Russell's defensive impact

[penbeast0]

DFG% comes from single game boxscore numbers, which has been collected for over a decade.

Fine.

The data you presented has data for all 3 teams from the 1965-66 to 1967-68 seasons, for some 60-80 games per season, a large sample size of data. And it shows an average defensive FG% allowed of .415 for Boston, .425 for Philadelphia, and .430 for San Francisco.

I'm sorry but over some 190-240 games the difference between these is very small. And those 3 years Wilt played 47 min/g (242 g), Bill 41 min/g (237 g), and Nate 42 min/g (189 g).

Even in 1963-64 with Boston at .409 and San Francisco at .414 that fact is there is simply very little difference. And that season Russell played 45 min/g (78 g) and Chamberlain 46 min/g (80 g).

Compare this to data from the first two years we actually have NBA team defensive data for - 1970-71 and 1971-72. Those two seasons the difference in team FG% allowed on defense ranged from .424 to .469 (4.7% diff) and .420 to .486 (6.6% diff).

So - again - any insistence on the Celtics being this overly great defensive team in the late 50s and early 60s because a website shows a team DRtg calculated from an erroneous calculation of game pace is false.

Celtics definitely played fast paced basketball, they were known for it and you can see that by all the footage we have.

You can claim this all you want, but their game pace was nowhere near as high - and their team defense nowhere near as great - as that listed on that website.

So any claim that Russell was a significantly better defender than the likes of Wilt and Nate because this website shows team defense that was unworldly for the time is unfounded, based off of their incorrect assumptions for team game pace.

Do you have any evidence that BB-Ref is wrong other than looking at other eras and saying other teams didn't have the degree of variance that the Celtics are claimed to have?

[coastalmarker99]

Defensive Win Shares (DWS) are based on DRtg;

OWS based on ORtg

The Celtics' DWS-OWS ratio was incredibly lopsided. For example, 55.7-6.8 in 1964

It's unbelievable that the 1964 Celtics with peak Sam Jones plus Heinsohn, Havlicek, Russell (good passer) totalled less than 7 OWS!

Something is wrong with that data.
Along with more ORB leading to fewer possessions than estimated, I suspect Celtics' offensive TOV rate was lower than typical.

A lower TOV rate would also make their offensive possessions fewer than expected.


Both ORB & TOB lead in same direction: fewer poss than estimated.


Usually, TOV and FT Rate (FTA per FGA) are positively correlated.

This empirical finding makes theoretical sense: Players who drive often, get fouled often but also commit turnovers more often.

In fact, referee's blown calls often turn TOV into "fouls" and fouls into "TOV."

Knowing that TOV correlated to FT Rate (FTA/FGA), let's look at Celtics.

Celtics had LOW FT rate in Russell era. In fact, 4x they were 1st in FGA & LAST in FTA!

FGA, FTA
1, 3 (1956: before Russ)

1, 5 (1957)
1, 7
1, 6
1, 8
1, 7
1, 8
1, 9
1, 9
1, 8
3, 8
6, 5
7, 9
3, 11

his very-low FT rate leads to conclusion that Celtics had (very?) low offensive TOV rate, too.

Low TOV also connects with low FG%: Celtics shot quickly before getting fouled or turning over the ball. And Russell often rebounded those offensive misses, keeping possession going.

Cousy was the Celtics' main ball-handler through 1963. From video and reputation, Mr. Highlight Film might be a high TOV.

But perhaps Cousy's TOV *rate* lowered as Celtics' pace skyrocketed: less half-court O, more fast-break shooting.

Poss/G:
108.6 in 1956
136.3 in 1960

[coastalmarker99]

While I do think that Russell's defensive impact is slightly overrated.

His offensive impact on the other hand is underrated.

As he led the playoffs in total assists twice.

And he led his team in scoring a couple of times in the finals.

[coastalmarker99]

Watch on YouTube

;t=584s

The footage we have of those Celtic teams just shows how good their defensive schemes were.

Russell didn't even have to do that much with all the help he had on that end of the floor.

As in the first 9 minutes of this video Wilt didn’t miss a shot, and was the only one on the warriors to score.

While His teammates were just getting locked up and could barely get the ball past halfcourt.


As a result of this, I would hesitate to give Russell all of the credit for Wilt's teammates' offensive woes.

Furthermore


In the 1966 ECFs, Wilt's teammates shot the following: Greer was at .32.5. Walker was at .37.5. Jackson was at .42.9. and Jones was at .32.5. plus Cunningham was at .16.1.


It wasn't as if Russell was guarding everybody and locking them down.

I mean the 76ers players were completely missing even when wide open.

Watch on YouTube

Take for example at the 18.33-minute mark of this video Wally jones horribly bricking a wide-open shot before Wilt grabs the offensive rebounds and slams it down.

[coastalmarker99]

It would have been very interesting to see how Russell's defence would have held up.

Had he to endure the workload of 77 Kareem or 64 Wilt or 94 Hakeem.

As Russell himself commented that Wilt could do a better job in his [Russell's] role than Russell could do in Wilt's.


I do believe that if you put Kareem and Wilt on those Celtic teams and told them to just do Russell's job.

That they would achieve the same success as he did on defence.


As Kareem led ATG defences on the Bucks from 1971 to 1974 while having to carry a far bigger load than Russell and dealing with a far worse coach and weaker defensive teammates.


Also In 75-76 Kareem that year led the league in rebounding, blocks, and his defensive impact was off the charts.

He had a DRtg of 90 compared to team DRtg of 98.8 (13th of 18 teams).

With him on the floor LA had the best defensive team in the league (DRtg of 90 would be #1 in the league...) and without him on the floor they were probably close to last.

[penbeast0]

I think that comment was politeness, or fluff if you will. I think Wilt could do Russell's job as a shotblocker and rebounder, and he was even stronger which could translate into better man defense. I don't see him, in even his best years, contesting as many shots away from the basket especially on players he wasn't guarding, nor do I see him creating as many turnovers. Wilt was also less instinctive, more intellectual, which often translates into slower decision making. More anecdotal than evidence based as I doubt I have seen 100 games of the two combined and most of those were from 65 on. Kareem would be similar with less of a strength advantage (and lower defensive rebounding rates). However, from everything I understand about the play styles, that's my take on Russell's primary advantage.

A similar exercise could be looking at Hakeem on defense v. Shaq. Shaq is great at holding post position, a strong shotblocker, and arguably the GOAT at intimidating others driving the lane (his percentage of And-1's allowed is further away from the 2nd best of his era than the 2nd best is from the 20th). Along with Wilt, the greatest physical freak to ever play the game. He was not Hakeem's equal at the quickness aspects of team defense including stepping out on pnr, recovering to his man after committing to stopping another player, etc.

[kcktiny]

Do you have any evidence that BB-Ref is wrong

Have you not read through this thread? What are you missing?

There is NO team defensive data prior to 1969-70. So then how could their DRtg and calculation of team possessions be right? It is an estimate. To properly calculate these you need BOTH offensive and defensive team data.

In lieu of no team defensive data they have obviously guesstimated the defensive team data in order to calculate both. How else are you going to get numbers for DRtg and game pace?

Even for ORtg you need to estimate turnovers.

For example in 1958-59 they list the Boston game pace as 128.7. The next fastest team is 121.0 (Syracuse). But the slowest team is Cincinnati at 116.4. That means that the difference in game pace between the two fastest teams (7.7) is greater than the second fastest team and the slowest team (5.6), i.e. the difference in game pace between Boston and Syracuse is significantly more than the difference in game pace between the other 7 teams.

Are you following this?

To estimate that high of a game pace for Boston they have obviously estimated that their opponents attempted a TON of FGAs, as the lower Boston's FG% allowed on defense is means more opponent FGAs and a higher game pace.

And that is why they have listed a very low DRtg for Boston.

In 1963-64 someone else has shown defensive FG% allowed data for Boston and San Francisco where they allowed similar FG%s of .409 and .414. Yet the website in question shows Boston with a DRtg of 83.8 and San Francisco at 88.6. And with a game pace of 125.0 to 115.1 for San Francisco. So not knowing this FG% allowed data they obviously assumed a much lower FG% allowed by Boston compared to San Francisco, which means far more missed FGAs by their opposition, and far more team possessions, and thus a much higher game pace.

[capfan33]

DFG% comes from single game boxscore numbers, which has been collected for over a decade.

Fine.

The data you presented has data for all 3 teams from the 1965-66 to 1967-68 seasons, for some 60-80 games per season, a large sample size of data. And it shows an average defensive FG% allowed of .415 for Boston, .425 for Philadelphia, and .430 for San Francisco.

I'm sorry but over some 190-240 games the difference between these is very small. And those 3 years Wilt played 47 min/g (242 g), Bill 41 min/g (237 g), and Nate 42 min/g (189 g).

Even in 1963-64 with Boston at .409 and San Francisco at .414 that fact is there is simply very little difference. And that season Russell played 45 min/g (78 g) and Chamberlain 46 min/g (80 g).

Compare this to data from the first two years we actually have NBA team defensive data for - 1970-71 and 1971-72. Those two seasons the difference in team FG% allowed on defense ranged from .424 to .469 (4.7% diff) and .420 to .486 (6.6% diff).

So - again - any insistence on the Celtics being this overly great defensive team in the late 50s and early 60s because a website shows a team DRtg calculated from an erroneous calculation of game pace is false.

Celtics definitely played fast paced basketball, they were known for it and you can see that by all the footage we have.

You can claim this all you want, but their game pace was nowhere near as high - and their team defense nowhere near as great - as that listed on that website.

So any claim that Russell was a significantly better defender than the likes of Wilt and Nate because this website shows team defense that was unworldly for the time is unfounded, based off of their incorrect assumptions for team game pace.

Do you have any evidence that BB-Ref is wrong other than looking at other eras and saying other teams didn't have the degree of variance that the Celtics are claimed to have?

As I stated earlier, BBreference isn't updated with Ben's latest formula so it is wrong technically. But the Celtics still have the greatest defensive dynasty ever either way.

[kcktiny]

Who is this Ben and how are his guesstimates for team defensive data any better than the website's?

[Dooley]

Who is this Ben

Ben Taylor (I assume)

and how are his guesstimates for team defensive data any better than the website's?

If it comes to that, why are your guesstimates for team defensive data better than the website's guesses?

[capfan33]

Who is this Ben and how are his guesstimates for team defensive data any better than the website's?

The website's numbers are his numbers. There the original numbers he came up with years ago, but recently he came up with a better way to estimate pace which isn't reflected yet on BBall reference. And these numbers do downrade the Celtics defenses a fair amount, still the GOAT defensive dynasty but not nearly as big an outlier. 64 goes from a -10.7 to -8.7 which is a pretty sizable difference.

[kcktiny]

why are your guesstimates for team defensive data better than the website's guesses?

Where have I made guesstimates for team data?

I have simply pointed out how huge of outliers the website's DRtg and team possession estimates were for Boston in the late 50s and early to mid-60s, ratings that some in this thread have used as evidence for Russell being the greatest defender ever, saying other great defenders of that time were not "on his level".

And these numbers do downrade the Celtics defenses a fair amount, still the GOAT defensive dynasty but not nearly as big an outlier. 64 goes from a -10.7 to -8.7 which is a pretty sizable difference.

So the question here is are those who flaunted Russell's defensive ability as being far better than anyone else going to change their minds. I can't say, but I bet even in the face of the evidence there are those that will continue to repeat those same opinions regardless of any new data.

[Dooley]

why are your guesstimates for team defensive data better than the website's guesses?

Where have I made guesstimates for team data?

I have simply pointed out how huge of outliers the website's DRtg and team possession estimates were for Boston in the late 50s and early to mid-60s, ratings that some in this thread have used as evidence for Russell being the greatest defender ever, saying other great defenders of that time were not "on his level".

You seem to be taking the positive constructive position that the correct DRTG for those Celtics teams would be significantly higher

You're not saying that the team DRTG is unknown and that the BBR guesstimates are unreliable and we don't know the truth at all, you're making a definite argument about the correct DRTGs

Well, what is this based on? Just the fact that the Celtics numbers given are an outlier?

[kcktiny]

You seem to be taking the positive constructive position that the correct DRTG for those Celtics teams would be significantly higher

Correct.

You're not saying that the team DRTG is unknown

Yes I am. There is no actual team defensive data for that time period.

and that the BBR guesstimates are unreliable

I would not use them to make definitive statements.

and we don't know the truth at all

Define truth here. The exact DRtgs? Correct, we do not nor will we likely ever know them. A better truth? Yes.

you're making a definite argument about the correct DRTGs

No. Once again - I am simply stating the listed DRtgs are too low and the team possession estimates are too high.

Well, what is this based on?

Read this thread. I have stated it multiple times.

Just the fact that the Celtics numbers given are an outlier?

Correct. Outliers over multiple seasons.

Look - apparently I am not the only one who has come to this conclusion.

recently he came up with a better way to estimate pace which isn't reflected yet on BBall reference. And these numbers do downrade the Celtics defenses a fair amount, still the GOAT defensive dynasty but not nearly as big an outlier.

This Ben fellow appears to have reached the same deduction, according to someone else in this thread.

[ZeppelinPage]

Thread on why Basketball Reference underestimates the Celtics offense:

Read on Twitter

Ben Taylor's own data supports this theory as he likely has differing pace estimations that account for the specific context of the Celtics touched on in the thread above.

According to Taylor's data, the Cousy-era Celtics are more often than not a plus offensive team (with 4 of 7 seasons ranked in the top 4):

1957 Celtics rORTG: +2.2 (2nd of 8)
1958 Celtics rORTG: +1.4 (3rd of 8)
1959 Celtics rORTG: +1.7 (4th of 8)
1960 Celtics rORTG: +2.5 (2nd of 8)
1961 Celtics rORTG: -1.4 (7th of 8)
1962 Celtics rORTG: +0.9 (5th of 9)
1963 Celtics rORTG: -0.6 (5th of 9)

Following Cousy's departure, the offensive outlook is split, but nonetheless more favorable compared to Basketball Reference:

1964 Celtics rORTG: -2.4 (8th of 9)
1965 Celtics rORTG: -0.4 (6th of 9)
1966 Celtics rORTG: -0.5 (6th of 9)
1967 Celtics rORTG: +2.3 (2nd of 10)
1968 Celtics rORTG: +0.5 (5th of 12)
1969 Celtics rORTG: +0.1 (8th of 14)

This data paints the Celtics as (for the most part) a more balanced team. A fraction of the credit the defense receives is simply given to the offense. They still had a fantastic defense for their era and on an all time scale--this defense was the backbone of their team and was key to their championships. But with a more accurate possession estimate, it's likely that the Celtics offense contributed to winning more than previously thought.

[70sFan]

Fine.

The data you presented has data for all 3 teams from the 1965-66 to 1967-68 seasons, for some 60-80 games per season, a large sample size of data. And it shows an average defensive FG% allowed of .415 for Boston, .425 for Philadelphia, and .430 for San Francisco.

I'm sorry but over some 190-240 games the difference between these is very small. And those 3 years Wilt played 47 min/g (242 g), Bill 41 min/g (237 g), and Nate 42 min/g (189 g).

Sorry, but the difference between Celtics and Sixers defense wasn't that massive during that period. I picked older Russell years, because the boxscore numbers are more complete for them but Russell was past his prime at that point, while Wilt peaked:

1966-68 Celtics: 41.5 DFG%, -5.4 rDRtg (basketball-reference), -3.9 rDRtg (backpicks)
1966-68 Sixers: 42.5 DFG%, -3.7 rDRtg (basketball-reference), -3.5 rDRtg (backpicks)

So yeah, the difference isn't drastic (although 1% is noticeable), but the difference in defensive ratings isn't huge either - we're comparing two absolutely elite teams and it wasn't Russell Celtics peak by any means.

Even in 1963-64 with Boston at .409 and San Francisco at .414 that fact is there is simply very little difference. And that season Russell played 45 min/g (78 g) and Chamberlain 46 min/g (80 g).

Again, that's because SFW had all-time defensive season as well:

1964 Celtics: 40.9 DFG%, -10.8 rDRtg (basketball-reference), -8.7 rDRtg (backpicks)
1964 Warriors: 41.4 DFG%, -6.0 rDRtg (basketball-reference), -5.5 rDRtg (backpicks)

The data doesn't show gigantic gap, because you're comparing him to the best defensive teams of the 1960s. If we go with 1962 comparison for example, the difference is significant:

1962 Celtics: 40.5 DFG% (54 games), -8.5 rDRtg (basketball-reference), -6.1 rDRtg (backpicks)
1962 Warriors: 42.2 DFG% (78 games), -1.2 rDRtg (basketball-reference), +0.0 rDRtg (backpicks)

Compare this to data from the first two years we actually have NBA team defensive data for - 1970-71 and 1971-72. Those two seasons the difference in team FG% allowed on defense ranged from .424 to .469 (4.7% diff) and .420 to .486 (6.6% diff).

What are you doing here? You're comparing the difference of three of the best defensive teams in the league to league wide range? That's ridiculous. If you want to do that, compare Celtics to bad defensive teams:

1962 Celtics: 40.5 DFG% (54 games)
1962 Lakers: 42.1 DFG% (55 games)
1962 Warriors: 42.2 DFG% (78 games)
1962 Nationals: 42.4 DFG% (38 games)
1962 Royals: 42.6 DFG% (80 games)
1962 Hawks: 43.6 DFG% (71 games)
1962 Pistons: 43.6 DFG% (36 games)
1962 Packers: 45.3 DFG% (36 games)
1962 Knicks: 47.2 DFG% (37 games)

The data is very incomplete, but if we use it then difference in DFG% ranged from 40.5% to 47.2% (6.7% diff.). It's actually the bigger difference than the two seasons you mentioned. Celtics have 1.6% advantage over the second place and it's not a small difference at all.

So - again - any insistence on the Celtics being this overly great defensive team in the late 50s and early 60s because a website shows a team DRtg calculated from an erroneous calculation of game pace is false.

We have more conservative estimations that still put Celtics as by far the best defensive team of their era. If you find a reliable model that would question it, then I can change my perspective, but for now all you're asking for is to stop doing any analysis, because our data isn't perfect.

You can claim this all you want, but their game pace was nowhere near as high - and their team defense nowhere near as great - as that listed on that website.

How can you know that? You don't have more data than us, you can assume it's wrong but you don't know it. You just think it's impossible for Celtics to be fast paced defensive team, but even if you listen Russell and Red that was their gameplan. It's very likely that Celtics were consistently among the fastest teams in the league.

So any claim that Russell was a significantly better defender than the likes of Wilt and Nate because this website shows team defense that was unworldly for the time is unfounded, based off of their incorrect assumptions for team game pace.

[/quote]
I already pointed out that these claims are not strictly based on pace estimations, but more on tracking data from the film we have.

[70sFan]

Thread on why Basketball Reference underestimates the Celtics offense:

Read on Twitter

Ben Taylor's own data supports this theory as he likely has differing pace estimations that account for the specific context of the Celtics touched on in the thread above.

According to Taylor's data, the Cousy-era Celtics are more often than not a plus offensive team (with 4 of 7 seasons ranked in the top 4):

1957 Celtics rORTG: +2.2 (2nd of 8)
1958 Celtics rORTG: +1.4 (3rd of 8)
1959 Celtics rORTG: +1.7 (4th of 8)
1960 Celtics rORTG: +2.5 (2nd of 8)
1961 Celtics rORTG: -1.4 (7th of 8)
1962 Celtics rORTG: +0.9 (5th of 9)
1963 Celtics rORTG: -0.6 (5th of 9)

Following Cousy's departure, the offensive outlook is split, but nonetheless more favorable compared to Basketball Reference:

1964 Celtics rORTG: -2.4 (8th of 9)
1965 Celtics rORTG: -0.4 (6th of 9)
1966 Celtics rORTG: -0.5 (6th of 9)
1967 Celtics rORTG: +2.3 (2nd of 10)
1968 Celtics rORTG: +0.5 (5th of 12)
1969 Celtics rORTG: +0.1 (8th of 14)

This data paints the Celtics as (for the most part) a more balanced team. A fraction of the credit the defense receives is simply given to the offense. They still had a fantastic defense for their era and on an all time scale--this defense was the backbone of their team and was key to their championships. But with a more accurate possession estimate, it's likely that the Celtics offense contributed to winning more than previously thought.

This data paints Celtics as more balanced team at the beginning of Russell's career (which is logical, they had more offensive talent then), but it was clear that their defense was the main reason of their success:

1957 Celtics: -2.5 rDRtg
1958 Celtics: -3.1 rDRtg
1959 Celtics: -3.5 rDRtg
1960 Celtics: -3.7 rDRtg
1961 Celtics: -5.5 rDRtg
1962 Celtics: -6.1 rDRtg
1963 Celtics: -6.3 rDRtg

1964 Celtics: -8.7 rDRtg
1965 Celtics: -7.2 rDRtg
1966 Celtics: -4.5 rDRtg
1967 Celtics: -4.3 rDRtg
1968 Celtics: -2.8 rDRtg
1969 Celtics: -4.6 rDRtg

During the heart of Russell career (1960-66), Celtics were slightly below average on offense (-0.3) and absurdly dominant on defense (-6.0). It does show that Celtics weren't nearly as one-dimensional as some may think based on basketball-reference estimates, but it's still very defensive oriented team. I mean, that's comparable offensive results to 1992-95 Knicks (-0.9 and -6.2) and nobody would call them a balanced team.

Also, this is how the Celtics did before and after Russell:

1955 Celtics: +4.2 rORtg, +4.3 rDRtg
1956 Celtics: +3.5 rORtg, +3.0 rDRtg

1970 Celtics: -1.2 rORtg, +0.3 rDRtg
1971 Celtics: +1.1 rORtg, +0.6 rDRtg

Celtics before Russell were all-time bad defensive team. After the retirement, they went from elite to slightly below average. I don't see any other reason than getting Russell for that change.

[70sFan]

Also In 75-76 Kareem that year led the league in rebounding, blocks, and his defensive impact was off the charts.

He had a DRtg of 90 compared to team DRtg of 98.8 (13th of 18 teams).

With him on the floor LA had the best defensive team in the league (DRtg of 90 would be #1 in the league...) and without him on the floor they were probably close to last.

These are not the actual on/off numbers, these are boxscore estimates which are completely useless.

[70sFan]

More about differentials:

1968/69

1969 Celtics (72 games): 40.8 DFG%
1969 Lakers (82 games): 45.0 DFG%
1969 Suns (82 games): 47.0 DFG%

6.2 percentage points differential (4.2 vs Wilt team)

1967/68

1968 Celtics (73 games): 42.8 DFG%
1968 Sixers (82 games): 43.0 DFG%
168 Sonics (82 games): 46.7 DFG%

3.9 percentage points differential

1965/66

1966 Celtics (69 games): 40.4 DFG%
1966 Sixers (80 games): 42.2 DFG%
1966 Knicks (86 games): 45.6 DFG%

5.2 percentage points differential (1.8 vs Wilt team)

I don't know, to me it looks like the differential is quite steady across years and it compares well to the older seasons from the 1970s. Are you still convinced that everything is off with these numbers?

[capfan33]

Thread on why Basketball Reference underestimates the Celtics offense:

Read on Twitter

Ben Taylor's own data supports this theory as he likely has differing pace estimations that account for the specific context of the Celtics touched on in the thread above.

According to Taylor's data, the Cousy-era Celtics are more often than not a plus offensive team (with 4 of 7 seasons ranked in the top 4):

1957 Celtics rORTG: +2.2 (2nd of 8)
1958 Celtics rORTG: +1.4 (3rd of 8)
1959 Celtics rORTG: +1.7 (4th of 8)
1960 Celtics rORTG: +2.5 (2nd of 8)
1961 Celtics rORTG: -1.4 (7th of 8)
1962 Celtics rORTG: +0.9 (5th of 9)
1963 Celtics rORTG: -0.6 (5th of 9)

Following Cousy's departure, the offensive outlook is split, but nonetheless more favorable compared to Basketball Reference:

1964 Celtics rORTG: -2.4 (8th of 9)
1965 Celtics rORTG: -0.4 (6th of 9)
1966 Celtics rORTG: -0.5 (6th of 9)
1967 Celtics rORTG: +2.3 (2nd of 10)
1968 Celtics rORTG: +0.5 (5th of 12)
1969 Celtics rORTG: +0.1 (8th of 14)

This data paints the Celtics as (for the most part) a more balanced team. A fraction of the credit the defense receives is simply given to the offense. They still had a fantastic defense for their era and on an all time scale--this defense was the backbone of their team and was key to their championships. But with a more accurate possession estimate, it's likely that the Celtics offense contributed to winning more than previously thought.

This data paints Celtics as more balanced team at the beginning of Russell's career (which is logical, they had more offensive talent then), but it was clear that their defense was the main reason of their success:

1957 Celtics: -2.5 rDRtg
1958 Celtics: -3.1 rDRtg
1959 Celtics: -3.5 rDRtg
1960 Celtics: -3.7 rDRtg
1961 Celtics: -5.5 rDRtg
1962 Celtics: -6.1 rDRtg
1963 Celtics: -6.3 rDRtg

1964 Celtics: -8.7 rDRtg
1965 Celtics: -7.2 rDRtg
1966 Celtics: -4.5 rDRtg
1967 Celtics: -4.3 rDRtg
1968 Celtics: -2.8 rDRtg
1969 Celtics: -4.6 rDRtg

During the heart of Russell career (1960-66), Celtics were slightly below average on offense (-0.3) and absurdly dominant on defense (-6.0). It does show that Celtics weren't nearly as one-dimensional as some may think based on basketball-reference estimates, but it's still very defensive oriented team. I mean, that's comparable offensive results to 1992-95 Knicks (-0.9 and -6.2) and nobody would call them a balanced team.

Also, this is how the Celtics did before and after Russell:

1955 Celtics: +4.2 rORtg, +4.3 rDRtg
1956 Celtics: +3.5 rORtg, +3.0 rDRtg

1970 Celtics: -1.2 rORtg, +0.3 rDRtg
1971 Celtics: +1.1 rORtg, +0.6 rDRtg

Celtics before Russell were all-time bad defensive team. After the retirement, they went from elite to slightly below average. I don't see any other reason than getting Russell for that change.

Yea as I said, the Celtics were clearly still a defensively oriented team but not to the extent that we previously thought. And I still think Russell is a clearly better defender than Wilt/Thurmond overall.