Was Kareem Definitely the King of the 70's?

[MrLurker]

Thank you, the decline of the Celtics from 1969 to 1970 being used as the biggest argument for Russell instead of how he played at his peak (62-65) is the weirdest trend and goes to show the unhealthy obsession the board has with WOWY.

Calling a 246 game sample(82 games a season) small because it suggests your priors are miles off is a "wierd" way of framing evidence.

As is the idea that data should be ignored because it is not a player's perceived "peak". Also known as cherrypicking.

And no, Kareem's 1975 is not really comparable unless you are worried about raw wins more than championships.

I think this board's worst habit is presuming that when beliefs are contradicted by evidence, the only suitable explanation is an epistemological failing. Up there with appealing to consensus when the simplest interpretation of said consensus suggests the exact opposite of what you're arguing.

If you aren't interested in admitting you're probably wrong, or responding in kind when your points are engaged with, you can at least avoid spewing nonsense(sample size, really?) to pretend posters whose conclusions actually track with evidence are using evidence in an "unhealthy" manner. Especially when you don't have any sort of evidentiary counter and have consequently spent the last three pages appealing to voting results as opposed to basketball.

The Celtics from 69 to 70 also lose Sam Jones, Bailey Howell becomes washed up (falls from .215 WS/48 to .072), they change coaches, and they weren’t even that dominant in 69 regular season, so I don’t see how falling to be a below average team (34-48, -1.6 SRS) is supposed to be mind blowing indicator for Russell, we know that Russell is an elite player so they’re supposed to get worse. I say that as someone who has made the argument before that the 60s Celtics being much more dominant than the 70s Celtics can’t have many other explanations than Russell being much more impactful than Cowens.

When a player gets injured or retires there is a multitude of factors like fit, effort/heart, etc. that may make one hold up more than the other, removing Walton from Blazers made jenga tower collapse, removing Kawhi from Raptors didn’t at all, yes that has something to do with Walton being good but I can’t get behind treating it as the end all, especially considering playoffs value is not the same as regular season value, the Raptors did great in regular season before and after Kawhi but needed him in the playoffs. This is before considering the other issues like a player getting injured for 20 games or something which is frequently used here is a small sample and in RealGM’s favorite stat ElGee literally just went through it and manually decided which games not to count cause other teammates were injured at the same time, if you think ElGee is a buddha who has total objectively fair methods of doing this that’s ok, but I see someone that had a long history of being things like “pro-Oscar/Malone” and “anti-Wilt/Dantley” therefore that he can decide what injured samples to use to make his favorite players look better and to confirm his previous takes makes it completely useless. ElGee is a great video editor but I trust his stats work about as much as I trust Malcolm Gladwell to tell accurate history.

There are a few points of confusion for me here.

First - I brought up 71 and 72 where the Celtics had already replaced what they'd lost and then some. Dutchball seems to not interpret your comment as a reply, but your initial comment came immediately after mine.

Second - My source was simply BasketballReference. Who is Elgee - and how does he come into this?

Third - The Celtics were a near +6 team in the regular-season and near the top of the standings by SRS. Dropping from there to missing the playoffs seems like a large drop - especially if we agree Russell was not at his best.

Fourth - Could I get some clarification on your initial point? What would you consider a big set of games. What would you consider a small one?

Please do not take this as me using a soapbox - but I think we've been talking past each other instead of to.

[penbeast0]

Ok, quit repeating the same points and disengage with each other. I would suggest you put each other on ignore for a day.

[Dr Positivity]

Third - The Celtics were a near +6 team in the regular-season and near the top of the standings by SRS. Dropping from there to missing the playoffs seems like a large drop - especially if we agree Russell was not at his best.

Fourth - Could I get some clarification on your initial point? What would you consider a big set of games. What would you consider a small one?

Walton misses 41 games in 77 and 78, Kareem misses like 43 in his prime, mostly in 75 and 78 neither of which are his best seasons. I consider these both small numbers personally where chance like their shooting % being bad and the opponents being good could make a big impact on the on/off. eg. The Heat last regular season seemed to have bad luck shooting all year, if this type of slump happens to an injured star’s teammates over a relative small number of 40 games it would affect his +/-. Secondly the small sample of it is only half my issue with it as it’s a team stat, that the rest of the team cannot adapt strategically or emotionally without their best player may say more about them than it does the star. If you have two teams with superstar PGs and one team invests in a good backup PG and the other barely has one, and they both get injured and the team without the backup tanks without him, does that say anything about the quality of the two star PGs or does it say something just about the quality of the backups?

[prolific passer]

Celtics had a bad year losing a guy who averaged 15 and 23 a game then ended up drafting a guy who gave them 18-19ppg and 14-15rpg and went back to their winning ways.

[OhayoKD]

Celtics had a bad year losing a guy who averaged 15 and 23 a game then ended up drafting a guy who gave them 18-19ppg and 14-15rpg and went back to their winning ways.

"Back to their winning ways" denotes a year where in a league with sky-high srs they posted a worse mark than the Jordan-less and grant-less 95 Bulls

When a player gets injured or retires there is a multitude of factors like fit, effort/heart, etc. that may make one hold up more than the other, removing Walton from Blazers made jenga tower collapse, removing Kawhi from Raptors didn’t at all, yes that has something to do with Walton being good but I can’t get behind treating it as the end all, especially considering playoffs value is not the same as regular season value, the Raptors did great in regular season before and after Kawhi but needed him in the playoffs. This is before considering the other issues like a player getting injured for 20 games or something which is frequently used here is a small sample and in RealGM’s favorite stat ElGee literally just went through it and manually decided which games not to count cause other teammates were injured at the same time, if you think ElGee is a buddha who has total objectively fair methods of doing this that’s ok, but I see someone that had a long history of being things like “pro-Oscar/Malone” and “anti-Wilt/Dantley” therefore that he can decide what injured samples to use to make his favorite players look better and to confirm his previous takes makes it completely useless. ElGee is a great video editor but I trust his stats work about as much as I trust Malcolm Gladwell to tell accurate history.

Really not sure where to start.

Elgee's stat is not what is being used here. You are trying very hard to undersell what was a nearly identical roster minus a 6th man whose best player jumped with an improved offense, Moreover, 71 and 72 are cited. Years when the roster is clearly better than what Russell had to work with in 69.


You are also taking the stand that larger samples than anything else we can produce for a season are small. All because you've conflated the possiblity of variance with the probablity that variance is flying in the flavor of a player who you assumed wasn't as good as the data suggests.

This is an excellent example of reverse-engineering logic to arrive at a conclusion.

[Dr Positivity]

You are trying very hard to undersell what was a nearly identical roster minus a 6th man whose best player jumped with an improved offense

Howell in 1969 - 20/9, .215 WS/48 (leads team in both this and raw WS), +4.5 TS% (2nd highest on team behind lower volume Nelson)
Howell in 1970: 12/8, 0.072 WS/48, -3.0 TS%

So even though he's still on the team, probably their best offensive player becoming washed is a huge difference in addition to losing Russell and Jones.

Moreover, 71 and 72 are cited. Years when the roster is clearly better than what Russell had to work with in 69.

Personally I do not think 71 Havlicek has a better supporting cast than 69 Russell, considering Cowens is a rookie with average stats. In 72 the Celtics are back to being a contender. But even beside that, I'd think everyone would agree Russell is better than Havlicek and Cowens, so I'm not sure what it really means that his teams did better for his talent or declined without him than them.

[OhayoKD]

You are trying very hard to undersell what was a nearly identical roster minus a 6th man whose best player jumped with an improved offense

Howell in 1969 - 20/9, .215 WS/48 (leads team in both this and raw WS), +4.5 TS% (2nd highest on team behind lower volume Nelson)
Howell in 1970: 12/8, 0.072 WS/48, -3.0 TS%

So even though he's still on the team, probably their best offensive player becoming washed is a huge difference in addition to losing Russell and Jones.

Their best offensive player was Hondo whose volume and efficiency jump coinciding with a 2-point jump in offensive-rating.
And then there is replacing Russell who was double-timing as a coach and player without an assistant with a full-time coach who is at least considered one of the best of the era.

Moreover, 71 and 72 are cited. Years when the roster is clearly better than what Russell had to work with in 69.

Personally I do not think 71 Havlicek has a better supporting cast than 69 Russell

Personally, I do not know why you are trying to separate Havliceck here when he was on both the 69 and 71 team. "71-Hondo" is not particularly relevant.

, considering Cowens is a rookie with average stats. In 72 the Celtics are back to being a contender. But even beside that, I'd think everyone would agree Russell is better than Havlicek and Cowens, so I'm not sure what it really means that his teams did better for his talent or declined without him than them.

You are dumbing down the point. It's not that they did worse, it is the degree to which they did worse. Being a few tenths behind best-in-the-league srs and beating a gauntlet(future champion knicks, the combination of the 2nd and 3rd best teams the prior year) in the playoffs to barely above average is a gigantic drop-off, and that roster is less similar than the 1970 one whose results you're pooh-poohing. And of course in Russell's percieved peak in a league where the best opposing teams were sometimes +2, from a "impact on championships" perspective, Russell's impact being anything like it looks as a retiree is scale-breaking.

95 Chicago stripped of their best player, third best player, and led by a 2nd-best player who didn't want to be there were a better force, especially relative to the league, than the 71 Celtics. If we account for srs-tresholds, 92 to 94 is not comparable if we use lineups featuring the best players.

The cold data does not merely support "Russell>Cowens". In fact pretty much none of his cold-data, be it games without teammates, the wierd small-sample rapm analogues, or these maxed out samples in the immediate or delayed aftermath support your picture of Russell.

Appealing to uncertainity does not justify assuming uncertainty works in Russell's favor. Unless you are assuming variance skews towards him, in any probablistic lense, Russell, both the winningest player, and the one with the best cold-data(favorable one-year interpretations exist for Lebron and Kareem but eh), and the one who played in a league where being worth *4 points of srs would break the corp scale in certain years, is the clear evidentiary pick for "most individually dominant".

Even if you did not care about him having the best resume, always winning, or him apparently leading POY voting as the least hyped and shortest tenured of the 4 major goat candidates.

Regardless, your initial epistemological critique is just not serious. "Sample-size" as a counter to 82-game samples. Never mind 82-game samples over 3-years with roster changes accounted for. Sample-size is an advantage for pure-signal analysis, an advantage it carries over any other type of data along with inclusiveness.

"small samples" as a criticism applies far more strongly to on/off or rapm where the per-season and per-game samples being leveraged are comparatively tiny.

[Dr Positivity]

Their best offensive player was Hondo whose volume and efficiency jump coinciding with a 2-point jump in offensive-rating.

Havlicek has better playmaking than Howell, but he is much less efficient at .459 TS instead of .539 TS (league average is .487). Havlicek has better raw ppg but because of playing 38 minutes than 32, Howell's per minute volume is higher. It's debatable which you rate higher offensively between Howell's more efficient scoring and Havlicek's passing, but either way I'd say losing the all-star caliber version of Howell is significant, as is Jones to an extent.

If Howell and Jones are on the 1970 team and have the same seasons as 1969, is it believable they could win 40-45 games and lose in 1st round? At that point the dropoff from the 69 Celtics is reduced and more analagous to the first Lakers first year without Magic or something.

Personally, I do not know why you are trying to separate Havliceck here when he was on both the 69 and 71 team. "71-Hondo" is not particularly relevant.

The best player on 71 Celtics is Havlicek, not rookie Cowens. So if referencing stronger supporting cast, it makes more sense to use him.

You are dumbing down the point. It's not that they did worse, it is the degree to which they did worse. Being a few tenths behind best-in-the-league srs and beating a gauntlet(future champion knicks, the combination of the 2nd and 3rd best teams the prior year) in the playoffs to barely above average is a gigantic drop-off, and that roster is less similar than the 1970 one whose results you're pooh-poohing. And of course in Russell's percieved peak in a league where the best opposing teams were sometimes +2, from a "impact on championships" perspective, Russell's impact being anything like it looks as a retiree is scale-breaking.

There aren't many situations in history like 69 Celtics where they lost an ATG player and leader, probably their 3rd best player in Howell for all intents and purposes and another relevant player (Jones being 3rd leading scorer), it doesn't seem that crazy to me that they dropped to 34 W level team, not even one of the worst teams of all time or anything. The 99 Bulls clearly lost more in Pippen/Phil/Rodman, however they were a much worse team (13 Ws -9 SRS), so is the Celtics dropoff more significant? I think how the Cavs played both times Lebron left says more to me than the Celtics or Bulls dropoffs because you can basically pin it all on Lebron, now that I would call "scale breaking".

The cold data does not merely support "Russell>Cowens". In fact pretty much none of his cold-data, be it games without teammates, the wierd small-sample rapm analogues, or these maxed out samples in the immediate or delayed aftermath support your picture of Russell.

I don't consider judging an individual by a team sport's results when a star retires (let alone 2 years after) when there's a multitude of other complex variables including simultaneous retirement, coaching change, players aging and emotional fallout of a dynasty ending as "cold data", but that's just me. Also, any +/- stat, even the ones I like more like RAPM, are imperfect because they're virtually all using regular season play and it's been proven before (eg. Robinson vs Hakeem) there is a playoff skillset factor that that doesn't quite capture. This doesn't hurt Russell's case as I think he's on the side of being better in the playoffs than regular season, but nonetheless.

[OhayoKD]

Their best offensive player was Hondo whose volume and efficiency jump coinciding with a 2-point jump in offensive-rating.

Havlicek has better playmaking than Howell, but he is much less efficient at .459 TS instead of .539 TS (league average is .487). Havlicek has better raw ppg but because of playing 38 minutes than 32, Howell's per minute volume is higher. It's debatable which you rate higher offensively between Howell's more efficient scoring and Havlicek's passing, but either way I'd say losing the all-star version of Howell is significant, as is Jones to an extent.

If Howell and Jones are on the 1970 team and have the same seasons as 1969, is it believable they could win 40-45 games and lose in 1st round?

Well the 6 minute difference would be a rather large tell I think, but whatever. Let's say they are truly(ignoring they also looked like a 35-win team in the games without russell that very year) a .500 team that exits in the first round. Taking an average team past

-> league best srs at the point where srs was the highest over the duration of your career

-> the next year's champions

-> last year's 2nd or 3rd best team + the best player from the other 2nd or 3rd best team

is something almost no player has a counter-example for, and it is something which occured with a player at the very end of his career.

At that point the dropoff from the 69 Celtics is reduced and more analagous to the first Lakers first year without Magic or something.

Magic whoby pretty much all the impact signals has a strong case for "most valuable of era" to go with being the literal #1 in winning percentage for the regular season or the playoffs? And of course the Lakers were not falling from a championship, never mind a championship against a gauntlet. If your "realist take" here is "69 Russell is a suped up Magic Johnson"...

Personally, I do not know why you are trying to separate Havliceck here when he was on both the 69 and 71 team. "71-Hondo" is not particularly relevant.

The best player on 71 Celtics is Havlicek, not rookie Cowens. So if referencing stronger supporting cast, it makes more sense to use him.

If Russell was on the 71 team Havelick would no longer be the "best player". This is just semantics. Unless you are using "performance without hondo", it is what they did with hondo that maps to russell's cast(which included a worse version of Hondo)

You are dumbing down the point. It's not that they did worse, it is the degree to which they did worse. Being a few tenths behind best-in-the-league srs and beating a gauntlet(future champion knicks, the combination of the 2nd and 3rd best teams the prior year) in the playoffs to barely above average is a gigantic drop-off, and that roster is less similar than the 1970 one whose results you're pooh-poohing. And of course in Russell's percieved peak in a league where the best opposing teams were sometimes +2, from a "impact on championships" perspective, Russell's impact being anything like it looks as a retiree is scale-breaking.

There aren't many situations in history like 69 Celtics where they lost an ATG player and leader, probably their 3rd best player in Howell and another relevant player (Jones being 3rd leading scorer), it doesn't seem that crazy to me that they dropped to 34 W level team, not even one of the worst teams of all time or anything.

The 99 Bulls clearly lost more in Pippen/Phil/Rodman, however they were a much worse team (13 Ws -9 SRS), so is the Celtics dropoff more significant?

The 95 Bulls clearly lost more in Jordan and Grant. Why are you jumping to 99? The Jordan-less 95 Bulls posted a higher raw srs and significantly closer to the top opponents than the 71 Celtics who at that point had a better version of Hondo, a better coach, and a replacement for Bill Russell.

The 94 Bulls were a more legitimate contender than the 72 Celtics despite experiencing a roster drop-off more similar to the 70 Celtics.

We can also run this with Bird's Celtics, the KD-Curry Warriors, ect and ect, Dynastic teams are not supposed to turn average, below average, or slightly above average in the absence of a fire-sale. Not even when they lose top-10 calibre players. Just like they are not supposed to stay the best team in the league when a great's best teammates miss a sizable sample of games(Russell - hondo), or keep winning when their two best effectively join forces, or when the core of their inital superteam is gutted.

The greatest winner with the greatest resume also looks like a statistical outlier in all the facets he conceivably could with the data available in the time period. You can shout "variance" at the abyss, but all that does is call into question certainty. The evidence still flows in one direction, and there is very little flowing the other.

I think how the Cavs played both times Lebron left says more to me than the Celtics or Bulls dropoffs because you can basically pin it all on Lebron, now that I would call "scale breaking".

Russell being worth half of what he looks in 69/70 during the middle of the 60's would get him years where he, on a random team, produces bigger regular-season outliers. And unlike Lebron's impact which is demonstrably more tied to situation and can be undone by fluke 3-point shooting, Russell's teams were the greatest metronomes.

Lebron is basically the only player you can make this kind of case for(with some steps, Kareem and Wilt perhaps), and you have to ignore surrounding years to do it.

The cold data does not merely support "Russell>Cowens". In fact pretty much none of his cold-data, be it games without teammates, the wierd small-sample rapm analogues, or these maxed out samples in the immediate or delayed aftermath support your picture of Russell.

I don't consider judging an individual by a team sport's results when a star retires (let alone 2 years after) when there's a multitude of other complex variables including simultaneous retirement, coaching change, players aging and emotional fallout of a dynasty ending as "cold data", but that's just me.

The emotional fallout of a dynasty? Really? Where was this emotional fallout in 94 and 95. Why didn't emotional fallout stop the Warriors from winning a title after they replaced durant, iggy, and klay's prime-drop off with wiggins and jordan poole. Why didn't it stop the Celtics from nearly making the conference without Larry Bird.

It is "cold data" because there is no directional bias. There are complex variables everywhere with everything, but in a sport where the goal of a player is to make teams better, directly looking at the largest possible samples regarding "did the team get better" is as "cold" as it gets.

Talking about "emotional fallout" to explain away a team's drop-off after everything was replaced is very much the opposite of cold data. Just as it is when you use box-score composites derived from entirely subjective values attached to entirely subjective decisions made by humans regarding what to count and what not to count.

When you bring up box-score stats as proof of how good players are, even when those players are playing on average offenses which only actually won because of defense, you are introducing a shitton of complex vairables. The difference is your complex variables are biased towards certain types playstyles/forms of production. Our complex variables are biased towards winning.

Which bias do you think is more relevant to assessing a player's ability to improve teams? 

[Dr Positivity]

The 95 Bulls clearly lost more in Jordan and Grant. Why are you jumping to 99? The Jordan-less 95 Bulls posted a higher raw srs and significantly closer to the top opponents than the 71 Celtics who at that point had a better version of Hondo, a better coach, and a replacement for Bill Russell.

The Bulls in 95 were 34-31 before Jordan came back compared to the 71 Celtics being 44-38, and in general, I'm just not there for comparing how the Bulls did 2 years after Jordan retirement to the Celtics 2 years after Russell retirement as meaningful information about Jordan vs Russell, sorry, judging individuals by team results is flawed in a normal scenario let alone on completely different teams 1.5-2 years later.

The emotional fallout of a dynasty? Really? Where was this emotional fallout in 94 and 95. Why didn't emotional fallout stop the Warriors from winning a title after they replaced durant, iggy, and klay's prime-drop off with wiggins and jordan poole. Why didn't it stop the Celtics from nearly making the conference without Larry Bird.

The Warriors last year "didn't have it" all year seemingly cause Draymond punched Poole in the face and it seemed to mess up their team mentally, they played worse in 18 and 19 regular seasons than 17 and 15 and 16 when they had a worse team cause Durant wasn't getting along with them as much anymore + they didn't care about regular season as much. The Bulls in 93 played worse in regular season than in 91 and 92 probably cause they had less to prove in regular season and were getting burned out, which affects the comparison with 94 and 95 Bulls. These are all examples of "emotional factors" in a game with humans. I don't know if the Celtics decline in 1970 had anything to do with losing Russell and Jones leadership in addition to their play, but it could be a factor among other ones I listed like player and coach changes.

It is "cold data" because there is no directional bias. There are complex variables everywhere with everything, but in a sport where the goal of a player is to make teams better, directly looking at the largest possible samples regarding "did the team get better" is as "cold" as it gets.

Talking about "emotional fallout" to explain away a team's drop-off after everything was replaced is very much the opposite of cold data. Just as it is when you use box-score composites derived from entirely subjective values attached to entirely subjective decisions made by humans regarding what to count and what not to count.

When you bring up box-score stats as proof of how good players are, even when those players are playing on average offenses which only actually won because of defense, you are introducing a shitton of complex vairables. The difference is your complex variables are biased towards certain types playstyles/forms of production. Our complex variables are biased towards winning.

Which bias do you think is more relevant to assessing a player's ability to improve teams? 

When it comes to box score, "skillset analysis", accolades, or +/-, in my opinion going 100% all in on any will lead to flaws, especially for 60s players when the data is worse such as having to use shoddy injured game +/- instead of modern day RAPM. If +/- was the end all, it would mean the Timberwolves traded for an MVP caliber player in Gobert. You can do this with any individual metric. Ultimately, it's not to say +/- is useless, but I'm not so in love with the stat that I'm not going to take how his team played in the played in the 40-50 games without Kareem vs Walton's 40-50 as more meaningful than all of the other information combined. When it comes to Russell I do think it's meaningful that the Celtics got worse without him, but it's one of several meaningful factors, such as how I value that he has more regular season MVPs than regular season god Wilt, but I'm not going to stake my entire Russell opinion on it.

[70sFan]

I think that focusing solely on 1969 can hurt Russell's overall case, because he did so much more in his career that you don't need to use only that one sample and leave the possibility of someone calling out a potential fluke. It's not the case with Russell, he has the most straightforward case for the GOAT among all candidates.

[OhayoKD]

When it comes to box score, "skillset analysis", accolades, or +/-, in my opinion going 100% all in on any will lead to flaws, especially for 60s players when the data is worse such as having to use shoddy injured game +/- instead of modern day RAPM. If +/- was the end all

Yeah, but good stat-use is not a democracy.

the "box-score" and "skillset analysis" are completely subjective and who they "favor" is entirely dependent on what you count which of course, can only be tied to winning if you use...winning. If I wanted to make a box-score that only counted defensive components, Russell would likely have the best defensive box-score. If I wanted to do skillset analysis that way, he would have the best skillset too. And there's no real reason to think that would be less accurate for the 60(or even the 70's really) than ws/48 or per or the like.

The same thing is also true for accolades to a degree though Russell covers so many bases that a majority of frames will come up with him as a distant top.

If we can agree that players are ultimately "good" based on how much they can improve a team, then the question of "who is likelier to improve a team by more" is an objective/falsifiable(if hard to answer) one which you can only answer with winning baked in.

On a more technical note, WOWY is not +/-. +/- works off much smaller samples per-game and per-season. WOWY has the largest possible ones. +/- is also not as inclusive as whatever effect a player can have on the bench or in terms of intangibles is also less accounted for, never mind it being prone to colinearity and such in a way WOWY largely isn't.

+/- sees all the top players near each other because it's based on what teams do in a few minutes without a guy. Force those teams to survive without a guy longer and suddenly we see big deltas; some players are easier to adapt for when you overcome the uh "emotional shock". Some players are harder to. And some players are so hard to adapt for you can literally replace them when they're retiring, see their teammates improve, see a hof-coach replace a player-coach and...

massive drop-off. Which gets us to...

These are all examples of "emotional factors" in a game with humans. I don't know if the Celtics decline in 1970 had anything to do with losing Russell and Jones leadership in addition to their play, but it could be a factor among other ones I listed like player and coach changes but it could be a factor among other ones I listed like player and coach changes.

They are all examples of emotional factors that could have had a bigger effect for teams that dropped off less than russell's celtics did. But you are not presenting that possiblity. And it's especially wierd doing this with the Bulls:

The Bulls in 93 played worse in regular season than in 91 and 92 probably cause they had less to prove in regular season and were getting burned out, which affects the comparison with 94 and 95 Bulls

The 94 and 95 Bulls had a best player who was far more discontent than he was in 93 with a very real hatred of one of the teammates they added. I do not recall Hondo filing a trade request after Russell's depature. I also do not recall hondo lambasting his coach as a racist(which by the way phil jackson absolutely was), or begging russell to come back.

Have you considered the possiblity that the "emotional factor" could have hindered Chicago more?

The actual reason the "the bulls were worn down in 93 and revitalized in 94" is trotted out is that a flat reading of what happened supports the(uncomfortable for some) idea that scoring titles and per and steals per game aren't worth nearly as much as people initially assumed.

It is not something built on what players actually did(filing trade requests, sitting out possessions, ect), or even a consistent application of the justifying theory(how are four title pushes in a row less emotionally exhausting than 3?)

The larger point here though is that by noting the possiblity of a hindrance for one set of teammates, and not the other, you are painting variance as factor that could only possibly have boosted Russell's impact signal when it also could have hindered it.

Evidence is not about proving beyond all shadow of a doubt. Even scientific truths have a kernel of uncertainty. It is about marking one thing as more likely than another. Then we look for replication to determine uncertainty(with the line of what is considered a fact being drawn based on precedent, not objective percentages). Variance can go both ways. Emotional factors can go both ways. The signal could underrate Russell relative to Steph and Jordan. Or it can overrate him. But because his signals are stronger, him being overrated(and overrated by enough) is a necessary for him to be worth less. It is not necessary that he is underrated.

But that's just probablistic thinking 101. What is worth considering is what we consider a fact in a field.

In basketball Lebron>Dereck Fisher is a fact. Jordan>Pippen is a fact. Jordan>Kobe is a fact. There is uncertainty here, and we don't really know the percentage of likelihood, but there are just too many correlates and too much replication for anyone to care.

Let's apply this to "is 70 the signal or the noise". What do the other data-points around 70 say?

69 has the celtics very bad without Russell over 8 games.
71 has the celtics as less above average than what realistically should have been a fire-sale in 95
72 has the celtics still as very weak 2018-raptors calibre contender

The celtics without russell's teammates during his tenure, paticularly in the 60's, has the team doing just fine regardless of who leaves and doesn't leave. Over a small 2.2 game/season sample, it has the team collapse from greatest outlier ever to bad

They are very good without him in his and before his rookie-year with a team whose roster would be depleted by 62, and before that we see russell win b2b ncaa titles with a team that did not ever make the ncaa tournament before his arrival.

The worst data-point from russell sees the most different roster and is the second most removed temporally from the year in question. Everything else pretty much corraborates the same narrative. The data closest in temporal proximity is the best for russell. The data that is tiny but covers his career says the same.

The only data where his team is even above average without him comes in circumstances nba history would suggest dynasties should be at least contention-worthy without.

Why are you taking 1970 to be the noise as opposed to the signal? 1970 is what has the corroboration. And it implies something wild. Because in certain years of russell's tenure, being worth 4-points of srs makes you the clear best team in the league. In other words, if Russell, who as a player-coach saw an 8-point full-strength drop off in 69 and 70, is merely worth half that on a random team, you are significantly more likely to win a title with him than an 8-point player is likely to win during the 90's or early 70's or 10's.
 
In other words, at face-value, the data suggests russell is scale-breaking. And that data has far more corroborating it than it has challenging it. Which leaves us with,

When it comes to Russell I do think it's meaningful that the Celtics got worse without him, but it's one of several meaningful factors, such as how I value that he has more regular season MVPs than regular season god Wilt, but I'm not going to stake my entire Russell opinion on it.

What exactly do you consider a "meaningful" factor which you can "stake your opinion" on which doesn't support Russell as a scale-breaker?

WS/48 is no more meaningful than a defensive only box-score with defensive only weightings describing how a team with an average offense and the best ever defense won 11 times. You could probably get more accurate title predictions those years just counting whose center deterred or deflecting more shots than you would get with per or ws/48 or whatever meaningful factor you are considering that suggests russell is worse than lebron or jordan or whoever.

That might not matter if you are playing the time-machine game, but the original contention was that Russell was lacking in individual dominance. It's all well and good saying "well it's not 100 percent because of variance", but if we are going to positively argue Russell was not the most dominant(or actually less individually dominant than Lebron, Kareem, and Jordan), then you I'd expect evidence that actually supports that, not simply a criticism of a stat as "flawed".