RealGM 2017 Top 100 List: #6

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

drza
Analyst
Posts: 3,518
And1: 1,861
Joined: May 22, 2001

Re: RealGM 2017 Top 100 List: #6 

Post#161 » by drza » Fri Jun 30, 2017 3:43 pm

penbeast0 wrote:
colts18 wrote:(Shaq but no alternative, need that for vote to count)



FWIW, TRex has been counting votes in cases where a person doesn't give an alternative. I asked him about it in thread 3, I believe, and he counted my votes in threads 3 & 4 when I didn't have an alternative picked.
Creator of the Hoops Lab: tinyurl.com/mpo2brj
Contributor to NylonCalculusDOTcom
Contributor to TYTSports: https://www.youtube.com/playlist?list=PLTbFEVCpx9shKEsZl7FcRHzpGO1dPoimk
Follow on Twitter: @ProfessorDrz
70sFan
RealGM
Posts: 30,133
And1: 25,419
Joined: Aug 11, 2015
 

Re: RealGM 2017 Top 100 List: #6 

Post#162 » by 70sFan » Fri Jun 30, 2017 3:44 pm

So, by DRAPM Duncan wasn't top 5 defender in 2007? It doesn't look very well to me...
Xherdan 23
Bench Warmer
Posts: 1,324
And1: 1,537
Joined: Apr 07, 2016
   

Re: RealGM 2017 Top 100 List: #6 

Post#163 » by Xherdan 23 » Fri Jun 30, 2017 3:56 pm

drza wrote:RAPM gets a mixed response aropund here, so take this for what it's worth. But, Defensive RAPM is one approach that was not at all surprised that KG was the best defensive player of 2008...because according to DRAPM, he was the top impact defender of 2007 as well.

2007 DRAPM
1. Kevin Garnett
2) Erick Dampier
3) Dikembe Mutombo
4) Jarron Collins
5) Shane Battier

2008 DRAPM
1) Kevin Garnett
2) Chuck Hayes
3) Jermaine O'Neal
4) Dikembe Mutombo
5) Jarron Collins
( 6) Tim Duncan )

*It has been pointed out in this thread, that in 2006 and 2007, the Wolves without KG played at a -12 level

*It has been pointed out, in this thread, that essentially every rotation player on the 2007 Wolves (except then-rookie Randy Foye) was out of the league shortly after that season, not for age reasons

*I've pointed out that, by every publicly available quantitative approach I can think of (e.g. 3 different boxscore methods & +/-), 2007 Garnett had very arguably the worst supporting cast of any star player in history

--> DRAPM, which regresses an individual's presence on the court with changes in the team's scoring margin, indicated that Garnett was pulling HUGE defensive weight to get that team defense to average. As in, he was pulling more defensive weight than any other player in the NBA in 2007, just to get that mess of a cast to defensively average. There was (and is) a tendency to scoff at that notion, but then in 2008 his cast went from historically terrible to defensively solid, and he takes that level of cast to a historic defense. Again, with his presence correlating to the biggest positive defensive impact in the league.

Looking only (or even primarily) at global team results isn't necessarily the best way to evaluate an individual in cases when we have a lot more context and specific data to utilize to come to our conclusions. My quick 2 cents, anyway.


Would you mind explaining to me how DRAPM is calculated? Meaning, would a better cast from the bench affect KG's numbers even if he's playing exactly the same?
If it does affect his numbers, would he look better or worse with a better defensive cast?

I've said it before, I'm not much of a stats guy, mostly because I don't understand them and never bothered to dig deep into how they're being calculated.
My perception of KG as a weaker defender in Minnesota (note: weaker in comparison to the defensive monster that was Boston KG, not a weak defender in general) comes from watching him these years and getting the same vibe I got from this year's Kawhi and LeBron - it's not that they're not capable defenders, it's that their offensive responsibilities are taking a toll on them and it shows on defense.

I'm not trying to argue that KG wasn't a great defender in Minny, only that he wasn't GOAT level like he was on Boston.
In theory, there is no difference between theory and practice. But, in practice, there is.
- Jan L. A. van de Snepscheut
drza
Analyst
Posts: 3,518
And1: 1,861
Joined: May 22, 2001

Re: RealGM 2017 Top 100 List: #6 

Post#164 » by drza » Fri Jun 30, 2017 4:02 pm

70sFan wrote:So, by DRAPM Duncan wasn't top 5 defender in 2007? It doesn't look very well to me...


A few things, in response.

1) Duncan actually had the highest RAPM overall in the NBA in 2007

2) However, in a seemingly odd twist, he measured out with by-far the best Offensive RAPM of his career in 2007 (ranking top-5 in NBA), while measuring out with the 20th-ranked DRAPM

3) RAPM isn't a perfect stat by any means, especially for any one year (e.g. there's definite value in repeated results, to strengthen overall conclusions) so it's possible that some of his defensive impact ended up showing up in offensive RAPM instead for that particular year. The fact that the year is an outlier, in this measure, somewhat supports this as a possibility

4) But, on the other hand, 82games.com has Duncan's on/off record for 2007. This is what they have for team offense and defense:

Offense: on-court 113 points/100 possessions, off-court 105.3 points/100 possessions (net +7.6)
Defense: on-court 99.2 points/100 possessions, off-court 105.8 points allowed/100 (net -6.5)

His offensive net positive was bigger than his defensive, in 2007, so perhaps there was something about that season where something he did was more necessary to the offense than normal, while the defense was slightly more resillient to his presence than usual? Shrugs

Point is...RAPM indicated that Duncan was the most impactful player in the NBA in 2007, overall. FWIW
Creator of the Hoops Lab: tinyurl.com/mpo2brj
Contributor to NylonCalculusDOTcom
Contributor to TYTSports: https://www.youtube.com/playlist?list=PLTbFEVCpx9shKEsZl7FcRHzpGO1dPoimk
Follow on Twitter: @ProfessorDrz
Dr Spaceman
General Manager
Posts: 8,575
And1: 11,211
Joined: Jan 16, 2013
   

Re: RealGM 2017 Top 100 List: #6 

Post#165 » by Dr Spaceman » Fri Jun 30, 2017 4:04 pm

Xherdan 23 wrote:
drza wrote:RAPM gets a mixed response aropund here, so take this for what it's worth. But, Defensive RAPM is one approach that was not at all surprised that KG was the best defensive player of 2008...because according to DRAPM, he was the top impact defender of 2007 as well.

2007 DRAPM
1. Kevin Garnett
2) Erick Dampier
3) Dikembe Mutombo
4) Jarron Collins
5) Shane Battier

2008 DRAPM
1) Kevin Garnett
2) Chuck Hayes
3) Jermaine O'Neal
4) Dikembe Mutombo
5) Jarron Collins
( 6) Tim Duncan )

*It has been pointed out in this thread, that in 2006 and 2007, the Wolves without KG played at a -12 level

*It has been pointed out, in this thread, that essentially every rotation player on the 2007 Wolves (except then-rookie Randy Foye) was out of the league shortly after that season, not for age reasons

*I've pointed out that, by every publicly available quantitative approach I can think of (e.g. 3 different boxscore methods & +/-), 2007 Garnett had very arguably the worst supporting cast of any star player in history

--> DRAPM, which regresses an individual's presence on the court with changes in the team's scoring margin, indicated that Garnett was pulling HUGE defensive weight to get that team defense to average. As in, he was pulling more defensive weight than any other player in the NBA in 2007, just to get that mess of a cast to defensively average. There was (and is) a tendency to scoff at that notion, but then in 2008 his cast went from historically terrible to defensively solid, and he takes that level of cast to a historic defense. Again, with his presence correlating to the biggest positive defensive impact in the league.

Looking only (or even primarily) at global team results isn't necessarily the best way to evaluate an individual in cases when we have a lot more context and specific data to utilize to come to our conclusions. My quick 2 cents, anyway.


Would you mind explaining to me how DRAPM is calculated? Meaning, would a better cast from the bench affect KG's numbers even if he's playing exactly the same?
If it does affect his numbers, would he look better or worse with a better defensive cast?

I've said it before, I'm not much of a stats guy, mostly because I don't understand them and never bothered to dig deep into how they're being calculated.
My perception of KG as a weaker defender in Minnesota (note: weaker in comparison to the defensive monster that was Boston KG, not a weak defender in general) comes from watching him these years and getting the same vibe I got from this year's Kawhi and LeBron - it's not that they're not capable defenders, it's that their offensive responsibilities are taking a toll on them and it shows on defense.

I'm not trying to argue that KG wasn't a great defender in Minny, only that he wasn't GOAT level like he was on Boston.


By all means I hope someone explains to you how DRAPM works and the specifics of it. But I don't think the specifics are important in this debate.

All we need to know from RAPM is that something huge is going on with KG. If you don't think it's right about him in a single year, then fine, but if 13 years of data say he's either at or near the highest-impact player in he league, well that should be taken seriously regardless of how you feel about the stat. Not that it should be the end-all, but...

The argument for Garnett was never specifically how mind-blowing his impact was. All these guys had mind blowing impact. What makes Garnett special is seemingly how portable this impact was. I don't know if there's another guy who literally did a 180 in the middle of their career and just excelled in their new role so much they won a title. LeBron may be as impactful as Garnett, but he wouldn't be playing the role Garnett played in Boston.

You're looking to get specific when the broad strokes are the important part. RAPM says something about KG that it's never said about anyone else.
“I’m not the fastest guy on the court, but I can dictate when the race begins.”
Doctor MJ
Senior Mod
Senior Mod
Posts: 53,520
And1: 22,528
Joined: Mar 10, 2005
Location: Cali
     

Re: RealGM 2017 Top 100 List: #6 

Post#166 » by Doctor MJ » Fri Jun 30, 2017 4:12 pm

Dr Positivity wrote:68 was an abnormally strong year for the Lakers. If you compare 69 Lakers to the other West and Baylor teams, Wilt's impact looks a little better, still not through the roof though. It's possible the pre 68 Lakers teams are a better comparison for a team that lost a lot of depth to get Wilt. Wilt's 2nd season with the Lakers where he plays 12 Gs is also a not bad measure for his value to them, they ended up having a noticeably worse season despite having a 21ppg Happy Hairston. On the whole I would say Wilt likely has value to the Lakers his first few seasons. But if it's just making them 55 W team instead of 45 W team, that's probably still below what other contenders do here. I side with Wilt's impact to Warriors and Sixers looking positive as well (Improves team in rookie season, is catalyst to PHI quickly rising to the top ), although the question again is whether it's enough to be put over some sick players impact wise


Well they implemented the Princeton in '68 so it's not random that the Lakers were stronger. They had finally figured something out with the the help of a new coach who would shortly have his career ruined by Wilt.


Sent from my iPhone using RealGM Forums
Getting ready for the RealGM 100 on the PC Board

Come join the WNBA Board if you're a fan!
70sFan
RealGM
Posts: 30,133
And1: 25,419
Joined: Aug 11, 2015
 

Re: RealGM 2017 Top 100 List: #6 

Post#167 » by 70sFan » Fri Jun 30, 2017 4:15 pm

Doctor MJ wrote:
Dr Positivity wrote:68 was an abnormally strong year for the Lakers. If you compare 69 Lakers to the other West and Baylor teams, Wilt's impact looks a little better, still not through the roof though. It's possible the pre 68 Lakers teams are a better comparison for a team that lost a lot of depth to get Wilt. Wilt's 2nd season with the Lakers where he plays 12 Gs is also a not bad measure for his value to them, they ended up having a noticeably worse season despite having a 21ppg Happy Hairston. On the whole I would say Wilt likely has value to the Lakers his first few seasons. But if it's just making them 55 W team instead of 45 W team, that's probably still below what other contenders do here. I side with Wilt's impact to Warriors and Sixers looking positive as well (Improves team in rookie season, is catalyst to PHI quickly rising to the top ), although the question again is whether it's enough to be put over some sick players impact wise


Well they implemented the Princeton in '68 so it's not random that the Lakers were stronger. They had finally figured something out with the the help of a new coach who would shortly have his career ruined by Wilt.


Sent from my iPhone using RealGM Forums


He ruined his career by his own decision - not bringing Wilt back in last moments of game 7.
mischievous
General Manager
Posts: 7,675
And1: 3,485
Joined: Apr 18, 2015

Re: RealGM 2017 Top 100 List: #6 

Post#168 » by mischievous » Fri Jun 30, 2017 4:44 pm

Xherdan 23 wrote:
mischievous wrote:
Dr Spaceman wrote:
Hakeem didn't peak on both ends at the same time. 90 is one of the best defensive seasons of all time. People rave about it. Why don't they rave about 91? 92? 88? 93-95 is one of the better offensive big man peaks in history. But it's widely acknowledged his defense slipped those years, especially in the RS.

I don't want Hakeem's 93-95 stretch to be projected over his whole career. He wasn't always that guy. His effort waned, he was a hothead (to the degree he had a religious conversion and added the 'H') he feuded with management, and he's admitted he was not a great guy to be around in his 20s. He could absolutely explode, like he did in the 86 playoffs, but he could also be over aggressive, biting on fouls and having extreme tunnel vision. He didn't really "get" how an offense should flow early in his career, and would demand the ball and shoot over double teams with alarming frequency.

I don't think it's incorrect to say Dirk had a better career based on his off-peak seasons even if Hakeem peaked higher.

But has any of that stuff ever prevented him from playing at a high level or succeeding? And does he have any seasons in there where his team got upset from his own poor play like Dirk in 07 or 06?

I can say that Dirk wasn't always the same guy that we saw in 06 and 2011. Still great, but he had his issues like weaker defense earlier in his career, and even when his defense improved I don't think he was much over neutral if at all.


'07 is a 6 game sample against a bad matchup.
True, Dirk played bad in the series but every player had at least one series like that (I don't hold Hakeem's 4 straight first round exits followed by a missed playoff against him, do you?).

I don't think it would be remembered that much if the Mavs hadn't overachieved in the regular season and Dirk wouldn't get the MVP in a suit.

As for '06, how did his team got upset from his poor play?
They came into the season, playoff and finals as the underdog with a team that had no business sniffing the playoffs without Dirk, let alone the finals.
How was Dirk supposed to stop D-Wade in that series? Especially from the free throw line.

Yes, he had trouble with shorter, stronger, aggressive defenders like Haslem and Captain Jack, but he did a fine job against Miami and wasn't his team's problem in any way.
The season after the "we believe" series he came back with a complete post game to punish these smaller defenders which is something I don't think gets enough credit for (same with MJ and LeBron improving their post game or Duncan improving free throws as opposed to Shaq not working on FT and KD still not having a post game).

The Mavs were clearly favored over the Heat thus = upset. Dirk couldn't do anything against Wade but that's far from the point.

Ultimately it's not about holding a series or 2 against a player, it's just Spaceman is trying to hold Hakeem's attitude, etc against him earlier in his career. He still performed well and had an excellent career, so I don't see how that's a legit reason to take Dirk over him. Hakeem wasn't perfect pre-peak but it's not like Dirk was god.
drza
Analyst
Posts: 3,518
And1: 1,861
Joined: May 22, 2001

Re: RealGM 2017 Top 100 List: #6 

Post#169 » by drza » Fri Jun 30, 2017 4:59 pm

Xherdan 23 wrote:
drza wrote:
Spoiler:
RAPM gets a mixed response aropund here, so take this for what it's worth. But, Defensive RAPM is one approach that was not at all surprised that KG was the best defensive player of 2008...because according to DRAPM, he was the top impact defender of 2007 as well.

2007 DRAPM
1. Kevin Garnett
2) Erick Dampier
3) Dikembe Mutombo
4) Jarron Collins
5) Shane Battier

2008 DRAPM
1) Kevin Garnett
2) Chuck Hayes
3) Jermaine O'Neal
4) Dikembe Mutombo
5) Jarron Collins
( 6) Tim Duncan )

*It has been pointed out in this thread, that in 2006 and 2007, the Wolves without KG played at a -12 level

*It has been pointed out, in this thread, that essentially every rotation player on the 2007 Wolves (except then-rookie Randy Foye) was out of the league shortly after that season, not for age reasons

*I've pointed out that, by every publicly available quantitative approach I can think of (e.g. 3 different boxscore methods & +/-), 2007 Garnett had very arguably the worst supporting cast of any star player in history

--> DRAPM, which regresses an individual's presence on the court with changes in the team's scoring margin, indicated that Garnett was pulling HUGE defensive weight to get that team defense to average. As in, he was pulling more defensive weight than any other player in the NBA in 2007, just to get that mess of a cast to defensively average. There was (and is) a tendency to scoff at that notion, but then in 2008 his cast went from historically terrible to defensively solid, and he takes that level of cast to a historic defense. Again, with his presence correlating to the biggest positive defensive impact in the league.

Looking only (or even primarily) at global team results isn't necessarily the best way to evaluate an individual in cases when we have a lot more context and specific data to utilize to come to our conclusions. My quick 2 cents, anyway.


Would you mind explaining to me how DRAPM is calculated? Meaning, would a better cast from the bench affect KG's numbers even if he's playing exactly the same?
If it does affect his numbers, would he look better or worse with a better defensive cast?

I've said it before, I'm not much of a stats guy, mostly because I don't understand them and never bothered to dig deep into how they're being calculated.
My perception of KG as a weaker defender in Minnesota (note: weaker in comparison to the defensive monster that was Boston KG, not a weak defender in general) comes from watching him these years and getting the same vibe I got from this year's Kawhi and LeBron - it's not that they're not capable defenders, it's that their offensive responsibilities are taking a toll on them and it shows on defense.

I'm not trying to argue that KG wasn't a great defender in Minny, only that he wasn't GOAT level like he was on Boston.


Best way I can try to explain it to have meaning, goes back before it goes forward, so bear with me. Let's go through the evolution of the +/- stats, a bit, as it will help the explanation make more sense.

1) Raw +/-. This was the first usage. Simple concept, like hockey: add up all the points a team scores and gives up while player is on court, use that as a measure. Gives some general info, but caliber of teammates on floor with player dominates the results

2) On/off +/-. This was an attempt to separate a player a bit more from his teammates. Measure +/- when player on the court, +/- when player off court, and subtract them. Still a very simple measurement...sometimes I mentally keep track of it during a game, much like you might keep track of a player's points or rebounds. It does help a bit with isolating player from teammates, but still get lots of colinearity possible within a unit. Also (on message), susceptible to what you describe...having a better/worse cast on bench would affect the "off" part of the measure.

3) APM, RAPM, etc. This one groups all the regressions together for brevity. In this approach, you no longer start with a player's on-court and/or off-court and subtract them. Instead, you use 5-man line-up data. You take every combination of 5 players that played in a given season; also take every combination of 5 players that a unit played against; and use those as inputs into your algorithm. Details vary based on which method you're using and how you implement it, but the punch line is that the regression results correlate a player's presence on the court with a certain amount of change in his unit's interractions with opponents. And this correlation is done for everyone, taking into account the strengths of the different 5-man units AND the strength sof the different 5-man opponents.

APM/RAPM come with their own sets of warts that need to be understood, but the point here is that the method is supposed to account for strength of teammates and opponents, such that the quality of your cast shouldn't materially affect the bottom line. Again, and I stress this every time we get into stat discussions, it isn't perfect...it can be very noisy, for example, and requires lots of data to increase accuracy. But, this approach at least minimizes the type of teammate dependence that you bring up better than any other impact approach that we currently have access to in the public.

Bringing it to 07 and 08 Garnett, he definitely had a much better defensive score in 2008 than he did in 2007. I don't want to give the impression that nothing at all changed, when he went to Boston and made defense more of a focus. He definitely had more impact in 2008, on defense, than he did in 2007. But he was already first in the league in measured defensive impact in 2007, so even if it got better, it wasn't like him being the best in 2008 came out of nowhere. He'd been near/at the top of the league defensively for much of his career in Minnesota, even when his offensive impact was huge. The difference in Boston, was, he did put less energy into offense but absolutely maximized on defense.
Creator of the Hoops Lab: tinyurl.com/mpo2brj
Contributor to NylonCalculusDOTcom
Contributor to TYTSports: https://www.youtube.com/playlist?list=PLTbFEVCpx9shKEsZl7FcRHzpGO1dPoimk
Follow on Twitter: @ProfessorDrz
Xherdan 23
Bench Warmer
Posts: 1,324
And1: 1,537
Joined: Apr 07, 2016
   

Re: RealGM 2017 Top 100 List: #6 

Post#170 » by Xherdan 23 » Fri Jun 30, 2017 5:02 pm

mischievous wrote:Ultimately it's not about holding a series or 2 against a player, it's just Spaceman is trying to hold Hakeem's attitude, etc against him earlier in his career. He still performed well and had an excellent career, so I don't see how that's a legit reason to take Dirk over him. Hakeem wasn't perfect pre-peak but it's not like Dirk was god.


It's not just Hakeem's attitude, it's his level of play.
He had a GOAT level offensive peak and a GOAT level defensive peak.
I think what Spaceman was trying to say is: These peaks didn't align and didn't last very long.

What happens is people tend to project his '90-'91 defense and his '93-'95 offense on to Hakeem's entire '85-'97 prime and that's not remotley close to the player he actually was.
In theory, there is no difference between theory and practice. But, in practice, there is.
- Jan L. A. van de Snepscheut
Xherdan 23
Bench Warmer
Posts: 1,324
And1: 1,537
Joined: Apr 07, 2016
   

Re: RealGM 2017 Top 100 List: #6 

Post#171 » by Xherdan 23 » Fri Jun 30, 2017 5:09 pm

drza wrote:
Spoiler:
Xherdan 23 wrote:
drza wrote:RAPM gets a mixed response aropund here, so take this for what it's worth. But, Defensive RAPM is one approach that was not at all surprised that KG was the best defensive player of 2008...because according to DRAPM, he was the top impact defender of 2007 as well.

2007 DRAPM
1. Kevin Garnett
2) Erick Dampier
3) Dikembe Mutombo
4) Jarron Collins
5) Shane Battier

2008 DRAPM
1) Kevin Garnett
2) Chuck Hayes
3) Jermaine O'Neal
4) Dikembe Mutombo
5) Jarron Collins
( 6) Tim Duncan )

*It has been pointed out in this thread, that in 2006 and 2007, the Wolves without KG played at a -12 level

*It has been pointed out, in this thread, that essentially every rotation player on the 2007 Wolves (except then-rookie Randy Foye) was out of the league shortly after that season, not for age reasons

*I've pointed out that, by every publicly available quantitative approach I can think of (e.g. 3 different boxscore methods & +/-), 2007 Garnett had very arguably the worst supporting cast of any star player in history

--> DRAPM, which regresses an individual's presence on the court with changes in the team's scoring margin, indicated that Garnett was pulling HUGE defensive weight to get that team defense to average. As in, he was pulling more defensive weight than any other player in the NBA in 2007, just to get that mess of a cast to defensively average. There was (and is) a tendency to scoff at that notion, but then in 2008 his cast went from historically terrible to defensively solid, and he takes that level of cast to a historic defense. Again, with his presence correlating to the biggest positive defensive impact in the league.

Looking only (or even primarily) at global team results isn't necessarily the best way to evaluate an individual in cases when we have a lot more context and specific data to utilize to come to our conclusions. My quick 2 cents, anyway.


Would you mind explaining to me how DRAPM is calculated? Meaning, would a better cast from the bench affect KG's numbers even if he's playing exactly the same?
If it does affect his numbers, would he look better or worse with a better defensive cast?

I've said it before, I'm not much of a stats guy, mostly because I don't understand them and never bothered to dig deep into how they're being calculated.
My perception of KG as a weaker defender in Minnesota (note: weaker in comparison to the defensive monster that was Boston KG, not a weak defender in general) comes from watching him these years and getting the same vibe I got from this year's Kawhi and LeBron - it's not that they're not capable defenders, it's that their offensive responsibilities are taking a toll on them and it shows on defense.

I'm not trying to argue that KG wasn't a great defender in Minny, only that he wasn't GOAT level like he was on Boston.


Best way I can try to explain it to have meaning, goes back before it goes forward, so bear with me. Let's go through the evolution of the +/- stats, a bit, as it will help the explanation make more sense.

1) Raw +/-. This was the first usage. Simple concept, like hockey: add up all the points a team scores and gives up while player is on court, use that as a measure. Gives some general info, but caliber of teammates on floor with player dominates the results

2) On/off +/-. This was an attempt to separate a player a bit more from his teammates. Measure +/- when player on the court, +/- when player off court, and subtract them. Still a very simple measurement...sometimes I mentally keep track of it during a game, much like you might keep track of a player's points or rebounds. It does help a bit with isolating player from teammates, but still get lots of colinearity possible within a unit. Also (on message), susceptible to what you describe...having a better/worse cast on bench would affect the "off" part of the measure.

3) APM, RAPM, etc. This one groups all the regressions together for brevity. In this approach, you no longer start with a player's on-court and/or off-court and subtract them. Instead, you use 5-man line-up data. You take every combination of 5 players that played in a given season; also take every combination of 5 players that a unit played against; and use those as inputs into your algorithm. Details vary based on which method you're using and how you implement it, but the punch line is that the regression results correlate a player's presence on the court with a certain amount of change in his unit's interractions with opponents. And this correlation is done for everyone, taking into account the strengths of the different 5-man units AND the strength sof the different 5-man opponents.

APM/RAPM come with their own sets of warts that need to be understood, but the point here is that the method is supposed to account for strength of teammates and opponents, such that the quality of your cast shouldn't materially affect the bottom line. Again, and I stress this every time we get into stat discussions, it isn't perfect...it can be very noisy, for example, and requires lots of data to increase accuracy. But, this approach at least minimizes the type of teammate dependence that you bring up better than any other impact approach that we currently have access to in the public.

Bringing it to 07 and 08 Garnett, he definitely had a much better defensive score in 2008 than he did in 2007. I don't want to give the impression that nothing at all changed, when he went to Boston and made defense more of a focus. He definitely had more impact in 2008, on defense, than he did in 2007. But he was already first in the league in measured defensive impact in 2007, so even if it got better, it wasn't like him being the best in 2008 came out of nowhere. He'd been near/at the top of the league defensively for much of his career in Minnesota, even when his offensive impact was huge. The difference in Boston, was, he did put less energy into offense but absolutely maximized on defense.


Thank you very much, you're a legend!
In theory, there is no difference between theory and practice. But, in practice, there is.
- Jan L. A. van de Snepscheut
User avatar
Dr Positivity
RealGM
Posts: 62,850
And1: 16,407
Joined: Apr 29, 2009
       

Re: RealGM 2017 Top 100 List: #6 

Post#172 » by Dr Positivity » Fri Jun 30, 2017 5:25 pm

Doctor MJ wrote:
Dr Positivity wrote:68 was an abnormally strong year for the Lakers. If you compare 69 Lakers to the other West and Baylor teams, Wilt's impact looks a little better, still not through the roof though. It's possible the pre 68 Lakers teams are a better comparison for a team that lost a lot of depth to get Wilt. Wilt's 2nd season with the Lakers where he plays 12 Gs is also a not bad measure for his value to them, they ended up having a noticeably worse season despite having a 21ppg Happy Hairston. On the whole I would say Wilt likely has value to the Lakers his first few seasons. But if it's just making them 55 W team instead of 45 W team, that's probably still below what other contenders do here. I side with Wilt's impact to Warriors and Sixers looking positive as well (Improves team in rookie season, is catalyst to PHI quickly rising to the top ), although the question again is whether it's enough to be put over some sick players impact wise


Well they implemented the Princeton in '68 so it's not random that the Lakers were stronger. They had finally figured something out with the the help of a new coach who would shortly have his career ruined by Wilt.


Sent from my iPhone using RealGM Forums


Maybe, or were they just more talented in 68 than previous Lakers years? In 68 Clark puts up 20/4/4 on +3.4 TS, makes the all-star team and finishes 18th in WS. That's the type of 3rd season the Lakers rarely had beside West and Baylor. That's before considering Goodrich strong per minute offense in 26mpg.
Liberate The Zoomers
User avatar
Narigo
Veteran
Posts: 2,796
And1: 882
Joined: Sep 20, 2010
     

Re: RealGM 2017 Top 100 List: #6 

Post#173 » by Narigo » Fri Jun 30, 2017 5:38 pm

eminence wrote:
I think the peak/prime gap between them is pretty dang large. Malone to me is a Harden level player at his best, KG is on the Curry/LeBron level (just using current guys as an example, and obviously many people disagree with those tiers, but I hope the point shines through).


Imo, I think Karl Malone was McGrady/Kobe level player at his best. But the thing is, he was performing at that level for 15+ seasons. Garnett was LeBron level for about three seasons at his best(03-05). But none of of his other seasons was performing at that level. Maybe 06 and 08 is close but thats about it.


eminence wrote:
I don't feel the star longevity gap is much if anything, '97 to '12 vs '87 to '03.


Karl Malone(87- 03)
51,004 minutes

Kevin Garnett(97-12)
43,486 minutes

Malone has a sizable edge in longevity.
Narigo's Fantasy Team

PG: Damian Lillard
SG: Sidney Moncrief
SF:
PF: James Worthy
C: Tim Duncan

BE: Robert Horry
BE:
BE:
User avatar
Jaivl
Head Coach
Posts: 7,105
And1: 6,757
Joined: Jan 28, 2014
Location: A Coruña, Spain
Contact:
   

Re: RealGM 2017 Top 100 List: #6 

Post#174 » by Jaivl » Fri Jun 30, 2017 5:47 pm

Vote: Kevin Garnett
Second vote: Hakeem Olajuwon


I would be okay with Shaq as my second vote, too, but Hakeem's turn it is.

Garnett compared to Russell (#4):
Spoiler:
-Around 15 years of proven superstar impact (Russell: 12 years) + additional years as a solid piece.

-Proven impact on both ends, on different roles:
*Played as a 4 and as a 5, maybe even the 3 at some point.
*Played as a jack-off-all-trades on offense, horizontal defender (mid-decade Wolves), with worse-O-better-D quasi LeBron levels of impact (average around +7, peaked as +10 -higher than Bron-). <- outside of LeBron, probably the better scrub-carrier of all time
*Played as a defensive anchor and secondary/tertiary scorer (Celtics) on a contender, with Mutombo/Russell-esque levels of impact, considering era (+0 off +6 def) - and arguably outside his prime!
*At close to 40-years old, still a valuable contributor off the bench on defensive duties with a reduced offensive role.

-Has the physical qualities and IQ to quasi-replicate the impact of Russell in his era (similar mobility, greater length but worse verticality, top-tier defensive instincts, nightmare on switches) while Russell doesn't have the additional tools KG has (better finishing, mid-range proficiency, better offensive orchestrator).

-Doesn't have the GOAT-tier leadership of Bill, but it's one of the ones that are closer.


Why I don't think Wilt is on this tier (or why I don't have Wilt in my top 10):
Spoiler:
It's already been discussed and I'm not going to expand much, but there is a legitimate concern with Wilt's impact.

Now, don't get me wrong, Wilt's talent is second to none.
- He is one of the most gifted athletes in sports history (great leaping ability and strength, good quickness -not only for a 7'1", overall-, giant frame.
- He showed proficiency at nearly every aspect of the game (scoring, rebounding, defense, passing).
- He showed massive impact at different roles.
- He has a massive peak in 1967, maybe GOAT. Aided by a great supporting cast (Walker, Greer, Cunningham...), he led one of the best teams ever in both sides of the ball.

BUT

Some (quite a few) years he didn't translate those talents into wins. His obsession about records definitely hurt his play, we already discussed late Warriors and early Lakers seasons. He needed a coach to reign him in benefit of the team... and that usually didn't last for very long.


KG and Hakeem's case over other strong contenders (Magic):
Spoiler:
- KG and Hakeem have more all-star impact years than Magic has total years.
- Magic's offensive career has been somewhat overblown, and by that I mean Magic wasn't really at that GOAT offensive level for his first 6 years or so, not all 12 of Magic's years are GOAT-offensive level years.
- KG and Hakeem peaked as both elite offensive and defensive players. Those peaks didn't overlap, but I'd guess that falls more into context and team-building than into skill.
The 04 Wolves looked absurdly good on both O and D (top 5 in the league) with Garnett on court, and looked complete trash without him (would-be bottom offense in the NBA by a mile, from top 5 defense to bottom 10). Acting as the lead scorer, the secondary creator and the defensive anchor at once. With really only two (one and a half?) valuable teammates: Sam Cassell: a 34-year old PG (who, to be fair, was very solid and close to all-star level for many years) and Latrell Sprewell: a 33-year old forward who had a deserved reputation of being a chucker.


Point for Magic: He can create a great offense from scratch, while Garnett had bad defenses in Minnesota.
Counterpoint: Garnett's teams were some of the worst in history, while Magic was always blessed with great teammates.
Yeah, but not when you're playing with sub-replacement level teammates. Look at those 07 Wolves. Ugh. Garnett missed 6 games that season. In those games, the Wolves were a -14 SRS team. That would be literally an all-time worst. With Garnett they were "only" a -2 team. That's massive impact.

On/off, RAPM, common sense, everything paints the same picture.
This place is a cesspool of mindless ineptitude, mental decrepitude, and intellectual lassitude. I refuse to be sucked any deeper into this whirlpool of groupthink sewage. My opinions have been expressed. I'm going to go take a shower.
User avatar
An Unbiased Fan
RealGM
Posts: 11,736
And1: 5,708
Joined: Jan 16, 2009
       

Re: RealGM 2017 Top 100 List: #6 

Post#175 » by An Unbiased Fan » Fri Jun 30, 2017 5:59 pm

Dr Spaceman wrote:
An Unbiased Fan wrote:
Dr Spaceman wrote:I hope it's okay if I can jump in at this point. Garnett is getting traction and he deserves it. If people are having him more than a couple of spots behind Duncan, I think there's a problem. Dirk is also a guy I'll make a case for, I have him over Bird and probably Magic. I don't see Wilt being a candidate for me for several more spots. He's way too problematic and the best we saw from him certainly isn't better than what we saw from Shaq, who doesn't exactly have a clean record himself.

I would be intersted in hearing arguments/evidence for Dirk over Magic/Bird, because I don't quite see any.


Well, it's going to depend on your criteria of course, which I'm not sure of at this time. If you've posted them anywhere on here, that would be helpful.

But, generally speaking, Dirk has a prime that should be up there with anyone. 11 straight All Star selections, 11 straight All-NBA selections in the single most talented position in the league (competing with Duncan and Garnett), league MVP, Finals MVP, and 11 straight 50 win seasons. Dirk gave his team 11 straight seasons of deep playoff contention, pretty much exactly what Bird and Magic gave you, and he did it without a best in league supporting cast (Bird) and an owner willing to spend unprecedented money and being paired with the second best player of all time according to this board (Magic). Dirk also held on until age 35, where he was scoring 22 a game on .603 TS and a 24 PER, at which age Bird and Magic had been sitting at home for years.

Dirk is definitely a Top 5 plyaer of his era, fantastic offensive anchor, and had great longevity. But does his prime really reach the level of a Bird/Magic?

All-NBA 1st team selections:
Kobe - 11
Magic - 9
Bird - 9
Shaq - 8
Wilt - 7
Hakeem - 6
Dirk - 4
KG - 4

Top 5 MVP finisher:
Kobe - 11
Wilt - 10
Magic - 9
Bird -9
Shaq - 8
Hakeem - 6
KG - 5
Dirk - 4

^
When looking at how these players were viewed by their contemporaries, DIrk falls quite a bit behind. His longevity is nice, but he didn't put together the same number of elite seasons that others did, despite playing more seasons. So we enter a quality over quantity issue here.

He led many 50-win teams, but if we count regular season team success, we also have to look at Dirk who had numerous playoff failures. Whether is be the 06 Finals, or losing as the #1 seed his MVP season, Dirk has a few scars to bare. Doesn't take away from him being an all-time great level offensive anchor, but with the players up right now, he seems lacking.

Now, re: Magic specifically, I think it's important to keep in mind Magic didn't truly become MAGIC until 1987, and maybe even afterward. It's tempting to look at his career as one big block, because of the offensive dynasty and the titles and such. Magic was not a godly half court player before 85. Godly transition player, godly passer, but not in offensive GOAT territory. He wasn't a great shooter, he didn't work with his back to the basket much, and I think people would be surprised how limited his half court role was compared to Kareem. We have to be careful not to project attributes of his whole career onto individual years.

Players do evolve, but lets be clear, Magic was already an elite player multiple seasons before 1985. He was already anchoring top level offenses, and winning rings. Dirk too was different early in his career, he didn't became the MVP Dirk until around 2005.

Now, Dirk. I understand I hold a little bit of a minority opinion on his offense. I think if you wanted to argue Dirk as the offensive GOAT, it would be pretty easy to do so. People generally go with the thought process, "well his individual scoring numbers are pretty comparable to your run of the mill top 10 scorer, and he's not a notable playmaker, so what's the big deal?" Well, the big deal is that despite not being a classic playmaker, his "help offense" perhaps surpasses any player we've ever seen. Consider the hyperbole with which people talk about Jordan, how he was a master at commanding the entire floor just by feigning a cut, how a screen set for Jordan meant instant defcon 0 reactions from the defense, how the defense had to bend to accommodate him more than any scorer in history. And then consider that all of those apply to Dirk.

I fail to see how Dirk is above Magic, Kobe, or Bird offensively, all of whom were dominant on that side of the ball. Not sure what "help offense" is, perhaps you can detail that a bit. 55.9% of Dirk's FGs were assisted, so he wasn't shot creating, he was being fed the ball. A consistent 20-25 ppg scorer on high efficiency, but not many assists. Not GOAT level offense

By his late prime, Dirk was a better midrange scorer than Jordan. I firmly believe that. Off the dribble, out of the post, fading away from one leg, the degree of difficulty on his shots is second to none, and he was hitting 50+ percent off those shots some seasons. That's SO absurd. Dirk broke pick and roll defense. Walk through the scenarios. Trapping the ball ahndler is not an option for obvious reasons. Switch? Dirk posts up your guard at the nail for an easy 2 points. Go under? Dirk hits a pick and pop 3. Go over? Dirk gets to attack the rim with a head of steam. He is an unsolvable problem, and plenty of coaches have said as much. He draws so much attention around the free throw line, that defenses are willing to help away from the two highest-leverage spots on the floor (the rim and the corner) to take those looks away. Year after year after year, Dallas is near the top in corner 3s. Year after year after year, Dallas is near the top in rim efficiency. Year after year after year, Dallas is tops in offensive efficiency despite a revolving door of personnel. Even in 2017, we still talk about the "Dirk effect", how random undersized guards have career years playing next to this guy.

Every great offensive anchor can claim the similar things. This is more narrative than actually evidence. Magic is the GOAt mismatch on offense. Bird's ability to score form outside to inside, while playmaking was exceptional. Kobe's ability to hit anywhere on the court and shot-making is perhaps unrivaled. The question though is is the sum of Dirk's offensive parts exceedes that of other greats.

You say he was an unsolvable problem, but in 2007 with the #1 seed and HCA, Dirk put up 17.8 ppg on 50.9% TS. The Warriors were #19 in DRtg that season, so its not like he was going against a historically great defense. Dirk's career is filled with moments, series, games like this. Considering the careers of his competition, his resume seems light.

Now one more thing I'll say, the "playoff choker" myth and the "bad defender" myth both need to die a painful death. Both are based on the "soft euro" stereotype and both don't apply. For one, Dirk (along with Jordan) is the most consistently resilient playoff performer we've seen against top defenses. His 28/13 playoff numbers are without condition. As far as defense, he obviously had specific flaws that could be picked on, be¡tú he brought so many positives to the table as well. I know getting back in transition and boxing out aren't sexy, and to some may not even be worth discussing in this project, which is a shame, but they matter. They matter a lot. If they didn't, your coach wouldn't have been drilling it into your head since the age of 5.

Both are true faults of Dirk, definitely not myths. Again, 17.8 ppg as the #1 seed and MVP.....I don't want to pile on and start naming his playoff mishaps, but this isn't a road to go down if you want to advocate for Dirk.

Defensively, Dirk was poor early in his career, and mediocre in his prime. He's slow-footed, and not very quick defensively. Other players lit him up often. Doesn't really hurt him in comparison with Bird or Magic though, since I don't view them as any better on defense.

Well, for most of those guys, the answer is that he has a serious argument as defensive GOAT, while also winning an MVP for offense at his peak. That type of portability and versatility is literally just a KG thing. As far as D Rob, KG was a much better offensive player and was better for longer to boot. As for Hakeem, well I see them as about even. Hakeem is amazing.

So again, like with Dirk, what evidence is there that KG is the defensive GOAT? A player who consistently led bad defenses most of his career until he joined a superteam.
7-time RealGM MVPoster 2009-2016
Inducted into RealGM HOF 1st ballot in 2017
Doctor MJ
Senior Mod
Senior Mod
Posts: 53,520
And1: 22,528
Joined: Mar 10, 2005
Location: Cali
     

Re: RealGM 2017 Top 100 List: #6 

Post#176 » by Doctor MJ » Fri Jun 30, 2017 6:02 pm

70sFan wrote:
Doctor MJ wrote:
Dr Positivity wrote:68 was an abnormally strong year for the Lakers. If you compare 69 Lakers to the other West and Baylor teams, Wilt's impact looks a little better, still not through the roof though. It's possible the pre 68 Lakers teams are a better comparison for a team that lost a lot of depth to get Wilt. Wilt's 2nd season with the Lakers where he plays 12 Gs is also a not bad measure for his value to them, they ended up having a noticeably worse season despite having a 21ppg Happy Hairston. On the whole I would say Wilt likely has value to the Lakers his first few seasons. But if it's just making them 55 W team instead of 45 W team, that's probably still below what other contenders do here. I side with Wilt's impact to Warriors and Sixers looking positive as well (Improves team in rookie season, is catalyst to PHI quickly rising to the top ), although the question again is whether it's enough to be put over some sick players impact wise


Well they implemented the Princeton in '68 so it's not random that the Lakers were stronger. They had finally figured something out with the the help of a new coach who would shortly have his career ruined by Wilt.


Sent from my iPhone using RealGM Forums


He ruined his career by his own decision - not bringing Wilt back in last moments of game 7.


That moment occurred within the context of that year. It's a mistake to try to understand the moment without all that led up to it.


Sent from my iPhone using RealGM Forums
Getting ready for the RealGM 100 on the PC Board

Come join the WNBA Board if you're a fan!
scrabbarista
RealGM
Posts: 20,257
And1: 17,961
Joined: May 31, 2015

Re: RealGM 2017 Top 100 List: #6 

Post#177 » by scrabbarista » Fri Jun 30, 2017 6:04 pm

6. Magic Johnson

7. Shaquille O'Neal


Will comment tonight. (I've always had Magic high, while I've continuously flip-flopped on Shaq and Wilt.)
All human life on the earth is like grass, and all human glory is like a flower in a field. The grass dries up and its flower falls off, but the Lord’s word endures forever.
User avatar
An Unbiased Fan
RealGM
Posts: 11,736
And1: 5,708
Joined: Jan 16, 2009
       

Re: RealGM 2017 Top 100 List: #6 

Post#178 » by An Unbiased Fan » Fri Jun 30, 2017 6:11 pm

Jaivl wrote:- KG and Hakeem have more all-star impact years than Magic has total years.
- Magic's offensive career has been somewhat overblown, and by that I mean Magic wasn't really at that GOAT offensive level for his first 6 years or so, not all 12 of Magic's years are GOAT-offensive level years.
- KG and Hakeem peaked as both elite offensive and defensive players. Those peaks didn't overlap, but I'd guess that falls more into context and team-building than into skill.

Uh...wait. LA was the #1 offense in 1980. Magic was at an elite level from 1982 and on. With KG we 're talking about maybe a 04-08 span, and even then that's dubious since his teams in Minny were bad defensively.
7-time RealGM MVPoster 2009-2016
Inducted into RealGM HOF 1st ballot in 2017
User avatar
Jaivl
Head Coach
Posts: 7,105
And1: 6,757
Joined: Jan 28, 2014
Location: A Coruña, Spain
Contact:
   

Re: RealGM 2017 Top 100 List: #6 

Post#179 » by Jaivl » Fri Jun 30, 2017 6:15 pm

An Unbiased Fan wrote:Uh...wait. LA was the #1 offense in 1980. Magic was at an elite level from 1982 and on. With KG we 're talking about maybe a 04-08 span, and even then that's dubious since his teams in Minny were bad defensively.

2. Kareem Abdul-Jabbar

????. Ricky Davis
This place is a cesspool of mindless ineptitude, mental decrepitude, and intellectual lassitude. I refuse to be sucked any deeper into this whirlpool of groupthink sewage. My opinions have been expressed. I'm going to go take a shower.
ElGee
Assistant Coach
Posts: 4,041
And1: 1,207
Joined: Mar 08, 2010
Contact:

Re: RealGM 2017 Top 100 List: #6 

Post#180 » by ElGee » Fri Jun 30, 2017 6:15 pm

Winsome Gerbil wrote:Well more than Wilt just arriving went on. And even Wilt arriving...I mean he was 32 years old by that point, and ate up all of 13.6 shots a game. That's hardly a deal breaker. But the larger issue is that they lost their depth. And the depth they lost actually took more shots than Wilt did when he arrived.


Right — we should always be thinking of the context. Wilt’s move to LA in 68 (and the remaining Philly team) was not some perfect science experiment. Be careful using his age against him though, because you’re minimizing his final 5 seasons that way, and you’re minimizing his previous MVP campaign from a few months prior.

What we’re comparing is:

LA 68 (+8) —>69 (+6):
Baylor (39), West (38), Clark (38), Hawkins (32), Imhoff (28), Goodrich (26), Counts (21), Crawford (20)
Wilt (45), Baylor (40) West (39), Ericsson (26), Counts (24), Egan (22), Crawford (21), Hawkins (20)

Philly 68 (+8) —> 69 (+7)
Wilt (47), Greer (40), Walker (32), Jackson (31), Cunningham (28), Jones (27), Guokas (20)
Cunningham (40), Greer (40), Walker (34), Jackson (34), Jones (29), Imhoff (29), Clark (26)

With a 3x reigning MVP changing hand, ask yourself what a reasonable effect on the offense/overall SRS of both teams would be. First, LA loses it’s big (Imhoff), an all-star level guard (Clark) and the young guard reserve (Goodrich). Wilt should be a massive upgrade at the big spot, so you’re left “patching” another wing spot (basically platooned by Egan and Ericsson). We can say VBK was stubborn (although you see why he would be after the 68 alchemy)…but so were Wilt’s skills. (Portability! Chapter 8!)

Meanwhile, back in Philly, we can also say Cunningham had a major improvement, and without Jackson Philly was a +3 team (pretty good for losing both bigs). Clark can’t really crack the rotation the same way, so it’s not really his value translating that’s keeping Philly good. It’s just…Philly looks like a really good team, even without Wilt.

And be very careful with your argument about how many shots he took. First, if you believe such a level of volume is insignificant, then you’re minimizing a 3x defending MVP at the heart of his career -- a stretch people point to when it’s brought to their attention that his early teams weren’t actually much different on offense with his volume scoring. Second, Wilt took 18.2 scoring attempts (TSA) per game. The following seasons were also at 18.2 or less:

Penny 96
Paul 11-13, 15-16
Dirk 01
Payton 98
Stockton 88-92, 94-97
Leonard 16
Garnett 08
Magic 82-83, 85, 90-91
Oscar 71
Gasol 09, 11
Pierce 08
Nash 06, 07
Duncan 98, 07
Frazier 69, 70

These players either anchored the offense or were integral parts; such volume is far from insignificant, and it doesn’t tell us what a player’s Global Impact was. Eating up space, eating up time of possession and ball-stopping can all be detrimental to an offense, and with a good offense, are extremely hard (or impossible) to offset.
Check out and discuss my book, now on Kindle! http://www.backpicks.com/thinking-basketball/

Return to Player Comparisons