Greatest Peaks series (Thinking Basketball/Ben Taylor)

Moderators: Doctor MJ, trex_8063, penbeast0, PaulieWal, Clyde Frazier

70sFan
RealGM
Posts: 29,599
And1: 24,915
Joined: Aug 11, 2015
 

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#81 » by 70sFan » Wed Nov 25, 2020 10:36 am

freethedevil wrote:1965, +7 srs, taken to 7 y the, -.25 srs, hawks.

Celtics didn't face Hawks in 1965 playoffs.

[b]No, lets recap, I said "it was not harder for russell induvdiaully to lift his team" then it would be for an average 90's superstar. That the --Volume-- of 60 win teams increased proportionally to the # of teams in the league supports my claim, not yours.

But lifting your team can't be measured by added wins. It's very crude approximation that needs a lot of additional work behind it. Also - the number of 60+wins teams clearly shows that there is correlation between the number of teams and the outlier RS results. If you take away Russell's Celtics teams, then we had total of two seasons in the 1960s with 60+wins pace. To make this number identical to 1990s in linear way, you'd have to expand the league to 40+ teams to reach 1990s level, even without counting Jordan teams.
Cleary better than you apparently.

Then you certainly didn't show that here.

That's longetvity, not relative dominance. The bulls were relative yto era dominant, the celtics were a yram that alternated between the 13 heat and the 18 cavs who won 11 titles because it turns out when most of your opponets are, realtive to the era, not even +4, you dont need very much to elevate a team who, without their best player, is already .500 past them.

Celtics were by far the best team in the league in 9 of 13 Russell years, some by ridiculous margin.

You want to talk about era relativity? Celtics won 62 games in 1965. The second best team in the league? 49 wins Lakers. It's constant pattern - Celtics were much better than any other team in the league to the larger degree than any other team in the NBA history.

it defaltes your record just as much as beating up on bad outliers would, therefore nuking your claim that srs was "easier" to rack up in later eras for inudviudal teams.

Stats don't back it up, good teams usually are a bit better than 50-50 against other good teams but they blow out average and below average teams. In the 1960s, teams played a lot more games against other good teams and there were less weak ones.

They come directly from playoing the competition of your era, so yes actually, they are, by definition, relative to era numbers. It would seem my understanding of stastistics surpasses your understandning of the word ---relative---. "There weren't as many 60 win teams" is specfically comparing eras and is hence not era-relative.

See how this works?

No, to make a relative term, you'd have to look at how big of probability of given event was in given era and adjust for inflation/deflation. Raw team success is just team success, it doesn't tell you anything about era differences at all. You need to look at trends inside studied era and make conclusions from that, otherwise it's not analysis.
It's like saying that raw TS% is relative stat because it's against given era competition, but rTS% isn't. Your argument is stupid.

you think russell's playoff lift is remotely comparable to lebron in 09, then make the case against lebron in 09. if you think its comparable to lebron in 15, 17, 20, 16, then make the case.

Hell you're wlecome to make the case for 2018 where not only was lebron clearly providing more lift, he also did it on a team more impressive than two of your supposed atg russell seasons.

Failure to do so serves as an implicit concession of what's obvious beyond your baseless weighting of versatily vs dominance:

Lebron was clearly the more valuable\ player, realtive to era.

Again, this isn't very difficult

Fine - let's look at 1964 Russell.

1964 Celtics finished with 6.93 SRS (+2.5 above second Royals) and 60 wins pace. In two games Russell missed Celtics were 1-1. Cetlics finished with -4.5 offense and -10.8 defense - by far the best of all-time.

Celtics had only one player who shot above league average in terms of efficiency in that season - Sam Jones with 101 TS+. They had only two players with above average FG% - Sam Jones and Bill Russell himself. Outside of Jones, every single Celtics player finished with negative TS Add in RS. This is all-time bad supporting cast offensively. In comparison - 2009 Cavs had 7 players above league average in terms of scoring efficiency.

Not only that, but this team had no playmaker on offense either - KC Jones was bad offensive player, Sam Jones played mostly off-ball and Havlicek wasn't close to his prime. Despite that, Celtics finished head and shoulders above the league in regular season.

Let's move into playoffs. Celtics faced 55 wins team Royals (2nd best in the league) in the first round. They blew them away in 5 games by a margin of 8.4 points per game. None of the losses were competitive. Royals were the best offense in the league and they posted 93.2 ppg compared to 114.7 ppg in RS which is - 21.5 in terms of raw difference. We don't have pace adjusted ratings for this series, but we'll move into this later.

Then they faced 48 wins Warriors team (53 expected wins and 3rd SRS in the league) which was built more in a mold of Celtics - excellent defense but terrible offense. Again - Celtics almost swept them with one odd blowout loss.

Celtics made a quick work of two very strong teams without much help on offense. You could say that Boston coasted in RS offensively though and they definitely improved in playoffs, right? Overall, Celtics finished in playoffs with -2.9 offense and -12.4 defense. They were absolutely terrible offensively, but their defense reached the level no other team could ever approach. They finished with +6.2 point differential despite facing two of the best opponents available.

So are you going to tell me now that 1964 Celtics were some kind of stacked team? They were absolutely terrible offensively and they had no depth at center position - they had 35 years old Lovellette for half of the season and he didn't play in playoffs basically. Given what we know about Celtics without Russell, it's unlikely that they would have been good without him on defense.

But sure, he didn't carry his team because he didn't have big OC or BPM right?

b]:/ My grammar is just fine compared to yours.[/b]

Well, I'm not native speaker so I'll let others decide.
Did you not see what you just did,

So they choked because they went 7 games against Royals and 6 games against Lakers? Then tell me how 2013 Heat choked or how 2008 Celtics choked or... basically half of title teams choked.

But sure, you'll be defending 2019 Bucks to death, even though they lost to clearly worse team on paper in 6 games.

Quantible arguments:
-> the celtics were .500 before russell and were legitmately challenged by .500 or slightly above .500 teams for half their titles, including titles under what you're saying is their supposed peak

Celtics didn't win a single playoff series before Russell. Most title teams were challenged by lesser ones.

You are also wrong, because Celtics teams were challenged by .500 teams only 4 times in 13 years:

- 1957 Hawks (Russell was a rookie),
- 1959 Nationals (+3.7 SRS and 45 expected wins, so they weren't truly a .500 team),
- 1963 Royals,
- 1965 Sixers (which got Wilt in the middle of the season and became the best team in the league in next season, so they certainly weren't .500 team).


-> The amount of 60 wins team in the league has increased proportionally to the amount of teams therefore indiciating it has not been harder for an inudvidual team to reach new heights

Except it's not a linear proportion which you seem to miss all the time.
-> The celtics srs dropped by 4-6 points for half of their postseasons which absoltely blows out the drop offs of say, the robinson spurs.

I don't think we have playoffs SRS for the 1960s, I'd appreciate if you post a source.
-> Lebron's cavs played 18 with basketball without him from 08-10 and 18 win basketball before blowing things up when he left

Sample of 2-5 games is not good to make big conclusions.
-> Multiple of the lebron teams we consider "not real contenders' were better than celtics title teams based on their mov against the compeititon of the era in the playoffs

Celtics were the best in the league, in the league that was different that it is now. Most LeBron teams were not the best in the league.
-> Russell's teams never even came close to the heights of teams like the 91 bulls,yhr 71 bucks or the 17 warriors in the rs or the playoffs, and its rather difficult to justidy that with supporitng cast when the supporitng cast - russell played close to most of the teams russell had to beat for the title.

He did in 1964, despite having much worse supporting cast than Jordan or Curry.
sansterre
Bench Warmer
Posts: 1,312
And1: 1,816
Joined: Oct 22, 2020

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#82 » by sansterre » Wed Nov 25, 2020 12:00 pm

Okay, I can supply data to address the following:

1) The Bill Russell Celtics regular season SRS vs Postseason SRS
2) The Standard Deviations of SRS through NBA History

Since it seems like these are topics of discussion.

1957: RSRS +4.78, PSRS +8.67
1958: RSRS +5.02, PSRS +8.69
1959: RSRS +5.84, PSRS +10.33
1960: RSRS +7.62, PSRS +9.28
1961: RSRS +4.94, PSRS +13.94
1962: RSRS +8.26, PSRS +8.53
1963: RSRS +6.38, PSRS +5.57
1964: RSRS +6.93, PSRS +10.83
1965: RSRS +7.46, PSRS +8.78
1966: RSRS +4.34, PSRS +8.60
1967: RSRS +7.24, PSRS +2.40
1968: RSRS +3.87, PSRS +7.10
1969: RSRS +5.35, PSRS +9.14

So the Celtics got better in the playoffs every single year except for 1967. And in most years by a lot. So I don't know how much accuracy there is to claiming that they got worse in the playoffs with any consistency.

As far as STD through history, I'm not going to post the whole list. But here are Russell's, Jordan's and modern-ish teams with similar standard deviations (for reference).

1955: -0.19
1956: +0.41
**** Add Russell****
1957: +2.16 (06 Mavs)
1958: +1.63 (97 Jazz)
1959: +1.63 (97 Jazz)
1960: +1.73 (85 Celtics)
1961: +2.08 (83 76ers)
1962: +1.73 (2012 Thunder)
1963: +1.39 (2004 Lakers)
1964: +1.70 (90 Pistons)
1965: +2.08 (83 76ers)
1966: +1.84 (2019 Raptors)
1967: +0.94 (2019 76ers)
1968: +1.32 (2019 Rockets)
1969: +1.57 (1996 Sonics)

Here are Jordan's:

1983: -0.91
1984: -1.36
*** Add Jordan ***
1985: -0.02
1987: +0.14
1988: +0.33
1989: +0.9
1990: +1.07
1991: +2.47
1992: +2.17
1993: +1.77
1996: +2.63
1997: +2.04
1998: +1.98

And heck, let's do LeBron (I'll just do through Cavs 1st ed):

2002: -0.82
2003: -2.16
*** Add LeBron ***
2004: -0.67
2005: +0.06
2006: +0.41
2007: +1.29
2008: +0.63
2009: +1.97
2010: +0.97

Basically, if you're evaluating Russell's teams by MoV and SRS, they show up decently but way short compared to more modern great teams.

If you're evaluating them in terms of how dominant they were relative to their league, they're extremely competitive. My Top 100 teams project, which mixes the two, has seven Russell teams in the #50-100 range, two Russell teams in the #25-49 range, and none above that. I will also add that for players who were the best player for the most teams in my top 100, the top 14 are:

Bill Russell: 9
LeBron James: 8
Magic Johnson: 7
Larry Bird & Michael Jordan: 6
Tim Duncan: 5 (this got weird because of Kawhi, Robinson and Manu so I had to split a few)
Kareem Abdul-Jabbar: 4
Kevin Durant & Steph Curry: 3.5
Isiah Thomas, Kobe Bryant, Wilt Chamberlain, Shaquille O'Neal, Kawhi Leonard: 3

And again, those are just by Win Shares and VORP (and fudged a little for judgment calls).

Russell deserves some serious love.
"If you wish to see the truth, hold no opinions."

"Trust one who seeks the truth. Doubt one who claims to have found the truth."
70sFan
RealGM
Posts: 29,599
And1: 24,915
Joined: Aug 11, 2015
 

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#83 » by 70sFan » Wed Nov 25, 2020 12:16 pm

sansterre wrote:Okay, I can supply data to address the following:

1) The Bill Russell Celtics regular season SRS vs Postseason SRS
2) The Standard Deviations of SRS through NBA History

Since it seems like these are topics of discussion.

1957: RSRS +4.78, PSRS +8.67
1958: RSRS +5.02, PSRS +8.69
1959: RSRS +5.84, PSRS +10.33
1960: RSRS +7.62, PSRS +9.28
1961: RSRS +4.94, PSRS +13.94
1962: RSRS +8.26, PSRS +8.53
1963: RSRS +6.38, PSRS +5.57
1964: RSRS +6.93, PSRS +10.83
1965: RSRS +7.46, PSRS +8.78
1966: RSRS +4.34, PSRS +8.60
1967: RSRS +7.24, PSRS +2.40
1968: RSRS +3.87, PSRS +7.10
1969: RSRS +5.35, PSRS +9.14

So the Celtics got better in the playoffs every single year except for 1967. And in most years by a lot. So I don't know how much accuracy there is to claiming that they got worse in the playoffs with any consistency.

As far as STD through history, I'm not going to post the whole list. But here are Russell's, Jordan's and modern-ish teams with similar standard deviations (for reference).

1955: -0.19
1956: +0.41
**** Add Russell****
1957: +2.16 (06 Mavs)
1958: +1.63 (97 Jazz)
1959: +1.63 (97 Jazz)
1960: +1.73 (85 Celtics)
1961: +2.08 (83 76ers)
1962: +1.73 (2012 Thunder)
1963: +1.39 (2004 Lakers)
1964: +1.70 (90 Pistons)
1965: +2.08 (83 76ers)
1966: +1.84 (2019 Raptors)
1967: +0.94 (2019 76ers)
1968: +1.32 (2019 Rockets)
1969: +1.57 (1996 Sonics)

Here are Jordan's:

1983: -0.91
1984: -1.36
*** Add Jordan ***
1985: -0.02
1987: +0.14
1988: +0.33
1989: +0.9
1990: +1.07
1991: +2.47
1992: +2.17
1993: +1.77
1996: +2.63
1997: +2.04
1998: +1.98

And heck, let's do LeBron (I'll just do through Cavs 1st ed):

2002: -0.82
2003: -2.16
*** Add LeBron ***
2004: -0.67
2005: +0.06
2006: +0.41
2007: +1.29
2008: +0.63
2009: +1.97
2010: +0.97

Basically, if you're evaluating Russell's teams by MoV and SRS, they show up decently but way short compared to more modern great teams.

If you're evaluating them in terms of how dominant they were relative to their league, they're extremely competitive. My Top 100 teams project, which mixes the two, has seven Russell teams in the #50-100 range, two Russell teams in the #25-49 range, and none above that. I will also add that for players who were the best player for the most teams in my top 100, the top 14 are:

Bill Russell: 9
LeBron James: 8
Magic Johnson: 7
Larry Bird & Michael Jordan: 6
Tim Duncan: 5 (this got weird because of Kawhi, Robinson and Manu so I had to split a few)
Kareem Abdul-Jabbar: 4
Kevin Durant & Steph Curry: 3.5
Isiah Thomas, Kobe Bryant, Wilt Chamberlain, Shaquille O'Neal, Kawhi Leonard: 3

And again, those are just by Win Shares and VORP (and fudged a little for judgment calls).

Russell deserves some serious love.

Thank you, especially for playoff SRS. It's great to have someone willing to make calculations when you don't have enough time recently. You bring a lot to our small community, keep doing your things!
freethedevil
Head Coach
Posts: 7,262
And1: 3,236
Joined: Dec 09, 2018
         

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#84 » by freethedevil » Wed Nov 25, 2020 7:29 pm

70sFan wrote:
freethedevil wrote:1965, +7 srs, taken to 7 y the, -.25 srs, hawks.

Celtics didn't face Hawks in 1965 playoffs.
[b]sorry, my bad, the -.25 sixers, SOOO much better.
No, lets recap, I said "it was not harder for russell induvdiaully to lift his team" then it would be for an average 90's superstar. That the --Volume-- of 60 win teams increased proportionally to the # of teams in the league supports my claim, not yours.

But lifting your team can't be measured by added wins. It's very crude approximation that needs a lot of additional work behind it.
It can be measured, and until you provide a better alternative, that it is 'crude' is meaningless when the best counter-evidence you can offer is that "omg its not perfect", a bunch of granular **** which you make ___zero attempt_ to tie to holistic impact, and this bizarre attempt to redefine the word "realtive" so that it stops meaning what players did against their opponents to "what players did against their opponents on this special curve I'm making so that 60's teams look better than they were"
Also - the number of 60+wins teams clearly shows that there is correlation between the number of teams and the outlier RS results. If you take away Russell's Celtics teams, then we had total of two seasons in the 1960s with 60+wins pace.
If i take out the top 1/8th or 1/13th of the teams in the modern league, the xact same thing happens. All you're pointg out here is less teams played. Again, irrelavant as an excuse for russell's inability to lead atg teams especially since, you say we should judge players --relative to era--, not --relative to my era adjusted difficulty curve--.

Cleary better than you apparently.

Then you certainly didn't show that here.
less posturing more arguing 70's
That's longetvity, not relative dominance. The bulls were relative yto era dominant, the celtics were a yram that alternated between the 13 heat and the 18 cavs who won 11 titles because it turns out when most of your opponets are, realtive to the era, not even +4, you dont need very much to elevate a team who, without their best player, is already .500 past them.

Celtics were by far the best team in the league in 9 of 13 Russell years, some by ridiculous margin.


they were not "by far" the best team in the league when they were being taken to 6 and 7 games in the playoffs. That's not how being the best works. If i split up the 90's league pool by 3 I cna easily form combinations where various teams posted similar dominance. This is a complete and utter red herring when discussing russell's relative to era lift. For half of these titles there was only one team in the league russell could realistically win with, he happened to be on that team, thay does not suddenly make him a jordan level player. It doesn't even make him a robinson+ player. That russell won titles and that he was not approaching peak jordan(let alone peak lebron)"s lift for the vast majority of his career are not mutually exclusive concepts.

You want to talk about era relativity? Celtics won 62 games in 1965.
WHo did they win those games against? Oh right, the compeition. That's era relativity. ANything else is breaking era realtivity and if you wnna play this "well there werne't as many contenders in a smaller league", then its fair game to lower everyone in the 60's because the talent pool was comaprative garbage to our current one. Either you accept era relatitvity or russell played against milkmen. I'm fine with nuking era relativity, but if u do so, everything is on the table.

it defaltes your record just as much as beating up on bad outliers would, therefore nuking your claim that srs was "easier" to rack up in later eras for inudviudal teams.

Stats don't back it up, good teams usually are a bit better than 50-50 against other good teams but they blow out average and below average teams. In the 1960s, teams played a lot more games against other good teams and there were less weak ones.

They come directly from playoing the competition of your era, so yes actually, they are, by definition, relative to era numbers. It would seem my understanding of stastistics surpasses your understandning of the word ---relative---. "There weren't as many 60 win teams" is specfically comparing eras and is hence not era-relative.

See how this works?

No, to make a relative term are these measures derived form the copetition of the era? Yes? Then its relative to era., you'd have to look at how big of probability of given event was in given era and adjust for inflation/deflation. Raw team success is just team success,

A player's chapmionship equity is NOT "what impact on odds they had on the team they played on", it is impact on odds on a random team, a direct product of how much _lift_ a player provides. This is why you track lift, because a player's ability to transalte their play into wins is what may transalte over to a random team, not what their team collectively did. What you still seem to not be getting here is that a smaller league doesn't just mean, there's"less teams that can dethrone you", what it means is "there's less teams" you can realsitically win a title with. When russell was being taken to 7 and 6 by teams that were avergage in his own league, the only reason he had any shot of a title those years is because he was on a team, that was nearly as good as his opponents. On a random team those odds plummet. It is compeltely meaningless to say "they wouldn't have won those titles without russell." Because russel won several titles in posrseasons wher ehe wasn't even able to lift his teams like 17 westbrook. If russell was actually consistently an atg level player, then that would have showed up with consistent playoff teams, alas ,his teams repeatedly got taken to 7 by teams only marginally better(If at all) to russell's supporitng cast. Simply being the straw on the camel's back is not near enough to suggest a player deserves parity with those who are elevating their team by 20-40 wins.

It's like saying that raw TS% is relative stat because it's against given era competition, but rTS% isn't. Your argument is stupid.
no your equivalency is stupid. TS has zero to do with who or what you're playing against. SRS comes from who you're playing against. TS is analous to who many points a team scores. RTS is anaglous to point differential.

This special curve you're using to inflate the value of everyone who played inthe 60's would be the equivalent of ditching rTS, and replacing it with

RrTS where we now adjust for how many players managed to reach +4rts in given season. Its focusing on the 'ease' of an era as opposed to what is actually achieved in said era


you think russell's playoff lift is remotely comparable to lebron in 09, then make the case against lebron in 09. if you think its comparable to lebron in 15, 17, 20, 16, then make the case.

Hell you're wlecome to make the case for 2018 where not only was lebron clearly providing more lift, he also did it on a team more impressive than two of your supposed atg russell seasons.

Failure to do so serves as an implicit concession of what's obvious beyond your baseless weighting of versatily vs dominance:

Lebron was clearly the more valuable\ player, realtive to era.

Again, this isn't very difficult

Fine - let's look at 1964 Russell.

1964 Celtics finished with 6.93 SRS (+2.5 above second Royals) and 60 wins pace. In two games Russell missed Celtics were 1-1. Cetlics finished with -4.5 offense and -10.8 defense - by far the best of all-time.
I do love how you said my 30 game sample for lebron, ot th emulti-season sample predating russell were all useless but this 2 game sample is meaningful.


But you know what, I'll give you this sample. 1-1 to 60 win gets you, checks notes, 20 wins which is,
looks at half of lebron's prime seasons
nowhere near as much lift as peak lebron. The cavs played 35 win basketball(over a much larger sample mind you) with kyrie and kevin love from 15-17, withotu those two they were beatign on a 67 win team befote said team pulled out the deathlineup, something that took them from 67 wins, to a record breaking 73 wins on a 70 win srs.

Maybee we should use 2016, where the hurt version of the warriors sitll beat a team that played 65 win basketball at full strength(okc) and who had just beated a 70 win srs spurs, and then proceeded to --lose-- to lebron's cavs with the aid of their second best player missing one game. And a couple minor injuries for a game to their worst starters. For context, the 68 and 69 lakers who russell cemented his legacy against were vastly more shorthanded teams.

So to recap, we haev 40 win lift on the early cavs, 25-30 win on the second stint cavs, and the best you can do for me from what you call russell's prime is a 2 game sample showcasing 20 win lift?
You cherrypkcing defense is meaningless, lest you think its ridiculous to not have nash on the same level as micheal jordan. What matters is the overall end, not whether the defensive rating o r the offensive rating under lineups built to optimize a player who couldn't do jack on one end of the floor happen to acheive on that one end of the floo

Actually forget lebron, lets do a peak I'm sure you consider comparable to lebron or jordan's, 15 curry. Given that 15 curry lite dhis team by 25 wins from 15-17 on a team more dominant than russell's celtics. When has russell ever proved he's as good as 15 curry?

And just so we're clear, useful sample sizes do not support your framing of things. The pre-russell celtics were a .500 team, russell was taken to 7 en route to titles by 4 different .500 teams in the playoffs. That does not happen if you're an unflinchingly mvp level player trhoughout your prime.


Celtics had only one player who shot above league average in terms of efficiency in that season - Sam Jones with 101 TS+. They had only two players with above average FG% - Sam Jones and Bill Russell himself. Outside of Jones, every single Celtics player finished with negative TS Add in RS. This is all-time bad supporting cast offensively. In comparison - 2009 Cavs had 7 players above league average in terms of scoring efficiency.
then clearly they made up in some other way the 09 cavs did not, becuase again, no matter how you try to **** on this supporitng cast, they were perfectly capable of staying near or above water without russell.

Also, has it occured to you, that maybe the effiency of these casts had something to do with one of them playing next to Lebron James, and the other cast playing next to bill russell?

Indeed, russell's teammate effiency dropped a bit upon his arrival.

It would appear the 40 win player who can anchor a defense be by far the best scorer in the league AND created the most offense for his teammates _just happened_ to play with players who shot better than russell's.

I'm completely clueless as to how that's possible :roll:


Not only that, but this team had no playmaker on offense either - KC Jones was bad offensive player, Sam Jones played mostly off-ball and Havlicek wasn't close to his prime. Despite that, Celtics finished head and shoulders above the league in regular season.
russell playmking so hard, ther'es like zero gap in their scoring effiency before he joins and after. This team that did perfectly fine without him for several seasons clearly needed russell to not be the timberwolves.

Let's move into playoffs. Celtics faced 55 wins team Royals (2nd best in the league)

being the second best team in an 8 teague doesn't mean you're good. You can keep listing your rankings, and I will keep ignoring them like the red herrings they are.
in the first round. They blew them away in 5 games by a margin of 8.4 points per game. None of the losses were competitive. Royals were the best offense in the league and they posted 93.2 ppg compared to 114.7 ppg in RS which is - 21.5 in terms of raw difference. We don't have pace adjusted ratings for this series, but we'll move into this later.

Then they faced 48 wins Warriors team (53 expected wins and 3rd SRS in the league) which was built more in a mold of Celtics - excellent defense but terrible offense. Again - Celtics almost swept them with one odd blowout loss.
.You're telling me that an atg player on a 40 win team was able to take care of a 53 win team. Yes, this absolutely proves russell is comaparable to a 40 win player. Why don't you do that bit where you actually compare it to lebron's best seasons and deonstrate he was comparably valuable. You claimed "russell is as valauble as Lebron and/or Jordan at their peaks. Please prove that now. Reminder 18 Lebron "swept" a better team than either you listed with a dumpsterfire in terms of cast by comparison.
Celtics made a quick work of two very strong teams without much help on offense.
If by very strong you mean the 17 celtics, the 18 raptors, the 18 pacers, then yes 'very strong'. :lol:

Reminder, the 20 heat, who the lakers beat, went 8-3 against the bucks and the 55 win celtics who themselves bat a raptors team playing at a 63 win pace. The 2020 Lkaers - lebron(and - ad) played like a 40-45 win team without either of their superstars. If you goal here was to show Bill Russell at his best is within range of Lebron in his 17th season, you did a very good job. If you were trying to argue russell was equal in value to lebron at his best, then no, your attempt at 'proving' russsell's parity with them pretty adequately proves he's not as valuable.(at least not on the court anyway which is what you've been trying to claim)


You could say that Boston coasted in RS offensively though and they definitely improved in playoffs, right? Overall, Celtics finished in playoffs with -2.9 offense and -12.4 defense. I'm sorry is there a reason you keep trying to paint them sucking as an offense --with russell-- as a good thing for russell? They were absolutely terrible offensivelywhich somehow has nothing to do with russell..., but their defense reached the level no other team could ever approachat the apprent cost of a remotely competent offense, gee its almost like cherrypikcing one side of the floor is an awful way of assessing players holsitic impact.. They finished with +6.2 point differential despite facing two of the best opponents available. they posted a strong differential against the 17 celtics. I am blown away. 30+ Lebron could never....(wait, no he did, with a weaker supporing cast. And since you like cherrypicking one side of the floor, he did it leading an offense BETTER than the 17 warriors. Lmao.

So are you going to tell me now that 1964 Celtics were some kind of stacked team?
I dont need to, beating 'good' teams with a .500 supporting cast is something we have seen several players do. nothing you've said constitutes compelling evidence, given what we've actually seen from the players you insist russell is a match for, this is rather pathetic. The celtics would need to be the kg-less timberwolves, for this to be lebron worthy lift. Off course, they weren't, so...
They were absolutely terrible offensively with russell and they had no depth at center positionwhich should inflate how much they drop off withou thim but, they were ,.500 sooo - they had 35 years old Lovellette for half of the season and he didn't play in playoffs basically. Given what we know about Celtics without Russell, it's unlikely that they would have been good without him on defense. and again, if you have to ignore that the offense also came with bill russell, you're implictly conceding that the actual lift russell presented wasn't nearly as impressive as you're acting like it is.

But sure, he didn't carry his team because he didn't have big OC or BPM right?
Did I say he never carried his team? I said he never came close to matching lebron's carry's jobs which you've done a very bad job rebutting.
b]:/ My grammar is just fine compared to yours.[/b]

Well, I'm not native speaker so I'll let others decide.
Did you not see what you just did,

Then tell me how 2013 Heat choked

They did choke, and you're choking right now trying to compare one of lebron's least valauble seasons's to what you consider russell's most valuable implicitly conceding, for as second time, Lebron is simply a much more valuable player. But you couldn't stop there, no you had to make it worse:


or how 2008 Celtics choked or...

You are, just so we're clear, comparing KG 8 years after what is by far his most valuable season to what is allegedly 'the most impressive' russell season?

Russell is both a rival for lebron and also a rival for kobe, crazy how that works.


basically half of title teams choked.
the vast majority of title teams sweep +1 and +2 and +0 srs teams or demolish them in 5. Russell does have seasons where his title teams behaved like normal title teams, so does Kobe bryant.
But sure, yo


[b]

2. Lmao at [/b]u'll be defending 2019 Bucks to death, even though they lost to clearly worse team on paper in 6 games.
What? The raptors, without a better version of Kevin Durant went 17-4. There is absolute zip to suggest either team you listed is as good as these raps without kawhi, forget adding MVP-level peak in the playoffs.

Lmao at you unironically trying to equate losing by a van vleet hot streak to a 60 win team+Kawhi Leonard to being takent o 7 by +1, +2, +3, and -srs teams.


And lamo at you deciding lift based on their rs with Giannis as oposed to how they played without him.

You were never willing to compare bar to the standard you claimed he was worthy of, and now you've progressively lowere dhim from one of lebron's worst postseasons to over the hill KG and 19 Giannis.

Quantible arguments:
-> the celtics were .500 before russell and were legitmately challenged by .500 or slightly above .500 teams for half their titles, including titles under what you're saying is their supposed peak

Celtics didn't win a single playoff series before Russell. Absolutely meaningless comment. Winning a playoff series would be the equivalent of making the conference final without him. MVP's do not need conference final teams to win titles, let alone players you're allegding must be treated as comparable to Lebron.

You are also wrong, because Celtics teams were challenged by .500 teams only 4 times in 13 years:

- 1957 Hawks (Russell was a rookie),
- 1959 Nationals (+3.7 SRS and 45 expected wins, so they weren't truly a .500 team),
- 1963 Royals,
- 1965 Sixers (which got Wilt in the middle of the season and became the best team in the league in next season, so they certainly weren't .500 team). Fair enough, they were really a +4.1 team. I am in awe.

Guess you forgot the

+1.77 Hawks in 1960
-> +2 warriors in 1960
-> The 62 Warriors AND the 62 Lakers (+2 and +1 respectively)
-> The 63 Lakers +2.67
-> The +1 66 Royals
-> The +2.76 66 Lkaers

So basically every playoffs during this strtch where russelll was supposedly a consistent mvp-atg level player.

To his credit the celtics finally faced a legit title level team in the 67 celtics(+8) and....

Got blown to bits. Pretty in-line with a team that usaully was getting challenged by the 18 raptors or the 91 Sixers. I understand why you want to break out this special curve.




-> The amount of 60 wins team in the league has increased proportionally to the amount of teams therefore indiciating it has not been harder for an inudvidual team to reach new heights

Except it's not a linear proportion which you seem to miss all the time.
-> The celtics srs dropped by 4-6 points for half of their postseasons which absoltely blows out the drop offs of say, the robinson spurs.

I don't think we have playoffs SRS for the 1960s, I'd appreciate if you post a source. I assumed massive drop offs because thwy were regular challenged by mediocre teams despite. This was wrong. It would seem it snot so much they choked int he playoffs, it would seem they just weren't that good in the first place:
they crossed +7, 4 times. Most title teams are +8, a mark they crossed once. Off course their first title level RS team did have a horrific +4 srs drop off, but to the celtics credit, rather than being great and dropping to merely good.. They started off as merely good and then fluctuated between mediocre, and legimately great.

IOW, Bill Russell's prime was not a consistent slog of similarly valuable seasons. Like your typical non- Lebron, Duncan, or Kareem atg, russell had a few seasons where he had atg level lift(that still did not touch lebron). The idea that he was consistently at that level is baseless and the idea that he was ever close to asvaluable as the most valuable lebron years is equally baseless AND heavily contradicted by even your presented eveidence.


-> Lebron's cavs played 18 with basketball without him from 08-10 and 18 win basketball before blowing things up when he left

Sample of 2-5 games is not good to make big conclusions. 35 games buddy, but i'm glad you now care about sample sizes after disregarding multiple full seasons of .500 non-russell celtics play for a 2 game sample.
-> Multiple of the lebron teams we consider "not real contenders' were better than celtics title teams based on their mov against the compeititon of the era in the playoffs

Celtics were the best in the league, in the league that was different that it is now. Most LeBron teams were not the best in the league. Because in a bigger league multiple teams can concievably contend. Again, when assessing induvidual lift, this works --both ways-- because just as russell's lift is likely to transalte to a title if he's on the one team in the league that is good enough for that, there are "less teams" potentially good enough for that.

All these playoff series i lsited above? There is only one team in the league good enough got him to survive those, and that is the team he happened to be drafted on. That there wer eless teams is a red herring when assessing championship dd list on a random team.

-> Russell's teams never even came close to the heights of teams like the 91 bulls,yhr 71 bucks or the 17 warriors in the rs or the playoffs, and its rather difficult to justidy that with supporitng cast when the supporitng cast - russell played close to most of the teams russell had to beat for the title.

He did in 1964, despite having much worse supporting cast than Jordan or Curry.
No, he pulled a less impressive version of the 2020 lakers and you can say "he had a much supporting cast" but you have failed to prove it. Russell's cast was .500 without him, you cherrypikcing defense doesn't change that.

70sFan
RealGM
Posts: 29,599
And1: 24,915
Joined: Aug 11, 2015
 

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#85 » by 70sFan » Wed Nov 25, 2020 8:54 pm

I'm not going to do this forever. You have your cute way of evaluating players, so stick with that if you feel so confident with results. You keep ignoring things I and other bring up in this thread, so I lost the belief that this discussion matter.
freethedevil
Head Coach
Posts: 7,262
And1: 3,236
Joined: Dec 09, 2018
         

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#86 » by freethedevil » Wed Nov 25, 2020 10:12 pm

70sFan wrote:I'm not going to do this forever. You have your cute way of evaluating players, so stick with that if you feel so confident with results. You keep ignoring things I and other bring up in this thread, so I lost the belief that this discussion matter.

You have no way of evaluating players, that's the problem. Hence why you keep trying to force all of them to be equals.
freethedevil
Head Coach
Posts: 7,262
And1: 3,236
Joined: Dec 09, 2018
         

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#87 » by freethedevil » Wed Nov 25, 2020 10:14 pm

sansterre wrote:Okay, I can supply data to address the following:

1) The Bill Russell Celtics regular season SRS vs Postseason SRS
2) The Standard Deviations of SRS through NBA History

Since it seems like these are topics of discussion.

1957: RSRS +4.78, PSRS +8.67
1958: RSRS +5.02, PSRS +8.69
1959: RSRS +5.84, PSRS +10.33
1960: RSRS +7.62, PSRS +9.28
1961: RSRS +4.94, PSRS +13.94
1962: RSRS +8.26, PSRS +8.53
1963: RSRS +6.38, PSRS +5.57
1964: RSRS +6.93, PSRS +10.83
1965: RSRS +7.46, PSRS +8.78
1966: RSRS +4.34, PSRS +8.60
1967: RSRS +7.24, PSRS +2.40
1968: RSRS +3.87, PSRS +7.10
1969: RSRS +5.35, PSRS +9.14

So the Celtics got better in the playoffs every single year except for 1967. And in most years by a lot. So I don't know how much accuracy there is to claiming that they got worse in the playoffs with any consistency.

As far as STD through history, I'm not going to post the whole list. But here are Russell's, Jordan's and modern-ish teams with similar standard deviations (for reference).

1955: -0.19
1956: +0.41
**** Add Russell****
1957: +2.16 (06 Mavs)
1958: +1.63 (97 Jazz)
1959: +1.63 (97 Jazz)
1960: +1.73 (85 Celtics)
1961: +2.08 (83 76ers)
1962: +1.73 (2012 Thunder)
1963: +1.39 (2004 Lakers)
1964: +1.70 (90 Pistons)
1965: +2.08 (83 76ers)
1966: +1.84 (2019 Raptors)
1967: +0.94 (2019 76ers)
1968: +1.32 (2019 Rockets)
1969: +1.57 (1996 Sonics)

Here are Jordan's:

1983: -0.91
1984: -1.36
*** Add Jordan ***
1985: -0.02
1987: +0.14
1988: +0.33
1989: +0.9
1990: +1.07
1991: +2.47
1992: +2.17
1993: +1.77
1996: +2.63
1997: +2.04
1998: +1.98

And heck, let's do LeBron (I'll just do through Cavs 1st ed):

2002: -0.82
2003: -2.16
*** Add LeBron ***
2004: -0.67
2005: +0.06
2006: +0.41
2007: +1.29
2008: +0.63
2009: +1.97
2010: +0.97

Basically, if you're evaluating Russell's teams by MoV and SRS, they show up decently but way short compared to more modern great teams.

If you're evaluating them in terms of how dominant they were relative to their league, they're extremely competitive. My Top 100 teams project, which mixes the two, has seven Russell teams in the #50-100 range, two Russell teams in the #25-49 range, and none above that. I will also add that for players who were the best player for the most teams in my top 100, the top 14 are:

Bill Russell: 9
LeBron James: 8
Magic Johnson: 7
Larry Bird & Michael Jordan: 6
Tim Duncan: 5 (this got weird because of Kawhi, Robinson and Manu so I had to split a few)
Kareem Abdul-Jabbar: 4
Kevin Durant & Steph Curry: 3.5
Isiah Thomas, Kobe Bryant, Wilt Chamberlain, Shaquille O'Neal, Kawhi Leonard: 3

And again, those are just by Win Shares and VORP (and fudged a little for judgment calls).

Russell deserves some serious love.

Does this adjust for pace?

Ie: losing by 7 with lots of possessions is different than losing by 7 with half as many possessions'

Because the celtics were taken to 7 or 6 by a inferioir regular seaosn srs team just about every postseason.
70sFan
RealGM
Posts: 29,599
And1: 24,915
Joined: Aug 11, 2015
 

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#88 » by 70sFan » Wed Nov 25, 2020 10:22 pm

freethedevil wrote:
70sFan wrote:I'm not going to do this forever. You have your cute way of evaluating players, so stick with that if you feel so confident with results. You keep ignoring things I and other bring up in this thread, so I lost the belief that this discussion matter.

You have no way of evaluating players, that's the problem. Hence why you keep trying to force all of them to be equals.

Based on what? I never said all of them are equal and I certainly have my own way of evaluation, it's just much less rigid than yours. Which is fine, as long as you are consistent and you respect other opinions we should have no problem with that. You clearly don't respect any opinion that is not similar to yours and you act like your way is the best way possible.

I won't discuss with someone who doesn't respect my opinion. This type of discussion is worthless.
freethedevil
Head Coach
Posts: 7,262
And1: 3,236
Joined: Dec 09, 2018
         

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#89 » by freethedevil » Wed Nov 25, 2020 10:54 pm

70sFan wrote:
freethedevil wrote:
70sFan wrote:I'm not going to do this forever. You have your cute way of evaluating players, so stick with that if you feel so confident with results. You keep ignoring things I and other bring up in this thread, so I lost the belief that this discussion matter.

You have no way of evaluating players, that's the problem. Hence why you keep trying to force all of them to be equals.

Based on what? I never said all of them are equal and I certainly have my own way of evaluation,

Nope. You have no way of holistically evaluating players, that's the problem, hence why you're always appealing to moderation when comparing players and ciring consensus. Whenever ur pressed on a stance you spend half of your response complaining about how dissenting opioins are being argued, and what the opinoins are, as opposed to whats being argued.

My "cute" method means I never have to do crap like this:
Sure, because in your imagination James wasn't close to his best in 2013. You base it not on his skillset, or ability to play basketball, but on raw results. James wasn't worse player in 2013 than in 2009.


^^^ You know what that is? That's an appeal to increduulity. It is not common practice among those who can back up their opinions. It is however common practice with you.

it's just much less rigid than yours

Oh most definitely, which is why you tried to potray me being willing to change my opinions as a bad thing:
The funniest thing is that you used to be very pro-Russell but now you decided to throw any analysis away and focus on raw RS team results without trying to contextualize anything

Finally

you respect other opinions we should have no problem with that.

Only one of us, the supposedly 'respectful', one has rpereatdly engaged in ad-hom.

Exhibit A:
You clearly don't respect any opinion that is not similar to yours and you act like your way is the best way possible.

Exhibit B:
[]Sure, because in your imagination James wasn't close to his best in 2013.

The majority of your last three posts are made of
-> appeals to incredulity on the ridiculousness of my stances
-> meta-commentary attempting to put a referendum on me
->ad hom
The vast majority of my posts are direct rebuttals


Which is reflective of 'respect'?

How many times has 'agenda' or 'bias' been the first thing you've brought up in a point of discussion.
[b]
And, your latest contribution:
I'm not going to do this forever. You have your cute way of evaluating players, so stick with that if you feel so confident with results. You keep ignoring things I and other bring up in this thread, so I lost the belief that this discussion matter.

Glass houses and all that.
sansterre
Bench Warmer
Posts: 1,312
And1: 1,816
Joined: Oct 22, 2020

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#90 » by sansterre » Thu Nov 26, 2020 12:42 am

freethedevil wrote:
sansterre wrote:Okay, I can supply data to address the following:

1) The Bill Russell Celtics regular season SRS vs Postseason SRS
2) The Standard Deviations of SRS through NBA History

Since it seems like these are topics of discussion.

1957: RSRS +4.78, PSRS +8.67
1958: RSRS +5.02, PSRS +8.69
1959: RSRS +5.84, PSRS +10.33
1960: RSRS +7.62, PSRS +9.28
1961: RSRS +4.94, PSRS +13.94
1962: RSRS +8.26, PSRS +8.53
1963: RSRS +6.38, PSRS +5.57
1964: RSRS +6.93, PSRS +10.83
1965: RSRS +7.46, PSRS +8.78
1966: RSRS +4.34, PSRS +8.60
1967: RSRS +7.24, PSRS +2.40
1968: RSRS +3.87, PSRS +7.10
1969: RSRS +5.35, PSRS +9.14

So the Celtics got better in the playoffs every single year except for 1967. And in most years by a lot. So I don't know how much accuracy there is to claiming that they got worse in the playoffs with any consistency.

As far as STD through history, I'm not going to post the whole list. But here are Russell's, Jordan's and modern-ish teams with similar standard deviations (for reference).

1955: -0.19
1956: +0.41
**** Add Russell****
1957: +2.16 (06 Mavs)
1958: +1.63 (97 Jazz)
1959: +1.63 (97 Jazz)
1960: +1.73 (85 Celtics)
1961: +2.08 (83 76ers)
1962: +1.73 (2012 Thunder)
1963: +1.39 (2004 Lakers)
1964: +1.70 (90 Pistons)
1965: +2.08 (83 76ers)
1966: +1.84 (2019 Raptors)
1967: +0.94 (2019 76ers)
1968: +1.32 (2019 Rockets)
1969: +1.57 (1996 Sonics)

Here are Jordan's:

1983: -0.91
1984: -1.36
*** Add Jordan ***
1985: -0.02
1987: +0.14
1988: +0.33
1989: +0.9
1990: +1.07
1991: +2.47
1992: +2.17
1993: +1.77
1996: +2.63
1997: +2.04
1998: +1.98

And heck, let's do LeBron (I'll just do through Cavs 1st ed):

2002: -0.82
2003: -2.16
*** Add LeBron ***
2004: -0.67
2005: +0.06
2006: +0.41
2007: +1.29
2008: +0.63
2009: +1.97
2010: +0.97

Basically, if you're evaluating Russell's teams by MoV and SRS, they show up decently but way short compared to more modern great teams.

If you're evaluating them in terms of how dominant they were relative to their league, they're extremely competitive. My Top 100 teams project, which mixes the two, has seven Russell teams in the #50-100 range, two Russell teams in the #25-49 range, and none above that. I will also add that for players who were the best player for the most teams in my top 100, the top 14 are:

Bill Russell: 9
LeBron James: 8
Magic Johnson: 7
Larry Bird & Michael Jordan: 6
Tim Duncan: 5 (this got weird because of Kawhi, Robinson and Manu so I had to split a few)
Kareem Abdul-Jabbar: 4
Kevin Durant & Steph Curry: 3.5
Isiah Thomas, Kobe Bryant, Wilt Chamberlain, Shaquille O'Neal, Kawhi Leonard: 3

And again, those are just by Win Shares and VORP (and fudged a little for judgment calls).

Russell deserves some serious love.

Does this adjust for pace?

Ie: losing by 7 with lots of possessions is different than losing by 7 with half as many possessions'

Because the celtics were taken to 7 or 6 by a inferioir regular seaosn srs team just about every postseason.


This does *not* adjust for pace.

I agree that adjusting it for pace would give a better per-possession measure of team dominance.

On the other hand, every ten-point win would have taken the same number of possessions to turn into a loss. Whether you won by 10 in a 120 possession game, or won by 10 in an 80 possession game, you won by about the same number of possessions, if that makes any sense.

Imagine comparing a +10 SRS team that plays at 120, and a +7 SRS team that plays at 60. In terms of per-possession dominance, the +7/60 team is better. And yet, because they play so slow, the amount of uncertainty in their results is increased, and they only end up winning by about 7 possessions instead of 10, like the worse but faster team does.

Basically, faster pace good teams are more likely to win than equally good slow paced teams by increasing the sample size from which the victor is determined.

So even though pace adjusted is a better measure of dominance, it makes the stat less predictive.

This is all theorycrafting - I haven't actually run the numbers. I just spent a week weighing the pros and cons of adjusting my SRS by pace and the above ended up guiding my decision not to.

If pace-adjusted SRS turns out to be more predictive than unadjusted SRS then that's very much worth knowing.
"If you wish to see the truth, hold no opinions."

"Trust one who seeks the truth. Doubt one who claims to have found the truth."
MyUniBroDavis
General Manager
Posts: 7,827
And1: 5,029
Joined: Jan 14, 2013

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#91 » by MyUniBroDavis » Thu Nov 26, 2020 1:10 am

sansterre wrote:
freethedevil wrote:
sansterre wrote:Okay, I can supply data to address the following:

1) The Bill Russell Celtics regular season SRS vs Postseason SRS
2) The Standard Deviations of SRS through NBA History

Since it seems like these are topics of discussion.

1957: RSRS +4.78, PSRS +8.67
1958: RSRS +5.02, PSRS +8.69
1959: RSRS +5.84, PSRS +10.33
1960: RSRS +7.62, PSRS +9.28
1961: RSRS +4.94, PSRS +13.94
1962: RSRS +8.26, PSRS +8.53
1963: RSRS +6.38, PSRS +5.57
1964: RSRS +6.93, PSRS +10.83
1965: RSRS +7.46, PSRS +8.78
1966: RSRS +4.34, PSRS +8.60
1967: RSRS +7.24, PSRS +2.40
1968: RSRS +3.87, PSRS +7.10
1969: RSRS +5.35, PSRS +9.14

So the Celtics got better in the playoffs every single year except for 1967. And in most years by a lot. So I don't know how much accuracy there is to claiming that they got worse in the playoffs with any consistency.

As far as STD through history, I'm not going to post the whole list. But here are Russell's, Jordan's and modern-ish teams with similar standard deviations (for reference).

1955: -0.19
1956: +0.41
**** Add Russell****
1957: +2.16 (06 Mavs)
1958: +1.63 (97 Jazz)
1959: +1.63 (97 Jazz)
1960: +1.73 (85 Celtics)
1961: +2.08 (83 76ers)
1962: +1.73 (2012 Thunder)
1963: +1.39 (2004 Lakers)
1964: +1.70 (90 Pistons)
1965: +2.08 (83 76ers)
1966: +1.84 (2019 Raptors)
1967: +0.94 (2019 76ers)
1968: +1.32 (2019 Rockets)
1969: +1.57 (1996 Sonics)

Here are Jordan's:

1983: -0.91
1984: -1.36
*** Add Jordan ***
1985: -0.02
1987: +0.14
1988: +0.33
1989: +0.9
1990: +1.07
1991: +2.47
1992: +2.17
1993: +1.77
1996: +2.63
1997: +2.04
1998: +1.98

And heck, let's do LeBron (I'll just do through Cavs 1st ed):

2002: -0.82
2003: -2.16
*** Add LeBron ***
2004: -0.67
2005: +0.06
2006: +0.41
2007: +1.29
2008: +0.63
2009: +1.97
2010: +0.97

Basically, if you're evaluating Russell's teams by MoV and SRS, they show up decently but way short compared to more modern great teams.

If you're evaluating them in terms of how dominant they were relative to their league, they're extremely competitive. My Top 100 teams project, which mixes the two, has seven Russell teams in the #50-100 range, two Russell teams in the #25-49 range, and none above that. I will also add that for players who were the best player for the most teams in my top 100, the top 14 are:

Bill Russell: 9
LeBron James: 8
Magic Johnson: 7
Larry Bird & Michael Jordan: 6
Tim Duncan: 5 (this got weird because of Kawhi, Robinson and Manu so I had to split a few)
Kareem Abdul-Jabbar: 4
Kevin Durant & Steph Curry: 3.5
Isiah Thomas, Kobe Bryant, Wilt Chamberlain, Shaquille O'Neal, Kawhi Leonard: 3

And again, those are just by Win Shares and VORP (and fudged a little for judgment calls).

Russell deserves some serious love.

Does this adjust for pace?

Ie: losing by 7 with lots of possessions is different than losing by 7 with half as many possessions'

Because the celtics were taken to 7 or 6 by a inferioir regular seaosn srs team just about every postseason.


This does *not* adjust for pace.

I agree that adjusting it for pace would give a better per-possession measure of team dominance.

On the other hand, every ten-point win would have taken the same number of possessions to turn into a loss. Whether you won by 10 in a 120 possession game, or won by 10 in an 80 possession game, you won by about the same number of possessions, if that makes any sense.

Imagine comparing a +10 SRS team that plays at 120, and a +7 SRS team that plays at 60. In terms of per-possession dominance, the +7/60 team is better. And yet, because they play so slow, the amount of uncertainty in their results is increased, and they only end up winning by about 7 possessions instead of 10, like the worse but faster team does.

Basically, faster pace good teams are more likely to win than equally good slow paced teams by increasing the sample size from which the victor is determined.

So even though pace adjusted is a better measure of dominance, it makes the stat less predictive.

This is all theorycrafting - I haven't actually run the numbers. I just spent a week weighing the pros and cons of adjusting my SRS by pace and the above ended up guiding my decision not to.


I mean in theory, it shouldnt be like that though, because variance sizes shoudl decrease with lower possessions too, although i get what youre saying and it might play out that way though
sansterre
Bench Warmer
Posts: 1,312
And1: 1,816
Joined: Oct 22, 2020

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#92 » by sansterre » Thu Nov 26, 2020 1:46 am

MyUniBroDavis wrote:
sansterre wrote:
freethedevil wrote:Does this adjust for pace?

Ie: losing by 7 with lots of possessions is different than losing by 7 with half as many possessions'

Because the celtics were taken to 7 or 6 by a inferioir regular seaosn srs team just about every postseason.


This does *not* adjust for pace.

I agree that adjusting it for pace would give a better per-possession measure of team dominance.

On the other hand, every ten-point win would have taken the same number of possessions to turn into a loss. Whether you won by 10 in a 120 possession game, or won by 10 in an 80 possession game, you won by about the same number of possessions, if that makes any sense.

Imagine comparing a +10 SRS team that plays at 120, and a +7 SRS team that plays at 60. In terms of per-possession dominance, the +7/60 team is better. And yet, because they play so slow, the amount of uncertainty in their results is increased, and they only end up winning by about 7 possessions instead of 10, like the worse but faster team does.

Basically, faster pace good teams are more likely to win than equally good slow paced teams by increasing the sample size from which the victor is determined.

So even though pace adjusted is a better measure of dominance, it makes the stat less predictive.

This is all theorycrafting - I haven't actually run the numbers. I just spent a week weighing the pros and cons of adjusting my SRS by pace and the above ended up guiding my decision not to.


I mean in theory, it shouldnt be like that though, because variance sizes shoudl decrease with lower possessions too, although i get what youre saying and it might play out that way though


Agreed. A lot of my reasoning came down to "BBR could have pace-adjusted this with almost no effort; that they didn't suggests that there was a reason for not doing so."
"If you wish to see the truth, hold no opinions."

"Trust one who seeks the truth. Doubt one who claims to have found the truth."
MyUniBroDavis
General Manager
Posts: 7,827
And1: 5,029
Joined: Jan 14, 2013

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#93 » by MyUniBroDavis » Thu Nov 26, 2020 2:08 am

sansterre wrote:
MyUniBroDavis wrote:
sansterre wrote:
This does *not* adjust for pace.

I agree that adjusting it for pace would give a better per-possession measure of team dominance.

On the other hand, every ten-point win would have taken the same number of possessions to turn into a loss. Whether you won by 10 in a 120 possession game, or won by 10 in an 80 possession game, you won by about the same number of possessions, if that makes any sense.

Imagine comparing a +10 SRS team that plays at 120, and a +7 SRS team that plays at 60. In terms of per-possession dominance, the +7/60 team is better. And yet, because they play so slow, the amount of uncertainty in their results is increased, and they only end up winning by about 7 possessions instead of 10, like the worse but faster team does.

Basically, faster pace good teams are more likely to win than equally good slow paced teams by increasing the sample size from which the victor is determined.

So even though pace adjusted is a better measure of dominance, it makes the stat less predictive.

This is all theorycrafting - I haven't actually run the numbers. I just spent a week weighing the pros and cons of adjusting my SRS by pace and the above ended up guiding my decision not to.


I mean in theory, it shouldnt be like that though, because variance sizes shoudl decrease with lower possessions too, although i get what youre saying and it might play out that way though


Agreed. A lot of my reasoning came down to "BBR could have pace-adjusted this with almost no effort; that they didn't suggests that there was a reason for not doing so."


I mean thats just the formula tho right? At the end of the day bball isnt really a set of 95 independent possessions but a set of possessions building upon each other, which is why per possession numbers make more sense, so equaung everything to per 100 isnt always the best, but at the same time doing it raw isnt fair in terms of averages either. Like its hard to explain but you do kind of get a boost when you see the other team scoring to keep pace and vice versa, which is why iirc a teams high scoring effeciency games correlate well withh their worst deffensive games, although other factors play a part


Random thought, id be curious on the importance of winning each quarter, ignoring garbage time. Winning the first and third quarters are indicitive of someting happening tactically and garbage time is a huge thing so its kinda hard to find which causes which and stuff though
s0ciety
Freshman
Posts: 83
And1: 77
Joined: Feb 26, 2020

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#94 » by s0ciety » Thu Nov 26, 2020 4:27 am

sansterre wrote:Okay, I can supply data to address the following:

1) The Bill Russell Celtics regular season SRS vs Postseason SRS
2) The Standard Deviations of SRS through NBA History

Since it seems like these are topics of discussion.

1957: RSRS +4.78, PSRS +8.67
1958: RSRS +5.02, PSRS +8.69
1959: RSRS +5.84, PSRS +10.33
1960: RSRS +7.62, PSRS +9.28
1961: RSRS +4.94, PSRS +13.94
1962: RSRS +8.26, PSRS +8.53
1963: RSRS +6.38, PSRS +5.57
1964: RSRS +6.93, PSRS +10.83
1965: RSRS +7.46, PSRS +8.78
1966: RSRS +4.34, PSRS +8.60
1967: RSRS +7.24, PSRS +2.40
1968: RSRS +3.87, PSRS +7.10
1969: RSRS +5.35, PSRS +9.14


Would you mind sharing the way you compute Post-Season SRS?

My model seems to undervalue the Celtics compared to yours :

Code: Select all

Season         sansterre       s0ciety        Difference
1957           8.67            6.70           1.97
1958           8.69            6.14           2.55
1959           10.33           9.66           0.67
1960           9.28            8.03           1.25
1961           13.94           14.06         -0.12
1962           8.53            7.02           1.51
1963           5.57            4.10           1.47
1964           10.83           10.62          0.21
1965           8.78            7.63           1.15
1966           8.60            8.26           0.34
1967           2.40            1.90           0.50
1968           7.10            7.07           0.03
1969           9.14            7.75           1.39
freethedevil
Head Coach
Posts: 7,262
And1: 3,236
Joined: Dec 09, 2018
         

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#95 » by freethedevil » Thu Nov 26, 2020 7:45 am

sansterre wrote:
freethedevil wrote:
sansterre wrote:Okay, I can supply data to address the following:

1) The Bill Russell Celtics regular season SRS vs Postseason SRS
2) The Standard Deviations of SRS through NBA History

Since it seems like these are topics of discussion.

1957: RSRS +4.78, PSRS +8.67
1958: RSRS +5.02, PSRS +8.69
1959: RSRS +5.84, PSRS +10.33
1960: RSRS +7.62, PSRS +9.28
1961: RSRS +4.94, PSRS +13.94
1962: RSRS +8.26, PSRS +8.53
1963: RSRS +6.38, PSRS +5.57
1964: RSRS +6.93, PSRS +10.83
1965: RSRS +7.46, PSRS +8.78
1966: RSRS +4.34, PSRS +8.60
1967: RSRS +7.24, PSRS +2.40
1968: RSRS +3.87, PSRS +7.10
1969: RSRS +5.35, PSRS +9.14

So the Celtics got better in the playoffs every single year except for 1967. And in most years by a lot. So I don't know how much accuracy there is to claiming that they got worse in the playoffs with any consistency.

As far as STD through history, I'm not going to post the whole list. But here are Russell's, Jordan's and modern-ish teams with similar standard deviations (for reference).

1955: -0.19
1956: +0.41
**** Add Russell****
1957: +2.16 (06 Mavs)
1958: +1.63 (97 Jazz)
1959: +1.63 (97 Jazz)
1960: +1.73 (85 Celtics)
1961: +2.08 (83 76ers)
1962: +1.73 (2012 Thunder)
1963: +1.39 (2004 Lakers)
1964: +1.70 (90 Pistons)
1965: +2.08 (83 76ers)
1966: +1.84 (2019 Raptors)
1967: +0.94 (2019 76ers)
1968: +1.32 (2019 Rockets)
1969: +1.57 (1996 Sonics)

Here are Jordan's:

1983: -0.91
1984: -1.36
*** Add Jordan ***
1985: -0.02
1987: +0.14
1988: +0.33
1989: +0.9
1990: +1.07
1991: +2.47
1992: +2.17
1993: +1.77
1996: +2.63
1997: +2.04
1998: +1.98

And heck, let's do LeBron (I'll just do through Cavs 1st ed):

2002: -0.82
2003: -2.16
*** Add LeBron ***
2004: -0.67
2005: +0.06
2006: +0.41
2007: +1.29
2008: +0.63
2009: +1.97
2010: +0.97

Basically, if you're evaluating Russell's teams by MoV and SRS, they show up decently but way short compared to more modern great teams.

If you're evaluating them in terms of how dominant they were relative to their league, they're extremely competitive. My Top 100 teams project, which mixes the two, has seven Russell teams in the #50-100 range, two Russell teams in the #25-49 range, and none above that. I will also add that for players who were the best player for the most teams in my top 100, the top 14 are:

Bill Russell: 9
LeBron James: 8
Magic Johnson: 7
Larry Bird & Michael Jordan: 6
Tim Duncan: 5 (this got weird because of Kawhi, Robinson and Manu so I had to split a few)
Kareem Abdul-Jabbar: 4
Kevin Durant & Steph Curry: 3.5
Isiah Thomas, Kobe Bryant, Wilt Chamberlain, Shaquille O'Neal, Kawhi Leonard: 3

And again, those are just by Win Shares and VORP (and fudged a little for judgment calls).

Russell deserves some serious love.

Does this adjust for pace?

Ie: losing by 7 with lots of possessions is different than losing by 7 with half as many possessions'

Because the celtics were taken to 7 or 6 by a inferioir regular seaosn srs team just about every postseason.


This does *not* adjust for pace.

I agree that adjusting it for pace would give a better per-possession measure of team dominance.

On the other hand, every ten-point win would have taken the same number of possessions to turn into a loss. Whether you won by 10 in a 120 possession game, or won by 10 in an 80 possession game, you won by about the same number of possessions, if that makes any sense.

Imagine comparing a +10 SRS team that plays at 120, and a +7 SRS team that plays at 60. In terms of per-possession dominance, the +7/60 team is better. And yet, because they play so slow, the amount of uncertainty in their results is increased, and they only end up winning by about 7 possessions instead of 10, like the worse but faster team does.

Basically, faster pace good teams are more likely to win than equally good slow paced teams by increasing the sample size from which the victor is determined.

So even though pace adjusted is a better measure of dominance, it makes the stat less predictive.

This is all theorycrafting - I haven't actually run the numbers. I just spent a week weighing the pros and cons of adjusting my SRS by pace and the above ended up guiding my decision not to.

If pace-adjusted SRS turns out to be more predictive than unadjusted SRS then that's very much worth knowing.

Well the discrepancy I'm seeing is on one hand, the russell celtics basically spent most of the playoffs being challenged by signifcantly worse rs teams to 6 and game 7 outcomes, but on the other hand, they somehow "increased" their point differential margin ogf victory?

You're saying the likelihood of winning is hurt if we pace adjust, but the likelihood of winnign a game seems to have no correlation with the mov you're listing for the celtics.
freethedevil
Head Coach
Posts: 7,262
And1: 3,236
Joined: Dec 09, 2018
         

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#96 » by freethedevil » Thu Nov 26, 2020 8:30 am

sansterre wrote:
MyUniBroDavis wrote:
sansterre wrote:
This does *not* adjust for pace.

I agree that adjusting it for pace would give a better per-possession measure of team dominance.

On the other hand, every ten-point win would have taken the same number of possessions to turn into a loss. Whether you won by 10 in a 120 possession game, or won by 10 in an 80 possession game, you won by about the same number of possessions, if that makes any sense.

Imagine comparing a +10 SRS team that plays at 120, and a +7 SRS team that plays at 60. In terms of per-possession dominance, the +7/60 team is better. And yet, because they play so slow, the amount of uncertainty in their results is increased, and they only end up winning by about 7 possessions instead of 10, like the worse but faster team does.

Basically, faster pace good teams are more likely to win than equally good slow paced teams by increasing the sample size from which the victor is determined.

So even though pace adjusted is a better measure of dominance, it makes the stat less predictive.

This is all theorycrafting - I haven't actually run the numbers. I just spent a week weighing the pros and cons of adjusting my SRS by pace and the above ended up guiding my decision not to.


I mean in theory, it shouldnt be like that though, because variance sizes shoudl decrease with lower possessions too, although i get what youre saying and it might play out that way though


Agreed. A lot of my reasoning came down to "BBR could have pace-adjusted this with almost no effort; that they didn't suggests that there was a reason for not doing so."

I mean they could also have retired useless stats liek PER but they haven't so maybe we shouldn't appeal to their authority.
sansterre
Bench Warmer
Posts: 1,312
And1: 1,816
Joined: Oct 22, 2020

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#97 » by sansterre » Thu Nov 26, 2020 10:42 am

s0ciety wrote:
sansterre wrote:Okay, I can supply data to address the following:

1) The Bill Russell Celtics regular season SRS vs Postseason SRS
2) The Standard Deviations of SRS through NBA History

Since it seems like these are topics of discussion.

1957: RSRS +4.78, PSRS +8.67
1958: RSRS +5.02, PSRS +8.69
1959: RSRS +5.84, PSRS +10.33
1960: RSRS +7.62, PSRS +9.28
1961: RSRS +4.94, PSRS +13.94
1962: RSRS +8.26, PSRS +8.53
1963: RSRS +6.38, PSRS +5.57
1964: RSRS +6.93, PSRS +10.83
1965: RSRS +7.46, PSRS +8.78
1966: RSRS +4.34, PSRS +8.60
1967: RSRS +7.24, PSRS +2.40
1968: RSRS +3.87, PSRS +7.10
1969: RSRS +5.35, PSRS +9.14


Would you mind sharing the way you compute Post-Season SRS?

My model seems to undervalue the Celtics compared to yours :

Code: Select all

Season         sansterre       s0ciety        Difference
1957           8.67            6.70           1.97
1958           8.69            6.14           2.55
1959           10.33           9.66           0.67
1960           9.28            8.03           1.25
1961           13.94           14.06         -0.12
1962           8.53            7.02           1.51
1963           5.57            4.10           1.47
1964           10.83           10.62          0.21
1965           8.78            7.63           1.15
1966           8.60            8.26           0.34
1967           2.40            1.90           0.50
1968           7.10            7.07           0.03
1969           9.14            7.75           1.39


Sorry man, I should have mentioned. I'm using a weird formula that updates regular season SRS as more data comes in. So a team that plays the '01 Lakers in the playoffs isn't credited with playing a decent regular season team, they're credited with playing a decent regular season team that turned into a juggernaut in the playoffs and so is somewhere between. It's a major component in my Top 100 Teams project (and the exact formula is in my master list post).

So your numbers are absolutely correct (I infer) if you're working off of only regular season SRS for establishing strength of schedule.
"If you wish to see the truth, hold no opinions."

"Trust one who seeks the truth. Doubt one who claims to have found the truth."
User avatar
Goudelock
Retired Mod
Retired Mod
Posts: 9,304
And1: 20,915
Joined: Jan 27, 2015
Location: College of Charleston
 

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#98 » by Goudelock » Mon Nov 30, 2020 4:41 pm

Devin Booker wrote:Bro.
70sFan
RealGM
Posts: 29,599
And1: 24,915
Joined: Aug 11, 2015
 

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#99 » by 70sFan » Mon Nov 30, 2020 5:23 pm

Can't wait to see Kareem video, even though I expect that I won't agree with his conclusions.

Walton's video is much better than the first one - this is what I expected from Ben. You can get a very solid idea of how Walton played from that video and what his limitations were. He could go a bit deeper with his scoring game, but overall it's fantastic.

I like that he used a lot of material and that both videos approached almost 20 minutes. I expect a lot of good work from him during this project.
User avatar
WestGOAT
Veteran
Posts: 2,589
And1: 3,497
Joined: Dec 20, 2015

Re: Greatest Peaks series (Thinking Basketball/Ben Taylor) 

Post#100 » by WestGOAT » Mon Nov 30, 2020 5:47 pm

sansterre wrote:
s0ciety wrote:
sansterre wrote:Okay, I can supply data to address the following:

1) The Bill Russell Celtics regular season SRS vs Postseason SRS
2) The Standard Deviations of SRS through NBA History

Since it seems like these are topics of discussion.

1957: RSRS +4.78, PSRS +8.67
1958: RSRS +5.02, PSRS +8.69
1959: RSRS +5.84, PSRS +10.33
1960: RSRS +7.62, PSRS +9.28
1961: RSRS +4.94, PSRS +13.94
1962: RSRS +8.26, PSRS +8.53
1963: RSRS +6.38, PSRS +5.57
1964: RSRS +6.93, PSRS +10.83
1965: RSRS +7.46, PSRS +8.78
1966: RSRS +4.34, PSRS +8.60
1967: RSRS +7.24, PSRS +2.40
1968: RSRS +3.87, PSRS +7.10
1969: RSRS +5.35, PSRS +9.14


Would you mind sharing the way you compute Post-Season SRS?

My model seems to undervalue the Celtics compared to yours :

Code: Select all

Season         sansterre       s0ciety        Difference
1957           8.67            6.70           1.97
1958           8.69            6.14           2.55
1959           10.33           9.66           0.67
1960           9.28            8.03           1.25
1961           13.94           14.06         -0.12
1962           8.53            7.02           1.51
1963           5.57            4.10           1.47
1964           10.83           10.62          0.21
1965           8.78            7.63           1.15
1966           8.60            8.26           0.34
1967           2.40            1.90           0.50
1968           7.10            7.07           0.03
1969           9.14            7.75           1.39


Sorry man, I should have mentioned. I'm using a weird formula that updates regular season SRS as more data comes in. So a team that plays the '01 Lakers in the playoffs isn't credited with playing a decent regular season team, they're credited with playing a decent regular season team that turned into a juggernaut in the playoffs and so is somewhere between. It's a major component in my Top 100 Teams project (and the exact formula is in my master list post).

So your numbers are absolutely correct (I infer) if you're working off of only regular season SRS for establishing strength of schedule.


Out of curiosity, how do you both adjust for strength of schedule in the playoffs? Do you adjust margin-of-victory per game to the opponent's (regular-season vs post-season) SRS? What is the exact equation?

I know you use a factor 7 to weigh play-off games more heavily sansterre, but I was wondering how to do a regular SRS calculation. It seems a bit circular to me intuitively based on what I read from sports-reference:
Everyone else's ratings just changed again, so we've got to run through the same procedure again. And again. And again. And eventually the numbers stop changing.

From: https://www.pro-football-reference.com/blog/index4837.html?p=37
Image
spotted in Bologna

Return to Player Comparisons