Page 8 of 14
Re: Vote for the all-time top 15 players
Posted: Fri Nov 08, 2013 2:25 am
by jbrocato23
permaximum wrote:Yeah I remember 2003 playoffs. David Robinson, Tony Parker, Stephen Jackson, Ginobili, Bowen and Steve Kerr's 3s... That prime Duncan deserves a spot but Iverson with Snow, Mckie and Mutombo doesn't deserve a spot? Yeah.... 2001 lakers would have swept that 2003 spurs. Shaq was unstoppable and I don't even talk about Kobe. Shaq is a good example of how a top-15 big at his prime plays like.
You realize Duncan's Spurs beat prime Shaq Kobe Lakers that year in the playoffs right? And Duncan averaged 28-11.8-4.8 in that series. Robinson was completely washed up at that point (he averaged 8.5ppg on the season and 7.8 in the playoffs), and Parker (20yo) and Ginobili weren't even close to their primes. I actually think Duncan's supporting cast was comparable if not worse than Iverson's in '01.
Funny note: RAPM is the best metric in theory and this is what we get. LeBron's RAPM doubles Jordan's... There are a lot of players who have better RAPM values than Jordan's. His highest is 5.5. Lebron's 11.9.
Are you getting those values from JE's site? Those pre-2001 values are from a completely different metric that isn't based on pbp data and you can't compare them to post 2001 values.
Re: Vote for the all-time top 15 players
Posted: Fri Nov 08, 2013 8:14 am
by permaximum
Shaq was injured most of that season and his peak years were 1999-2000-2001. If I remember right Robinson made critical moves in playoffs. I wasn't saying his supporting cast was at their prime. It was just his supporting cast was excellent. As for Snow, Mckie and Mutombo, I don't think their primes were even noticable for most of the people.
Yes, I got those values from JE's site. No, they're not a completely different metric and they're surely comparable. Let's pretend that they're not comparable to 2000+ years. What about Jordan's values such as +3, +4, +5 when there are +7 +8 +9's in those seasons with Jordan?
JE can enlighten us on this one but he wouldn't put something that not comparable to anything or something that even not comparable within, online.
Re: Vote for the all-time top 15 players
Posted: Fri Nov 08, 2013 11:27 am
by Mike G
schtevie wrote:
(1) Coincident with Russell's arrival and departure, the Celtics took a discrete and large step up and down in defensive performance.
(2) Team defensive performance approximates the typical arc of a player aging curve.
(3) Team offensive performance throughout the dynasty was strictly mediocre and showed no obvious correlation with the passage of time.
(4) Bill Russell was the only constant of the era..
Agreed on all counts. It's a very cozy scenario. However, it seems to imply that the Other Celtics were an average group; and I still contend that, especially in their heyday, they were notably above avg.
Let's look at one of those years picked at random, 1963. I've sorted players by team and ranked them by their
sqrt(PER*WS). Sometimes one stat covers something the other doesn't; and WS is dependent on minutes played.
Supporting players, abbreviated:
Code: Select all
PWS #3 guy tm PER WS ws/48
11.4 Dick Barnett LAL 16.3 8.0 .150
11.3 Tom Heinsohn* BOS 18.7 6.8 .163
10.1 Wayne Embry CIN 18.2 5.6 .108
8.9 Larry Costello SYR 13.8 5.7 .132
8.6 Zelmo Beaty STL 14.3 5.2 .131
7.5 Dave Budd NYK 13.8 4.1 .113
7.3 Ray Scott DET 15.1 3.5 .066
6.9 Bill McGill CHZ 21.5 2.2 .176
6.0 Al Attles SFW 10.6 3.4 .086
PWS #5 guy tm PER WS ws/48
9.3 Bob Cousy* BOS 16.6 5.2 .126
7.9 Paul Neumann SYR 13.4 4.6 .139
7.0 Lenny Wilkens* STL 12.1 4.0 .075
6.7 Adrian Smith CIN 13.1 3.4 .107
6.4 Frank Selvy LAL 11.5 3.6 .074
6.2 Bob Ferry DET 12.6 3.1 .061
5.4 Don Nelson CHZ 13.8 2.1 .094
3.5 Donnie Butcher NYK 11.4 1.1 .043
3.5 George Lee SFW 10.1 1.2 .048
PWS #7 guy tm PER WS ws/48
7.4 Len Chappell SYR 17.2 3.2 .123
6.8 Frank Ramsey* BOS 12.8 3.6 .113
6.2 Mike Farmer STL 12.0 3.2 .089
5.1 Leroy Ellis LAL 11.4 2.3 .067
4.7 Hub Reed CIN 11.2 2.0 .074
4.1 Willie Jones DET 13.1 1.3 .043
4.0 Larry Staverman CHZ 13.5 1.2 .093
1.0 Howie Montgomery SFW 9.9 .1 .018
.0 Gene Conley NYK 10.7 -.1 -.002
PWS #9 guy tm PER WS ws/48
5.8 Dolph Schayes* SYR 14.2 2.4 .081
5.1 Clyde Lovellette BOS 16.2 1.6 .139
2.6 Phil Jordon STL 9.5 .7 .024
2.1 HotRod Hundley LAL 11.2 .4 .027
.0 Maury King CHZ 8.3 -.1 -.005
.0 Dave Piontek CIN 6.4 -.3 -.031
.0 Johnny Egan DET 9.8 .0 .001
.0 Paul Hogue NYK 7.4 -1.0 -.037
.0 Wayne Hightower SFW 7.8 -.8 -.029
Zeroes are given, when the PWS is negative: Sub-replacement player is one interpretation.
Syracuse was even deeper than Boston. Other than that, no team is close after #7.
Wilt is especially bereft of support, out on the west coast. (This was the year with no team in Philly).
The Celts beat the Nats at players #1 thru #6 (Russell v Kerr, thru Sanders v Shaffer), so held a 3 PPG advantage in SRS. And this was a 'down year' for Boston.
I realize it's possible and even likely that Russell is undervalued in both PER and WS -- some credit thus flowing to his teammates. And he won MVP for several years, despite being overshadowed by others, statistically. And
still, other Celtics regularly were AllStar and All-League.
Re: Vote for the all-time top 15 players
Posted: Fri Nov 08, 2013 2:27 pm
by schtevie
Mike G wrote:schtevie wrote:
(1) Coincident with Russell's arrival and departure, the Celtics took a discrete and large step up and down in defensive performance.
(2) Team defensive performance approximates the typical arc of a player aging curve.
(3) Team offensive performance throughout the dynasty was strictly mediocre and showed no obvious correlation with the passage of time.
(4) Bill Russell was the only constant of the era..
Agreed on all counts. It's a very cozy scenario. However, it seems to imply that the Other Celtics were an average group; and I still contend that, especially in their heyday, they were notably above avg.
I agree it is a cozy scenario because...well...it's the story most consistent with the facts. As for your reluctance to acknowledge that the Other Celtics (as a collective) group were an average group, I just don't get it. If you accept my estimates of possessions (the basis of points (1), (2), and (3) above, and if you don't, I would encourage you to create a preferred estimate to convince yourself that the underlying story won't change) it is extraordinarily unlikely that they weren't an average group. Why?
On offense, because their offensive performance was strictly average over the time period in question. If they weren't an average group it would only be because Bill Russell was a below-average offensive player. And he may have been. But if you look at his offensive box score stats and what we can reasonably believe were his offensive rebounding stats, it isn't reasonable to believe that he was such a drag.
And then on defense, the story you are obliged to tell (because we don't have any valuable statistical evidence besides the estimated DRtg) is in total contradiction to accepted narrative: that Russell actually wasn't all that great of a defensive player. And then it just so happened that a changing group of above-average defensive Other Celtics just happened to be his teammates over 13 years and not the year before or after.
Mike G wrote:Let's look at one of those years picked at random, 1963. I've sorted players by team and ranked them by their
sqrt(PER*WS). Sometimes one stat covers something the other doesn't; and WS is dependent on minutes played.
Supporting players, abbreviated:
Code: Select all
PWS #3 guy tm PER WS ws/48
11.4 Dick Barnett LAL 16.3 8.0 .150
11.3 Tom Heinsohn* BOS 18.7 6.8 .163
10.1 Wayne Embry CIN 18.2 5.6 .108
8.9 Larry Costello SYR 13.8 5.7 .132
8.6 Zelmo Beaty STL 14.3 5.2 .131
7.5 Dave Budd NYK 13.8 4.1 .113
7.3 Ray Scott DET 15.1 3.5 .066
6.9 Bill McGill CHZ 21.5 2.2 .176
6.0 Al Attles SFW 10.6 3.4 .086
PWS #5 guy tm PER WS ws/48
9.3 Bob Cousy* BOS 16.6 5.2 .126
7.9 Paul Neumann SYR 13.4 4.6 .139
7.0 Lenny Wilkens* STL 12.1 4.0 .075
6.7 Adrian Smith CIN 13.1 3.4 .107
6.4 Frank Selvy LAL 11.5 3.6 .074
6.2 Bob Ferry DET 12.6 3.1 .061
5.4 Don Nelson CHZ 13.8 2.1 .094
3.5 Donnie Butcher NYK 11.4 1.1 .043
3.5 George Lee SFW 10.1 1.2 .048
PWS #7 guy tm PER WS ws/48
7.4 Len Chappell SYR 17.2 3.2 .123
6.8 Frank Ramsey* BOS 12.8 3.6 .113
6.2 Mike Farmer STL 12.0 3.2 .089
5.1 Leroy Ellis LAL 11.4 2.3 .067
4.7 Hub Reed CIN 11.2 2.0 .074
4.1 Willie Jones DET 13.1 1.3 .043
4.0 Larry Staverman CHZ 13.5 1.2 .093
1.0 Howie Montgomery SFW 9.9 .1 .018
.0 Gene Conley NYK 10.7 -.1 -.002
PWS #9 guy tm PER WS ws/48
5.8 Dolph Schayes* SYR 14.2 2.4 .081
5.1 Clyde Lovellette BOS 16.2 1.6 .139
2.6 Phil Jordon STL 9.5 .7 .024
2.1 HotRod Hundley LAL 11.2 .4 .027
.0 Maury King CHZ 8.3 -.1 -.005
.0 Dave Piontek CIN 6.4 -.3 -.031
.0 Johnny Egan DET 9.8 .0 .001
.0 Paul Hogue NYK 7.4 -1.0 -.037
.0 Wayne Hightower SFW 7.8 -.8 -.029
Zeroes are given, when the PWS is negative: Sub-replacement player is one interpretation.
Syracuse was even deeper than Boston. Other than that, no team is close after #7.
Wilt is especially bereft of support, out on the west coast. (This was the year with no team in Philly).
The Celts beat the Nats at players #1 thru #6 (Russell v Kerr, thru Sanders v Shaffer), so held a 3 PPG advantage in SRS. And this was a 'down year' for Boston.
I realize it's possible and even likely that Russell is undervalued in both PER and WS -- some credit thus flowing to his teammates. And he won MVP for several years, despite being overshadowed by others, statistically. And
still, other Celtics regularly were AllStar and All-League.
Repeating a point just made, if it is in fact true (and it is) that the Celtics overall had an average offensive efficiency over the years in question (and actually in 1963, by my estimate they were below average, -1.5, compared to the NBA average, their fourth worst year of the thirteen) how can they as a group be above average by another statistical measure? They simply can't. There is an adding-up constraint that must be observed. So, what this implies is that your listed measures are, er, imperfect.
Again, this does not mean that there weren't one or more individual standouts, even, very likely, HoF-worthy (in the meritocratic sense) teammates. But, no, they cannot have been collectively above average (in a +/- sense) unless you are prepared to say that Russell wasn't what we think he was on defense and that he was a distinctly bad offensive player. That's just the way it is.
P.S. And just to be specific about the offense in the "juggernaut" years of the Russell dynasty (the six years/championships from 1960 to 1965) by my estimate, the Celtics offense averaged 1.1 fewer points per 100 possessions than the average NBA team over that period (year by year: +1.7, -2.2, +0.1, -1.5, -3.4, and -1.3), what was a relatively worse performance that their "non-juggernaut" years. Again, if you want to fiddle with the possession formula, you might be able to make the offense a little bit better, but you won't be able to make it above-average.
Re: Vote for the all-time top 15 players
Posted: Fri Nov 08, 2013 3:48 pm
by Mike G
I don't actually give a whit about estimating possessions. Nor do I care to divide talents into 'offensive' vs 'defensive'. Too many plays are both.
If there's a mad scramble for the ball, everybody gets two hands on it a time or two, one team or the other eventually gets it -- They may say it was possessed and surrendered 5 times by each team, or no times. Adding another 5 possessions (and/or 5 turnovers) doesn't really change either team's actual performance on either offense or defense.
If a team has a below-avg FG% and TS%, yet outscores the league by 5 to 10 points per game, there are 3 possible sources for their scoring advantage: Good shot defense (even worse Opp FG%), turnover advantage, and rebounding advantage.
If you have an advantage of 5 in TO and another 5 in Reb, you have essentially 10 free possessions.
Stuff that difference in the denominator, and you could be either the best or worst offense in the league, in getting your 115 points.
We can assume Russell was quite the defensive marvel, relative to those who came before him. I don't recall anyone dwelling on the defensive prowess of Mikan, Foust, Johnston.
Rather like Oscar was to offense, among guards. Revolutionary. Previously unthinkable. But there have been several comparable in subsequent decades.
Don't take just my word on those Other Celtics. Russell was great, and Others were merely allstar and all-league. No other team was close to having as many in the interval. Without pbp and comprehensive video evidence, that (and their stats) is what we have to go by.
You do know that in Russell's rookie season, Tom Heinsohn was ROY and an allstar.
After Russell's finale, Sam Jones (1 year removed from allstar, 2 from all-NBA) also quit.
Re: Vote for the all-time top 15 players
Posted: Sat Nov 09, 2013 7:01 am
by Bobbofitos
I think I'm going to only vote for players who played in the past 20 years henceforth to counter the old timer bias. It's beyond absurd.
Re: Vote for the all-time top 15 players
Posted: Sat Nov 09, 2013 9:08 am
by permaximum
I'm surprised nobody voted for Nowitzki.
Oddly, there were votes for Wade and Paul for top 5. I think that voter hasn't voted yet.
BTW, I would give more value to simple box score stats than PER, WS, WP...
Re: Vote for the all-time top 15 players
Posted: Sat Nov 09, 2013 11:06 am
by Mike G
When we're voting for 5 or 10 or so, and 50 are offered, the idea is that there won't be any additional write-in votes. After all, we're now working from a list that 'we' voted for.
The 'outlier' votes are useful for breaking ties. It's not startling that people have a preference or favorite they just can't seem to shake. Like a favorite old song that was playing when you had a great time one night. Nobody can tell you it's not that great a song, or that better stuff has come along since.
It's a good thing to have strong faith in your own perceptions. It's what makes us unique and entertaining and likeable. We can't just be automatons in thinking, any more than players can always be at their most efficient.
You don't have to try and balance the field by limiting your votes to only one era or another. Then you've subbed out your whole starting lineup of rational decision-making for a bunch of hothead emotions 'with upside'. Instead, maybe try subbing in one or two atypical decisions -- Undecided between Baylor and Nowitzki? Go with your gut feeling about their eras.
I can't tell that there's an overall bias for or against any era. One commenter swears Havlicek said he'd not have been good enough to even play now. Another swears Russell would dominate now.
I don't happen to believe either one is more likely than the other.
There's even more intense partisanship out there, that we haven't seen here. On the old APBR board, we'd see things like -- some guy who couldn't get playing time in college in the '60s would certainly be an NBA starter today. Someone else said Mikan was about = Greg Kite (classic '80s stiff).
We're a pretty reasonable bunch here, in general.
Re: Vote for the all-time top 15 players
Posted: Sat Nov 09, 2013 11:13 am
by Mike G
permaximum wrote:I would give more value to simple box score stats than PER, WS, WP...
Agree with this, to an extent. Averaging 30 in 121-120 games is like averaging 25 in 101-100 games, though, isn't it?
Scaling by a simple ratio like 5/6 is not that complex a procedure. It's like noting that someone scored 30, but it was in a triple-overtime game and 60 minutes.
In that 191-188 game (Det-Den sometime in the '80s), if you didn't get 35 points per position, you were out of the running.
In that Nuggets season when they allowed 131 ppg, they had to have 26 pts per position to be competitive. Contrast to a playoff series where the avg score is 82-78.
Re: Vote for the all-time top 15 players
Posted: Sat Nov 09, 2013 3:01 pm
by schtevie
Mike G wrote:I don't actually give a whit about estimating possessions.
Urk!
Mike G wrote:Nor do I care to divide talents into 'offensive' vs 'defensive'.
Ack!
Mike G wrote:Too many plays are both.
What the hey? This is awesome (and not in a good way). I am prepared to believe that this diversion we engage in is not terribly progressive - when push comes to shove, sentiment and prior belief quite evidently are a stronger force than rationality - but retrogressive?
Mike G wrote:If there's a mad scramble for the ball, everybody gets two hands on it a time or two, one team or the other eventually gets it -- They may say it was possessed and surrendered 5 times by each team, or no times. Adding another 5 possessions (and/or 5 turnovers) doesn't really change either team's actual performance on either offense or defense.
Is it the case that over the course of a season there are some marginal calls in scorekeeping, where given a mad scramble for the ball, there are judgements made about whether possession was actually transferred and how many times? Sure. Is there any reason to believe that the judgement calls are typically or systematically wrong? Please. Is there any underlying conceptual controversy? Absolutely not. Whenever a player loses possession on such occasions, it is a turnover, a possession ended without a point scored, thereby implying a diminution of average offensive efficiency. And yes, it is most certainly the appropriate designation because maintaining possession of the ball is the precondition for scoring. But more importantly for, ya know, the actual discussion at hand, is there any plausible way to believe that these events are empirically relevant over the course of an entire season? No. Way. And just to get more specific still, the numbers I was using to estimate "TO - OR" hence possessions were representative not of just one season but many many.
Mike G wrote:If a team has a below-avg FG% and TS%, yet outscores the league by 5 to 10 points per game, there are 3 possible sources for their scoring advantage: Good shot defense (even worse Opp FG%), turnover advantage, and rebounding advantage. If you have an advantage of 5 in TO and another 5 in Reb, you have essentially 10 free possessions.
Stuff that difference in the denominator, and you could be either the best or worst offense in the league, in getting your 115 points.
Volte face? I thought we weren't interested in possessions, but now we reintroduce all the components that determine same. Alright. Decomposition it is. So yes, suppose a team has a below average FG% and TS%, say, like the 1963 Celtics. No wait, they weren't just below average; they were
the worst team in the NBA on these measures. But we proceed. How to possibly rehabilitate such an offense? Turnovers (they might have committed fewer of them) or so-called Offensive Rebounds (more plays per possession). And so we return to a possession-based perspective. The average is what it was (plus or minus just a little bit).
Mike G wrote:Don't take just my word on those Other Celtics. Russell was great, and Others were merely allstar and all-league. No other team was close to having as many in the interval. Without pbp and comprehensive video evidence, that (and their stats) is what we have to go by.
And herein lies the error. Accolades dispensed by sportswriters or whomever offer no relevant, primary evidence to the basic question at hand. Was the Celtics offense on average average (or below)? Yes, there is no reasonable question of this, and it doesn't require pbp or comprehensive video evidence just well-established basic accounting. Was the Celtics defense on average average? No. There is no reasonable question of this. The only question is partitioning such success between Russell and the Other.
Mike G wrote:You do know that in Russell's rookie season, Tom Heinsohn was ROY and an allstar.
After Russell's finale, Sam Jones (1 year removed from allstar, 2 from all-NBA) also quit.
I do believe I already made ironic mention of Heinshohn's rookie season, in the sense that his (and Frank Ramsey's) arrival coincided with the apparent deterioration in the offense. And as for Sam Jones, my understanding is that he was a very good defensive guard. But based upon all else we (should) know about the structural aspects of defense in the NBA, the drop in the retirement year could only be credited in lesser part to his departure.
Re: Vote for the all-time top 15 players
Posted: Sat Nov 09, 2013 8:11 pm
by jbrocato23
I'm curious where people have David Robinson. The box score says he's one of the greatest ever (at least per min or per poss) and regression based plus minus missed most of his career but had him as elite defensively in his twilight years. But his playoff failures are well noted, and this along with his not-so-stellar longevity seem to drop him quite a bit on most lists.
Re: Vote for the all-time top 15 players
Posted: Sat Nov 09, 2013 8:54 pm
by bchaikin
I'm curious where people have David Robinson.
from the ages of 24-30 not many were better. like sidney moncrief a great player whose career was stunted by injury. his peak value is among the very best...
imho in peak value a top 5 C, with chamberlain, russell, jabbar, and olajuwon, just a bit better than shaq, malone, mikan, johnston, gilmore...
Re: Vote for the all-time top 15 players
Posted: Sat Nov 09, 2013 10:41 pm
by Bobbofitos
Hakeem > Shaq for peak, what? Shaq's peak was unrivaled. The watermark was injuries, not per minute domination.
Re: Vote for the all-time top 15 players
Posted: Sun Nov 10, 2013 3:45 am
by bchaikin
Hakeem > Shaq for peak, what? Shaq's peak was unrivaled.
note the qualifier "...just a bit better than...", plus the "...imho...". if someone wanted to argue that shaq was the best C of all-time, he'd have a lot of evidence on his side...
it all depends on how you define peak value...
Re: Vote for the all-time top 15 players
Posted: Sun Nov 10, 2013 5:44 pm
by Need To Argue
jbrocato23 wrote:Need To Argue wrote:
Not picking Oscar is far worse. 3 for 6 on West is out there as well. Only 1 vote for Baylor and for Moses is strange too.
I have Duncan as my 16 so just missing, Hakeem at 20 and Shaq at 27.
I understand people seeing things differently, but if they didn't see those guys it is hard to understand how great they were.
A computer spitting out numbers isn't going to explain Elgin Baylor. At forward only Bird was better for me. Julius was close to Elgin and now LeBron is making it real interesting.
Pettit is right there behind that group. I get others not picking Havlicek or Barry, but Elgin doesn't make sense to me. He was amazing just as LeBron is now.
My two cents.
Your 60s-70s bias is getting absurd. Shaq at 27? What? Did you see Shaq play from 95-03 (and esp 00-01)?? He was completely unstoppable. And riddle me this: how can you have West and Baylor both ahead of both Shaq and Kobe when West and Baylor played together on the Lakers for 11 seasons (i'm not counting '72 for obvious reasons) and won exactly zero titles when Shaq and Kobe played together on the Lakers for 8 years and won 3?
Look, I love NBA history. I watch as many old games as I can find. I've seen Baylor and West (and Hondo, Barry, etc) play, they were great. But they weren't as good as you're making them out to be. How is it that neither West nor Baylor ever won an MVP (and Baylor only finished 2nd once....Havlicek's best finish was 4th btw)? It just doesn't make sense to me in voting for the best players of all time how you can just brush of players with track records of extreme success who were the obvious primary drivers of that success (Shaq, Duncan, Olajuwon) in favor of players with much much less success AND worse stats.
Shaq at 27 is no insult. He is Larry Holmes, a champion when there aren't any other centers to challenge him. He missed out on the other centers from the 80's (who were better) because he wasn't ready for them when they were in their primes. He did great when there was no one left, a la Holmes who no one is picking as the greatest of all time for good reason.
West and Baylor and titles can be compared to Ernie Banks, does his lack of a title make him any less of a shortstop?
Havlicek is Derek Jeter, yes A-Rod can be better on paper to stats guys, but if it's about winning I'll take Jeter and Havlicek every time.
Bias, I have Duncan at 16 and Kobe at 17, how is that outrageous? They are two of the best players I've ever seen. I never get to say that because I always have to remind people of other greats from the days gone by. This is why basketball has become so similar to football, we forget the past with no sense of history. Ace Parker a HOF player just passed and all the NFL could say is the oldest HOFer died. Why do kids not no about Johnny Unitas and Wilt Chamberlain? Some here have Chris Paul in their top 100, how is that possible? He is a great player today, but let him grow a few whiskers before we elevate him to legendary status.