years first. I remember us looking back on data that went back I believe to 2001 wondering if Stockton would continue to hold this lead when we had more data - back to the time period when the Jazz were contenders, with Malone absolutely at the fulcrum of their offense (compared to earlier in their run together) and winning MVPs that nobody was taking Stockton seriously for.

The thought wasn't so much "X couldn't possibly be true" so much as that if it turned out Stockton just kept being the guy leading this stat all the way back, then Stockton really should have been the MVP candidate from the team.

But as soon as the '90s data came for those critical years, they largely favored Malone, which made it all the harder to argue against conventional wisdom.

We still don't have access to the earlier years where Stockton put up his biggest box scores and so time will tell on that, but I do think the argument of the MVP of those Jazz finals teams is settled in favor of Malone.

And that makes it all the more interesting to consider that while Stockton was ranking #1 in DRAPM circa '00-02, Malone was ranking 576th and with an overall of 138th. Did Malone seriously get that much worse that quickly?

Well, it's definitely possible. These things can absolutely happen even without glaring injuries when you're talking about high primacy guys. It's like you're playing the same poker strategy you always were, but all of a sudden it's effectiveness goes from net gain to net loss...multiplied by however long you keep at it.

The thing here in relation to Malone & Stockton is that we're largely talking about - from my eyeballing of the lineup data - is a team that wasn't staggering the stars so much as they were just playing Stockton as much as they felt they could handle, and then Malone & company would play the rest.

I would say when they started doing this - cutting Stockton's minutes - the lineups with Malone without Stockton still mostly gave the Jazz a net gain, but eventually Malone reached a point that without a partner as good as Stockton, it was a net loss.

I would also note then the aspect of it being possible for Player A to be the Raw +/- leader for the team while Player B is the higher RAPM player - all it takes is a situation where the team really functions best with Player B out there next to Player A, but then Player B needs to rest and so Player A has to slog through without his partner. In such a case, who should be the MVP? Context matters of course, but in general I'd side with Player A.

For the Malone vs Stockton comparison specifically though, we need to remember that this isn't about Stockton holding steady in RAPM while Malone dropped, but about Stockton actually seeing his RAPM numbers here rise as Malone and the Jazz fell from contention, which is precisely the opposite of the prediction of those numbers getting better as we go back in time.

Projecting superior performance in a metric based on the continuation of a trend makes sense...but the thing about Stockton is that continuing the trend we see won't actually lead to better performance in the deeper past. That's why it's so weird, and why we should think really hard about what to make of it.

I think these are all fair points/questions. And I wouldn’t say there’s concrete answers to a lot of the questions you’ve raised.

What I’d say though is that, in order for Nash to be better than Stockton, we’d basically need *all* of the following to be true:

1. Old Stockton’s DRAPM materially overrates his defense enough to negate the advantage old Stockton has over prime Nash overall in RAPM measures.

2. Old Stockton’s DRAPM overrating his defense is not just offensive impact bleeding over into DRAPM.

3. Prime Stockton was not more impactful than old Stockton.

It is actually possible that all those things are true. You raise some circumstantial reasons to wonder about it. But I’m just inclined to think it’s not most likely. Basically I think “RAPM over large samples is probably generally right” and “players are more impactful in their primes than in their late 30’s” are both pretty good baseline assumptions, and I definitely wouldn’t bet on the idea that they’re both wrong. That said, I do think it’s possible. Stockton’s defensive RAPM numbers do *feel* a little high, and “offensive impact bleeding over into DRAPM” is pretty speculative. Meanwhile, old Stockton being as impactful as prime Stockton feels more plausible when we recognize that old Stockton had lower MPG—we might imagine that old Stockton scaled down in impact by playing fewer minutes at a similar level of per-possession impact, whereas most players scale down in impact with age by playing similar minutes with lower levels of per-possession impact. It’s all plausible, but on balance I lean towards thinking it’s likely the case that either old Stockton’s RAPM is real or prime Stockton was more impactful than old Stockton (or both).

One thing I do want to also note on Malone and Stockton is that we do know that Stockton performs substantially better than Malone in the limited pre-1997 impact metrics we have. Malone performs very badly in Squared’s 1985-1996 RAPM, while Stockton does well. And Stockton performs better than Malone in Engelmann’s quarter-by-quarter RAPM approximation for the 1990s. Those are both not conclusive. Squared doesn’t have a particularly large amount of Jazz data, and Engelmann’s quarter-by-quarter RAPM approximation is inherently less precise than RAPM. But that data is at least consistent with Stockton being the higher-impact player. We do know that Malone had better on-off in 1994 and 1995, while Stockton had better on-off in 1996, so that probably gives us some insight into their RAPM in those years and it looks a bit better for Malone. But on balance I look at the pre-1997 data and lean towards a conclusion that Stockton was more impactful in those years too, especially when we know that he was more impactful in those later years.

[Doctor MJ]

I think it’s reasonable to be a little skeptical of just how good Stockton’s defensive impact numbers are. A few points I’d make on this though:

1. My point about this is more a directional one than one that relies on the precise numbers. Even if we did think old Stockton wasn’t quite as good defensively as the DRAPM data indicates, he’d still be quite a lot better defensively than Nash. And even discounting Stockton’s DRAPM a good bit, that gap would be quite significant.

2. Your point about the distinction between ORAPM and DRAPM not being super clean is a good one and might be at play here to at least some degree. I think one aspect that may be relevant here is that Stockton being on the court likely made the Jazz more slow and methodical on the offensive end, particularly when the other team was making runs. Basically, a calming influence from a veteran PG—which can have a significant effect on the defense, since panicking teams often end up being killed in transition and whatnot. Just a theory, but this is the type of thing that you’re talking about where the distinction between ORAPM and DRAPM isn’t clean. I think it’s perfectly plausible that things Stockton was doing on the offensive end were bleeding into DRAPM instead of ORAPM, contributing to DRAPM that might feel a little too high.

3. While Nash’s numbers in the NBArapm data set do look about as good as old Stockton, I think it’s reasonable to infer that old Stockton was probably not Stockton’s RAPM peak though, so that makes me pretty bullish on the idea that prime Stockton was more impactful than prime Nash.

4. I also note that there’s other RAPM sets that are even better for old Stockton. For instance, in TheBasketballDatabase, old Stockton has higher five-year RAPM than Nash ever did. As another example, the NBArapm website actually has its separate “6Factor” RAPM. That starts at 2000, so it only has four years for Stockton, but Stockton’s four-year RAPM in that measure is 7.7, which is higher than Nash’s best four-year RAPM of 6.6. Similarly, in Cheema’s five-year RAPM, old Stockton has two five-year spans that are above any five-year span for Nash, and another span that is above all but one of Nash’s spans. Overall, I think the RAPM picture is better for old Stockton than it is for prime Nash. And it’s hard for me to look at that and conclude that prime Stockton wouldn’t be ahead of prime Nash. We don’t have much data for prime Stockton, so it’s possible, but the RAPM snippets we have from those earlier years look good for Stockton, so it doesn’t feel likely.

Okay by point:

1. Even if Stockton's defense is a "quite a lot better" than Nash, if it's less than the DRAPM indicates, then it's less than the overall indicators we have for Nash, which renders a prime superiority argument supporting Stockton based on the assumption that whatever Stockton's numbers were late in his career, they must have been better in his prime (point 3).

2. It would absolutely make sense if old man Stockton slowed the Jazz down, but the Pace data if anything suggests the opposite.

3. Okay so while the logic of expecting a player's impact numbers to be better nearer in his prime absolutely makes sense, I'm not suggesting this might not be the case simply because we don't know it. Here's a simple stat I track:

Raw +/- leaders for a team each year, here's what we have for Utah during Stockton's career:

2003 - Ostertag
2002 - Kirilenko
2001 - Stockton
2000 - Stockton
1999 - Malone
1998 - Malone
1997 - Malone
1996 - Stockton (RS only)
1995 - Malone (RS only)
1994 - Malone (RS only)

So we note that that '00-02 samples where Stockton's DRtg really spikes coincide with the only time frame we know of where Stockton leads the team in +/- twice in a row.

My perspective here is shaped because we got access to the later years first. I remember us looking back on data that went back I believe to 2001 wondering if Stockton would continue to hold this lead when we had more data - back to the time period when the Jazz were contenders, with Malone absolutely at the fulcrum of their offense (compared to earlier in their run together) and winning MVPs that nobody was taking Stockton seriously for.

The thought wasn't so much "X couldn't possibly be true" so much as that if it turned out Stockton just kept being the guy leading this stat all the way back, then Stockton really should have been the MVP candidate from the team.

But as soon as the '90s data came for those critical years, they largely favored Malone, which made it all the harder to argue against conventional wisdom.

We still don't have access to the earlier years where Stockton put up his biggest box scores and so time will tell on that, but I do think the argument of the MVP of those Jazz finals teams is settled in favor of Malone.

And that makes it all the more interesting to consider that while Stockton was ranking #1 in DRAPM circa '00-02, Malone was ranking 576th and with an overall of 138th. Did Malone seriously get that much worse that quickly?

Well, it's definitely possible. These things can absolutely happen even without glaring injuries when you're talking about high primacy guys. It's like you're playing the same poker strategy you always were, but all of a sudden it's effectiveness goes from net gain to net loss...multiplied by however long you keep at it.

The thing here in relation to Malone & Stockton is that we're largely talking about - from my eyeballing of the lineup data - is a team that wasn't staggering the stars so much as they were just playing Stockton as much as they felt they could handle, and then Malone & company would play the rest.

I would say when they started doing this - cutting Stockton's minutes - the lineups with Malone without Stockton still mostly gave the Jazz a net gain, but eventually Malone reached a point that without a partner as good as Stockton, it was a net loss.

I would also note then the aspect of it being possible for Player A to be the Raw +/- leader for the team while Player B is the higher RAPM player - all it takes is a situation where the team really functions best with Player B out there next to Player A, but then Player B needs to rest and so Player A has to slog through without his partner. In such a case, who should be the MVP? Context matters of course, but in general I'd side with Player A.

For the Malone vs Stockton comparison specifically though, we need to remember that this isn't about Stockton holding steady in RAPM while Malone dropped, but about Stockton actually seeing his RAPM numbers here rise as Malone and the Jazz fell from contention, which is precisely the opposite of the prediction of those numbers getting better as we go back in time.

Projecting superior performance in a metric based on the continuation of a trend makes sense...but the thing about Stockton is that continuing the trend we see won't actually lead to better performance in the deeper past. That's why it's so weird, and why we should think really hard about what to make of it.

I think these are all fair points/questions. And I wouldn’t say there’s concrete answers to a lot of the questions you’ve raised.

What I’d say though is that, in order for Nash to be better than Stockton, we’d basically need *all* of the following to be true:

1. Old Stockton’s DRAPM materially overrates his defense enough to negate the advantage old Stockton has over prime Nash overall in RAPM measures.

2. Old Stockton’s DRAPM overrating his defense is not just offensive impact bleeding over into DRAPM.

3. Prime Stockton was not more impactful than old Stockton.

It is actually possible that all those things are true. You raise some circumstantial reasons to wonder about it. But I’m just inclined to think it’s not most likely. Basically I think “RAPM over large samples is probably generally right” and “players are more impactful in their primes than in their late 30’s” are both pretty good baseline assumptions, and I definitely wouldn’t bet on the idea that they’re both wrong. That said, I do think it’s possible. Stockton’s defensive RAPM numbers do *feel* a little high, and “offensive impact bleeding over into DRAPM” is pretty speculative. Meanwhile, old Stockton being as impactful as prime Stockton feels more plausible when we recognize that old Stockton had lower MPG—we might imagine that old Stockton scaled down in impact by playing fewer minutes at a similar level of per-possession impact, whereas most players scale down in impact with age by playing similar minutes with lower levels of per-possession impact. It’s all plausible, but on balance I lean towards thinking it’s likely the case that either old Stockton’s RAPM is real or prime Stockton was more impactful than old Stockton (or both).

One thing I do want to also note on Malone and Stockton is that we do know that Stockton performs substantially better than Malone in the limited pre-1997 impact metrics we have. Malone performs very badly in Squared’s 1985-1996 RAPM, while Stockton does well. And Stockton performs better than Malone in Engelmann’s quarter-by-quarter RAPM approximation for the 1990s. Those are both not conclusive. Squared doesn’t have a particularly large amount of Jazz data, and Engelmann’s quarter-by-quarter RAPM approximation is inherently less precise than RAPM. But that data is at least consistent with Stockton being the higher-impact player. We do know that Malone had better on-off in 1994 and 1995, while Stockton had better on-off in 1996, so that probably gives us some insight into their RAPM in those years and it looks a bit better for Malone. But on balance I look at the pre-1997 data and lean towards a conclusion that Stockton was more impactful in those years too, especially when we know that he was more impactful in those later years.

Good rundown.

Hey, can you link to the Squared and Englemann quarter studies?


Sent from my iPhone using Tapatalk

[lessthanjake]

Okay by point:

1. Even if Stockton's defense is a "quite a lot better" than Nash, if it's less than the DRAPM indicates, then it's less than the overall indicators we have for Nash, which renders a prime superiority argument supporting Stockton based on the assumption that whatever Stockton's numbers were late in his career, they must have been better in his prime (point 3).

2. It would absolutely make sense if old man Stockton slowed the Jazz down, but the Pace data if anything suggests the opposite.

3. Okay so while the logic of expecting a player's impact numbers to be better nearer in his prime absolutely makes sense, I'm not suggesting this might not be the case simply because we don't know it. Here's a simple stat I track:

Raw +/- leaders for a team each year, here's what we have for Utah during Stockton's career:

2003 - Ostertag
2002 - Kirilenko
2001 - Stockton
2000 - Stockton
1999 - Malone
1998 - Malone
1997 - Malone
1996 - Stockton (RS only)
1995 - Malone (RS only)
1994 - Malone (RS only)

So we note that that '00-02 samples where Stockton's DRtg really spikes coincide with the only time frame we know of where Stockton leads the team in +/- twice in a row.

My perspective here is shaped because we got access to the later years first. I remember us looking back on data that went back I believe to 2001 wondering if Stockton would continue to hold this lead when we had more data - back to the time period when the Jazz were contenders, with Malone absolutely at the fulcrum of their offense (compared to earlier in their run together) and winning MVPs that nobody was taking Stockton seriously for.

The thought wasn't so much "X couldn't possibly be true" so much as that if it turned out Stockton just kept being the guy leading this stat all the way back, then Stockton really should have been the MVP candidate from the team.

But as soon as the '90s data came for those critical years, they largely favored Malone, which made it all the harder to argue against conventional wisdom.

We still don't have access to the earlier years where Stockton put up his biggest box scores and so time will tell on that, but I do think the argument of the MVP of those Jazz finals teams is settled in favor of Malone.

And that makes it all the more interesting to consider that while Stockton was ranking #1 in DRAPM circa '00-02, Malone was ranking 576th and with an overall of 138th. Did Malone seriously get that much worse that quickly?

Well, it's definitely possible. These things can absolutely happen even without glaring injuries when you're talking about high primacy guys. It's like you're playing the same poker strategy you always were, but all of a sudden it's effectiveness goes from net gain to net loss...multiplied by however long you keep at it.

The thing here in relation to Malone & Stockton is that we're largely talking about - from my eyeballing of the lineup data - is a team that wasn't staggering the stars so much as they were just playing Stockton as much as they felt they could handle, and then Malone & company would play the rest.

I would say when they started doing this - cutting Stockton's minutes - the lineups with Malone without Stockton still mostly gave the Jazz a net gain, but eventually Malone reached a point that without a partner as good as Stockton, it was a net loss.

I would also note then the aspect of it being possible for Player A to be the Raw +/- leader for the team while Player B is the higher RAPM player - all it takes is a situation where the team really functions best with Player B out there next to Player A, but then Player B needs to rest and so Player A has to slog through without his partner. In such a case, who should be the MVP? Context matters of course, but in general I'd side with Player A.

For the Malone vs Stockton comparison specifically though, we need to remember that this isn't about Stockton holding steady in RAPM while Malone dropped, but about Stockton actually seeing his RAPM numbers here rise as Malone and the Jazz fell from contention, which is precisely the opposite of the prediction of those numbers getting better as we go back in time.

Projecting superior performance in a metric based on the continuation of a trend makes sense...but the thing about Stockton is that continuing the trend we see won't actually lead to better performance in the deeper past. That's why it's so weird, and why we should think really hard about what to make of it.

I think these are all fair points/questions. And I wouldn’t say there’s concrete answers to a lot of the questions you’ve raised.

What I’d say though is that, in order for Nash to be better than Stockton, we’d basically need *all* of the following to be true:

1. Old Stockton’s DRAPM materially overrates his defense enough to negate the advantage old Stockton has over prime Nash overall in RAPM measures.

2. Old Stockton’s DRAPM overrating his defense is not just offensive impact bleeding over into DRAPM.

3. Prime Stockton was not more impactful than old Stockton.

It is actually possible that all those things are true. You raise some circumstantial reasons to wonder about it. But I’m just inclined to think it’s not most likely. Basically I think “RAPM over large samples is probably generally right” and “players are more impactful in their primes than in their late 30’s” are both pretty good baseline assumptions, and I definitely wouldn’t bet on the idea that they’re both wrong. That said, I do think it’s possible. Stockton’s defensive RAPM numbers do *feel* a little high, and “offensive impact bleeding over into DRAPM” is pretty speculative. Meanwhile, old Stockton being as impactful as prime Stockton feels more plausible when we recognize that old Stockton had lower MPG—we might imagine that old Stockton scaled down in impact by playing fewer minutes at a similar level of per-possession impact, whereas most players scale down in impact with age by playing similar minutes with lower levels of per-possession impact. It’s all plausible, but on balance I lean towards thinking it’s likely the case that either old Stockton’s RAPM is real or prime Stockton was more impactful than old Stockton (or both).

One thing I do want to also note on Malone and Stockton is that we do know that Stockton performs substantially better than Malone in the limited pre-1997 impact metrics we have. Malone performs very badly in Squared’s 1985-1996 RAPM, while Stockton does well. And Stockton performs better than Malone in Engelmann’s quarter-by-quarter RAPM approximation for the 1990s. Those are both not conclusive. Squared doesn’t have a particularly large amount of Jazz data, and Engelmann’s quarter-by-quarter RAPM approximation is inherently less precise than RAPM. But that data is at least consistent with Stockton being the higher-impact player. We do know that Malone had better on-off in 1994 and 1995, while Stockton had better on-off in 1996, so that probably gives us some insight into their RAPM in those years and it looks a bit better for Malone. But on balance I look at the pre-1997 data and lean towards a conclusion that Stockton was more impactful in those years too, especially when we know that he was more impactful in those later years.

Good rundown.

Hey, can you link to the Squared and Englemann quarter studies?


Sent from my iPhone using Tapatalk

Sure.

The Squared one is here: https://squared2020.com/2025/01/26/historical-rapm-1985-1996/

Engelmann’s quarter-by-quarter thing is here: https://web.archive.org/web/20150329072440/http://stats-for-the-nba.appspot.com/ratings/90s.html

On the Squared one, I do want to note two things:

- Squared has updated this in the time since I made the OP in this thread, so I notice that Stockton’s ranking is very slightly different than I described there.

- If you scroll down to the very bottom, it tells you how these teams did in the sampled games. The Jazz went 40-60 in the sampled games, during a set of years in which they went 604-380. So they did quite a lot worse in the sampled games than they did overall. I imagine that probably means both Stockton and Malone are underrated in this measure, though we don’t know for sure (and the sample of games for them is small enough that it’s also just noisy in general).

[Doctor MJ]

Sure.

The Squared one is here: https://squared2020.com/2025/01/26/historical-rapm-1985-1996/

Engelmann’s quarter-by-quarter thing is here: https://web.archive.org/web/20150329072440/http://stats-for-the-nba.appspot.com/ratings/90s.html

On the Squared one, I do want to note two things:

- Squared has updated this in the time since I made the OP in this thread, so I notice that Stockton’s ranking is very slightly different than I described there.

- If you scroll down to the very bottom, it tells you how these teams did in the sampled games. The Jazz went 40-60 in the sampled games, during a set of years in which they went 604-380. So they did quite a lot worse in the sampled games than they did overall. I imagine that probably means both Stockton and Malone are underrated in this measure, though we don’t know for sure (and the sample of games for them is small enough that it’s also just noisy in general).

Awesome, thank you!

I'm going to look more into Squared's data right now, and yeah, I had already noted that with his selected games the Jazz are a 40% winning team but in reality they were more than a 60% winning team, which I think is a big deal.

Re: Englemann. Where does it say this was quarter-by-quarter? I had thought his '90s XRAPM work was based on a box score hybrid.

To be clear, I ask on that specifically because I'm very interested in a quarter-by-quarter analysis, but generally I'm not all that interested in his XRAPM work.

[lessthanjake]

Sure.

The Squared one is here: https://squared2020.com/2025/01/26/historical-rapm-1985-1996/

Engelmann’s quarter-by-quarter thing is here: https://web.archive.org/web/20150329072440/http://stats-for-the-nba.appspot.com/ratings/90s.html

On the Squared one, I do want to note two things:

- Squared has updated this in the time since I made the OP in this thread, so I notice that Stockton’s ranking is very slightly different than I described there.

- If you scroll down to the very bottom, it tells you how these teams did in the sampled games. The Jazz went 40-60 in the sampled games, during a set of years in which they went 604-380. So they did quite a lot worse in the sampled games than they did overall. I imagine that probably means both Stockton and Malone are underrated in this measure, though we don’t know for sure (and the sample of games for them is small enough that it’s also just noisy in general).

Awesome, thank you!

I'm going to look more into Squared's data right now, and yeah, I had already noted that with his selected games the Jazz are a 40% winning team but in reality they were more than a 60% winning team, which I think is a big deal.

Re: Englemann. Where does it say this was quarter-by-quarter? I had thought his '90s XRAPM work was based on a box score hybrid.

To be clear, I ask on that specifically because I'm very interested in a quarter-by-quarter analysis, but generally I'm not all that interested in his XRAPM work.

My understanding is just based on others linking to it and describing it that way. Not sure exactly when I first had it described to me, but here is an example of another poster talking about it: https://forums.realgm.com/boards/viewtopic.php?f=64&t=2298723&p=107408289&hilit=Appspot#p107408289. I will note that it’s certainly possible my understanding of it is wrong—I’m just portraying it based on my understanding of posts from longer-standing posters who were apparently around when it was created.

[eminence]

I believe we had a discussion that touched on JEs quarter by quarter thing in the #1 thread of the '23 top 100.

And here is an apbr thread from JE: https://www.apbr.org/metrics/viewtopic.php?f=2&t=8067

[Doctor MJ]

I believe we had a discussion that touched on JEs quarter by quarter thing in the #1 thread of the '23 top 100.

And here is an apbr thread from JE: https://www.apbr.org/metrics/viewtopic.php?f=2&t=8067

Okay, yeah it doesn't sound like what I was hoping for.

I was hoping that he'd had the minute totals for each player for each quarter along with the total score from each quarter. Sounds like he's trying to estimate player minute totals, which is a reasonable next best thing, but makes the whole study that much more iffy.

[lessthanjake]

I believe we had a discussion that touched on JEs quarter by quarter thing in the #1 thread of the '23 top 100.

And here is an apbr thread from JE: https://www.apbr.org/metrics/viewtopic.php?f=2&t=8067

Okay, yeah it doesn't sound like what I was hoping for.

I was hoping that he'd had the minute totals for each player for each quarter along with the total score from each quarter. Sounds like he's trying to estimate player minute totals, which is a reasonable next best thing, but makes the whole study that much more iffy.

Yeah, agreed. I’m not sure I quite appreciated the fact that he didn’t have each player’s minutes by quarter and was instead doing an estimate (though I *think* I have actually seen that apbr thread before, so I probably should’ve internalized that fact prior to reading through it again now). While I never regarded it as more than a rough/flawed measure, this does make the whole thing have even more potential error in it than I was thinking it had. I don’t think it’s useless, particularly given that there’s not much to work with from that era, but I don’t think it should be regarded as super accurate. For purposes of this thread, I think that ratchets down my confidence that Stockton was more impactful than Malone in the pre-1997 era. On balance, I still do lean that direction, but obviously if you make one of the pro-Stockton data points less reliable than I was thinking it was, then my confidence in a pro-Stockton conclusion has to go down at least a bit.

[Doctor MJ]

Okay, I've done some looking specifically at Malone & Stockton +/- data for everything we have.

A caveat: As far as I can tell, Squared has not released a complete list of all years he's using separately.

So based on '84-85, '87-88, '89-90, '90-91, '92-93 & '95-96, his tallies would be:

Stockton +78
Malone -77

However, in his '85-96 cume, he's got:

Stockton +114
Malone -69

I have no skepticism here about his data, I'll just note, I wasn't able to reverse engineer his total tally just by looking at the specific years I see on his page.

It's worth noting that we have the actual (Pollack) regular season data for '95-96, so we can compare that with the sample Squared has.

Squared '95-96:

Malone -23
Stockton -14

Pollack '95-96:

Malone +659
Stockton +673

So in that case, Stockton has the edge in both cases, but clearly the Pollack sample is vastly bigger, and it's also WAY positive for both which again indicates what kind of Jazz games he's been likely to scout so far - he seems to be picking matchups with elite teams, where the Jazz tend to lose, despite the Jazz generally being a winning team.

Okay, if I just drop Squared '95-96, and use the rest of his tally with the data we have from Pollack and PBP era, here are the career tallies we have for them:

Malone +4720
Stockton +4866

So, edge to Stockton.

If we break it down without Squared's data:

Malone +4766
Stockton +4738

Slight edge to Malone.

If we break it down by decade:

1980s (pure Squared):

Malone 0
Stockton +48

1990s (all 3 sources minus Squared '95-96):

Malone +3699
Stockton +3316

Edge to Malone.

2000s

Malone +1013
Stockton +1473

Major edge to Stockton.

I think it's also good to specifically separate out '93-94 to '96-97 because that's the full season data we have in the seasons where Stockton is playing star minutes close to Malone's.

If we do that we get a tally of:

Malone +2732
Stockton +2395

Edge again to Malone.

I don't want to talk as if all of this data should lead to any one indisputable conclusion, I hope people look at it for themselves.

For myself, just in this comparison between the Jazzmen, it's hard for me not to focus on the mid-to-late '90s era as what really matters the most, because that's when the Jazz were actual contenders. In that period we first see Malone & Stockton playing major minutes, and Malone leading the raw +/-, followed by some time of continued contention with Stockton playing more limited minutes, and Malone continuing to lead the raw +/-, followed by the end of Jazz contention, where Malone's numbers really fall off but Stockton's at the very least continue to look good.

I think from this it's entirely possible that Stockton was the more impactful player both early in their Stockton/Malone run and late in the run with Malone only surpassing Stockton's impact in the middle...but again, it's that middle that made the Jazz really matter.

Now all the debate around Stockton tends to focus on his comparison with other point guards and I'm not trying to avoid that. Further, I'll say in theory it's possible for Stockton to have been more impactful than any of the other point guards we've been discussing, but not as much so as Malone in prime.

But I think it's important to remember that with a lot of the later era data, what we end up seeing is something like:

Stockton > Other Elite PGs > Malone

And when we see this, I think it's important to recognize that if you're ranking Stockton ahead of other point guards based on this data, but ranking Malone them, then you've got an awkward triangle because of Malone's advantage over Stockton during the contending years.

Additionally, if you grant Malone slack because he was out of prime in order to justify something like:

Malone > Stockton > Other Elite PGs

Then I do think you have to grapple with why a prime Stockton with a teammate better than himself didn't lead to absolutely insane teams. I understand people tend to think "They played the Jordan Bulls, what are you going to do?", but I think that tends to presuppose that the Jazz performed as well as you could possibly expect any team to do against the Bulls in the playoffs, but I would not say this was so as in neither of those Chicago-Utah finals years did we see the Jazz put up ORtg's that were high relative to other Bulls opponents.

Just generally, my view is that the Jazz peaked when Sloan maxed out his offensive scheme against illegal defense rules, but this led to a predictable offense that a great defense in the playoffs could really stifle. While we can always wonder what might have been if Stockton had been given carte blanche to do whatever he wanted, as things actually happened, the Jazz at their peak were a Malone-based machine that didn't look that amazing offensively against top competition.

Circling back to the data I posted before here:

None of this means that I doubt that Stockton had a greater RS RAPM VORP than the point guards my spreadsheet is telling me he did, but in terms of indicators that would lead me to conclude that at the very best of Utah, Stockton's offensive impact was ultra-elite, not so much. What I see is a core that at their best used an offensive scheme based not-on-Stockton, and which didn't translate super-well against elite playoff competition.

And that's why I struggle with what to do with Stockton appearing to have night & day impact as an old man, playing in limited minutes, on a team that isn't a contending threat. Fundamentally, it doesn't seem to be reasonable at all to assume that the Jazz figured something out in the '00s with old Stockton that had held them back in the '90s, because the team really did achieve less in the '00s.

This then leads me to conclude that the most likely explanation for Stockton's RAPM data spiking super-late in his career has more to do with all the strategy being focused on an aging Malone, and his loss in effectiveness giving top teams the edge they need so that didn't need to focus on attacking Stockton like it sure seems like they could have.

Based on what's been postulated before, this has me in the camp where I'd thinking old Stockton's DRAPM is probably legit in the sense that there was no magical defensive lineup surrounding Stockton that moved his offensive impact into the DRAPM category, but questionable in the sense that I just don't believe that old man Stockton was a bullet proof defender.

Were it the case that we saw similar Stockton DRAPM numbers prior to that then the focus on "old man Stockton" wouldn't make the same kind of sense, but as it is, we really do have to grapple with why old Stockton's DRAPM spikes.

Now by contrast, I expect that we haven't necessarily seen peak Stockton ORAPM - my guess is that his arc there looks more like a normal player - and that peak might make up for the fact that his prime DRAPM looks pretty mortal...but this isn't something I feel comfortable just assuming.

[lessthanjake]

For myself, just in this comparison between the Jazzmen, it's hard for me not to focus on the mid-to-late '90s era as what really matters the most, because that's when the Jazz were actual contenders. In that period we first see Malone & Stockton playing major minutes, and Malone leading the raw +/-, followed by some time of continued contention with Stockton playing more limited minutes, and Malone continuing to lead the raw +/-, followed by the end of Jazz contention, where Malone's numbers really fall off but Stockton's at the very least continue to look good.

I think from this it's entirely possible that Stockton was the more impactful player both early in their Stockton/Malone run and late in the run with Malone only surpassing Stockton's impact in the middle...but again, it's that middle that made the Jazz really matter.

As an initial matter, I’ll say that there’s a lot of good info in your post and some great discussion. I’m just going to respond to a few things that I don’t necessarily agree with.

The first one is this thing about the time period where the Jazz were the best being what matters most. I’m not really sure I understand why we should take that view. The Jazz were at their best in those years because they actually had a pretty good supporting cast for once (mostly because of Hornacek). I’m not sure why we should largely discount Stockton’s/Malone’s impact in years where they just happened to have a worse supporting cast. If it’s the case that Malone happened to be better than Stockton in the few years where the supporting cast happened to be pretty good, and Stockton was typically better than Malone otherwise, then why wouldn’t we conclude that Stockton was generally better than Malone? The years where the Jazz had a good supporting cast may have mattered the most from the perspective of the team trying to win rings, but it shouldn’t really matter more from the perspective of identifying who was better over the course of their careers.

Then I do think you have to grapple with why a prime Stockton with a teammate better than himself didn't lead to absolutely insane teams. I understand people tend to think "They played the Jordan Bulls, what are you going to do?", but I think that tends to presuppose that the Jazz performed as well as you could possibly expect any team to do against the Bulls in the playoffs, but I would not say this was so as in neither of those Chicago-Utah finals years did we see the Jazz put up ORtg's that were high relative to other Bulls opponents.

A few things here:

1. The Jazz did actually have an incredible team when their supporting cast wasn’t garbage. As mentioned in my OPs, prior to getting Hornacek, the Jazz supporting cast did not have a single significant player who has a positive career BPM, nor did any of them have a positive BPM for the Jazz in that 1988-1994 time period where Stockton/Malone were stars and Hornacek wasn’t there (well, he arrived near the end of 1994, but mostly wasn’t there that year). Once they actually got a decent team, they won 60, 55, 64, and 62 games in the next four seasons. They had SRS of 7.76, 6.25, 7.97, and 5.73 (and that latter number was much higher in the games Stockton played). In the playoffs, those Jazz put up the two highest three-year playoff relative net ratings of any team that did not win a title at some point in the three-year span. They were a great team! Probably the easiest answer for best team to not win a title!

2. As for how they played the Bulls, I disagree that they didn’t perform as well as we could expect. The 1997 Jazz were barely outscored by the Bulls in the series. The Bulls won one of those games on a Jordan buzzer-beater, and another off of a Kerr game-winner with a few seconds left. I recently posted a ranking of the Finals SRS (relative to opponent regular season SRS) for each Finals loser (https://forums.realgm.com/boards/viewtopic.php?f=64&t=2462708) and the 1997 Jazz came out in the #2 spot. Given that, it’s hard to say they didn’t do as well as one could reasonably expect. The 1998 Jazz didn’t do so great in that particular measure, but that’s basically just the result of one blowout loss. As it was, the 1998 Jazz were just some poetic Jordan heroics away from hosting a Game 7. And I can just say anecdotally that when I watched that series at the time, near the end of Game 6 was the only time in any of those Bulls Finals wins that I genuinely thought the Bulls were likely to lose the series. To me, how the Jazz did against the Bulls is actually a feather in their cap.

3. You mention offense specifically against the Bulls, but I think the whole picture is important here. That’s both because playoff series take on their own tenor in terms of physicality and whatnot, but also because Stockton was not a one-way star. If the Jazz did really well overall against the Bulls (and, as described above, I think they did), then I’m not sure that it really matters whether we think that was led by their offense or their defense, since Stockton was an impactful player on both sides of that.

And that's why I struggle with what to do with Stockton appearing to have night & day impact as an old man, playing in limited minutes, on a team that isn't a contending threat. Fundamentally, it doesn't seem to be reasonable at all to assume that the Jazz figured something out in the '00s with old Stockton that had held them back in the '90s, because the team really did achieve less in the '00s.

I don’t think there’s any reason that the data suggests the 2000s Jazz figured anything out. It’s not like the Jazz did better with Stockton on the floor in the 2000s than they did with him on the floor in the 1990s. In Stockton’s seasons in the 2000s, they had a +7.9 net rating with Stockton on the floor. If we look at the 1994-96 on-off data we have, as well as the normal on-off data from 1997-1999, we can see the Jazz averaging being about +10 with Stockton on the floor in those years. So the 1990s Jazz were very likely better with Stockton on the floor than the 2000s Jazz were. They seem to have done worse with Stockton off too. Basically, it seems like the 1990s Jazz were better with Stockton on the floor than the 2000s Jazz were, and were also considerably better with Stockton off the floor than the 2000s Jazz were. They were just less good in general in the 2000s, though they might well have fallen off more with Stockton off the court in the 2000s than they did in the 1990s (which probably does have to do with Malone declining).

This then leads me to conclude that the most likely explanation for Stockton's RAPM data spiking super-late in his career has more to do with all the strategy being focused on an aging Malone, and his loss in effectiveness giving top teams the edge they need so that didn't need to focus on attacking Stockton like it sure seems like they could have.

Malone declining is almost certainly a thing. But I don’t really see the same “RAPM data spiking super-late in [Stockton’s] career” that you do. Stockton’s RAPM data looks really good throughout. Sure, we could look at something like two-year RAPM having his best years be 2000-2001 and say it’s a bit odd. But Stockton was ranked highly in RAPM prior to that. For instance, his three-year RAPM from 1997-1999 was ranked 5th in the NBA by one measure and 9th by another. He does later go up to 1st, but it already started at a good-looking spot prior to that.

And I think we should always just keep in the back of our minds that RAPM is noisy, especially when not in large samples (and these aren’t large samples, since we’re cutting up into 2 or 3 year values here). So I personally wouldn’t draw too much of a conclusion based on a guy going from top 5-10 in one span to #1 in another span. I’d tend to just smooth that out in my brain a bit, to conclude that the truth of his impact probably didn’t *really* change much between those years and that there likely was just some variance at play. Put differently, what happened there is basically just downstream of him having his two best on-offs in 2001 and 2000 (but, I should note, not actually better net ratings while on than he had had the prior two years). I guess we could question why that was, but I think randomness is usually a perfectly reasonable explanation for changes in on-off in individual years, especially when the difference comes exclusively from changes in the OFF value (which has the smaller sample). Randomness may not be the only explanation, but I don’t think the fact pattern here actually requires us to take any unflattering conclusion about Stockton.

[Doctor MJ]

For myself, just in this comparison between the Jazzmen, it's hard for me not to focus on the mid-to-late '90s era as what really matters the most, because that's when the Jazz were actual contenders. In that period we first see Malone & Stockton playing major minutes, and Malone leading the raw +/-, followed by some time of continued contention with Stockton playing more limited minutes, and Malone continuing to lead the raw +/-, followed by the end of Jazz contention, where Malone's numbers really fall off but Stockton's at the very least continue to look good.

I think from this it's entirely possible that Stockton was the more impactful player both early in their Stockton/Malone run and late in the run with Malone only surpassing Stockton's impact in the middle...but again, it's that middle that made the Jazz really matter.

As an initial matter, I’ll say that there’s a lot of good info in your post and some great discussion. I’m just going to respond to a few things that I don’t necessarily agree with.

The first one is this thing about the time period where the Jazz were the best being what matters most. I’m not really sure I understand why we should take that view. The Jazz were at their best in those years because they actually had a pretty good supporting cast for once (mostly because of Hornacek). I’m not sure why we should largely discount Stockton’s/Malone’s impact in years where they just happened to have a worse supporting cast. If it’s the case that Malone happened to be better than Stockton in the few years where the supporting cast happened to be pretty good, and Stockton was typically better than Malone otherwise, then why wouldn’t we conclude that Stockton was generally better than Malone? The years where the Jazz had a good supporting cast may have mattered the most from the perspective of the team trying to win rings, but it shouldn’t really matter more from the perspective of identifying who was better over the course of their careers.

Then I do think you have to grapple with why a prime Stockton with a teammate better than himself didn't lead to absolutely insane teams. I understand people tend to think "They played the Jordan Bulls, what are you going to do?", but I think that tends to presuppose that the Jazz performed as well as you could possibly expect any team to do against the Bulls in the playoffs, but I would not say this was so as in neither of those Chicago-Utah finals years did we see the Jazz put up ORtg's that were high relative to other Bulls opponents.

A few things here:

1. The Jazz did actually have an incredible team when their supporting cast wasn’t garbage. As mentioned in my OPs, prior to getting Hornacek, the Jazz supporting cast did not have a single significant player who has a positive career BPM, nor did any of them have a positive BPM for the Jazz in that 1988-1994 time period where Stockton/Malone were stars and Hornacek wasn’t there (well, he arrived near the end of 1994, but mostly wasn’t there that year). Once they actually got a decent team, they won 60, 55, 64, and 62 games in the next four seasons. They had SRS of 7.76, 6.25, 7.97, and 5.73 (and that latter number was much higher in the games Stockton played). In the playoffs, those Jazz put up the two highest three-year playoff relative net ratings of any team that did not win a title at some point in the three-year span. They were a great team! Probably the easiest answer for best team to not win a title!

2. As for how they played the Bulls, I disagree that they didn’t perform as well as we could expect. The 1997 Jazz were barely outscored by the Bulls in the series. The Bulls won one of those games on a Jordan buzzer-beater, and another off of a Kerr game-winner with a few seconds left. I recently posted a ranking of the Finals SRS (relative to opponent regular season SRS) for each Finals loser (https://forums.realgm.com/boards/viewtopic.php?f=64&t=2462708) and the 1997 Jazz came out in the #2 spot. Given that, it’s hard to say they didn’t do as well as one could reasonably expect. The 1998 Jazz didn’t do so great in that particular measure, but that’s basically just the result of one blowout loss. As it was, the 1998 Jazz were just some poetic Jordan heroics away from hosting a Game 7. And I can just say anecdotally that when I watched that series at the time, near the end of Game 6 was the only time in any of those Bulls Finals wins that I genuinely thought the Bulls were likely to lose the series. To me, how the Jazz did against the Bulls is actually a feather in their cap.

3. You mention offense specifically against the Bulls, but I think the whole picture is important here. That’s both because playoff series take on their own tenor in terms of physicality and whatnot, but also because Stockton was not a one-way star. If the Jazz did really well overall against the Bulls (and, as described above, I think they did), then I’m not sure that it really matters whether we think that was led by their offense or their defense, since Stockton was an impactful player on both sides of that.

And that's why I struggle with what to do with Stockton appearing to have night & day impact as an old man, playing in limited minutes, on a team that isn't a contending threat. Fundamentally, it doesn't seem to be reasonable at all to assume that the Jazz figured something out in the '00s with old Stockton that had held them back in the '90s, because the team really did achieve less in the '00s.

I don’t think there’s any reason that the data suggests the 2000s Jazz figured anything out. It’s not like the Jazz did better with Stockton on the floor in the 2000s than they did with him on the floor in the 1990s. In Stockton’s seasons in the 2000s, they had a +7.9 net rating with Stockton on the floor. If we look at the 1994-96 on-off data we have, as well as the normal on-off data from 1997-1999, we can see the Jazz averaging being about +10 with Stockton on the floor in those years. So the 1990s Jazz were very likely better with Stockton on the floor than the 2000s Jazz were. They seem to have done worse with Stockton off too. Basically, it seems like the 1990s Jazz were better with Stockton on the floor than the 2000s Jazz were, and were also considerably better with Stockton off the floor than the 2000s Jazz were. They were just less good in general in the 2000s, though they might well have fallen off more with Stockton off the court in the 2000s than they did in the 1990s (which probably does have to do with Malone declining).

This then leads me to conclude that the most likely explanation for Stockton's RAPM data spiking super-late in his career has more to do with all the strategy being focused on an aging Malone, and his loss in effectiveness giving top teams the edge they need so that didn't need to focus on attacking Stockton like it sure seems like they could have.

Malone declining is almost certainly a thing. But I don’t really see the same “RAPM data spiking super-late in [Stockton’s] career” that you do. Stockton’s RAPM data looks really good throughout. Sure, we could look at something like two-year RAPM having his best years be 2000-2001 and say it’s a bit odd. But Stockton was ranked highly in RAPM prior to that. For instance, his three-year RAPM from 1997-1999 was ranked 5th in the NBA by one measure and 9th by another. He does later go up to 1st, but it already started at a good-looking spot prior to that.

And I think we should always just keep in the back of our minds that RAPM is noisy, especially when not in large samples (and these aren’t large samples, since we’re cutting up into 2 or 3 year values here). So I personally wouldn’t draw too much of a conclusion based on a guy going from top 5-10 in one span to #1 in another span. I’d tend to just smooth that out in my brain a bit, to conclude that the truth of his impact probably didn’t *really* change much between those years and that there likely was just some variance at play. Put differently, what happened there is basically just downstream of him having his two best on-offs in 2001 and 2000 (but, I should note, not actually better net ratings while on than he had had the prior two years). I guess we could question why that was, but I think randomness is usually a perfectly reasonable explanation for changes in on-off in individual years, especially when the difference comes exclusively from changes in the OFF value (which has the smaller sample). Randomness may not be the only explanation, but I don’t think the fact pattern here actually requires us to take any unflattering conclusion about Stockton.

More good thoughts!

Re: why should period with better supporting cast mean more when comparing star teammates? What a great question to specifically question.

I think if you're thinking about this in terms of pure goodness at basketball, ideally it shouldn't.

If you're talking about value and achievement, I'd say it does, and in practice, it's an affordance we can't really function without. I certainly approve though the spirit of trying to normalize all winning bias away, and yes, perhaps it would have been easier to build a champion around Stockton & Malone when they were younger and they just kept blowing it until they were older.

Re: better supporting cast, synthetic vs analytic approach. So a note here that occurred to me when writing this post which may be irrelevant to the dialogue between the two of us - thanks for bearing with me:

Basketball is a fluid team sport where so much of being good at it is about being able to synergize with teammates to elevate their effectiveness, and so choosing to prioritize based on how he played when his teams weren't that great over the times when he did have quality teammates will miss out on that synergy.

Re: better...how? So clearly the big get for the Jazz was Jeff Hornacek, and we actually do seem to have indicators possibly suggesting that Stockton was more impactful for a Jazz lineup than Malone before & after Hornacek, but Malone was the more impactful player when Hornacek was there.

It's obviously oversimplistic to say "It's all because of Hornacek's presence or absence", but I think what we can acknowledge is that a player's irreplaceability from within roster very much depends on who is on said roster, and such changes might have swung and swung back the Jazz hierarchy of irreplaceability.

Re: Jazz played well against the Bulls, outscored them in the 1997 series. True, but Stockton did not. I get saying that both teams offenses were struggling so it may not be that meaningful to point to the Jazz offense struggling...but the Jazz offense was specifically struggling with Stockton.

Small sample size of course, and Stockton actually has a higher ORtg than Malone the next year (when Stockton was starting to play less, and thus less without Malone).

Re: disagreement about existence of spike in late career Stockton RAPM data.

Well if we look at nbarapm's 2 year RAPM, Stockton goes from a 4.5 in 1998 to a 7.6 in 2001. There are some other related stats, but that's what I'm pointing to.

Is it just noise? I mean, if it is, then why would we use an extreme outlier value for an old man during a time of team irrelevance over the more age-appropriate data from the earlier span? If it's a meaningful improvement, we should feel a need to explain it. If it's noise, then we should not take it very seriously. Either way, we shouldn't be grabbing on to the weird outlier as if it represents the natural trendline for Stockton's career.

[lessthanjake]

A lot here I don’t disagree with, so I’ll just respond on a few things:

Re: better...how? So clearly the big get for the Jazz was Jeff Hornacek, and we actually do seem to have indicators possibly suggesting that Stockton was more impactful for a Jazz lineup than Malone before & after Hornacek, but Malone was the more impactful player when Hornacek was there.

It's obviously oversimplistic to say "It's all because of Hornacek's presence or absence", but I think what we can acknowledge is that a player's irreplaceability from within roster very much depends on who is on said roster, and such changes might have swung and swung back the Jazz hierarchy of irreplaceability.

I do think it is plausible that Stockton’s impact may have gone down when they brought on Hornacek, given that Hornacek was a capable playmaker. I don’t know that we truly have enough data to conclude in the first instance that Stockton’s impact *did* go down in those years, but you’ve brought up some circumstantial evidence to that effect, and I don’t think it’d be all that surprising an outcome.

That said, I don’t think we can really conclude Stockton’s impact tumbled a large amount—after all, if it did then I don’t think we’d expect the Jazz to have improved as much as they did when Hornacek arrived. The Jazz became *very* good! So this is probably more a point at the margins. And I’ll note that, even in 1997-1998, a pretty old Stockton on the Jazz with Hornacek was putting up great RAPM near the top of the league. So if that was his impact nadir, I’d say that’s really good!

Re: disagreement about existence of spike in late career Stockton RAPM data.

Well if we look at nbarapm's 2 year RAPM, Stockton goes from a 4.5 in 1998 to a 7.6 in 2001. There are some other related stats, but that's what I'm pointing to.

Is it just noise? I mean, if it is, then why would we use an extreme outlier value for an old man during a time of team irrelevance over the more age-appropriate data from the earlier span? If it's a meaningful improvement, we should feel a need to explain it. If it's noise, then we should not take it very seriously. Either way, we shouldn't be grabbing on to the weird outlier as if it represents the natural trendline for Stockton's career.

I think my response to this would be that I’m not really “us[ing] an extreme outlier value” for Stockton. For sake of completeness, I did list his rank in two-year RAPM in my OP, but I’ve focused more on longer-term RAPM than that, and definitely haven’t made arguments centered on the impact data from those specific years by themselves. Once we are looking at 4-year or 5-year RAPM, I don’t really think that probably-noisily-good on-off in two years is much of a problem for the data, since at that point we have enough data that the whole thing is not very noisy anymore.

In other words, I’d say Stockton being #1 in 2000-2001 two-year RAPM with an abnormally high value probably is in part a result of noise going in his favor in those years. But it’s a lot harder to argue that noise is a big factor in Stockton being #2 in 1997-2001 five-year RAPM. I don’t think the arguments for Stockton rely on noise, because we actually have enough great data for him to not be relying on particularly noisy samples. In contrast, I think an argument against Stockton that centers on changes in his impact in different small samples has a good chance of being an argument based around noise.

[lessthanjake]

.

One other thing I want to note about using the raw on-off data for Stockton/Malone in 1994-1996 (or even for the later years) is that the Jazz basically had all of Hornacek’s minutes be with Malone. Like, from 1997-1999, Hornacek only averaged about 100-120 minutes a year on the court without Malone. He had *way* more time on the court without Stockton.

This is presumably a big part of why Stockton has better RAPM than Malone in those late 1990s years, even though Malone has a better net rating on the court and a significantly better on-off. Malone was put on the court with better teammates. Just to put a point on this, from 1997-1999, the Jazz were +10.4 with Stockton on the court, and he had a +9.7 on-off. Meanwhile, in those years, the Jazz were +10.6 with Malone on, and he had a +18.1 on-off. Looking at that, we’d definitely expect Malone to have a better RAPM. And yet we look at 3-year RAPM on the NBArapm website for 1997-1999, and Karl Malone is at +3.2 (which was ranked 24th in the league), while Stockton was at +4.8 (which was ranked 9th in the league), and it’s a similar story with 1997-1998 two-year RAPM. They both fare a bit better in TheBasketballDatabase’s RAPM, but its three-year 1997-1999 RAPM has Stockton ranked 5th in the league and Malone ranked 8th. Heck, TheBasketballDatabase’s one-year RAPM even has Stockton ahead in 1997, when Stockton had a +8.7 ON and +7.6 on-off, while Malone had a +11.7 ON and a +21.9 on-off (though I’ll note that Engelmann’s single-season PI RAPM still has Malone a little ahead of Stockton).

So RAPM definitely thinks that Malone was put on the court with much better teammates to a degree that makes Stockton come out ahead despite having notably worse on-off. And that definitely checks out when it comes to the Jazz rotation strategy for Hornacek. And it extends to more than just Hornacek. Indeed, a quick check for Bryon Russell shows something pretty similar to Hornacek (though not quite as pronounced, he definitely did play much more with Malone than with Stockton). And it’s true for their starting centers too (whether Ostertag or Keefe). I don’t know for sure if the same rotation strategy was true in Hornacek’s pre-1997 years with the Jazz, but I imagine it was. In which case, looking at Pollack’s on-off and inferring what that’d mean about Stockton and Malone’s RAPM might actually give us a really skewed picture. If those years were anything like the years just after that, Stockton’s RAPM would probably be better in 1994-1996, even though the on-off looks a bit better for Malone.

I also imagine that this has a significant effect on your point about the Jazz not doing so great with Stockton on the court against the Bulls in those Finals.

[penbeast0]

If I remember right, the Jazz in those Hornacek years did try to have Stockton on with the reserves to give some shot creation, but I'm confused. Shouldn't Stockton playing with reserves give him a disadvantage for those impact numbers as the opposing teams could really focus on shutting him down without Malone and Hornacek there to provide secondary creation?

[lessthanjake]

If I remember right, the Jazz in those Hornacek years did try to have Stockton on with the reserves to give some shot creation, but I'm confused. Shouldn't Stockton playing with reserves give him a disadvantage for those impact numbers as the opposing teams could really focus on shutting him down without Malone and Hornacek there to provide secondary creation?

I think that is probably true in terms of affecting how effective Stockton could be, though the flip side of that is that it’s probably easier to be impactful on deficient lineups. I was making a more mechanical point than that though—which was simply that Stockton’s RAPM is better than Malone’s in those years even though his raw on-off data looks substantially worse. That’s generally not how it goes, and the reason for that in this case is surely because of the differences in teammates they were on the court with. This is notable because Doctor MJ and I had been discussing some points regarding how the Jazz did in raw plus-minus terms when Stockton was on the court compared to when Malone was. That included in 1994-1996 data we have from Pollack, as well as their respective plus-minuses in those Finals against the Bulls. When I looked at that 1994-1996 on-off data, it looked to me like a reasonable assumption that Malone’s RAPM would be better than Stockton’s in those years, because his on-off looks a bit better overall. But if there was such a difference in the lineups they were used with that Stockton had better RAPM in 1997-1999 despite having substantially less good-looking on-off, then there’s a good chance that the same would be true in those earlier years.