migya wrote:ShotCreator wrote:He almost certainly had the most career value. He was probably a top 10 guy all-time on CV but most fans are comfortably excluding him because of the lack of an explosive, dominant peak, traditionally speaking anyway.
Many here the last few years have gone on about these impact stats and values and that doesn't seem to be used with Stockton. His impact was huge.
Well to be clear, those of us who have been paying close attention to +/- stats since they started being available online, Stockton's intriguing numbers were one of the biggest drivers of curiosity for getting access to earlier seasons.
If memory serves, for a good while we only had data going back to '01-02, and that gradually propagated backwards to '96-97 because that's when the NBA first kept league-wide play-by-play stats. We then got access to the raw +/- data for all teams going back to '93-94 and the 76ers back to '76-77, but beyond that, any data we've gotten have been labors of love of current people with the most ambitious project being what Squared2020 is doing on his site. (Though shout out to ElGee for going all in on game-by-game WOWY into the deep past over a decade ago.)
So with that in mind, I'll say what my perspective on this has been - which isn't consensus, but has come from conversations with others:
The initial 2000s +/- data we had painted a picture of Stockton having far more RAPM-style "impact" than teammate Malone, and raised a clear set of questions relating to what we'd see if we had access to earlier data.
However, the fact that Stockton at the time was playing way less minutes than Malone is why I put impact in quotes above, not to discredit it, but just note that regression isn't going to account for a situation where Star A in limited minutes gets to play with Star B, but Star B has to play a lot without Star A. Star A in such circumstances will absolutely have better RAPM numbers, but this doesn't mean that he's necessarily adding more value than the guy fighting the good fight paying big minutes while the other guys rests.
Thus, what many of us wanted to see was a) how did the Jazz +/- data look specifically in their two finals years, and b) how did the Jazz +/- data look going back at least since the Jazz became the S&M gang in the late '80s.
Additionally, just from a real basketball perspective, there's the matter that because Stockton & Malone play different roles with different levels of defensive attention, it was very possible that their respective viabilities to add value in those roles fell off at very different rates. Volume scorers add value directly often by having decent efficiency (rather than great) multiplied by that massive volume, which means that sometimes even a small decrease in their efficiency can take them from major positive impact to outright negative impact. By contrast, a pass-first point guard who minimizes risk effectively can remain a net positive offensive player indefinitely.
This then to say, we wanted to see what prime Malone's data actually said both because of his role, and because it just didn't make logical sense to not think he was super-valuable when he was playing big minutes on those finals teams.
What do we now see? Well, I'll start of with just the raw +/- which I think is a great starting point - though others disagree. Here are the team leaders for the S&M Jazz in raw +/- going back to '93-94 (though pre-96-97 we don't have playoff data):
'02-03: Greg Ostertag
'01-02: Andrei Kirilenko
'00-01: John Stockton
'99-00: John Stockton
'98-99: Karl Malone
'97-98: Karl Malone
'96-97: Karl Malone
'95-96: John Stockton
'94-95: Karl Malone
'93-94: Karl Malone
So then I'll say, I'm afraid that from a S vs M perspective, this data to me told a story that contemporary thought from the '90s was correct to elevate Malone as the MVP candidate over Stockton, and I'd really hate to try to convince people that Stockton was actually the team's MVP in their finals years regardless of what RAPM data tells us. Analytically we know that a coach's lineup staggering can make the MVP have a lesser raw +/- so this data isn't the end-all be-all, but from a question of whether Stockton's massive impact edge over Malone in the '00s continued back to when they were a contending team, it didn't.
Now, from what I've seen, Squared2020's tracking from earlier than this looks quite good for Stockton, so we'll see what more is revealed with the blood, sweat & tears of guys like Squared.
Additionally, some have pushed back saying that we shouldn't necessarily focus on the best Jazz years to evaluate who the better player was between S & M. Logically, it's possible for one guy (Malone) to be more impactful with the right supporting cast acquisitions (most notably Hornacek), but for the other guy to be more impactful in most situations. I personally still tend to focus on the best years of a core to analyze the core, but I'll concede, this isn't an undebatable thing.
Final thing I'm going to speak to is what more specific data tells us about where Stockton's impact was coming from in those later years. If we go to nbarapm.com and look at the 4y RAPM that the player view defaults too, here's a list of some of the great guards with the best span we have for them:
Chris Paul (2015-2018) Off: 6.4, Def: 2.7, Tot: 9.1
Steve Nash (2005-2008) Off: 9.3, Def: -0.6, Tot: 8.8
John Stockton (1998-2001) Off: 4.7, Def: 3.6, Tot: 8.3
Jason Kidd (2002-2005) Off: 4.3, Def: 1.5, Tot: 5.9
I'll emphasize up front that we shouldn't lose site of the fact that Stockton being up there at all, so far ahead of peak Kidd, is incredibly impressive and I don't mean to imply otherwise.
But the defense is extremely high for any point guard, and we're talking about Stockton age 35 to 38 being compared with overall peak Paul & Kidd, and this raises a big concern for me when judging actual playing ability:
Do we really think that any player in the back half of his 30s couldn't have been exploited by opposing offenses if they were in a contested playoff series? I'm afraid I'm skeptical.
To be clear, I think Stockton was a great defensive point guard for his age his whole career, so this isn't mean to be a general-Stockton skepticism, but I'm afraid when I see data like this, it makes me think what regular season offenses just weren't geared toward using Stockton's age against him.
To some degree, this remains a truth today - it's only in the playoffs where opponents really focus in trying to find any and all weaknesses on your team - but there's also the whole Illegal Defense rules that basically defined Stockton's entire career, because even after the 2001 rule change, we didn't really see seriously see offensive adjustments until D'Antoni's Suns in 2004, by which time Stockton was retired.
Thing is, so long as the offense is built around manipulating the Illegal Defense rule, every defender who wasn't guarding the offenses alpha got to basically hang out outside of that action rather than being targeted themselves, and I'd say that was a boon for point guards on defense in no small part because very few teams used score-first point guards at the time.
By contrast, in the time since, offensive strategy has emphasized spacing & targeting through switches, and it's largely just become a truth that anyone with limitations in rapid deceleration (and re-acceleration) is going to need help when they are facing a super-agile ballhandler, and realistically, anyone in their 30s is going to be more limited than they were in their 20s.
This doesn't mean I think old man Stockton "would get played off the court because of defense" to be clear, because the fact that basically everyone needs help to defend against elite drivers in a spaced-out game, but it does mean that when I see Stockton's defense at this age being the thing that's helping him make up for not being an ultra-outlier on offense relative to other floor generals, I still have concerns as to limitations compared to what a face-value read of the data appears to say.
As I say all of this, I should always note that the overall RAPM is generally the thing to focus on above ORAPM & DRAPM specifically because ORAPM & DRAPM can get skewed based on a coach's lineup choices. As such, it's possible that weird lineup choices were causing old Stockton's DRAPM to get inflated at the cost of his ORAPM. If that's the case, then we shouldn't be taking Stockton's DRAPM at face value.
But what I'd maintain is that if old Stockton's overall value was built on having actual defensive impact, it would be my basketball assessment that this overrates what we'd expect from him when targeted as we'd expect him to have to deal with today. He's far from alone in this regard, but we're discussing him because he was such an outlier generally, and when I see the offensive vs defensive splits, it's the defense where he looks like the true outlier relative to guard-norms, and for the reasons I've given above, I think that warrants some consideration.
