Talk to me about the guy we name the important trophy after

[Texas Chuck]

[

Though to counter your point more directly, I don't see the point in comparing two athletes who are from different eras solely based on the merit of what they've done against athletes of their time. At that point, you're not actually comparing them. There needs to be some hypothetical in the equation, some form of imagination, other wise we're just arbitrarily assuming that if someone dominated their era more they're better which is silly.

The problem with this of course is almost everyone automatically assumes the modern athlete is superior. They never give the guys from Russell's era the same benefits of better nutrition, coaching, weight-training, not having to have full-time employment in many cases in the off-season, etc...

They assume a guy like Russell couldnt dominate in the modern era even tho a similar-sized guy in Ben Wallace was a dominant player without many of Russell's gifts. Multiple people itt for example are taking KG ahead of him. KG was better offensively to be sure, but KG isnt a dominant enough offensive player imo to make up for the vast difference in defensive impact. I think part of the reason they do this is because of the bias against the older generation when there is no reason to believe those players would be worse athletically.

But its the different ways people deal with this generational gap and lack of detailed data that was my inspiration for the thread.

[JordansBulls]

I used to have Russell in my top 15, but I honestly think he is top ten material after thinking about it objectively. But there is another problem.

I think a lot of Russell's defensive strengths, and I have a hard time separating it from Kevin Garnett. To those who have Russell over KG (which I'm sure most do), what makes Bill a better defender in KG? Numbers are welcome, but I'd also like to see reasoning for those numbers. What does Russell truly do better? I suppose rebounding would be one, though KG is probably the best defensive rebounder of his era.

It is more of his overall impact as a player. But that has more to do with the type of guys in the league as well. IIRC Wilt averaged something like 34 and 24 vs the rest of the league but 28 and 28 vs Russell. What do you think Wilt would average on someone like Hakeem or Shaq or someone like that?

[Doctor MJ]

1) ElGee gets quote here because b-r doesn't list their attempts at this on the season pages, but they have made attempts and their numbers don't tend to be drastically different. If you're debating about whether these numbers or raw PPG is more useful, no need to debate that. These are much better.

2) Still b-r doesn't list them because we don't have all the data so some estimates are required. b-r as a business is being cautious with what they attach their name too. ElGee as an analyst is simply getting and giving the best estimates available.

Do you know if there's a (relatively) complete listing on their site of their attempts, or just for a few teams/seasons here and there (I'll check the blog archives later today when I have a chance)? I have files for ElGee's and Oliver's attempts already, but B-R's would be a great addition. Both are good, but it's always nice to add more for comparison's sake.

This link is a good starting point:

http://www.basketball-reference.com/blog/?p=7239

[ceiling raiser]

Thanks, the post links to the offense methodology, which notes the following:

Possessions ~ -4.05*Wins - 3.96*Losses + 0.97*FG + 0.75*FGA + 0.70*FTA - 1.37*OReb + 0.53*TotReb + 0.31*Fouls - 0.50*Points +0.19*Opp. Pts

which is all that's needed to create a spreadsheet for ORtg/Drtg/Pace.

[Doctor MJ]

Thanks, the post links to the offense methodology, which notes the following:

Possessions ~ -4.05*Wins - 3.96*Losses + 0.97*FG + 0.75*FGA + 0.70*FTA - 1.37*OReb + 0.53*TotReb + 0.31*Fouls - 0.50*Points +0.19*Opp. Pts

which is all that's needed to create a spreadsheet for ORtg/Drtg/Pace.

My pleasure. I look forward to seeing the things you come up with.

[ceiling raiser]

Thanks, the post links to the offense methodology, which notes the following:

Possessions ~ -4.05*Wins - 3.96*Losses + 0.97*FG + 0.75*FGA + 0.70*FTA - 1.37*OReb + 0.53*TotReb + 0.31*Fouls - 0.50*Points +0.19*Opp. Pts

which is all that's needed to create a spreadsheet for ORtg/Drtg/Pace.

My pleasure. I look forward to seeing the things you come up with.

When I get home I'll plug the formula into my working spreadsheet from the Oliver calculations:

viewtopic.php?f=64&t=1239896#p35420740

though it seems I'll need to add wins and losses as columns. I expect B-R's attempt to produce similar results to ElGee/Oliver, but there might be some interesting cells with markedly different values.

[ElGee]

1) ElGee gets quote here because b-r doesn't list their attempts at this on the season pages, but they have made attempts and their numbers don't tend to be drastically different. If you're debating about whether these numbers or raw PPG is more useful, no need to debate that. These are much better.

2) Still b-r doesn't list them because we don't have all the data so some estimates are required. b-r as a business is being cautious with what they attach their name too. ElGee as an analyst is simply getting and giving the best estimates available.

Do you know if there's a (relatively) complete listing on their site of their attempts, or just for a few teams/seasons here and there (I'll check the blog archives later today when I have a chance)? I have files for ElGee's and Oliver's attempts already, but B-R's would be a great addition. Both are good, but it's always nice to add more for comparison's sake.

I have Neil's regression method in a spreadsheet so I have all his numbers if anyone ever wants them. Oliver uses a method as well and I've replicated it recently. Basically here's the short of it:

-my method pretty strongly outperforms either regression method when tested upon 1974-1976 data. Now, that is not a holy grail of testing, nor does it tell us much the farther back in time we go, but just consider the origin of the estimates and you can decide which one you like better:

**The regression methods are looking at (essentially low-grade) correlations between the few box score stats kept in the past and their "effect" on pace. This is counterintuitive to me since the correlations are so low (what does good rebounding have to do with pace? Fouling might have something to do with it, etc.) which is why I developed the Complex Method

**The Complex Method starts by calculating pace from exactly what we know and then using historical estimations or historical trends in rebounding (for eg) to estimate how a team would perform filling in those areas. (We really need to fill in TOV and OREB primarily.) The only thing we can really never be too sure of is turnovers, which has no causal or correlative relationship to any of the early box score metrics, and that's the true "guess" (it uses a constant that changes and tapers off as we move away fro m1974) in the equation. Still, there is a margin of error in any of these estimations, and I used to have some greater detail on that and some hardcore detail on the method on my blog.

Basically, when you use this method on data we already have, you will get more accurate that basic regression will ever get you.

In short: I trust the Complex 1971-1973 estimations A LOT. I trust the late 60's estimations a little, and I find the earlier stuff helpful but fuzzy -- I always think in terms of 2-3 point swings from offense to defense when we are in the heart of the 60's. That said,

(1) O/DRtg pace splits today still have 2-3 points shifts that need to be accounted for based on strategy and
(2) anyone who runs all 3 methods will notice the general trends are all the same.

Boston has an outlying defense. Wilt's teams go up and down and are never impressive on offense until 67. Oscar and West play on some monster offenses. Kareem/Oscar Bucks look great.

[ElGee]

Im curious now how to hear how some of you deal with Bill Russell. Ive been involved in a nice discussion in the 2000s thread regarding KG and how some good posters analyze players.

So what do you make of a guy whose strongest cases for being in the top 5,10, maybe lower for some you? are defense and winning. The defense is harder to evaluate because we dont have nearly the data we have for the modern era and winning is obviously more about the team and some solid posters seem to completely seperate that out.

So I guess my questios is how highly do you think of Russell the player and what do you base your opinion on?

drza, ElGee, and Doctor MJ are three guys Im really interested in hearing from since they take a significantly different approach from me, but Im also interested in hearing anyone and everyone elses thoughts as well.

I think this is my first OP on here, lol

I care about how much the player impacts the game (changes the odds of winning) within his era. I'll leave the cross-era comparisons to tsherkin.

In Russell's case, I grew up in Boston and heard tale after tale about his dominance and the 11 championships. I also heard about the "balanced Celtics" and all the Hall of Famers and at a young age, when Michael Jordan was coming to prominence, realized it was nonsense to just deify Russell because of his great teams. Russell was overrated to me and Wilt was the one getting the shaft -- how much can one man do?

I held this belief for a while, pretty much until a few summers ago when I was able to use the Internet to watch the handful of available Russell games (most referenced earlier in this thread). Now I had new information and it was pretty shocking to see -- Russell was a freak athlete. He was quick as all hell, fast and had crazy hops. By this point, I had also covered enough basketball games to know the angle of the camera makes a subtle difference in how players look -- the low-definition of the 60's films is not as impressive to our minds. But man, chase-down blocks from halfcourt...can you imagine how he would look in slow-mow hi-def?

I digress -- by then I was on realgm and needed to expand my historical understanding of the game. I could name every MVP off the top of my head but I deliberately limited my historical study to the 3-point era because (a) it was too hard to get info earlier in the old days and (b) the 1970's were just about the least sexy thing in basketball to me, save for Blazer Mania. So I started digging on the 60's and realized people were shooting in the dark back then -- they didn't account for pace, they didn't track anything, and they espoused what are strategically pretty crazy ideas. That was the genesis of the Pace Estimation. (Which in honesty was borne out of an afternoon of playing with TS% and asking "why can't we know more?")

And what happened? Russell's Celtics made me think my calculations were wrong. I thought the thing was broken. The team was entirely defense-oriented, and all the tales of these "HOF" supporting cast players looked borderline silly, especially when you realize that in the 9-team league, a club with Jones/Hondo on the wing, a shooter and some decent defenders (KC Jones and my personal underrated defender of the 60's, Satch Sanders) looks downright average. Then I dug way deeper -- it turns out in 1963 when Cousy retired everyone wrote Boston off (!) and labeled them the "old Celtics" (5 more titles to go!). It also turns out that in all the games Russell missed in the decade, Boston worsens by almost exactly their defensive edge over the league. And it turns out, in Second Wind (which I then had to read) that Russell considered 64 Boston's best defensive team...which it is by most estimations.

So basically I realized without the 3-point line, that this 6-10 athletic freak (who had a Global understand of basketball as a game of efficiency when people 30 years later still viewed it as a 1-on-1 game of volume) was imparting basically ALL of his value of one side of the ball, and that value created a margin of error so large that Boston won 11 titles. I consider his defensive value so large it rivals peak Wilt's offense and defense put together, and Wilt's an all-time level defender. Do I care about the 11 titles? Not really -- they got lucky on a few occasions. 6 would have been pretty unlucky. 8 probably expected. But their margin of error was so large that they were going to be on the north side of close series more often than not because they had to underperform while others had to over-perform just to make it close. Their SRS (team strength) was lapping the field almost entirely from Russell's defense.

Cousy leaves? Team improves. Jones has limited minutes then ages? Doesn't matter. Havlicek misses games or plays less? Doesn't matter. Russell misses games? Team falls apart. Russell retires? Team, and specifically the defense, falls apart.

That's Bill Russel to me in broad strokes.

[tsherkin]

[quote=
I care about how much the player impacts the game (changes the odds of winning) within his era. I'll leave the cross-era comparisons to tsherkin.

Will you, now? And why is that?

[ThaRegul8r]

I started digging on the 60's and realized people were shooting in the dark back then -- they didn't account for pace, they didn't track anything, and they espoused what are strategically pretty crazy ideas. That was the genesis of the Pace Estimation. (Which in honesty was borne out of an afternoon of playing with TS% and asking "why can't we know more?")

And what happened? Russell's Celtics made me think my calculations were wrong. I thought the thing was broken.

I used to be a member of APBR, and Dean Oliver said the same thing in 2002 when he was talking about the numbers he came up with for the 60's Celtics defense a couple of years before he published Basketball on Paper. They were, quote, "so good that I have been having trouble believing them."

[Gideon]

To me, impact is what it's all about... both impact within era and also hypothetical impact in other situations (different eras, different teammates, different systems, different opponents, etc.). However, when I judge players from previous eras, I feel it's only fair to imagine they would have access to modern nutrition and training -- taking somebody from 1965 straight from the time machine to the arena would probably lead to all sorts of problems, but give a great player the courtesy of at least a supervised offseason and a training camp and I'm sure most would adjust quite well.

Using that criteria, I consider Russell the GOAT. His impact was unparalleled. The thing about the "RINGZZ!" is not that they matter in a literal sense by themselves, but that Russell's impact actually was so great that he really did lead his team to 11 titles while the entire roster changed around him. Even if his "real" title average would only have been 8 or 9, that's still unique in NBA history. Rings might not matter, but helping your team to win is everything (regardless of actual outcome); that's a rather subtle distinction, but I really believe it... and I also really believe Russell helped his team win more than anybody else ever has.

The advanced defensive stats from the 60s that are starting to emerge -- (as some posters who are much more knowledgeable than I am about these stats and their intricacies have referenced above) -- provide more evidence as to Russell's amazing impact. Just watching him has also been "evidence" for me... I feel like I've never seen anyone affect games more than Russell does.

I also disagree with the common argument that the center position (and, thus, Russell's competition) was much weaker than it is nowadays. A quick look at rosters in the mid-to-late 60s will prove otherwise, especially when you consider that the smaller # of teams meant Russell matched up with guys like Wilt, Thurmond, and Reed 7-10 times a year.

Russell is the #1 player on my all-time list... I truly think he both literally had the biggest impact in league history and also would have had the biggest impact in the majority of circumstances compared to any other player in league history.

[ceiling raiser]

1) ElGee gets quote here because b-r doesn't list their attempts at this on the season pages, but they have made attempts and their numbers don't tend to be drastically different. If you're debating about whether these numbers or raw PPG is more useful, no need to debate that. These are much better.

2) Still b-r doesn't list them because we don't have all the data so some estimates are required. b-r as a business is being cautious with what they attach their name too. ElGee as an analyst is simply getting and giving the best estimates available.

Do you know if there's a (relatively) complete listing on their site of their attempts, or just for a few teams/seasons here and there (I'll check the blog archives later today when I have a chance)? I have files for ElGee's and Oliver's attempts already, but B-R's would be a great addition. Both are good, but it's always nice to add more for comparison's sake.

I have Neil's regression method in a spreadsheet so I have all his numbers if anyone ever wants them. Oliver uses a method as well and I've replicated it recently. Basically here's the short of it:

-my method pretty strongly outperforms either regression method when tested upon 1974-1976 data. Now, that is not a holy grail of testing, nor does it tell us much the farther back in time we go, but just consider the origin of the estimates and you can decide which one you like better:

**The regression methods are looking at (essentially low-grade) correlations between the few box score stats kept in the past and their "effect" on pace. This is counterintuitive to me since the correlations are so low (what does good rebounding have to do with pace? Fouling might have something to do with it, etc.) which is why I developed the Complex Method

**The Complex Method starts by calculating pace from exactly what we know and then using historical estimations or historical trends in rebounding (for eg) to estimate how a team would perform filling in those areas. (We really need to fill in TOV and OREB primarily.) The only thing we can really never be too sure of is turnovers, which has no causal or correlative relationship to any of the early box score metrics, and that's the true "guess" (it uses a constant that changes and tapers off as we move away fro m1974) in the equation. Still, there is a margin of error in any of these estimations, and I used to have some greater detail on that and some hardcore detail on the method on my blog.

Basically, when you use this method on data we already have, you will get more accurate that basic regression will ever get you.

In short: I trust the Complex 1971-1973 estimations A LOT. I trust the late 60's estimations a little, and I find the earlier stuff helpful but fuzzy -- I always think in terms of 2-3 point swings from offense to defense when we are in the heart of the 60's. That said,

(1) O/DRtg pace splits today still have 2-3 points shifts that need to be accounted for based on strategy and
(2) anyone who runs all 3 methods will notice the general trends are all the same.

Boston has an outlying defense. Wilt's teams go up and down and are never impressive on offense until 67. Oscar and West play on some monster offenses. Kareem/Oscar Bucks look great.

Thanks for the response. I've been a fan of your work since I first noticed your site, and use your estimates in most of my spreadsheets:

https://docs.google.com/spreadsheet/lv? ... rpage=1309

I probably prefer your calculations to Oliver's (as they seem valid & reliable), and will to B-R's as well, but just wanted to add more data for completion's sake.

On Topic: Dr. MJ's and ElGee's posts are both excellent, and I have to echo the recommendations to read Second Wind (Go Up For Glory is outstanding as well).

[Johnlac1]

When people put down Russell because of his mediocre scoring skills, they fail to realize he won 11 titles in 13 years. He didn't see the need to improve his offense. I doubt he picked up a basketball in the off-season much after he turned pro. Thrust him into today's game where he'd need to have better scoring skills, and I have no doubt he'd improve. He didn't look bad shooting a jump shot, he just didn't see the need to spend time grooving something he thought he didn't have to groove. But even given his lack of off. scoring skills, he'd still be an excellent def. center today. Probably the best. Nerlins Noel is the first center I've seen who reminds me of Russell's def. skills. But Russell was also one of the smartest players to play the game. He wasn't just a shot-blocker.

[lorak]

I care about how much the player impacts the game (changes the odds of winning) within his era.

So for you Russell>Jordan?

And also: how good was Mikan?

[penbeast0]

Yes, Russell>Jordan; his GOAT level defense and rebounding are superior to MJ's GOAT level scoring.

Mikan was very good but not as dominant or for as long despite having the best team around him in his day (All-Stars Slater Martin, Vern Mikkelson, and Jim Pollard . . . of whom Martin and Mikkelson reportedly played quite well after Mikan's retirement). He also clearly suffered during his career when they changed to the 24 second rule while Russell adapted from the 50's era basketball of his first couple of years to the late 60s without a hitch. Mikan also didn't dominate the Globetrotters to the same degree which also lessens his domination of his era. He has a real case for being a top 20 player of all time, possibly top 15, probably not top 10.

[ceiling raiser]

Yes, Russell>Jordan; his GOAT level defense and rebounding are superior to MJ's GOAT level scoring.

Mikan was very good but not as dominant or for as long despite having the best team around him in his day (All-Stars Slater Martin, Vern Mikkelson, and Jim Pollard . . . of whom Martin and Mikkelson reportedly played quite well after Mikan's retirement). He also clearly suffered during his career when they changed to the 24 second rule while Russell adapted from the 50's era basketball of his first couple of years to the late 60s without a hitch. Mikan also didn't dominate the Globetrotters to the same degree which also lessens his domination of his era. He has a real case for being a top 20 player of all time, possibly top 15, probably not top 10.

OT: Kurland (who dominated with the Phillips 66ers to a comparable extent as Mikan did to the NBL/BAA/NBA) opting for the AAU probably robbed the pro cage scene of what could've been an amazing early rivalry.

EDIT: I can't hold it against him though, with the state of the league(s) back then, he opted for financial security and seemingly was set for life. A recent article paints him as having no regrets.

[penbeast0]

Yeah, Groza opted out too by getting caught accepting bribes. He was the other guy who looked like a serious competitor for Mikan though a bit undersized but only had 2 pro years before being banned for life. Neil Johnston came in and quickly put up big offensive numbers but had a rep for weak defense which was even more important then.

[lorak]

He also clearly suffered during his career when they changed to the 24 second rule

He was injured that ONE year and after full year without basketball. At the time (no trainings like today) it was lethal for basketball ability, especially when you were so big.

Mikan also didn't dominate the Globetrotters to the same degree which also lessens his domination of his era.

He once even scored like +40 points on them, so big, big deal at the time, because of no 24s clock

Besides we don't know how Russell, or any other stars from different eras, would perform vs equivalent of globetrotters (so the best black players of that player's era), so it's not fair to use it, because we don't have point of reference (or benchmark? which word would be proper in this context?) to make that kind of comparison.

[penbeast0]

Both admittedly lesser points; didn't realize the injury was particularly serious which indeed could be the reason for his failure to cope with the changes. I also don't know how dominant he was before the BAA/NBA so his career length could be longer than I recognize. I've read a couple of biographies on him but neither were really focused on the basketball as much as the personal story.

[DEEP3CL]

Let's say we can freeze Russell's skills for each era starting from the 70's thru the 90's. What type of player would he have been in each era ? His skill level would fit 2 of them fine, the 90's is the one that concerns me. The game had gotten much rougher. I don't doubt he could have played and did well, but in that eastern conference was no joke in the 90's. Teams like the Pistons and Knicks could have neutralized him quite well.