So for peaks I tend to mildly separate bigs and perimetre players. Basketball for me is a sport which intuitively favours height. There tend to be skill tradeoffs, but a 6’9”+ player has more capacity to become a decent to good passer/dribbler/shooter than a guard has to become a decent to good rim protector at any real volume. Where that can be messy is that this advantage tends to carry over to what we tend to assess as “replacement players”. There are plenty of circumstantial exceptions based on team construction, but typically I would expect a replacement-level big to maintain more latent basketball value than a replacement-level guard. And I think this can ring intuitively true in a peaks project when you look at the players sampled: there are comfortably more 6’9”+ players, even though the population is a lot smaller.
Now, the other side of this is that bigs become both harder to evaluate relative to each other… and less likely to clearly separate themselves from the pack. I think back to 1992, when Magic’s forced retirement propelled Clyde Drexler into the spotlight as Jordan’s new top perimetre rival. Drexler has never appeared in one of these projects or come especially close. And while Barkley in 1993 was a more legitimate contender, he peaked at #23 in 2012 and has slid with every iteration; as of 2022, there was a 30-spot gap between him and Jordan. Reggie Miller was another moderate competitor, and he is an all-time playoff riser; he has also never appeared in one of these projects. Penny Hardaway, something of a rival from 1995-97, was #28 in 2012, not top forty in 2015/19, and #39 in 2022. Grant Hill was a marketed rival in 1997; never appeared in one of these projects. Even looking at some of the shooting guards who followed Jordan, we have no appearances from Vince Carter, Allen Iverson, Ray Allen… Jordan was a monolithic outlier from the rest of his “small” competition. People could debate the bigs, and while Magic was around they could debate the best (nominal) guard, but after 1991, there was no debate. We can fairly question Jordan’s modern translation, but in his era, Jordan was a historic outlier for his size in a way we had not seen since at Oscar and West (or
maybe ABA Erving) and would not see again until Curry.
Of course, while it would be easy enough to argue players like Oscar and West were more separated from other guards than Russell and Wilt were from other centres, that does not mean we qualify them as greater peaks. Or going back even further, Mikan versus Bob Cousy. We can then amend the idea a bit to acknowledge era dynamics, where the playmaking of Oscar and West had more of a capped value than the rim protection and overall defence of Russell and Wilt in terms of how those skills could be applied to the rest of the league. For Jordan, while I have seen comments about his playmaking being somewhat similarly limited by the constraints of his era, that was not anywhere near enough of an element of his game for it to overwrite the immense value of being a still-unparalleled volume scoring hub against defences which were far more limited in their ability to reliably force him to relinquish volume to his teammates.
The other important amendment in the context of a peaks project is that it does not necessarily matter when you were the clearest outlier relative to the league. Kareem was more of an outlier prior to the merger, but that does not mean he suddenly became a worse player during the 1976 offseason — and indeed, the history of this project suggests that the majority consistently assesses Kareem’s peak as that first post-merger year, even though he was not exhibiting the same suggestions of “impact” as what he had on the Bucks, nor as what Bill Walton was exhibiting at the same time. And part of why is that we all clearly do some amount of filtering for both league context and the quality of opposition, e.g. 1948-50 George Mikan never receiving any consideration for this spot.
Jordan is not my #2 perimetre peak over Magic because I think he was demonstrably more impactful; there is plenty of evidence that he was not (although I suppose it depends on the degree to which box scores are conflated with actual impact). And Magic is not my #3 perimetre peak over someone like Curry because I think Curry was inherently less impactful than Magic. What I am generally looking for is expected resilience as the foundation of a title team, and what separates Magic from Curry is that Magic consistently maintains in the postseason, and what separates Jordan from Magic is that Jordan consistently improves in the postseason. That improvement is also largely why I have Hakeem atop my list of bigs. Elephant in the room there is that Bill Russell won more consistently than anyone in the history of team sports, but I think his dominance was disproportionately tied to the dynamics of his era being comparatively skewed toward defence. Again, like with Mikan, context and environment do influence these choices.
“Confidence” is an essential element of these discussions, and I am generally pleased to see it come to the forefront in this thread (even if the concept behind it can be and has been abused or misrepresented). Saying, “this player is a better scorer because they score more,” is not drastically different from, “I am more confident that this player is a better scorer because they score more,” but it is an important distinction nevertheless. For Hakeem, there is a long-term consistency in the postseason which gives me confidence I do not have from other bigs, and that in turn makes me more confident in his peak postseason value. I am not sure that 1993/94 Hakeem is better than 2002/03 Duncan in their respective postseasons… but I am more sure of it in the sense that I feel more sure of Hakeem as a postseason performer generally.
If that tepid confidence is present with a cross-big comparison, it is even cooler with bigs versus perimetre players. For example, I am more confident that perimetre players like Jordan/Magic/Curry could be more “impactful”, via value over replacement, than the best bigs, than I am that the best perimetre players are better on their own merits independent of replacement value “impact” advantages. And when it comes to Hakeem versus Jordan, well, the simplest way to put it is that I am more confident in Hakeem’s ability to rise through adversity and win a series or game which he should not, and that matters more to me than the idea that Jordan is more likely to win by a larger amount and therefore less likely to be placed in adverse situations at all. Jordan only ever lost one series where he had the lead (1989 Pistons), which I think is second only to the undefeated-with-a-lead Russell, and he never lost a home series or a series where he was favoured (Russell’s exception for both is his injured series against the 1958 Hawks). Hakeem also only lost one series with a lead (1998 Jazz), but he did lose two series at home and as the favourite (1985 Jazz and 1987 Sonics). However, he was arguably the best ever at managing an upset*, going 10-10 as an SRS underdog (as of this year, slightly outpacing Lebron’s 10-11) and 7-10 as a road SRS underdog (slightly behind Lebron’s 8-10). And as a minor additional note, he is one of only three players (Russell and Lebron) to win a series down 3-1 on the road. To paraphrase while slightly inverting f4p’s comment about splitting hairs between similarly good players with different strengths, while I am happy with peak Jordan if I am in an adverse situation, and I am happy with Hakeem if I am in a favourable situation, I prefer the player I trust more in the adverse situation because I think adverse situations are more likely.
*Here I will note that Russell maintained a winning record on the road (4-1), as an SRS underdog (3-1), and as a road SRS underdog (2-1). He was also 11-0 in single-elimination games. Like I said, seems extremely specific to his era given his skillset, but in his era there is no one you could justify picking over him.
Alright, I suppose I have stalled long enough. No one cares about comparisons with Hakeem or Magic in the #1 thread, and frankly I do not expect there to ever be much movement on either Hakeem as a #2 peak or Magic as Jordan’s near-equal, because neither have the fanbase, available film, or box score profile needed to gain any type of foothold on that front.
I have more confidence in Lebron than in any player. We have discussed how he regularly sits atop long-term RAPM signals, with the sole current exception being an age-agnostic calculation that elevates one player averaging six fewer minutes per game across a sample that is roughly 40% the size. Indeed, anyone we see as a regular top ten competitor is drawing from a much smaller sample in total and per game. And focusing on the postseason advantages Lebron even further in every playoff study I have seen, with something like a 15% increase from his already top regular season baseline on average. From talking with people who have run RAPM, that effect is especially pronounced once he leaves Miami, although there we can acknowledge part of the improvement is because of a reduced regular season baseline from his all-time regular season peak. And this is “despite” (I say this sarcastically) Lebron not having outlier on/off in the regular season or postseason, for as much as a certain subset seems interested in equating the two when convenient. This is also despite (not sarcastic this time) Lebron playing in the most developed — in strategy, skill, medical advancement, breadth of player pool, etc. — league, and keeping pace as it changes and continues to change the skills it emphasises and requires for team success. 2004 to 2025 might be as developmentally different as 1964 to 1985.
This is a peaks project though, so while that career-long dominance makes me confident in Lebron’s impact generally, it does not do much to narrow the selection of year. Maintaining that focus on playoffs, we have one RAPM set which ran
5-year calculations from 1997-2021 (which does essentially cut out Jokic, but we can come back to him later). That gives us 21 separate five-year ranges (starting year 1997-2017, ending year 2001-21). Of those 21 ranges, Lebron is in first place for 11 of them. Again, confidence. He also has the “top” two values in the set (2012-16 and 2013-17), although I am not sure whether this set regularised the outputs such that values are intended to be compared across spans. His 2006-10 span is a narrow (6.15 versus 6.17) fourth place behind 2003-07 Garnett, with Garnett only drawing from ~80% of Lebron’s possession count. Jordan does lead the 1997-01 sample, albeit with the noise that comes from having only two years of on-court value: he drops to ninth in the sample where 1997 is replaced with 2002.
Looking at a different dataset, NBARAPM.com now has included the postseason in its 2-5 year spans. There, Lebron is the leader in five-year spans nine times (so slight step down from the previous set), including seven consecutive samples from 2007-11 to 2013-17. This set has him second to Garnett in 2006-10 (the previous set had Garnett a more distant third). Also, to whatever extent these samples are regularised for cross-sample comparison, 2006-10 is Lebron’s sixth highest-value here, as opposed to his third highest in the prior set, and his new highest value is 2007-11 (2013-17 second, 2008-12 third, 2012-16 fourth, 2009-13 fifth). Lebron and Garnett are the only two players hitting +10 in this set, with Garnett’s coming primarily from samples set around his trade to the Celtics (2006-10 and 2007-11). Jordan’s 1997-2001 (+9.5) again looks excellent and again drops down when replacing 1997 with 2002 (+6.7, but now in sixth place rather than ninth).
In four-year spans, Lebron leads eight times, with his best values occurring in 2009-12 (+10.5), 2007-10 (+10.2), 2008-11 (+9.8), and 2013-16 and 2014-17 (both +9.7). Jordan is interesting here because now we do not get a Bulls year with a Wizards year. 1997-2000 (+9) remains strong, and the hit from losing 1997 is present but less damaging than including 2002 (+6.9), now up to third place in the 1998-2001 span. Also of note is that 2022-25 Jokic (+10.1) briefly joins Lebron and Boston Trade Garnett in the +10 group, before falling back down in the three-year spans.
Lebron leads the three-year spans seven times, with his best values in 2009-11 (+10), 2015-17 (+9.9), and 2008-10 (+9.3). 2007-09 Garnett (+10.6) is an increasingly massive outlier. 1997-99 Jordan (+8) is a shrinking first in that span, but removing 1997 continues to be less damaging as 1998-2000 is second — and then their spots, if not their values, flip in the two-year spans, with 1997/98 (+6.7) placing second behind Shaq in its span while 1998/99 (+5.8) places first in its span. Garnett also sees a notable change in the two-year spans, with his typically identified peak in 2003/04 (+9.4) now his best value. Lebron and Garnett agree to rename the +10 club to the +9 club, as Lebron’s 2016/17 falls just shy at 9.9. Lebron still leads in seven spans, but deflation has struck the rest, with regular season peak 2009/10 (+9.0) his next highest value.
To the question of Lebron’s peak, RAPM generally seems to point to either 2016 (primarily a postseason peak) or 2009. In fact, I would go so far as to say I am confident that Lebron’s regular season peak is 2009/10. I did not include this earlier because the available years are smaller (1998-2012), but in 2012/13 someone published at attempt to normalise seasonal RAPM by standard deviations, and they calculated that Lebron’s 2010 regular season was the most significant outlier in the set and his 2009 regular season was the fourth most significant outlier in the set (after 1998 Shaq and 2004 Garnett). So why has 2012/13 continued to be the choice in the past two projects? Did the collective voters simply not understand that his impact in Miami was less than his impact in Cleveland? Or maybe, did they understand that there is more to a player’s quality than what their raw impact is in a certain situation? Why do so many of these threads refer to how much more productive Lebron was in the postseason when he did not need to share the court with Wade (especially the often injured Wade)? Okay, that can be a criticism of his lack of reliable shooting at volume and the extent to which he maintains his maximum offensive value next to other high volume slashers, but for most players that is not unique. It has been said before, but how much offensive value does Jordan retain if he is placed next to someone significantly eating into his shot volume? Why was Charles Barkley the standout on the 1992 Dream Team? Why could Jordan not fit well with Rip Hamilton on the Wizards? Why would we expect better results from adding him to two top ten scorers than we saw with Lebron? Why would we ever be more
confident in Jordan’s value independent of his situation, when he changed his playstyle and general roster construction so much less? I saw discussion about Jordan needing to “buy in” to the triangle, but his buy-in let him stay the most prolific shot-taker in league history!
But that is fine for “confidence” in terms of whether Jordan may have had higher “impact” on the 1990/91 Bulls than Lebron did on the 2012/13 Heat. Facile argument in any serious comparative sense — why would we be particularly confident in it relative to other title seasons (after all, easier to be confident in a title when you know it happened) like 2002-03 Duncan, 1993-94 Hakeem, 2000/01 Shaq, 1997/88 Magic, Bucks Kareem, pick a Russell season, etc. — but not something I would feel driven to strongly dispute in isolation. Unfortunately not everyone chose to stop there, or even just to take refuge in the ever-reliable box composite defence (with 2009 discounted as a fluke or as insufficiently accomplished). No, instead I have seen derisible claims of confidence that based on the selective samples we do have, Jordan should be presumed to have an impact advantage outright. On the “direct” RAPM front, there is no cohesive RAPM for 1990/91 (and likely never will be), and the gap in confidence needed to go from “Jordan is probably the RAPM leader in that span” to “it is probably the most impactful season ever” is so laughably wide that only someone with the most pitifully superficial understanding of RAPM would ever try to claim it. Same with any longer span in which a) your likelihood of sample completion is even more minimal, and b) the nearly two full seasons where Pippen led a +4 team without Jordan, and where Horace Grant significantly spiked the record and net rating of the Orlando Magic, would need to be incorporated. This player comparison lends itself easily enough that I see even Jordan stans making it, so surely his advoocates must know it hurt Kawhi’s RAPM for the Raptors to be so successful without him in 2018-20?
As we acknowledged, Jordan is a postseason riser, and while again there will never be full playoff RAPM for the 1990s, we do have reasonably tracked plus/minus for Jordan’s postseason career which, while skewed in an on/off sense by the pre-title period, has revealed a historically high postseason on/off that I have also seen claimed should make people “confident” in Jordan’s edge, even with a cursory acknowledgment that RAPM is different. But Lebron never had outlier postseason on/off for his career, yet somehow is a career postseason outlier regardless (nor has anyone voting for 2016/17 Lebron been tied to his postseason on/off as a reason). Steph Curry generally has more impressive postseason averages on that front, yet somehow he trails Lebron pretty comfortably — because RAPM is built around lineup results. Someone sincerely interested in RAPM would notice that Pippen and Grant also fare extremely well in Squared’s sample, and that is without much (in Pippen’s case, any?) tracking of their results with Jordan entirely absent from the game. No cohesive RAPM sample is going to overlook the 1994 Bulls performing at such a strong level in the postseason (or the 1995/96 Magic performing so much better with Grant). Nor is Jordan going to look like as much of an outlier with other Bulls starters included in lineup results. None of that is to say I think it is impossible or even grossly improbable that Jordan
could end up looking like a true impact outlier in a hypothetical where we had total lineup tracking through the entire history of the league, but I could say similar of Magic, Russell, Kareem, Oscar, etc. To be “fairly confident” in someone without that evidence over someone who is a demonstrated outlier over the past 25-30 years reflects a total lack of interest in the data or the process in favour of ideological puffery (and here I will note plenty of the obloquial posts against Jordan in this thread have also overplayed their sense of rational confidence
against his case as an all-time peak, but neither exaggeration justifies the other).
Anyway, settling on a specific year for Lebron does feel a bit arbitrary or otherwise predicated on giving disproportionate weight to different accomplishments or blemishes, but because it needs to be done:
I have strong confidence in 2009-10 as the all-time greatest regular season performance. I agree there are questions about hypothetically different matchups in the postseason, and for as superhuman as Lebron was in Games 1-5 of the Magic series, Game 6 was a costly blemish which I do not think later Lebrons would have given that he averaged 35/10.6/7.9 on 61% efficiency in 20 elimination games from 2012-24 (looks awfully 2009-esque to me
).
2012 versus 2013 feels like a functional tie, with any disparity in postseason production and plus/minus more a product of opponents (2013 Spurs the best team, 2013 Pacers the best defence) and team injury or variance (better bench in 2013, injured Wade a bigger drain than injured Bosh). 2013 had the better regular season and closeout game, 2012 was not a single shot away from losing a title. I also think there is a strong possibility 2014 is the better postseason performer than both, but that veers a little too deeply into pure hypotheticals to be worth anything here aside from providing the faintest of temporal edges to 2013.
And then 2016 has imo the greatest three-game stretch of play in NBA history, on a team far more dependent on Lebron than the Heat ever were (both on the court and in the locker-room). That said, up until those final three games, no one would consider 2016 as his secret peak. If I needed a single game or series win, I think I would pick
this version of Lebron… but for the past forty years, a team has needed four series and 15-16 games, and that is enough for me to edge toward 2013.
1. LeBron James (2013 > 2012 > 2016 > 2009)
2. Hakeem Olajuwon (1993 = 1994)
3. Michael Jordan (1990 > 1991)Will tally votes shortly. Any final votes, make them now, and any final edits, please make a separate post to that effect notifying me what you edited. Moderately obvious winner removes the need for Condorcet tallying of the players themselves, but unfortunately I expect I will still need to calculate the choice of year.
:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/15140115/20130620_ajl_su8_305.0.jpg)
