A Study in the Dependence on Spacing of Rim Gravity Players

[lessthanjake]

One tool that is often looked at here to assess offensive players is playoff relative ORTG when a player is on the court. A potential counterpoint to that sort of data is often that one player enabled their team to play more defense-minded lineups. I’ve often made this point regarding Steph Curry—the idea being that his unique offensive profile allowed his team to play with fewer spacers while still doing well offensively. In a sense, some of the offensive impact of a player like that is manifesting itself in the rDRTG, because the team is able to play better defensive players (which are commonly players who do not provide spacing on the other end) without collapsing offensively.

In terms of a basketball explanation for why some players might be better than other players at keeping the offense afloat with defense-minded lineups, I think a lot of it comes down to where the player’s gravity is operating on the floor. Players tend to produce shots for others in parts of the court that their gravity pulls defenders from. So if your gravity is at the rim, then you’re going to be pulling defenders away from the perimeter, which will tend to produce long-range shots for teammates. Conversely, if your gravity is on the perimeter, then you’re going to be pulling defenders away from the basket, which will produce short-range shots for teammates. In general, highly impactful defensive players in the NBA tend to be guys who are relatively better at making shots near the basket than they are at making shots on the perimeter. So if you have rim gravity, having defense-minded non-spacers will hurt the team’s offense a lot more than it would if you had perimeter gravity instead.

This makes intuitive sense, but I thought I’d test it out, using a case study of LeBron James and Steph Curry—since they’re the quintessential examples of players with huge rim gravity and perimeter gravity.

What I aimed to do was look at the on-court playoff rORTG (using PBPstats data) that these guys had when there were only 2 or fewer spacers on the court with them. This gets at situations where there were two non-shooters on the court. Drawing the line as to what should be defined as a “spacer” is inherently arbitrary. But in order to reduce any bias and to save my own time, I came up with a bright-line rule to apply before I looked at any data. Under that bright-line rule, I defined someone as a spacer if they shot at least 0.5 3PA per 36 minutes that season with at least a 32% 3P%. Reasonable minds can differ on what the line should be, and a different line would of course make the numbers look a little different, but this seemed sensible to me at the outset.

So what did I find?

Well, below is LeBron James’ on-court playoff rORTG with only 2 or fewer spacers in some different timeframes, with the overall rORTG in those time periods (i.e. not just limited to situations with 2 or fewer spacers) in parentheses. Please note that the career sample here is 10,441 possessions, so this is a considerable amount of data.

LeBron James on-court playoff rORTG with 2 or fewer spacers (overall rORTG in parentheses)

Career: +2.57 (overall: +6.01)
2006-2020: +3.09 (overall: +6.41)
2007-2020: +3.62 (overall: +8.06)
2009-2020: +5.29 (overall: +8.54)
2009-2018: +5.48 (overall: +8.78)

As we can see here, LeBron’s on-court playoff rORTG is quite a bit lower with 2 or fewer spacers on the floor than it is overall.

By contrast, let’s look at the same thing for Steph Curry. Here, the career sample is 5,524 possessions, so it’s a smaller sample than for LeBron, but not all that small:

Steph Curry on-court playoff rORTG with 2 or fewer spacers (overall rORTG in parentheses)

Career: +8.49 (overall: +7.34)
2014-2023: +9.31 (overall: +7.42)

This looks very different than the LeBron numbers. Steph’s playoff offenses are not worse when there’s less spacing on the floor with him. In fact, they actually did a bit better in those situations!

___________

Conclusion

So what is the takeaway here?

Well, at least in this case study, it supports the idea that players with perimeter gravity are much more resilient offensively when there’s not much spacing on the floor with them. In fact, this case study at least indicates that they may even be helped a bit by it (which may not be so crazy, since a perimeter gravity player is opening up space at the basket, and spacers might not actually be the best at scoring in those areas).

I think this has some important implications for assessing rORTG numbers. There’s generally a negative relationship between how well players shoot and how good they are defensively. That relationship may be breaking down to some degree nowadays (see, for instance, Wembanyama being a spacer), but there’s been that relationship throughout the vast majority of NBA history. Thus, the implication here is that a player with perimeter gravity can play with good/bigger defensive teammates without seeing a drop off offensively (indeed, it might even improve their offense!). On the other hand, guys that rely on rim gravity seem to present a real trade off for their teams. If you don’t have lots of spacers, then the offense suffers a lot, but if you do have lots of spacers then your defense will probably suffer a lot.

I think this gets to the concept of ceiling raising. If you have a guy who can stack with defense-minded teammates without having the offense suffer much (if at all), then you can actually maximize the team’s quality on both ends of the floor a lot more easily than you could with a guy who presents much starker trade offs for the team. And it’s teams that are fantastic on both ends of the floor that end up being the very greatest teams ever.

Caveats

I do want to list a few caveats about this analysis:

1. This is only a case study about two players. This data took a while to compile, so I did not go ahead and do it for more than just those two players. While the data backs up the intuition we might have just from thinking about how basketball works, there’s still no guarantee that this would hold up across lots of players. I may go back and do this analysis for more players, but wanted to at least post what I had before taking the time to do even more.

2. There is a confounding factor at play here regarding whether a player’s co-stars are considered “spacers” in this analysis. All else being equal, a guy whose offensive co-star is a spacer will tend to do worse in this kind of analysis than a guy whose offensive co-star is not a spacer. This is because, if the offensive co-star is not defined as a spacer, then they’re not counted against the limit of spacers, so you can have the co-star *and* two spacers on the court, while you’d otherwise only be able to have the co-star and one spacer. Obviously, we’d expect the former to lead to higher rORTGs. And we’d also expect there to be more possessions being counted here in years where a co-star isn’t a spacer. That said, this generally operates to the benefit of LeBron in this analysis. In particular, Wade was defined as a non-spacer every year, and Bosh was defined as a non-spacer all but one year. Meanwhile, Durant and Klay were always defined as spacers. Draymond was defined as a spacer in three years, and he’s not really an offensive co-star anyways.

3. When it comes to comparing the raw rORTG values when the guys had non-spacers on the floor, obviously there’s other factors at play. Steph with Durant and Klay and no other spacer is obviously in a better position offensively than LeBron with Daniel Gibson and Mo Williams and no other spacer. Over the course of the long-term, I don’t really think LeBron had less offensive talent on his teams, but that’s a rabbit hole that is not exactly on point here, since the main thrust of the point is to compare how each of these players’ offenses did with a lot of spacing to how their offenses did without a lot of spacing.

4. Because of the nature of this analysis, the samples are not the same each year. These guys had some years where they had a lot fewer playoff possessions with 2 or fewer spacers than other years. So, to take the most extreme example, in the 2017 playoffs, LeBron did not play a single possession with 2 or fewer spacers. So that year is just not part of the 2-or-fewer-spacers rORTG data at all, while it’s obviously in the overall rORTG data. On one hand, we might say that that means the data is not including one of his best offensive years and is skewed against him for that reason. But on the other hand, this analysis suggests that perhaps it was one of his best offensive years precisely because he always had spacing. Basically, there’s a chicken-and-egg issue here, particularly for LeBron. Not sure exactly where I land on that, but I figured I’d flag the issue.

[DraymondGold]

Cool stuff jake! This broadly agrees with my takeaways on scalability and chemistry, though as you say it's limited by just two players and the specific situations they happened to have, so I'd be careful making claims that are too general to an overall archetype.

I think it's also broadly consistent with some of the gravity discussion at the rim and perimeter in the recent Thinking Basketball podcasts that focused on the top 6 (I'd definitely recommend basically anyone on this site give those a listen -- you don't have to agree with the exact rankings to learn). I their final podcast, they talked about the importance of different skillsets for different archetypes, and shooting with heliocentrism/rim-driving was an obvious one they pointed out. They also discussed throughout the Wade podcast, Shaq/Curry podcast, LeBron podcast, and the final podcast about how different scoring threats create in different ways. For example, rim drivers seem to up their teammate's 3 point shooting volume (so arguably better shot selection), but often don't produce as much of a boost on the efficiency of those shots; compared to perimeter spacers (+ Shaq oddly enough) who increase the volume and efficiency of their teammate's rim scoring by a fair margin. It makes sense that if the heliocentric player doesn't have good spacers, that their kick-outs would produce much worse offensive results overall.

The reverse relationship between defense and shooting is an interesting one. I suppose it's just saying something like 3+D is a rare archetype for most of NBA history? It would be interesting to see if there's any evidence or past analysis on this, since I hadn't heard the claim spelled out before. I suppose another way to say this is that players who have a lot of defensive value are typically big man or highly athletic, and those players on average are less likely to be clearly above average in hand-eye coordination or shooting skill (and if someone does manage to combine both... they're likely an offensively great star). It does match with the trends I found in my scalability thread, where there does seem to be some sort of statistically significant trend with star player scalability and the best teams by overall SRS, but that the trend is far clear if you look at the team results with offense+defense (so it could be the scalable player is lifting defensively-slanted rosters via good offensive chemistry).

All that said, to make these conclusions or analysis a bit more compelling, it would be interesting to see the sample size increase slightly, even if we're just looking at a few hand-picked stars.
For rim-gravity players: Shaq and Duncan (for rim-gravity from the post) and Wade and Giannis (for rim-gravity from drives) seem like natural choices. I do see some differences in terms of gravity and how one's offensive threat warps the defenses for guys who operate primarily in the most valuable area on the court (post-up players) vs guys who cut or drive into the most valuable area on the court, even if both archetypes would prefer to get their shot in a similar area.
For perimeter gravity, maybe Nash (with his shooting, although his gravity style's a bit different than the motion style of Curry's) and Durant? Kobe or Jokic could also be interesting; although Kobe is pretty midrange-centric, and Jokic acts both as a spacer and as a post up player.

Since it may be labor-intensive to do the relative adjustment (is that done manually, or is there a way to do that naturally on pbpstats.com? I haven't found it yet if so)... it could also be interesting to do slightly smaller increments across a player's prime (maybe 5-year or 7-year, e.g. to compare how LeBron's performance evolves as he became more of a shooter).

[lessthanjake]

The reverse relationship between defense and shooting is an interesting one. I suppose it's just saying something like 3+D is a rare archetype for most of NBA history? It would be interesting to see if there's any evidence or past analysis on this, since I hadn't heard the claim spelled out before. I suppose another way to say this is that players who have a lot of defensive value are typically big man or highly athletic, and those players on average are less likely to be clearly above average in hand-eye coordination or shooting skill (and if someone does manage to combine both... they're likely an offensively great star). It does match with the trends I found in my scalability thread, where there does seem to be some sort of statistically significant trend with star player scalability and the best teams by overall SRS, but that the trend is far clear if you look at the team results with offense+defense (so it could be the scalable player is lifting defensively-slanted rosters via good offensive chemistry).

Yeah, the bolded basically spells out my intuition on this. I’ve not done any rigorous data analysis to support this point, but in general I think it’s pretty intuitive that non-spacers tend to be better defensive players. For one thing, a lot of times there’s two non-spacers on the floor, it’s going to be because there’s two bigs, and that’s surely good defensively. As it relates to these particular examples, for instance, I’ve long talked about Steph enabling the Warriors to have Draymond and another non-spacing big (Bogut, Looney, Pachulia, etc.) on the court at the same time. And a lot of LeBron’s minutes with 2+ non spacers were things like Mozgov/Tristan Thompson lineups. That’s probably the most important/common aspect of this. But also to the extent there’s a perimeter player that isn’t a spacer at all, it’s usually (though not always!) going to be a guy who is in the NBA for his defense (for instance, someone like Eric Snow or old Iguodala). There’s exceptions, though—for instance Dwyane Wade is an exception. So yeah, I definitely wouldn’t say there’s a perfect inverse relationship between spacing and defense, but I really do think there’s a significant relationship there. Being able to play two classic bigs or a pure defensive-specialist wing player without losing much (or maybe anything at all) on the offensive end is surely going to help the team have better defense!

All that said, to make these conclusions or analysis a bit more compelling, it would be interesting to see the sample size increase slightly, even if we're just looking at a few hand-picked stars.
For rim-gravity players: Shaq and Duncan (for rim-gravity from the post) and Wade and Giannis (for rim-gravity from drives) seem like natural choices. I do see some differences in terms of gravity and how one's offensive threat warps the defenses for guys who operate primarily in the most valuable area on the court (post-up players) vs guys who cut or drive into the most valuable area on the court, even if both archetypes would prefer to get their shot in a similar area.
For perimeter gravity, maybe Nash (with his shooting, although his gravity style's a bit different than the motion style of Curry's) and Durant? Kobe or Jokic could also be interesting; although Kobe is pretty midrange-centric, and Jokic acts both as a spacer and as a post up player.

Yeah, I think I am going to aim to try to do more players in the future. I think Wade and Giannis are good examples of rim-gravity players. It’s hard to really think of other guys who are pure perimeter-gravity players that we’d really have data on. Like, Reggie Miller would be a good example IMO, but we don’t have almost any data for him in this regard, so I can’t use him. And other guys like Durant/Kobe/etc. seem a bit more of a hybrid to me. There’s an element to which Steph’s gravity is pretty unique so it’s hard to expand out the sample to lots of players who are similar. Maybe Ray Allen is a similar player I could use? But he didn’t play many playoff games until he went to Boston, and once he got there I don’t really think he was the main thing driving Boston’s offense. I guess I could go for regular-season+playoff data for him instead to have a decent sample. I’ll have to think about it.

Since it may be labor-intensive to do the relative adjustment (is that done manually, or is there a way to do that naturally on pbpstats.com? I haven't found it yet if so)... it could also be interesting to do slightly smaller increments across a player's prime (maybe 5-year or 7-year, e.g. to compare how LeBron's performance evolves as he became more of a shooter).

Yeah, the thing that makes it labor intensive is that I have to go through each season and pick out the players to be controlled for, run PBPstats data for each playoff series that year individually, and then calculate the rORTG for each series (and then I get the full number for a given timeframe by doing a possession-weighted average of each series in that timeframe). I could definitely do it for shorter timeframes for other players, but it’s a bit of a slog regardless.

[f4p]

While it might be intuitive that this might help a team, the obvious caveat here would be that one of Steph's non spacers is basically always going to be Draymond, a nearly unique defense/non-spacer guy who impact stats show produces positive offensive impact, while I'm guessing most non-spacing bigs around other guys are probably negative. So Steph is essentially getting spacing impact from a non spacer.

Also, a bit strange Steph would do even better with non-spacers. That doesn't really make any sense. Like spacers should presumably only be good as they can presumably just make layups and dunks like non-spacers and then also make 3's.

[lessthanjake]

While it might be intuitive that this might help a team, the obvious caveat here would be that one of Steph's non spacers is basically always going to be Draymond, a nearly unique defense/non-spacer guy who impact stats show produces positive offensive impact, while I'm guessing most non-spacing bigs around other guys are probably negative. So Steph is essentially getting spacing impact from a non spacer.

So this is something worth thinking about, but I think a really important thing to note here is that the two years that impact stats say Draymond really had notably positive offensive impact (2015 and 2016) are two of the three years where he was actually counted as a spacer for purposes of this analysis. Impact stats indicate he was otherwise a pretty neutral offensive player. But yeah, even in the merely pretty-neutral offensive years where he wasn’t counted as a spacer, Draymond is going to be at least a bit better offensively than most non-spacers. Of course, the same is true to a far larger degree for Dwyane Wade and Chris Bosh, and to at least some degree for Anthony Davis (who only counted as a spacer in 2020), though.

Also, a bit strange Steph would do even better with non-spacers. That doesn't really make any sense. Like spacers should presumably only be good as they can presumably just make layups and dunks like non-spacers and then also make 3's.

I think it makes a lot of sense actually. Consider a couple things:

1. You say spacers can make layups and dunks too, but you’re kind of handwaving away that they probably aren’t going to be as good at that as non-spacers. This is primarily because non-spacers tend to be bigger, and bigger guys obviously find it easier to score at the basket. Meanwhile, as to non-bigs, it’s probably the case that non-spacers tend to be more athletic than spacers (after all, if they’re guards or wings who can’t shoot *and* are unathletic, it’s hard to see how they’re getting any minutes in the playoffs). Of course, being athletic tends to be better for finishing at the basket.

2. Just in general, you don’t want players crowding the same areas, because it becomes easier to help on players and recover. Intuitively, I don’t think this is a particularly big deal on the perimeter, since there’s just a lot of real estate out there for multiple players to be out there, but putting a bunch of spacers with Steph may actually make it marginally easier to help on him in the area of the court he’s most dangerous.

All that said, I’ll note that while the sample here was significant, it’s not so large that it’s not subject to noise. So, even though it actually does make some intuitive basketball sense that they’d be better with less spacing (for the reasons mentioned above), maybe Steph’s team doing *better* offensively with less spacing is just a product of a bit of randomness. Which is why my OP didn’t really lean hard into the fact that they were better. I think the more important insight here is that they weren’t worse and certainly not significantly worse, and that’s a fact that it’s harder to imagine is a product of randomness.

[Cavsfansince84]

I've suspected that Steph had a high correlation to at rim scoring after the 15/16 finals where that was really the difference in teh games the Warriors were winning. They were getting way more easy baskets at the rim that they barely worked for compared to the Cavs. LeBron was able to turn that around in the 16 finals with his ability to break down their defense.

[penbeast0]

...
Yeah, I think I am going to aim to try to do more players in the future. I think Wade and Giannis are good examples of rim-gravity players. It’s hard to really think of other guys who are pure perimeter-gravity players that we’d really have data on. Like, Reggie Miller would be a good example IMO, but we don’t have almost any data for him in this regard, so I can’t use him. And other guys like Durant/Kobe/etc. seem a bit more of a hybrid to me. There’s an element to which Steph’s gravity is pretty unique so it’s hard to expand out the sample to lots of players who are similar. Maybe Ray Allen is a similar player I could use? But he didn’t play many playoff games until he went to Boston, and once he got there I don’t really think he was the main thing driving Boston’s offense. I guess I could go for regular-season+playoff data for him instead to have a decent sample. I’ll have to think about it....

Lilliard would be a pretty clearly perimeter guy, maybe more than Reggie Miller who drove a lot and had a high foul draw.

[EmpireFalls]

This portends incredibly well for LaMelo Ball if and when they put an actual team around him.

[falcolombardi]

I mean yeah, a rim attacker/post up player benefits (and so does the team) when surrounded by spacing and shooters (and shooters benefit from them. Is not a unexpected correlation

Somethingh similar happens with off ball roamers beneffititng (and viceversa) from good passers

Steph offense stats are a lot better with draymond than without

[lessthanjake]

Okay, so I think I have some good ideas for players that might be useful to run similar analysis for. Giannis is a good rim-gravity example, and Wade is a pretty good one too. No one quite has the perimeter gravity of Steph, but Lillard might be a good example to look at, and I could potentially try Ray Allen as well, though the sample is small. I realize now, though, that adding new players would be a lot more work than I previously realized, because in order to compare the non-spacers rORTG to the overall rORTG, I have to run the numbers to get the overall rORTG too. For Steph and LeBron, that wasn’t extra work, because I had done that a long time ago. But I’ve not run those numbers for these other players before. So it would be a significant undertaking to do it (and, like 70sFan, I’ve got a baby at home, which takes up a lot of time!). It’s something I may try to chip away at over time, but I don’t think I’ll be able to pump it out anytime soon. Of course, others are free to run the numbers sooner if they’re interested, though!

[Warriors Analyst]

I've suspected that Steph had a high correlation to at rim scoring after the 15/16 finals where that was really the difference in teh games the Warriors were winning. They were getting way more easy baskets at the rim that they barely worked for compared to the Cavs. LeBron was able to turn that around in the 16 finals with his ability to break down their defense.

If I can add to that, the 16 Finals really went south for the Warriors when Barnes turned into a pumpkin and the Cavs Tony Allen'd him. The Warriors got way less clean looks at the rim when that happened I think Kerr was too slow to get Barnes off of the court in favor of say, Barbosa or Speights or Rush to try and open up the paint.

[NO-KG-AI]

Without making it about any particular player, I have always felt it was an underrated aspect. Of course you want maximum spacing, but no team can afford or will be able to be 15 deep with guys who are switchable defenders all over, and legitimate 3 point bombers. If you can only be optimized when the offense is 4 or 5 out, you handcuff your teams ability in a lot of aspects and lineups, even if you’re still pretty good. Being able to dominate with sub optimal spacing, and being able to lean heavy to defensive lineups is a massive advantage.

Dirk Nowitzki comes to mind for how some of those Dallas teams were constructed.

[Djoker]

Responding to lessthanjake's opening paragraph, another factor that should be considered when looking at ON Court rORtg is how many minutes the player is playing. A guy posting a +8 rORtg in 45 mpg is way more impressive than a guy posting +10 rORtg in 35 mpg. With the first guy, your team is prartically guaranteed to have a really good offense for the entire game; 3 minutes of suckiness can't hurt that much...

With regards to the results of the study, I mean it is expected. And as OP pointed out, being able to produce better offenses with lack of spacing can be a boon because often times the poor spacers are bigs that add a lot of defensive value.

[tsherkin]

I'd be interested to see the data on this for other stars.

This portends incredibly well for LaMelo Ball if and when they put an actual team around him.

Does it? He needs to learn how to score effectively before that seems as if it would be any sort of relevant. He doesn't. He's been fairly bad as a scorer, and very clearly not #1 option-level in every season of his career to date. His playmaking is undeniable, but him in the role of volume scorer has been clearly not a good choice to date. Not like Charlotte has better choices, but valid #1 options don't look that bad even under adverse conditions.

[lessthanjake]

So I decided to run these numbers for Ray Allen. Despite the title of the thread being focused on the flip side of this, I actually realize the more interesting insight here is probably the idea that perimeter gravity players are really resilient to a lack of spacing. The opposite is less of a surprise, since we know spacing is generally good for offense. Because of that, I wanted to run the numbers for another perimeter gravity guy as opposed to a rim gravity guy. I decided to do Ray Allen, since he seemed a really good play-by-play era example. A slight problem with him, though, is that the playoff sample is tiny in the years where he was leading his teams (i.e. before he got to Boston). And running the playoff data is a real chore, because I have to do separate calculations for every series. So, in order to have a usable sample size and to limit my own work, I decided to just look at Ray Allen in the regular season from 2001-2007 (note: I would’ve included 2000 as well, but PBPstats data starts at 2001).

Here’s what I came up with:

Ray Allen on-court RS rORTG with 2 or fewer spacers (overall rORTG in parentheses)

2001-2007: +5.62 (overall: +6.17)

So yeah, I think this supports the thesis. The rORTG with 2 or fewer spacers is not above the overall rORTG, but it is super close. And that’s despite the fact that his best offensive teammates were basically always on the spacers list (guys like Rashard Lewis, Michael Redd, Radmanovic, Cassell). As the guy leading his team, Ray Allen’s offenses were very resilient without much spacing.

[FrodoBaggins]

Curry making lineups with Draymond, Butler, Kuminga, and Horford viable.

[lessthanjake]

So I wanted to validate Steph’s playoff numbers here a bit by looking at regular season numbers for him. The playoff samples for him are not super small, but if you cut up a 155-game sample, it can start to be noisy. Looking at regular-season samples should give us some idea of whether his resilience to a lack of spacing in the playoffs is just a product of noise.

Here’s what I got:

Steph Curry’s on-court RS rORTG with 2 or fewer spacers (overall rORTG in parentheses)

2013-2025: +6.67 (overall: +7.17)
2014-2023: +7.81 (overall: +8.36)
2015-2019: +11.75 (overall: +11.70)

So yeah, I think this provides some good validation that the results in the playoffs were not just a product of noise. These numbers are very close. Unlike in the playoffs, Steph’s regular-season on-court rORTG isn’t generally *better* with 2 or fewer spacers, though it actually *was* slightly better in the core dynasty years of 2015-2019. The fact that his playoff on-court rORTG with fewer spacers was a decent bit above his overall playoff on-court rORTG is probably at least partially due to noise. But we can definitely see that Steph being very resilient to lack of spacing in the playoffs was not just a product of a low sample size. He shows such resiliency across massive samples too. And that is similar to what we saw with Ray Allen, as per my prior post.

[Ainosterhaspie]

Curry is a spacer, so Curry plus two spacers means the team has three spacers. If LeBron is not a spacer, then him plus two spacers means the team has two spacers. Isn't this just saying it's better to have three spacers than two spacers?

[lessthanjake]

Curry is a spacer, so Curry plus two spacers means the team has three spacers. If LeBron is not a spacer, then him plus two spacers means the team has two spacers. Isn't this just saying it's better to have three spacers than two spacers?

LeBron would be defined as a spacer by the criteria in this thread in the vast majority of the years of his career. I defined spacer very permissively, so that this was really capturing situations where there were two or more guys on the court that really were not a threat from three at all.

In any event, this thread is really not “just saying it’s better to have three spacers than two spacers.” The primary point of the thread is that players with perimeter gravity like Steph do not seem to have their teams’s offenses be particularly elastic to their team having a lot of spacers on the floor, while rim-gravity players do. So the primary question being analyzed here isn’t whether Steph + two spacers is better than LeBron + two spacers. Rather, the point is to compare the drop off (or lack thereof) these guys’ offenses have when they don’t have much spacing from their teammates. So yeah, whether Steph’s or LeBron’s offenses are better with less than three teammates that are spacers isn’t even really the point. The point is about the delta between each of these guys’ overall rORTG and their rORTG with low spacing, rather than which of LeBron or Steph had the higher rORTG with low spacing. To use made-up numbers, if LeBron’s overall rORTG was +10 and his rORTG with less than three teammates that are spacers was +7 and Steph’s overall rORTG was +6 and his rORTG with less than three teammates that are spacers was also +6, then it would support the primary thesis of the thread, even though LeBron’s rORTG with three spacers would be higher than Steph’s—because that would show more of a drop-off for the rim-gravity player.

Anyways, to the extent you’re wanting to look at this thread as a direct comparison between LeBron’s and Steph’s offenses (again, not the primary point here), I would note that in a vacuum it is true that three spacers is generally going to be better than two spacers, but I think a lot of people would bristle at the notion that Steph + two spacers is better than LeBron + two spacers. Just saying “it’s better to have three spacers than two spacers” is basically simplifying things to the point where you’re effectively presupposing that Steph is a better offensive player than LeBron and then remarking that evidence supporting that conclusion is totally expected. If you believe that, then I guess okay, but it seems like an odd push back for someone to make on this.

[Lebronnygoat]

First off idk how you quantified spacers like that. Very weak criteria. When I tried doing what you did for 14-23 span, I only get 3000 possessions. Are you removing Draymond? By your definition he’s 31.7% on a lot more 3PA’s than 0.5 in the RS. Sounds like a spacer. With Dray I see similarly the amount of possessions you wrote down. The only huge non Dray at the 5 lineups Curry leads good offensively is with Looney at the 5. The most used Bogut at the 5 lineup they ran in this span was a horrible offense. 103 ORTG. In fact, without Durant in this span the offense falls to 111.5 ORTG. League average in this span was 110.0. That’s a measly +1.5 offense. Not to mention, Dray at the 5 provides a huge amount more spacing versus a traditional big, especially against non elite defenses. Dray is also not a bad shooter for a 5 during these times, especially 2016 to 2017. He carries all time passing and facilitation/connecting for a 5, seriously watch the passes he makes. Plenty of the times Draymond is mobile and brings the 5 out the rim, then he has pinpoint passing accuracy to dart the interior even without Curry having to gravitate anyone to him. Draymond is also the best screener you can put for the Warriors motion system, helps Curry get open and others move off the ball. We talk about Jokic with his h-offs and screening, Draymond is really a pleasure to watch doing those actions as well. GP2 and Livingston playing are the weak links offensively, Draymond is actually a clear positive offensively at the 5.

Another issue we don’t talk about enough is that LeBron James deserves to play with more spacers, why? Because he can play the 4 as an amazing defender. He’s literally in the conversation as the goat wing defender. He’s allowed to have more offensive pieces 1-3 as opposed to Curry 2-4, that’s one more offensive position freed up. That’s huge. Not to mention, Curry being a lackluster perimeter defender causes for certain lineups to be played as well, it’s also the fact Draymond really hides Curry's flaws on defense a lot. Warriors want to play winning basketball, not lead amazing offenses like Dantoni. Sure as hell they led amazing offenses. But to act as if Curry’s offensive impact is better than LeBron’s, or even to hint at Curry's offensive impact leading to overall player impact being in LeBron’s tier is disgusting going off the fact Curry can simply do better with worse spacing which essentially allows better d players. Even the results without KD, Curry isn’t doing much on offense, which I don’t blame him but it’s not impressive as you were trying to make it. Curry without Durant leads about a +5 offense from 13-23 in the PS. LeBron absolutely is better than that from 08 or 09 till 15.