Statistically Speaking, Who Had Better Teammates, Overall? Lebron or MJ?

[ty 4191]

So....I ran some numbers (similar to what I ran for Wilt, Kareem, Shaq, and Hakeem, recently).

Career Teammates' True Shooting Added, to get an idea of who had the better teammates, overall.

For this exercise, I also ran, for defense, all of their respective teammates' defensive box plus minus.

I removed MJ and Lebron's numbers from all data.

All numbers are regular season:

Code: Select all

Year   Lebron Teammates True Shooting Addded
2004   -380
2005   -83
2006   -101
2007   -335
2008   -442
2009   49
2010   163
2011   269
2012   46
2013   424
2014   351
2015   220
2016   97
2017   215
2018   192
2019   -160
2020   78
2021   -102
   
Sum   501

Code: Select all

Year   Jordan Teammates True Shooting Added
1985   12
1986   -33
1987   -297
1988   -300
1989   -83
1990   43
1991   115
1992   200
1993   -148
1995   -6
1996   27
1997   38
1998   -159
2002   -25
2003   3
   
Sum   -613

Code: Select all

Year   Lebron Teammates DBPM
2004   -10
2005   -43
2006   8
2007   7
2008   -16
2009   2
2010   -15
2011   2
2012   4
2013   -15
2014   -4
2015   -8
2016   -2
2017   -1
2018   -20
2019   4
2020   0
2021   11
   
Sum   -96

Code: Select all

Year   Jordan Teammates DBPM
1985   -10
1986   -22
1987   2
1988   -2
1989   -12
1990   -19
1991   -2
1992   -1
1993   0
1995   7
1996   4
1997   9
1998   13
2002   -6
2003   -13
   
Sum   -52

Unfortunately, I can only find regular season numbers for right now for True Shooting Added. Maybe someone can help me out with the postseason? I'll run playoffs impact metrics next for defense and offense, where I can find it.

Thanks so much in advance.

What does this prove, if anything? Would you use different metrics? Why, if so?

[DQuinn1575]

So....I ran some numbers (similar to what I ran for Wilt, Kareem, Shaq, and Hakeem, recently).

Career Teammates' True Shooting Added, to get an idea of who had the better teammates, overall.

For this exercise, I also ran, for defense, all of their respective teammates' defensive box plus minus.

I removed MJ and Lebron's numbers from all data.

All numbers are regular season:

Code: Select all

Year   Lebron Teammates True Shooting Addded
2004   -380
2005   -83
2006   -101
2007   -335
2008   -442
2009   49
2010   163
2011   269
2012   46
2013   424
2014   351
2015   220
2016   97
2017   215
2018   192
2019   -160
2020   78
2021   -102
   
Sum   501

Code: Select all

Year   Jordan Teammates True Shooting Added
1985   12
1986   -33
1987   -297
1988   -300
1989   -83
1990   43
1991   115
1992   200
1993   -148
1995   -6
1996   27
1997   38
1998   -159
2002   -25
2003   3
   
Sum   -613

Code: Select all

Year   Lebron Teammates DBPM
2004   -10
2005   -43
2006   8
2007   7
2008   -16
2009   2
2010   -15
2011   2
2012   4
2013   -15
2014   -4
2015   -8
2016   -2
2017   -1
2018   -20
2019   4
2020   0
2021   11
   
Sum   -96

Code: Select all

Year   Jordan Teammates DBPM
1985   -10
1986   -22
1987   2
1988   -2
1989   -12
1990   -19
1991   -2
1992   -1
1993   0
1995   7
1996   4
1997   9
1998   13
2002   -6
2003   -13
   
Sum   -52

Unfortunately, I can only find regular season numbers for right now for True Shooting Added. Maybe someone can help me out with the postseason? I'll run playoffs impact metrics next for defense and offense, where I can find it.

Thanks so much in advance.

What does this prove, if anything? Would you use different metrics? Why, if so?

At a minimum you would need to weight the Def BM by minutes played - LBJ's 2005 team has guys with under 200 minutes played coming in at -8.0 and -9.1, which in your math counts as much as a season with Wade or Pippen when they played 3,000 minutes.

The more generic question is that you look at how anyone played that year, and it doesn't show that Cartwright didnt play as well with Jordan or Wade the first year with James - it says they are not as good last year, but at least some of it is they didnt play as well with Michael or LeBron. THe reverse is true for someone like a Paxson, who fit well with Jordan, so he shows up better than maybe he was.

[ty 4191]

At a minimum you would need to weight the Def BM by minutes played - LBJ's 2005 team has guys with under 200 minutes played coming in at -8.0 and -9.1, which in your math counts as much as a season with Wade or Pippen when they played 3,000 minutes.

The more generic question is that you look at how anyone played that year, and it doesn't show that Cartwright didnt play as well with Jordan or Wade the first year with James - it says they are not as good last year, but at least some of it is they didnt play as well with Michael or LeBron. THe reverse is true for someone like a Paxson, who fit well with Jordan, so he shows up better than maybe he was.

Hi Quinn,
Thanks for the suggestions.

That's true, I didn't weight it by MP, but I used the same methodology for both players, so it would even out over the course of a career, no?

How exactly (literally) would you weight it? And compute it?

Also would using defensive rating for all teammates be better than using DBPM?

[eminence]

Simple enough to do for oneself, but I think adding an average/season with the sum would be helpful.

[DQuinn1575]

At a minimum you would need to weight the Def BM by minutes played - LBJ's 2005 team has guys with under 200 minutes played coming in at -8.0 and -9.1, which in your math counts as much as a season with Wade or Pippen when they played 3,000 minutes.

The more generic question is that you look at how anyone played that year, and it doesn't show that Cartwright didnt play as well with Jordan or Wade the first year with James - it says they are not as good last year, but at least some of it is they didnt play as well with Michael or LeBron. THe reverse is true for someone like a Paxson, who fit well with Jordan, so he shows up better than maybe he was.

Hi Quinn,
Thanks for the suggestions.

That's true, I didn't weight it by MP, but I used the same methodology for both players, so it would even out over the course of a career, no?

How exactly (literally) would you weight it? And compute it?

Also would using defensive rating for all teammates be better than using DBPM?

Since the total number of the career is -52 and -96, then one guy or the other having 3 more guys who play under 200 minutes being -8 would be over half your difference - so since the individual differences are rather large compared to the total, it could impact it.

you would need to multiply the minutes played x bpm = extended bpm, add up the minutes played and the extended bpm and then divide - that will give you a weighted average.

I'm kind of concerned the defensive BPM is somewhat of a circular thing as it conforms to the total team defense - so if MJ's average defense is a +5 and LeBron is a +4, and then MJ's Def BPM is a +2 and LeBron a +3, then their teammates by default and design will be +3 for MJ and +1 for LeBron. So you are basically relying on Def BPM to determine how good LeBron and MJ is- and then attributing the rest of the team's success to their teammates - that works to the degree you have confidence in BPM - if it is 100% then the method is 100% accurate, if 60% then maybe 60% accurate (there may be some heavier math here, but Ill ignore it- its definitely not 100%), etc.

[ty 4191]

Simple enough to do for oneself, but I think adding an average/season with the sum would be helpful.

You got it, brother.

Code: Select all

Year   Lebron Teammates True Shooting Addded
2004   -380
2005   -83
2006   -101
2007   -335
2008   -442
2009   49
2010   163
2011   269
2012   46
2013   424
2014   351
2015   220
2016   97
2017   215
2018   192
2020   78
2021   -102

Average/Season   27.8

Code: Select all

Year   Jordan Teammates True Shooting Added
1985   12
1986   -33
1987   -297
1988   -300
1989   -83
1990   43
1991   115
1992   200
1993   -148
1995   -6
1996   27
1997   38
1998   -159
2002   -25
2003   3

Average/Season   -41

Code: Select all

Year   Lebron Teammates DBPM
2004   -10
2005   -43
2006   8
2007   7
2008   -16
2009   2
2010   -15
2011   2
2012   4
2013   -15
2014   -4
2015   -8
2016   -2
2017   -1
2018   -20
2020   0
2021   11

Average/Season   -5.3

Code: Select all

Year   Jordan Teammates DBPM
1985   -10
1986   -22
1987   2
1988   -2
1989   -12
1990   -19
1991   -2
1992   -1
1993   0
1995   7
1996   4
1997   9
1998   13
2002   -6
2003   -13

Average/Season   -3.5

[ty 4191]

Since the total number of the career is -52 and -96, then one guy or the other having 3 more guys who play under 200 minutes being -8 would be over half your difference - so since the individual differences are rather large compared to the total, it could impact it.

you would need to multiply the minutes played x bpm = extended bpm, add up the minutes played and the extended bpm and then divide - that will give you a weighted average.

You want to give this a whirl? You're clearly better with basic math than me!

[ty 4191]

How about only looking at both player's teammates, with their own stats removed, for teammates with 1000 MP or more (each/every season), using either DBPM or Defensive Rating?

[eminence]

How about only looking at both player's teammates, with their own stats removed, for teammates with 1000 MP or more (each/every season), using either DBPM or Defensive Rating?

Should give you something at least fairly similar to the minutes weighted version.

[DQuinn1575]

How about only looking at both player's teammates, with their own stats removed, for teammates with 1000 MP or more (each/every season), using either DBPM or Defensive Rating?

Should give you something at least fairly similar to the minutes weighted version.

Yes that should be pretty good

[LukaTheGOAT]

We have AuPM (the 2nd latest version) supporting cast data for PS teams, going back to 1996-1997 (as it uses plus-minus data), and therefore a majority of MJ's career is not covered.

However, looking at the RS Cast (Relative AuPM value of the 2nd through the 8th-best player on a team, among players who logged at least 40 percent of team’s minutes. Value is relative to +0.75.) we can try and estimate the strength of those last 2 Bulls teams' supporting cast.

MJ's RS Cast in 97 was a 8.3 per the last version of AuPM and a 8.6 in 98.

97: 8.3
98: 8.6

If we look at Lebron's RS Cast from 08 to 2020:

07: -0.3
08: 2.2
09: 4.8
10: 1.2
11: 3.4
12: 4.2
13: 4.9
14: 1.8
15: 1.9
16: 2.2
17: 0.9
18: -1.9
19: Not a Playoff team
20: 1

Based off the 2nd lastest version of AuPM at least, it would seem as if MJ had a stronger RS supporting cast in 97 and 98 than Lebron had any point in his career.

[Owly]

An issue I'd have with using TSA - as a measure of teammates - is that assumes each's playmaking was equal. If they aren't then the better playmaker is seen as indirectly "worse" for having played with "better" teammates.


Edit: Trex had a superstar teammate average "cast" thread with this sort of stuff for a bunch of metrics ... couldn't find it with a quick Google ... but imagine mentioning it might jog someones memory who can.

[falcolombardi]

if i had to rank them completely subjectively and only full and mostly healthy seasons

tier 1
1 2020 davis
2 2011 wade
3 2012 wade

tier 2
4-8 pippen (any order of 91,92,93,96,97)
small gap

tier 3
9- 98 pippen
10-13 2011-2014 bosh

tier 4

kyrie irving 2015-2017

tier 5

rodman 96-98 and grant 91-93, love 2015-2016

lebron has the best teammates but for very few seasons then jordan has the next best ones, then is a tie~ again

granted this is only for top 2 teammates at any given season and doesnt consider fit, coach, competition, role players

[Colbinii]

2011 Wade and 2020 AD are at the top. 2012 Wade and most of prime Pippen are in the next group. Grant/Bosh/Rodman/Kyrie round out the following tier.

Difficult to say as LeBron had higher highs and lower lows in terms of roster construction around him while in their primes and Jordan had a steady "Arguably best cast in NBA" for most of his competitive prime.

[VanWest82]

I said this in the Finals thread but I think you're going to have a hard time accurately quantifying this given that post 89, MJ played in the triangle which was designed to promote/feature different options besides MJ.

Outside of 2011 when Lebron and Wade ran some variation of my turn, your turn, Lebron has been playing Lebron ball for the majority of his career which is naturally going to supress the statistical contributions of his teammates both in terms of individual stats and +/- derived ones given the main drawback of Lebron ball is that his teams are often left with nothing to fall back on when he isn't in the game doing everything. I don't know how you account for that bias but it's surely there.

To answer the question, I'd argue Lebron has had better teammates on average due to all the HOFer vets and former stars / sub all stars rounding out his deep rosters. The list is substantial. MJ had really deep teams in 96 and 97 which compete with the best of Lebron's rosters but outside of that I don't think so. Too many Dwights, Shaqs, Ben Wallaces, Rondos, Gasols, Ray Allens, etc., let alone the RJs, Melos, Birdmans, Battiers, Mike Millers, Fryes, Wes Matthews, Danny Greens, Antawns, Wallys, Korvers, Verajaos, DRoses, etc., etc.

[Lost92Bricks]

You have to look at the playoffs. Lebron had teammates that had some incredible individual playoff performances.

Kyrie had some amazing games against the Warriors for example beyond anything any of MJ's teammates were capable of.

[falcolombardi]

You have to look at the playoffs. Lebron had teammates that had some incredible individual playoff performances.

Kyrie had some amazing games against the Warriors for example beyond anything any of MJ's teammates were capable of.

if you only look at scoring strictly then sure, kyrie 2016 finals (note he was not as good in 2017 and didnt play in 2015 finals) was a better scoring performance thsn may jordan teammate

i still would pick any pippen version between 91-97 over thst kyrie when co sidering their all around game (rebounding. playmaking, defense) and not only scoring

[Lost92Bricks]

if you only look at scoring strictly then sure, kyrie 2016 finals (note he was not as good in 2017 and didnt play in 2015 finals) was a better scoring performance thsn may jordan teammate

i still would pick any pippen version between 91-97 over thst kyrie when co sidering their all around game (rebounding. playmaking, defense) and not only scoring

As an overall player for a season yeah I agree but not in a playoff series.

[DCasey91]

We have AuPM (the 2nd latest version) supporting cast data for PS teams, going back to 1996-1997 (as it uses plus-minus data), and therefore a majority of MJ's career is not covered.

However, looking at the RS Cast (Relative AuPM value of the 2nd through the 8th-best player on a team, among players who logged at least 40 percent of team’s minutes. Value is relative to +0.75.) we can try and estimate the strength of those last 2 Bulls teams' supporting cast.

MJ's RS Cast in 97 was a 8.3 per the last version of AuPM and a 8.6 in 98.

97: 8.3
98: 8.6

If we look at Lebron's RS Cast from 08 to 2020:

07: -0.3
08: 2.2
09: 4.8
10: 1.2
11: 3.4
12: 4.2
13: 4.9
14: 1.8
15: 1.9
16: 2.2
17: 0.9
18: -1.9
19: Not a Playoff team
20: 1

Based off the 2nd lastest version of AuPM at least, it would seem as if MJ had a stronger RS supporting cast in 97 and 98 than Lebron had any point in his career.

I’d say because Pippen, Rodman, Harper (WS 5.7,7.2 97/98 and you can say he was the Bulls 5th best player lol) were all ++ defenders along with Kukoc providing elite offense for his size off the bench.

MJ always had a top3 cast or higher basically for the whole duration relative to his era.

[DCasey91]

Lebron has played with some poor rosters/overrated by fans and people here alike teams.

Names asides even that Heat team wasn’t as good as the whole 4 years, depth and injury was an issue.

Also just to add quickly it’s ironic that the East was “weak” in Lebron’s era (it was far from weak defensively) when MJ Bulls teams had more difficulty with defensively minded teams. Jazz, Sonic, Knicks, Heat, Pacers though they had Miller whose scoring efficiency was god tier pre age 37 in the playoffs averaged 23.5ppg (109 game sample size) on up to 11+ RTS is dumb efficient).

It’s a funny conundrum that one that I can’t get my head around.

Defense wins championships.

I’ll always say give Prime Pippen, Prime Grant with younger LBJ I’m sure he’d be successful. Better than Mo, Big Z or whatever the trash that the early Cavs FO came up with lmao.