What impact metrics show MJ as a GOAT candidate?

[ceiling raiser]

Do any?

[colts18]

The better question is what impact metrics do not show MJ as a GOAT candidate? Literally all available data we have has MJ as a GOAT level candidate.

-Every single Plus/Minus stat we have shows MJ is an Elite

-Every single RAPM shows MJ as an Elite (I'm not counting Wizards MJ)

-He played on some of the best teams in NBA history. 3 10+ SRS teams, 6 teams with a 6+ SRS

[DraymondGold]

Do any?

Ha, love the "Do any?" post... simple and straight to the point. Anyway, to answer your question:

....

There's definitely impact metrics that show MJ is a GOAT candidate. But before we begin, I think it's worth defining two terms....

(1) What do you mean by impact metrics? If by impact metrics you mean "only actual plus minus data", then our arguments for Jordan will be limited by the obvious fact that Jordan came before the plus-minus era. By that line of thinking, Kareem, Russell, Wilt, Bird, Magic, and the rest all don't have an impact-metric-driven GOAT case, since they too came before the plus minus era. To be clear, we do have partial plus-minus data for Jordan, which I'll get to in a bit. But not complete plus minus data. Me personally, I don't like the idea of discounting players' GOAT cases simply because they played before 1997. But in order for that to happen, in order to make a statistical GOAT case for pre-1997 players, we have to be more flexible with the stats we use (i.e. including box-stats) and accept that the plus minus data we do have comes with smaller samples and thus more uncertainty.

(2) And what do you mean by GOAT candidate? You might mean GOAT peak, GOAT prime, or GOAT career. People typically use stats to argue the first two (GOAT peak or GOAT prime), as this avoids the question of how to integrate value over time. And MJ absolutely has GOAT-tier peak or prime stats. If you mean GOAT career, well, very few people argue GOAT candidacy using stats that actually span players' full careers. For example, people rarely count up "career cumulative RAPM" or "career cumulative PIPM" or "career cumulative WOWY" to argue GOAT candidacy (with the exception of career CORP, which is more of Thinking Basketball's personal estimation than a normal stat). Instead, what they usually do is look at some combination of stats for peak or prime, then make longevity arguments or career cumulative arguments from there.

If you're specifically interested in GOAT career candidates with a focus on cumulative value or longevity, then this is probably one of Jordan's weakest areas as a GOAT candidate. Jordan played 15 seasons total... including 86 when he played 18 regular season games plus playoffs, 95 when he played 17 regular season games plus playoffs, and both Wizard years. Meanwhile LeBron has played 18 seasons and counting, and Kareem played 21 seasons. Now I generally think total seasons aren't the best barometer for longevity because:
(1) they don't account for era differences (e.g. Jordan and Kareem were required to go to college, unlike LeBron; and sports medicine has improved through the years, enabling better longevity).
(2) They don't account for the relative value of different seasons (Kareem's last seasons, Jordan's Wizard seasons, and LeBron's first seasons barely move the needle at all in their GOAT case).
... but still, Kareem and LeBron have the clear longevity advantage, so you have to rate Jordan's prime that much more than Kareem's and LeBron's in order to pick Jordan as the #1 GOAT from a career-value perspective. Those 3 separate retirements hurt Jordan here, particularly the lost 94/95 seasons. But if you value Jordan's prime highly, you still might have him in the GOAT tier, even if you think LeBron or Kareem have passed him from a career value perspective.

Now let's get to the stats that have Jordan as a GOAT candidate in peak / prime. For some of these stats, I'll be comparing Jordan with LeBron, not because I want this to be another LeBron vs Jordan thread (I actively do not), but because I think most would agree that LeBron is a GOAT candidate. So if Jordan compares well relative to LeBron, that means he probably has a GOAT case, even if you still side with LeBron.

A. Plus-Minus Stats (and box estimates of plus-minus stats)
1. 3-Year Postseason Augmented Plus Minus: 2nd all-time. Using Jordan's actual playoff on/off data, Jordan is a hair behind Duncan at 1st (well within uncertainty ranges to put Jordan first) and above every 3-year stint from LeBron.
2. 3-Year Postseason PIPM: 3rd all time.* Jordan's ahead of Miami LeBron, which is usually considered LeBron's peak (although LeBron has other samples that creep ahead).
3. 3-year Postseason RAPTOR: 1st all time.*
4. 3-year Postseason On/off: 4th all time. This is a noisy stat that's clearly worse than RAPM/etc. However, we don't have playoff RAPM yet. In raw playoff on/off per 48 mins, Jordan ranks 5th all time (after Ray Allen, Robinson, Duncan, Garnett; better than LeBron). See here for other sample sizes (viewtopic.php?p=104317081#p104317081).
5a. Regular Season RAPM, single season: 8th all time. This comes from Squared2020's fantastic historical research into Jordan's actual plus minus data. However, note that this comes from a half-season sample in 1988, a two-thirds-season sample in 1991 where the Bulls drastically underperformed (i.e. missing many of the Bulls' best games), and one-third-season sample in 1996. With larger samples in 1991, and with actual data in 1989, 1990, 1992, and 1993, it's very reasonable to expect Jordan to rank higher here.
Edit: 5b. Regular Season RAPM, multi-season: Tier 1 all time. It's hard to do a fair multi-year RAPM comparison for Jordan, given we don't have every season, the seasons we do have are incomplete, and those incomplete samples are likely off from the full-season value. However, I tried to do an apples-to-apples comparison of LeBron vs Jordan in RAPM. LeBron's best years seem higher than Jordan's available years, but Jordan's more consistent and his multi-year average is better. See Post #59 Part 2 (https://forums.realgm.com/boards/viewtopic.php?p=103920388#p103920388) later in this thread for details.
6. Regular Season 3-year PIPM: 1st all time.* Jordan's 1st in PIPM across a full season (1st-Cavs-stint LeBron creeps ahead in PIPM per-possession).
7. Regular Season 3-year RAPTOR: 1st all time.* Jordan's 1st in RAPTOR across a full season and per possession.

*Note that some of these stats use box-estimates compared to real plus minus data for more modern players. No, this is not a perfect comparison. But. It is the best we have, and the box estimates are designed to mimic the real plus-minus data as closely as possible. It's as a fair a statistical comparison as we can make.
*Note that some of these stats do not cover all NBA seasons. Russell and Wilt are notably missing from all of them. However, PIPM/RAPTOR estimates (#2, #3, #6, #7) go back to the NBA merger in 1977.

B. WOWY-based stats
Plus minus stats give a good sense of value within a particular role. But they usually only go back to 1997 (excepting the historical work of Squared2020, thinking basketball). However, we can create a plus-minus-like stat using games as the sample... so our "on" sample is games where someone played, and our "off" sample is games when a player missed. This is a noisy statistic (even noisier than plus/minus data). However, it allows us to measure a player's value in a role going all the way back to 1955!

Jordan performs worse in raw WOWY ("only" 32nd all time). However, just like we have reason to distrust raw on/off -- as it doesn't fully correct for the value of who you're playing with, against, and who's replacing you in a lineup -- we have a similar reason to distrust raw WOWY. And just like we can "adjust" raw plus-minus to create APM and RAPM to correct for the other players, we can do the same for WOWY.
8. 10-year prime WOWYR: 4th all time.* This is like "Adjusted WOWY", and it places Jordan's 10-year prime over Russell's, Kareem's, LeBron's, Wilt's, and Duncan's.
9. 10-year prime GPM: 8th all time.* This is an similar stat to WOWYR, calculated in a different way. It places Jordan just below Russell, but ahead of Kareem, LeBron, Wilt, Duncan.

C. Box stats
10a. 1-year Postseason BPM: 1st all time.
10b. 3-year Postseason BPM: 1st all time.
10c. 5-year Postseason BPM: 1st all time.
11a. 1-year Regular Season BPM: 1st all time.
11b. 3-year Regular Season BPM: 1st all time.
11c. 5-year Regular Season BPM: 1st all time.
12. 3-year Postseason Basketball Reference BPM: 1st all time.
13. 3-year Postseason WS/48: 1st all time
This is pretty clear-cut. Jordan's a GOAT candidate in the box stats.
*Note: I'm using Thinking Basketball BPM for 10a and 11b, as it tests as the most accurate box-stat for predicting wins. I'm hesitant to continue too far down the box-stat rabbit hole (e.g. looking at PER, etc.), but those also portray Jordan as a GOAT candidate. And to its credit, BPM is one of the best all-in-one stats we have going back to Russell's time.

So: Across these 13 fairly industry-standard stats so far, Jordan comes out: 1st all-time 7x, 2nd all-time 1x, 3rd all-time 1x, 4th all-time 2x, 8th all-time 2x. That level of dominance is absolutely deserving of a statistical GOAT-case.

D. Team Stats
We can also use team stats to help infer the value of a player. Now let me be clear: teammates matter, coaching matters, context matters. Team stats alone cannot rank players. But... at the GOAT level, it's reasonable to expect quite a bit of lift. We can look for statistical evidence of clear floor raising when a GOAT-tier player has a poor supporting cast, and for cases of all-time dominance when a GOAT-tier player has a good supporting cast. Jordan had a good supporting cast. Do his teams show all-time dominance, i.e. might we infer GOAT-level ceiling raising from Jordan during his prime? Absolutely.
14. Playoff SRS: 4th, 5th all time. Jordan's team was 5th in 1991 during Jordan's 1-year peak, higher than LeBron/Russell/Wilt/Shaq/Kareem/Duncan/etc. in years that usually go as their 1-year peak.
15. Playoff common-opponent Net Rating: 1st, 4th all time. They were 4th in 1991 during Jordan's 1-year peak.
16. Playoff record: 5th, 11th all time. They were 5th in 1991 during Jordan's 1-year peak.
17. Regular-Season / Playoff ELO: 2nd, 3rd, 9th, 10th, 11th all time.
18. Regular Season SRS: 2nd, 5th, 9th all time.
19. Regular Season Record: 2nd, 5th, 9th all time.
*Edit: see post #159 for more details here (https://forums.realgm.com/boards/viewtopic.php?p=104242045#p104242045)

During Jordan's 10-year prime, Jordan's teams showed peaks reached levels that almost no other team did. Again, this crude method doesn't distinguish between Jordan and his supporting cast. But these statistics do show that Jordan can be the clear-cut best player in a ceiling-raising role on one of the most dominant dynasties ever.

In short: Yes, Jordan absolutely has a statistical case for GOAT peak and GOAT prime, and thus (arguably) for GOAT career. While we do not have all the data we want for Jordan (mainly no full-career actual plus-minus data), the stats we do have absolutely paint him as having a GOAT-tier peak and prime.

This is not to say the statistics universally favor Jordan over other GOAT candidates like LeBron or Kareem or Russell. They don't. I'm sure you could find a similar array of stats to support LeBron. Taking different sample sizes (e.g. 2-year, 4-year, 8-year, even 10-year playoff runs) would similarly shake up the order. And as above, the more you focus on longevity, the better the case LeBron and Kareem have.

But if you're just asking do the impact metrics we have portray Jordan as a GOAT candidate, then yes. Absolutely. At his best, Jordan was absolutely GOAT-tier player.

[SpreeS]

There are plenty of them. What cant I find where is Kobes impact metrics for TOP10 conversation.

[LukaTheGOAT]

Watch on YouTube

https://fivethirtyeight.com/features/why-michael-jordan-was-the-best/

https://www.espn.com/nba/story/_/id/23456720/is-lebron-james-michael-jordan-greatest-nba-player-all

https://www.reddit.com/r/nba/comments/xy2c3n/500_greatest_careers_of_alltime_101_oc/

https://matthewmurphy1.wordpress.com/2020/06/22/417/

viewtopic.php?f=64&t=1971138&p=83324715#p83324715 (The Old BPM)

viewtopic.php?f=64&t=2105498&p=92682446&hilit=Jordan#p92682446

Watch on YouTube

[LukaTheGOAT]

There are plenty of them. What cant I find where is Kobes impact metrics for TOP10 conversation.

Someone wrote this a bit ago, as a case for Kobe.

viewtopic.php?f=64&t=1445779

[OhayoKD]

Since mj-topics tend to boil over, i'd like to preface this with a request that we all try our best to be nice to each other

Now, to jump into the deep end...

Do any?

My short answer is no. But getting to that short answer is a more involved process. Let's start by establishing what should qualify as "GOAT" lvl data:

The better question is what impact metrics do not show MJ as a GOAT candidate? Literally all available data we have has MJ as a GOAT level candidate.

-Every single Plus/Minus stat we have shows MJ is an Elite

-Every single RAPM shows MJ as an Elite (I'm not counting Wizards MJ)

-He played on some of the best teams in NBA history. 3 10+ SRS teams, 6 teams with a 6+ SRS

Sure. Jordan looks elite. And if we consider being elite the same as being a goat candidate, then there's no metric I'm aware of that suggests he isn't. However, the G in "GOAT" stands for greatest not elite. I'd argue that "elite" isn't really the bar we should set here. And it's worth noting that none of what Colts cited has him scoring the "greatest" even when many of these metrics he references don't include historical candidates like kareem and russell.

The only "winning" metric where he actually looks like "the greatest" is if you go by team success as colts does in point 3, but consider that "impact" is usually considered distinct from "total success". Impact, as its commonly used, denotes "isolated influence" on winning, not the whole pie, and even then, if we focus on winning championships, as opposed to srs(which can fluctuate in how indicative it is of championship probablity from era to era), Jordan quite clearly loses to Russell who also seems to outpace him in all the available impact data we have.

So in short, no. I don't think there are impact metrics which generally suggest Jordan as "the greatest." There are box-metrics which put him in the range, but these are not extrapolated from winning. Typically "impact" denotes when you isolate "individual influence on winning". Not when you look at different box-stuff and ascribe this or that many points to this or that category working with assumptions such as "blocks from a guard must be more valuable than blocks from a center." With that in mind...

A. Plus-Minus Stats (and box estimates of plus-minus stats)
1. 3-Year Postseason Augmented Plus Minus: 2nd all-time. Using Jordan's actual playoff on/off data, Jordan is a hair behind Duncan at 1st (well within uncertainty ranges to put Jordan first) and above every 3-year stint from LeBron.
2. 3-Year Postseason PIPM: 3rd all time.* Jordan's ahead of Miami LeBron, which is usually considered LeBron's peak (although LeBron has other samples that creep ahead).
3. 3-year Postseason RAPTOR: 1st all time.*
4. 3-year Postseason On/off: 9th all time. This is a noisy stat that's clearly worse than RAPM/etc. However, we don't have playoff RAPM yet. In raw on/off, Jordan still places high, having a better "on" rating than all but 2 of the players ahead of him and having a better on/off than Miami and 2nd-stint-Cavs LeBron.

The data being referenced here does not have plus-minus before 1997 and even then, whatever data is present is heavily informed by priors from seasons. These metrics, as they exist for Jordan are effectively variants of stuff like PER. They are looking at Jordan's box-stuff and then extrapolating offensive and defensive value. This is notable as using available impact data as opposed to box-stuff, Jordan's defense does not compare well to Lebron's at any point in his prime. As Lebron and Jordan are virtually tied on the offensive portion of all these metrics, simply replacing the defensive component with actual impact data, knocks Jordan off his perch. And remember, this is not including Kareem whose defenses were 4 points better, or Russell who won the most, by a landslide, on the strength of his team's defense.

Furthermore, if we use a more general frame as opposed to a specific one(3 year consecutive)...

Ben only lists 1 three year sample for jordan

Jordan's average from 89-91 is +7.7 in backpicks bpm and +7 in aupm, averaging to +7.35

For Lebron Ben lists three different three year samples(08-10, 12-14, 15-17) in AUPM:

08-10:+6.8, 16-18:+6.7, 12-14:+5.1

He only lists 12-14 for both bpm and aupm:

12-14: +5.8

So if we just use his three year averages, we see that lebron has a bunch of three-year stretches close to MJ but MJ's 89-91 scores highest. However it's interesting that lebron has three 3 year stretches that rank so high(2nd, 3rd, and 7th) and the lowest score came from the years ben rated as Lebron's peak. If we take a look at lebron's best two years from each of those three year stretches...(aupm/bpm is averaged)

Lebron's 09, 10, 12, 13, 16 and 17 are at +10, +8, +7.5, +7, +7.5, and +7.7 respectively.

You get 5 different seasons which would boost the average of MJ's 89-91 peak.

As draymond has covered with PIPM, even using a three-year frame, Lebron has 2 better samples, and the gap naturally widens when we just look at the best years as opposed to a 3-year lens. Notably, DPIPM's weightings are most closely tied to DRAPM making it the most "impacty" of the metrics in question. It shouldn't come as a surprise then that Jordan does the worst here looking at these box-score proxies. Playoff PIPM is also notably the one box-metric where Hakeem, arguably the most valuable defender since Bill Russell, has the second highest career average after...Lebron.


If we go onto actual plus-minus data, Jordan falls short in playoff on/off with 88-93 coming significantly behind 16-21 Lebron and being on par with the on/off for shaq and curry. Notably, Lebron's teams were as good with Lebron on the court as the Bulls were with Jordan. Keep in mind we don't have the data for players like Kareem, Hakeem, Bill, Magic, Bird, Walton, Wilt, or Russell. Jordan is competing in a very, very narrow field here and still doesn't look the best.

We have regular season on/off for Jordan's 97/98 and again, this doesn't look GOAT-worthy. Using Lebron as a reference, 98 scores lower than 18 different Lebron seasons. 97 scores lower than 17.

What we have from Jordan in RAPM, including data from 2 years often included in GOAT regular season conversations, also doesn't compare well to what we have for Lebron:

From the peaks project...
LeBron:

+8.84 in 2009 (would be 4th all time), +9.73 in 2011 (would be 2nd all time), +9.5 in 2012 (would be 2nd all time), +6.4 in 2013 (would be 13th all time), 6.79 in 2014, 8.7 in 2015, +8.62 in 2016 (would be 4th all time), 6.62 in 2017, 1.56 in 2018 (holy coasting! wow!), 3.44 in 2019.

Jordan:

+7.47 in 1988 (would be 8th all time. 43 game sample where Bulls just barely performed better than their average season level), +6.40 in 1991 (57 games where Bulls drastically underperformed their average season level), +7.17 in 1996 (21 games sample where Bulls performed at their average season level), +5.85 in 1997 (full season sample), +6.15 in 1998.```

Finally we have WOWY, which offers us the largest and most inclusive samples of data per-game and per-season and which isn't subject to in-era bias as long as you keep in-era srs fluctuation in mind(The Celtics were the greatest team ever at 50-60 wins a season while the Bulls, Lakers, and Warriors were not close at 60-70 wins a season). And here, when pre-97 greats finally get their shot, Jordan looks significantly worse. With an optimistic appraisal of 23 win lift using an 82 game sample in 84(going by record instead of srs and assuming no improvement despite Oakley spiking the Bulls D-rating), Jordan comes out about even with post-prime Duncan(30 games 04/05), somewhat behind the best stuff we see from Hakeem(25 and 30 game lift in 20 game samples in 88 and 90), consistently behind Kareem throughout the 70's(30 win lift in 75, a 29 win improvement with a player similar to oakley as a rookie, 62 wins without his co-star, and takes the depleted remnants of a 30 win team to 45 wins in 77), and a pretty sizable gap compared to Lebron who has multiple 40 win signals for 09 and 10, 30 win signals in his second cavs stint, and is mostly operating at, at least 20+ win lift throughout his prime leading multiple teams to 60 or near-60 win basketball without co-stars on top-heavy rosters(cavs, heatles).

Then we have Russell, leader of the greatest team ever, who, using an 82 game WOWY sample, seems to have beat two superteams on his last-legs with subpar help in 69, who has the best pre-nba impact alongside kareem in a more stacked league, saw the highest point differential at the olympics, and was able to consistently succeed with teammates going in and out of the lineup.

There are various other players who look alright in different frames, but to keep this succinct, WOWY(and the various derivations you can use to estimate it) really marks the "purest" family of impact signals and Jordan just doesn't look like the best, or even close to the best here. This holds true even if you insist on operating with the tiniest possible samples.

And just like we can "adjust" raw plus-minus to create APM and RAPM to correct for the other players, we can do the same for WOWY.

[/quote]
Except you can't, because WOWYR does not use lineup-level data, it utilizes game-level data. Even RAPM has various limitations that makes treating it like a "better" version of WOWY misguided. But this is just malpractice with WOWYR:

Instead of using results from lineups within a game (play-by-play data) like traditional APM, game-level plus-minus uses final scores from game to game for the players from that game. This allows for a historical, apple-to-apples comparison of per-game impact from before play-by-play was available (1997).

What this means is that you only really get data for a player when they aren't in the starting lineup. The adjustment's "correction" is marginal, and the sample of data we're working with here is realtively tiny. With Jordan, even including the 82 game samples from 1984 and 1994(not included in the WOWYR dray lists), we only have 8 games without MJ per season to work off. Take out that 164 game boost, and it's probably more like 4 games per season. For Bill, we only have 2.2 games a season. You are making the sample exponentially smaller, for a marginal improvement in "noise". This is a much, much smaller sample to work with than the unfavorable partial rapm data Dray quickly dismisses, and even then, Jordan does not really look like the greatest. Notably, if we take WOWYR seriously, Bill Russell led the greatest team ever with 35 win help throughout his prime while Jordan barely won half as much with 40-50 win help. While Jordan looks marginally better than Lebron, he's not really within range of GOAThood.

This is not to say the statistics universally favor Jordan over other GOAT candidates like LeBron or Kareem or Russell. They don't. I'm sure you could find a similar array of stats to support LeBron.

This is a wierd characterization. The statistics you've referenced near-universally favor Lebron in the majority of comparative frames with the vast majority of nba history not included in your sample. Moreover the impacty data we have consistently favors Lebron by virtually any frame and, when it is available, also favors Kareem and Russell with similar consistency. Notably the gap widens the more "impacty" data becomes with Jordan's relative statistical profile looking worse relative to other greats the more inclusive the data becomes.

Even if we stick to what you characterize as box-metrics(not really relevant but fine), Lebron wins out in a variety of frames(looks better in the playoffs with most frameworks), and Kareem remains competitive despite being bogged down with incomplete data.

If you want to define "GOAT TIER" as broadly as Colts has, fine. But otherwise, I don't really see how you can say the impact data is Goaty. Or let me put this another way...

But if you're just asking do the impact metrics we have portray Jordan as a GOAT candidate, then yes. Absolutely. At his best, Jordan was absolutely GOAT-tier player.

...Duncan scores higher in aupm despite aupm being partially constructed with BPM, scores as high as a pretty optimistic MJ WOWY appraisal in injury plagued 04/05, looks similarly dominant in RAPM stuff(though this gets very noisy with different scales), and won 57 and 62 wins at his most valuable looking stretch as opposed to 50 for Jordan in 1988.

Hakeem looks better if you use his very best WOWY samples, looks better in his first three years, and looks similarly impactful throughout his prime, while scoring higher in postseason PIPM(the box metric which most closely is tied to actual defensive impact.(remember that pre 97, none of the "plus-minus" stats you reference have plus minus(or film tracking)). Hakeem also scores similarly in 97/98 on/off despite arguably being further from his peak than Jordan was those years.


IIRC, you have dismissed both Duncan and Hakeem as having GOAT-level data on multiple occasions. If Jordan's impact stats potray him as "absolutely GOAT-Level at his best", why don't you extend that for Hakeem and Duncan who do just as well if not better using data which actually has "impact" in it.

[1993Playoffs]

None

Edit - none that I’ve seen

[Dutchball97]

Pretty sure it depends what you can even find. All I've seen is data starting from 97 or updates with data slightly before that. AuPM shows him as having All-time level impact in both 97 and 98 despite that not being him at his best anymore. I've seen RAPM for 1996 where Jordan ranks as the best in the league. The WOWY and WOWYR I've seen show MJ as one of the top guys historically. Ahead of LeBron but behind the likes of Magic, Stockton, Robinson and actual GOAT candidate Serge Ibaka.

I know thinkingbasketball has quite a few historical stats but that's behind a paywall. It'd be a great help to basketball discussions if impact metrics like this were as easily and openly available as box metrics on bkref.

[ShaqAttac]

There are plenty of them. What cant I find where is Kobes impact metrics for TOP10 conversation.

Someone wrote this a bit ago, as a case for Kobe.

viewtopic.php?f=64&t=1445779

yo, i aint read all of it but dude went nuclear.

never thought id see kobe being as good as duncan here

[ShaqAttac]

Pretty sure it depends what you can even find. All I've seen is data starting from 97 or updates with data slightly before that. AuPM shows him as having All-time level impact in both 97 and 98 despite that not being him at his best anymore. I've seen RAPM for 1996 where Jordan ranks as the best in the league. The WOWY and WOWYR I've seen show MJ as one of the top guys historically. Ahead of LeBron but behind the likes of Magic, Stockton, Robinson and actual GOAT candidate Serge Ibaka.

I know thinkingbasketball has quite a few historical stats but that's behind a paywall. It'd be a great help to basketball discussions if impact metrics like this were as easily and openly available as box metrics on bkref.

dont wowy have bron waay higher

[Vox Populi]

This clever video dissects the GOAT argument into 5 categories.

- Accomplishments
- Longevity
- Winning
- Statistics
- The Eye Test

The segment on statistics includes a portion on analytics.

Spoiler Alert/Trigger Warning: One clearly wins the argument over the other.

Watch on YouTube

[Dutchball97]

Pretty sure it depends what you can even find. All I've seen is data starting from 97 or updates with data slightly before that. AuPM shows him as having All-time level impact in both 97 and 98 despite that not being him at his best anymore. I've seen RAPM for 1996 where Jordan ranks as the best in the league. The WOWY and WOWYR I've seen show MJ as one of the top guys historically. Ahead of LeBron but behind the likes of Magic, Stockton, Robinson and actual GOAT candidate Serge Ibaka.

I know thinkingbasketball has quite a few historical stats but that's behind a paywall. It'd be a great help to basketball discussions if impact metrics like this were as easily and openly available as box metrics on bkref.

dont wowy have bron waay higher

https://thinkingbasketball.net/metrics/wowyr/

I got it from here and MJ seems to come up ahead of LeBron in most of the categories. I'm not making any sweeping statements about how this now means MJ is definitely more impactful than LeBron though because it's very possible there are older and/or newer versions that say the opposite.

[AEnigma]

^ WOWY is something Ben paywalls, although you can find some of the raw numbers on his player profiles.

[falcolombardi]

Pretty sure it depends what you can even find. All I've seen is data starting from 97 or updates with data slightly before that. AuPM shows him as having All-time level impact in both 97 and 98 despite that not being him at his best anymore. I've seen RAPM for 1996 where Jordan ranks as the best in the league. The WOWY and WOWYR I've seen show MJ as one of the top guys historically. Ahead of LeBron but behind the likes of Magic, Stockton, Robinson and actual GOAT candidate Serge Ibaka.

I know thinkingbasketball has quite a few historical stats but that's behind a paywall. It'd be a great help to basketball discussions if impact metrics like this were as easily and openly available as box metrics on bkref.

dont wowy have bron waay higher

Wowy is different that the stat ben made (wowyr)

Wowy looks at a team with and without a player over long stretches of time (example when he gets injured and misses a season compared to the prior season, or when he misses 10-15-20 games in a year)

Ben wowyr as i understand just takes whatever it cam from spare seasons which is much smaller and much more "noisy" sample.

As instead of using full season samples (like comparing a year with jordan and one without him cause injury or retirement) or long stretches of 10-20 games you are using 1-2 games a year

In actual wowy (with ot without you) where players miss significant time lebron lift is bigger than almost anyone ever, jordan included

I think ohayokd explained the difference more in depth and i bolded the more relevant paragraphs

Basically in actual "how much your teams improve with you vs without you" jordan falls short of lebron, and it requires a bit of a ultra noisy/small sample stat to get jordan ahead when everythingh else amd better sample leans clearly lebron in that area

Since mj-topics tend to boil over, i'd like to preface this with a request that we all try our best to be nice to each other

Now, to jump into the deep end...

Do any?

My short answer is no. But getting to that short answer is a more involved process. Let's start by establishing what should qualify as "GOAT" lvl data:

The better question is what impact metrics do not show MJ as a GOAT candidate? Literally all available data we have has MJ as a GOAT level candidate.

-Every single Plus/Minus stat we have shows MJ is an Elite

-Every single RAPM shows MJ as an Elite (I'm not counting Wizards MJ)

-He played on some of the best teams in NBA history. 3 10+ SRS teams, 6 teams with a 6+ SRS

Sure. Jordan looks elite. And if we consider being elite the same as being a goat candidate, then there's no metric I'm aware of that suggests he isn't. However, the G in "GOAT" stands for greatest not elite. I'd argue that "elite" isn't really the bar we should set here. And it's worth noting that none of what Colts cited has him scoring the "greatest" even when many of these metrics he references don't include historical candidates like kareem and russell.

The only "winning" metric where he actually looks like "the greatest" is if you go by team success as colts does in point 3, but consider that "impact" is usually considered distinct from "total success". Impact, as its commonly used, denotes "isolated influence" on winning, not the whole pie, and even then, if we focus on winning championships, as opposed to srs(which can fluctuate in how indicative it is of championship probablity from era to era), Jordan quite clearly loses to Russell who also seems to outpace him in all the available impact data we have.

So in short, no. I don't think there are impact metrics which generally suggest Jordan as "the greatest." There are box-metrics which put him in the range, but these are not extrapolated from winning. Typically "impact" denotes when you isolate "individual influence on winning". Not when you look at different box-stuff and ascribe this or that many points to this or that category working with assumptions such as "blocks from a guard must be more valuable than blocks from a center." With that in mind...

A. Plus-Minus Stats (and box estimates of plus-minus stats)
1. 3-Year Postseason Augmented Plus Minus: 2nd all-time. Using Jordan's actual playoff on/off data, Jordan is a hair behind Duncan at 1st (well within uncertainty ranges to put Jordan first) and above every 3-year stint from LeBron.
2. 3-Year Postseason PIPM: 3rd all time.* Jordan's ahead of Miami LeBron, which is usually considered LeBron's peak (although LeBron has other samples that creep ahead).
3. 3-year Postseason RAPTOR: 1st all time.*
4. 3-year Postseason On/off: 9th all time. This is a noisy stat that's clearly worse than RAPM/etc. However, we don't have playoff RAPM yet. In raw on/off, Jordan still places high, having a better "on" rating than all but 2 of the players ahead of him and having a better on/off than Miami and 2nd-stint-Cavs LeBron.

The data being referenced here does not have plus-minus before 1997 and even then, whatever data is present is heavily informed by priors from seasons. These metrics, as they exist for Jordan are effectively variants of stuff like PER. They are looking at Jordan's box-stuff and then extrapolating offensive and defensive value. This is notable as using available impact data as opposed to box-stuff, Jordan's defense does not compare well to Lebron's at any point in his prime. As Lebron and Jordan are virtually tied on the offensive portion of all these metrics, simply replacing the defensive component with actual impact data, knocks Jordan off his perch. And remember, this is not including Kareem whose defenses were 4 points better, or Russell who won the most, by a landslide, on the strength of his team's defense.

Furthermore, if we use a more general frame as opposed to a specific one(3 year consecutive)...

Ben only lists 1 three year sample for jordan

Jordan's average from 89-91 is +7.7 in backpicks bpm and +7 in aupm, averaging to +7.35

For Lebron Ben lists three different three year samples(08-10, 12-14, 15-17) in AUPM:

08-10:+6.8, 16-18:+6.7, 12-14:+5.1

He only lists 12-14 for both bpm and aupm:

12-14: +5.8

So if we just use his three year averages, we see that lebron has a bunch of three-year stretches close to MJ but MJ's 89-91 scores highest. However it's interesting that lebron has three 3 year stretches that rank so high(2nd, 3rd, and 7th) and the lowest score came from the years ben rated as Lebron's peak. If we take a look at lebron's best two years from each of those three year stretches...(aupm/bpm is averaged)

Lebron's 09, 10, 12, 13, 16 and 17 are at +10, +8, +7.5, +7, +7.5, and +7.7 respectively.

You get 5 different seasons which would boost the average of MJ's 89-91 peak.

As draymond has covered with PIPM, even using a three-year frame, Lebron has 2 better samples, and the gap naturally widens when we just look at the best years as opposed to a 3-year lens. Notably, DPIPM's weightings are most closely tied to DRAPM making it the most "impacty" of the metrics in question. It shouldn't come as a surprise then that Jordan does the worst here looking at these box-score proxies. Playoff PIPM is also notably the one box-metric where Hakeem, arguably the most valuable defender since Bill Russell, has the second highest career average after...Lebron.


If we go onto actual plus-minus data, Jordan falls short in playoff on/off with 88-93 coming significantly behind 16-21 Lebron and being on par with the on/off for shaq and curry. Notably, Lebron's teams were as good with Lebron on the court as the Bulls were with Jordan. Keep in mind we don't have the data for players like Kareem, Hakeem, Bill, Magic, Bird, Walton, Wilt, or Russell. Jordan is competing in a very, very narrow field here and still doesn't look the best.

We have regular season on/off for Jordan's 97/98 and again, this doesn't look GOAT-worthy. Using Lebron as a reference, 98 scores lower than 18 different Lebron seasons. 97 scores lower than 17.

What we have from Jordan in RAPM, including data from 2 years often included in GOAT regular season conversations, also doesn't compare well to what we have for Lebron:

From the peaks project...
LeBron:

+8.84 in 2009 (would be 4th all time), +9.73 in 2011 (would be 2nd all time), +9.5 in 2012 (would be 2nd all time), +6.4 in 2013 (would be 13th all time), 6.79 in 2014, 8.7 in 2015, +8.62 in 2016 (would be 4th all time), 6.62 in 2017, 1.56 in 2018 (holy coasting! wow!), 3.44 in 2019.

Jordan:

+7.47 in 1988 (would be 8th all time. 43 game sample where Bulls just barely performed better than their average season level), +6.40 in 1991 (57 games where Bulls drastically underperformed their average season level), +7.17 in 1996 (21 games sample where Bulls performed at their average season level), +5.85 in 1997 (full season sample), +6.15 in 1998.```

Finally we have WOWY, which offers us the largest and most inclusive samples of data per-game and per-season and which isn't subject to in-era bias as long as you keep in-era srs fluctuation in mind(The Celtics were the greatest team ever at 50-60 wins a season while the Bulls, Lakers, and Warriors were not close at 60-70 wins a season). And here, when pre-97 greats finally get their shot, Jordan looks significantly worse. With an optimistic appraisal of 23 win lift using an 82 game sample in 84(going by record instead of srs and assuming no improvement despite Oakley spiking the Bulls D-rating), Jordan comes out about even with post-prime Duncan(30 games 04/05), somewhat behind the best stuff we see from Hakeem(25 and 30 game lift in 20 game samples in 88 and 90), consistently behind Kareem throughout the 70's(30 win lift in 75, a 29 win improvement with a player similar to oakley as a rookie, 62 wins without his co-star, and takes the depleted remnants of a 30 win team to 45 wins in 77), and a pretty sizable gap compared to Lebron who has multiple 40 win signals for 09 and 10, 30 win signals in his second cavs stint, and is mostly operating at, at least 20+ win lift throughout his prime leading multiple teams to 60 or near-60 win basketball without co-stars on top-heavy rosters(cavs, heatles).

Finally you have Russell, leader of the greatest team ever, who, using an 82 game WOWY sample, seems to have beat two superteams on his last-legs with subpar help in 69, who has the best pre-nba impact alongside kareem in a more stacked league, saw the highest point differential at the olympics, and was able to consistently succeed with teammates going in and out of the lineup.

There are various other players who look alright in different frames, but to keep this succinct, WOWY(and the various derivations you can use to estimate it) really marks the "purest" family of impact signals and Jordan just doesn't look like the best, or even close to the best here. This holds true even if you insist on operating with the tiniest possible samples.

And just like we can "adjust" raw plus-minus to create APM and RAPM to correct for the other players, we can do the same for WOWY.

Except you can't, because WOWYR does not use lineup-level data, it utilizes game-level data. Even RAPM has various limitations that makes treating it like a "better" version of WOWY misguided. But this is just malpractice with WOWYR:

Instead of using results from lineups within a game (play-by-play data) like traditional APM, game-level plus-minus uses final scores from game to game for the players from that game. This allows for a historical, apple-to-apples comparison of per-game impact from before play-by-play was available (1997).

What this means is that you only really get data for a player when they aren't in the starting lineup. The adjustment's "correction" is marginal, and the sample of data we're working with here is realtively tiny. With Jordan, even including the 82 game samples from 1984 and 1994(not included in the WOWYR dray lists), we only have 8 games without MJ per season to work off. Take out that 164 game boost, and it's probably more like For Bill, we only have 2.2 games a season. You are making the sample exponentially smaller, for a marginal improvement in "noise". This is a much, much smaller sample to work with than the unfavorable partial rapm data Dray quickly dismisses, and even then, Jordan does not really look like the greatest. Notably, if we take WOWYR seriously, Bill Russell led the greatest team ever with 35 win help throughout his prime while Jordan barely won half as much with 40-50 win help. While Jordan looks marginally better than Lebron, he's not really within range of GOAThood.

This is not to say the statistics universally favor Jordan over other GOAT candidates like LeBron or Kareem or Russell. They don't. I'm sure you could find a similar array of stats to support LeBron.

This is a wierd characterization. The statistics you've referenced near-universally favor Lebron in the majority of comparative frames with the vast majority of nba history not included in your sample. Moreover the impacty data we have consistently favors Lebron by virtually any frame and, when it is available, also favors Kareem and Russell with similar consistency. Notably the gap widens the more "impacty" data becomes with Jordan's relative statistical profile looking worse relative to other greats the more inclusive the data becomes.

Even if we stick to what you characterize as box-metrics(not really relevant but fine), Lebron wins out in a variety of frames(looks better in the playoffs with most frameworks), and Kareem remains competitive despite being bogged down with incomplete data.

If you want to define "GOAT TIER" as broadly as Colts has, fine. But otherwise, I don't really see how you can say the impact data is Goaty. Or let me put this another way...

But if you're just asking do the impact metrics we have portray Jordan as a GOAT candidate, then yes. Absolutely. At his best, Jordan was absolutely GOAT-tier player.

...Duncan scores higher in aupm despite aupm being partially constructed with BPM, scores as high as a pretty optimistic MJ WOWY appraisal in injury plagued 04/05, looks similarly dominant in RAPM stuff(though this gets very noisy with different scales), and won 57 and 62 wins at his most valuable looking stretch as opposed to 50 for Jordan in 1988.

Hakeem looks better if you use his very best WOWY samples, looks better in his first three years, and looks similarly impactful throughout his prime, while scoring higher in postseason PIPM(the box metric which most closely is tied to actual defensive impact.(remember that pre 97, none of the "plus-minus" stats you reference have plus minus(or film tracking)). Hakeem also scores similarly in 97/98 on/off despite arguably being further from his peak than Jordan was those years.


IIRC, you have dismissed both Duncan and Hakeem as having GOAT-level data on multiple occasions. If Jordan's impact stats potray him as "absolutely GOAT-Level at his best", why don't you extend that for Hakeem and Duncan who do just as well if not better using data which actually has "impact" in it.

[/quote]

[rk2023]

Hi Draymondgold, thank you for sharing all you did! For sure some great points made, some questions and personal 2 cents I have for your case(s) made (will link data, articles, or points at the end of this Quote exchange).

(1) What do you mean by impact metrics? If by impact metrics you mean "only actual plus minus data", then our arguments for Jordan will be limited by the obvious fact that Jordan came before the plus-minus era. By that line of thinking, Kareem, Russell, Wilt, Bird, Magic, and the rest all don't have an impact-metric-driven GOAT case, since they too came before the plus minus era. To be clear, we do have partial plus-minus data for Jordan, which I'll get to in a bit. But not complete plus minus data. Me personally, I don't like the idea of discounting players' GOAT cases simply because they played before 1997. But in order for that to happen, in order to make a statistical GOAT case for pre-1997 players, we have to be more flexible with the stats we use (i.e. including box-stats) and accept that the plus minus data we do have comes with smaller samples and thus more uncertainty.

Agreed, particularly the fact that other GOAT candidates shouldn't be diminished at first glance because "one can't speak upon their careers". From a film standpoint I get why people say this, but feel there is a point where one can see enough / contextualize / and make valid assumptions through missing pieces in order to gain a better assessment.

(2) And what do you mean by GOAT candidate? You might mean GOAT peak, GOAT prime, or GOAT career. People typically use stats to argue the first two (GOAT peak or GOAT prime), as this avoids the question of how to integrate value over time. And MJ absolutely has GOAT-tier peak or prime stats. If you mean GOAT career, well, very few people argue GOAT candidacy using stats that actually span players' full careers. For example, people rarely count up "career cumulative RAPM" or "career cumulative PIPM" or "career cumulative WOWY" to argue GOAT candidacy (with the exception of career CORP, which is more of Thinking Basketball's personal estimation than a normal stat). Instead, what they usually do is look at some combination of stats for peak or prime, then make longevity arguments or career cumulative arguments from there.

It is more worthwhile to have a criteria that converges into opinions rather than opinions that diverge into a criteria (eg. agenda pushing). Since GOAT candidate is open ended as a term, it boils down to what you value. for me, I factor prime quality > longevity >= peak - while having an extra consideration to adjust longevity for era (see TB article below) and being more adamant on players' playoff track-records and translation. Would also like to re-iterate as I see candidate as one who might not be "the basketball GOAT" with one's criteria but has a key argument shaped by part of it. Will touch on this more, but I see (chronologically) Russell Jabbar Jordan and James to be GOAT candidates - not really much wiggle room after them in terms of ranking players in a fourth spot high-end. When this board does greatest careers in fall (presumably?), I'll get more keen into doubling-down on a formal ranking of these four.

If you're specifically interested in GOAT career candidates with a focus on cumulative value or longevity, then this is probably one of Jordan's weakest areas as a GOAT candidate. Jordan played 15 seasons total... including 86 when he played 18 regular season games plus playoffs, 95 when he played 17 regular season games plus playoffs, and both Wizard years. Meanwhile LeBron has played 18 seasons and counting, and Kareem played 21 seasons. Now I generally think total seasons aren't the best barometer for longevity because:
(1) they don't account for era differences (e.g. Jordan and Kareem were required to go to college, unlike LeBron; and sports medicine has improved through the years, enabling better longevity).
(2) They don't account for the relative value of different seasons (Kareem's last seasons, Jordan's Wizard seasons, and LeBron's first seasons barely move the needle at all in their GOAT case).
... but still, Kareem and LeBron have the clear longevity advantage, so you have to rate Jordan's prime that much more than Kareem's and LeBron's in order to pick Jordan as the #1 GOAT from a career-value perspective. Those 3 separate retirements hurt Jordan here, particularly the lost 94/95 seasons. But if you value Jordan's prime highly, you still might have him in the GOAT tier, even if you think LeBron or Kareem have passed him from a career value perspective.

Agreed, I think when you are looking at aggregate CORP (Ben's Approach) - he's still a candidate but I think has one of the weaker arguments of the four.

Per my view, here is how many >= strong MVP seasons each of the candidates have (going with more of a conservative approach):

Russell: 10
Jabbar: 11
Jordan: 9
James: 11

Anyone can feel free to correct if they feel different; there certainly could be more seasons for each player (Jordan less-so unless there's a different view than I have on 1987) and I'm happy to list which ones I factored in.

A. Plus-Minus Stats (and box estimates of plus-minus stats)

*placeholder*

1. 3-Year Postseason Augmented Plus Minus: 2nd all-time. Using Jordan's actual playoff on/off data, Jordan is a hair behind Duncan at 1st (well within uncertainty ranges to put Jordan first) and above every 3-year stint from LeBron.

Where is this coming from? I may not have the full knowledge of data sources but using BackPicks/Thinking-Basketball - the only AuPM 3-year run I see from Jordan is from 96-98 (of course, this is far from his apex and general prime) where he grades out with a 4.8 - tied with the Likes of 13-16 Westbrook and 03-05 Ginobili amongst others. Five of LeBron's 3-year runs grade out on the top 11 in his database and 4.6 grades out as his lowest-end within this span.

4. 3-year Postseason On/off: 9th all time. This is a noisy stat that's clearly worse than RAPM/etc. However, we don't have playoff RAPM yet. In raw on/off, Jordan still places high, having a better "on" rating than all but 2 of the players ahead of him and having a better on/off than Miami and 2nd-stint-Cavs LeBron.

Some food for thought here, linked both videos because I remember some good insights brought up regarding Jordan v James on impact and some on/off numbers. Can circle back to these in this general thread upon a more focused rewatch.

Watch on YouTube

;t=1425s

Watch on YouTube

;t=761s

5. Regular Season RAPM, single season: 8th all time. This comes from Squared2020's fantastic historical research into Jordan's actual plus minus data. However, note that this comes from a half-season sample in 1988, a two-thirds-season sample in 1991 where the Bulls drastically underperformed (i.e. missing many of the Bulls' best games), and one-third-season sample in 1996. With larger samples in 1991, and with actual data in 1989, 1990, 1992, and 1993, it's very reasonable to expect Jordan to rank higher here.

I see the latter most point as an assumption at the end of the day, but it is one rooted with reasonable "filling in the holes based on what we know logic" so it's cool with me. If you would have a source and a number in this instance I would appreciate it, there is a site I have bookmarked for RAPM pre 2013 but the numbers don't look right to me. From using Englemann's Google-Site (Copied Both at end):

1991-2014 xRAPM/100: James comes out second all time, 7.65
1997-2014 RAPM: James comes out first all time, 5.63
2009 and 2010 Single Season RAPM: James tops the site-base with values of 9.3 & 9.8 respectively.

I also found Squared2020 / Justin Jacobs' RAPM project for 1991 - where Jordan comes out a 7.71. If anyone on this board has some other post 2014 or more accurate RAPM years from pre 2000, definitely feel free to share

2. 3-Year Postseason PIPM: 3rd all time.* Jordan's ahead of Miami LeBron, which is usually considered LeBron's peak (although LeBron has other samples that creep ahead).
6. Regular Season 3-year PIPM: 1st all time.* Jordan's 1st in PIPM across a full season (1st-Cavs-stint LeBron creeps ahead in PIPM per-possession).

I'm unsure of PS PIPM data, but am curious regarding RS. The spread-sheet I have in that instance has these as the top years between the two with RS/PS games used (can link through PM if need be):

2009 James - 10.45 , 2010 James - 8.79 , 1988 Jordan - 8.74 , 1991 Jordan - 8.63 , 2017 James - 8.55 , 2016 James - 8.34 , 2013 James - 8.33 , 1996 Jordan - 8.13 , 1989 Jordan - 7.97. After that, there is a drop-off between the two; the seasons I listed are all in the top 18 single season marks.

3. 3-year Postseason RAPTOR: 1st all time.*
7. Regular Season 3-year RAPTOR: 1st all time.* Jordan's 1st in RAPTOR across a full season and per possession.

Where are the RAPTOR estimates coming from? I have yet to see anything outside of 2014 onwards

*Note that some of these stats use box-estimates compared to real plus minus data for more modern players. No, this is not a perfect comparison. But. It is the best we have, and the box estimates are designed to mimic the real plus-minus data as closely as possible. It's as a fair a statistical comparison as we can make.
*Note that some of these stats do not cover all NBA seasons. Russell and Wilt are notably missing from all of them. However, PIPM/RAPTOR estimates (#2, #3, #6, #7) go back to the NBA merger in 1977.

If you have data sources or links for some of these data-points, it would be immensely appreciated. Am getting more into advanced stats myself, so my working knowledge of what is available isn't down to a T by any means.

B. WOWY-based stats
Plus minus stats give a good sense of value within a particular role. But they usually only go back to 1997 (excepting the historical work of Squared2020, thinking basketball). However, we can create a plus-minus-like stat using games as the sample... so our "on" sample is games where someone played, and our "off" sample is games when a player missed. This is a noisy statistic (even noisier than plus/minus data). However, it allows us to measure a player's value in a role going all the way back to 1955!

Jordan performs worse in raw WOWY ("only" 32nd all time). However, just like we have reason to distrust raw on/off -- as it doesn't fully correct for the value of who you're playing with, against, and who's replacing you in a lineup -- we have a similar reason to distrust raw WOWY. And just like we can "adjust" raw plus-minus to create APM and RAPM to correct for the other players, we can do the same for WOWY.
8. 10-year prime WOWYR: 4th all time.* This is like "Adjusted WOWY", and it places Jordan's 10-year prime over Russell's, Kareem's, LeBron's, Wilt's, and Duncan's.
9. 10-year prime GPM: 8th all time.* This is an similar stat to WOWYR, calculated in a different way. It places Jordan just below Russell, but ahead of Kareem, LeBron, Wilt, Duncan.

I don't have anything here, as the difference between WOWY vs WOWYR - and the case of derivation in the latter isn't something I can speak on too firmly. If you would happen to know, some intel as for that derivation would be appreciated. I find it interesting that there is a decently smaller gap in the two's WOWYR (8.3 v 7.8) compared to the WOWY scores in their 2018 Backpicks profiles. The general premise represents my thoughts, as there is noise and I would trust more attributable-to-player metrics more-so.

C. Box stats
10a. 1-year Postseason BPM: 1st all time.
10b. 3-year Postseason BPM: 1st all time.
10c. 5-year Postseason BPM: 1st all time.
11a. 1-year Regular Season BPM: 1st all time.
11b. 3-year Regular Season BPM: 1st all time.
11c. 5-year Regular Season BPM: 1st all time.
12. 3-year Postseason Basketball Reference BPM: 1st all time.
13. 3-year Postseason WS/48: 1st all time
This is pretty clear-cut. Jordan's a GOAT candidate in the box stats.
*Note: I'm using Thinking Basketball BPM for 10a and 11b, as it tests as the most accurate box-stat for predicting wins. I'm hesitant to continue too far down the box-stat rabbit hole (e.g. looking at PER, etc.), but those also portray Jordan as a GOAT candidate. And to its credit, BPM is one of the best all-in-one stats we have going back to Russell's time.

If you're interested - here's a descending single-season distribution of how both's BPM seasons (in Elgee's model) look like in RS (first) and Play-offs (To Follow). Note that this is up to the 2020-21 NBA season.

Jordan: 9, 8.5, 8.4, 8.3, 7.8, 7.6, 7.2, 6.8, 5.4, 6, 4.7, 4.1, 2.8, 1.7, 1.4

James, 8.8, 8.6, 8.4, 7.2, 6.9, 6.8, 6.8, 6.2, 6.1, 6, 6, 5.8, 5.7, 5.6, 5.3, 5.1, 5, 1.5

Jordan: 12.1, 9.4, 9.3, 9.3, 8.5, 8.3, 7.5, 6.9, 6.7, 6.1, 5.9, 5.9, 5.5

James: 12.0, 9.5, 9.4, 9.3, 8.4, 8.4, 8.3, 8.2, 8.0, 7.0, 6.5, 5.9, 5.6, 4.8, 4.2

When you mention first all-time, It is by a sliver in both Regular Season and Playoffs - where James gains a clear upper hand in total BPM seasons at a star threshold in the former and comes out with a general higher top quartile and playoff track record in the latter.

I would agree with PER and WS/48 as something I'd be less likely to use.. per the DunksAndThrees article I linked below, those are the least predictive at measuring team success - and how an individual is lifting team success i'm assuming is central to a lot on here's criteria.

So: Across these 13 fairly industry-standard stats so far, Jordan comes out: 1st all-time 7x, 2nd all-time 1x, 3rd all-time 1x, 4th all-time 1x, 8th all-time 2x, 9th all-time 1x. That level of dominance is absolutely deserving of a statistical GOAT-case.

I think deserving of a case, but I wouldn't go as far to crown him the GOAT statistically. Looking at the two common choices used on BBR and (partially because of how sports media frames Bron vs. Jordan - some analytical pieces resulting as such from Pelton, Hollinger, 538, so on), on a year over year basis, I could give you a more consistent year over year prime from Jordan's 88-93 than James from 09-14 (what I see as his best six years), but I think LBJ reaches a slightly higher apex within that time span at different points then gains more ground with hyper-impactful surrounding years such as 16/17/20 and a better career longevity. All in all, feel Russell and Kareem are up there too. Stats (well we only really have WS/48 in Bill's case) are far from the measuring stick of Russell's impact too, just my thought on the subject.

D. Team Stats
We can also use team stats to help infer the value of a player. Now let me be clear: teammates matter, coaching matters, context matters. Team stats alone cannot rank players. But... at the GOAT level, it's reasonable to expect quite a bit of lift. We can look for statistical evidence of clear floor raising when a GOAT-tier player has a poor supporting cast, and for cases of all-time dominance when a GOAT-tier player has a good supporting cast. Jordan had a good supporting cast. Do his teams show all-time dominance, i.e. might we infer GOAT-level ceiling raising from Jordan during his prime? Absolutely.
14. Playoff SRS: 4th, 5th all time. Jordan's team was 5th in 1991 during Jordan's 1-year peak, higher than LeBron/Russell/Wilt/Shaq/Kareem/Duncan/etc. in years that usually go as their 1-year peak.
15. Playoff common-opponent Net Rating: 1st, 4th all time. They were 4th in 1991 during Jordan's 1-year peak.
16. Playoff record: 5th, 11th all time. They were 5th in 1991 during Jordan's 1-year peak.
17. Regular-Season / Playoff ELO: 2nd, 3rd, 9th, 10th, 11th all time.
18. Regular Season SRS: 2nd, 5th, 9th all time.
19. Regular Season Record: 2nd, 5th, 9th all time.

During Jordan's 10-year prime, Jordan's teams showed peaks reached levels that almost no other team did. Again, this crude method doesn't distinguish between Jordan and his supporting cast. But these statistics do show that Jordan can be the clear-cut best player in a ceiling-raising role on one of the most dominant dynasties ever.

I support the general premise, that Jordan was phenomenally impactful in his floor raising and lesser O-Load roles on championship calibre teams (moreso 1996-98). As I linked earlier, from the Jordan on/off tracking video - there was a clear-cut pattern going all the way from 1988-1993 where Jordan's "on moments" got higher - the supporting cast with Jordan "off" numbers showed even more of a substantial improvement. Due to noise in sample and nature of the PS, the "on" is what I take into consideration more - but cast I feel also played a part in catapulting to these heights. Basketball Reference in 1997 and 1998 grades him high, with an 8.8 on court net-rating / 100 and overall -18.2 net-swing from 97 and 98 (although these #s are significantly less in the Regular seasons where the Bulls won ~130 games).

In short: Yes, Jordan absolutely has a statistical case for GOAT peak and GOAT prime, and thus (arguably) for GOAT career. While we do not have all the data we want for Jordan (mainly no full-career actual plus-minus data), the stats we do have absolutely paint him as having a GOAT-tier peak and prime.

This is not to say the statistics universally favor Jordan over other GOAT candidates like LeBron or Kareem or Russell. They don't. I'm sure you could find a similar array of stats to support LeBron. Taking different sample sizes (e.g. 2-year, 4-year, 8-year, even 10-year playoff runs) would similarly shake up the order. And as above, the more you focus on longevity, the better the case LeBron and Kareem have.

But if you're just asking do the impact metrics we have portray Jordan as a GOAT candidate, then yes. Absolutely. At his best, Jordan was absolutely GOAT-tier player.

I can second this general premise, I don't have him as my greatest player of all time personally (if you haven't gotten that impression), but there can be arguments made if you are in the camp that he has the 3 best seasons played or that his 6 year prime is outlier good. I'm far more firm on the concept of a "Mount Rushmore" and tiers within some pyramid/pantheon, rather than outright crowning a GOAT. The most important thing to consider is that everyone's criteria differs.

Pre-2013 RAPM: https://web.archive.org/web/20131025011748/http:/stats-for-the-nba.appspot.com/
Englemann RAPM G-Site: https://sites.google.com/site/rapmstats/xrapm-per-100-91-14
TB Longevity / Older Players: https://backpicks.com/2017/08/01/are-older-players-getting-better-aging-throughout-nba-history/
Dunks and Threes Impact Metrics: https://dunksandthrees.com/blog/metric-comparison?s=09

[SpreeS]

There are plenty of them. What cant I find where is Kobes impact metrics for TOP10 conversation.

Someone wrote this a bit ago, as a case for Kobe.

viewtopic.php?f=64&t=1445779

yo, i aint read all of it but dude went nuclear.

never thought id see kobe being as good as duncan here

Pl link

[ceoofkobefans]

Pretty much all of them (don’t feel like posting them rn but if needed I can)

[ceoofkobefans]

There are plenty of them. What cant I find where is Kobes impact metrics for TOP10 conversation.

Off the dome I know VORP RS+PO has Bean at 11 (.3 points behind dirk for 10th) and RAPTOR WAR has him 8th all time. 5Yr RAPM has his peak at 10th as well

To be fair VORP isn’t pre 74 so I’m sure wilt or bill could be above him which puts him at 13 and same with RAPTOR WAR which is 77- so Kareem also is below him (this would have him 11th as a low end) WS RS+PO has him at 15th (13 NBA only) and these aren’t weighting PO higher or anything (where Kobe is one of the best PO risers and Performers ever) this is just adding RS and PO stats so that should be accounted as well 5 Yr RAPM has his peak at 10th (not career because there isn’t really a cumulative version of RAPM). This doesn’t include MJ Magic Bird Kareem Wilt ot Russell so even if we assume he’s worst than all of them that would still have him at 15th for *peaks* so I think that’s gives him a pretty solid t10 argument there

PIPM has him 3rd in Wins added for the PO idk about RS or RS+PO but PIPM is pretty low on Kobe so I could see him being outside the t10-15 in PIPM for the RS (isn’t pre 74 iirc)

But yea impact metrics love Kobe in the PO especially compared to the RS. He’s 8th in WS 5th in VORP 3rd in PIPM wins added (although I believe that’s the PIPM with On/off data incorporated not just the box model so could he closer to 10 than 1 easily).

[SpreeS]

There are plenty of them. What cant I find where is Kobes impact metrics for TOP10 conversation.

Off the dome I know VORP RS+PO has Bean at 11 (.3 points behind dirk for 10th) and RAPTOR WAR has him 8th all time. 5Yr RAPM has his peak at 10th as well

To be fair VORP isn’t pre 74 so I’m sure wilt or bill could be above him which puts him at 13 and same with RAPTOR WAR which is 77- so Kareem also is below him (this would have him 11th as a low end) WS RS+PO has him at 15th (13 NBA only) and these aren’t weighting PO higher or anything (where Kobe is one of the best PO risers and Performers ever) this is just adding RS and PO stats so that should be accounted as well 5 Yr RAPM has his peak at 10th (not career because there isn’t really a cumulative version of RAPM). This doesn’t include MJ Magic Bird Kareem Wilt ot Russell so even if we assume he’s worst than all of them that would still have him at 15th for *peaks* so I think that’s gives him a pretty solid t10 argument there

PIPM has him 3rd in Wins added for the PO idk about RS or RS+PO but PIPM is pretty low on Kobe so I could see him being outside the t10-15 in PIPM for the RS (isn’t pre 74 iirc)

But yea impact metrics love Kobe in the PO especially compared to the RS. He’s 8th in WS 5th in VORP 3rd in PIPM wins added (although I believe that’s the PIPM with On/off data incorporated not just the box model so could he closer to 10 than 1 easily).

These are all counting stats except Ramp and Pipm where he isnt TOP10 material. If we adjust longevity with previous eras, these counting stats aren’t top10 too