ESPN's Real Plus Minus Confirms That Kobe Is A Net Negative

[RoundMoundOfReb]

http://espn.go.com/nba/statistics/rpm/_/sort/RPM/position/2

Real plus minus is plus/minus adjusted for opponents/players on the court etc. Kobe is at a -1.34 this year...to put that into perspective (and to show this isn't a stat concocted to make Kobe look bad) he was a +2.22 and likely a a big plus in his prime.

BTW Notable Kobe teammates (scapegoats):

Lin: -0.67
Boozer: -2.78
Ed Davis: +0.56


Kobe is done. Why can't Laker fans just admit it?

[riseagainst]

if he is done, why do you constantly do research on him and make threads on him?

[RoundMoundOfReb]

if he is done, why do you constantly do research on him and make threads on him?

To educate. Because an overwhelming majority of people on here seem to think that Kobe is a good player.

[riseagainst]

if you want to educate, you would break this down, maybe explain in depth why he is a negative impact overall. Not just google "kobe, sucks, rpm, nba" and copy and paste paragraphs. That's called plagiarizing.

[kennethgriffin]

this league is 1 on 1 now? cause i remember it being a 5 on 5 league

give kobe a few above average team mates. then talk sh*t

[RoundMoundOfReb]

this league is 1 on 1 now? cause i remember it being a 5 on 5 league

give kobe a few above average team mates. then talk sh*t
I already explained that this adjusts for teammates.

[Artillery]

Everybody already knows this. People aren't joking when they say Kobe puts up empty stats - this is some of the most egregious stat-padding I've ever seen in all my time watching the NBA. It's absolutely disgusting to see what a cancer he's been on the court. No effort on defense(worse than Harden at his worst) and plays like a ball-stopper on O. Just an awful, awful player chasing meaningless stats while his peers(Duncan, Dirk) are actually contributing to their team despite being just as old as Kobe.

+6.52 - Duncan(5th overall)
-1.34 - Kobe(238th overall)

[BigBoss]

He's playing on a d league team. Thread backfire

[fpliii]

this league is 1 on 1 now? cause i remember it being a 5 on 5 league

give kobe a few above average team mates. then talk sh*t
RPM adjusts for quality of teammates/opponents faced (uses ridge regression, a peer-reviewed mathematical process for solving overdetermined/underdetermined systems of equations...well RAPM does, that's how KG rates so high there; RPM involves some box score stats too). Still early in the season though, maybe Kobe will improve as the season progresses.

[Quizno]

I already explained that this adjusts for teammates.
so you would agree that lebron james is the 14th best player in the NBA, correct?

[Myth]

if you want to educate, you would break this down, maybe explain in depth why he is a negative impact overall. Not just google "kobe, sucks, rpm, nba" and copy and paste paragraphs. That's called plagiarizing.

He is giving ESPN credit and isn't trying to take credit himself.

[Bandito]

if he is done, why do you constantly do research on him and make threads on him?
Because he is living rent free, even though Kobe is a millionaire and could pay rent. That just shows how Alpha he is:rockon:

[RoundMoundOfReb]

Also not surprised to see Jonas as a net negative. He's overrated as hell.

[ArbitraryWater]

if you want to educate, you would break this down, maybe explain in depth why he is a negative impact overall. Not just google "kobe, sucks, rpm, nba" and copy and paste paragraphs. That's called plagiarizing.

:biggums:

[RoundMoundOfReb]

so you would agree that lebron james is the 14th best player in the NBA, correct?
Based on the 20 or so games played so far and no prior evidence? That seems about right. He's been adjusting to the new coach and roster. Expect him to be top 5 by the end of the season.

[fpliii]

Based on the 20 or so games played so far and no prior evidence? That seems about right. He's been adjusting to the new coach and roster. Expect him to be top 5 by the end of the season.
RPM does though. The initial seed is 2001, and each year uses the RPM from the previous season in constructing the prior.

As I said though, RPM includes box score stats, and isn't pure RAPM (just based on scoring margin from play-by-play). LeBron will rate better in standard RAPM.

[RoundMoundOfReb]

RPM does though. The initial seed is 2001, and each year uses the RPM from the previous season in constructing the prior.

As I said though, RPM includes box score stats, and isn't pure RAPM (just based on scoring margin from play-by-play). LeBron will rate better in standard RAPM.
Good to know :applause:

You got a link to standard RAPM rankings? Been trying to find it.

[Artillery]

+4.74 Kawhi Leonard(12th overall)
+2.92 Manu Ginobili(29th overall)
-2.46 - Tony Parker(352nd overall)

:oldlol: at how crappy TP is. Spurs fans weren't kidding when they said this guy has been awful the past two years. They practically won the title last year without him. Dude is so lucky to have played next to two legends in Duncan and Ginobili his whole career.

Most overrated point guard in the NBA.

[fpliii]

Good to know :applause:

You got a link to standard RAPM rankings? Been trying to find it.
Problem is, they're not standardized.

97-00: ascreamingcomesacrossthecourt.blogspot.com
08-14: gotbuckets.com

For 03-07, there's nothing super reliable, but J.E. (maker of RPM) posted these awhile back:

https://sites.google.com/site/rapmstats/

NPI doesn't use a prior, standard does.

I compute my own RAPM as well, when I have some new results (trying to clean the data up to produce prior-informed 01 and 02, which don't exist yet...J.E. computed 02, but used in-sample numbers as a prior) I'll PM them for you.

In one place though, this is the best link I think:

https://docs.google.com/spreadsheet/ccc?key=0Am56xyn_resAdHV2M185TGhoazBTcm5WMFV0eXZqa FE#gid=0

The owner of the sheet normalized them to 1.0 SD. Normally, RAPM corresponds to raw impact (i.e. how much better or worse a guy's impact is per 100 possessions).

Hope that helps.

[fpliii]

+4.74 Kawhi Leonard(12th overall)
+2.92 Manu Ginobili(29th overall)
-2.46 - Tony Parker(352nd overall)

:oldlol: at how crappy TP is. Spurs fans weren't kidding when they said this guy has been awful the past two years. They practically won the title last year without him. Dude is so lucky to have played next to two legends in Duncan and Ginobili his whole career.

Most overrated point guard in the NBA.
Could not agree more. Makes me legitimately angry when people hype him up, as it's ALWAYS in an attempt to diminish Duncan.

People try and put him on CP3's or Nash's level. :oldlol: :oldlol: :oldlol:

[Quizno]

Based on the 20 or so games played so far and no prior evidence? That seems about right. He's been adjusting to the new coach and roster. Expect him to be top 5 by the end of the season.
fair enough. i don't understand this statistic enough to really comment on it, but you're crazy if you think kobe is actually a below average player. he's still easily an above average player and a borderline all-star in terms of how well he's playing. he's definitely not done. he just can't lead a team of scrubs to the playoffs or even to be really competitive. offensively, he's still a capable first option but would thrive more in a second option role at this point. he just needs a team around him. he'd be PERFECT for a team like the bulls, for example, where he has teammates that can create looks for each other, cover for him on defense, but still has an opportunity to iso and get buckets.

on top of that, you have to consider that kobe's coming off an achilles injury and he's still adjusting to the game. it's still early in the season and his shooting percentages will go up. his defense is definitely his biggest problem at this point, and it's particularly bad because there's also no interior defense at all so he can't lead offensive players to bynum or pau back when he was a decent rim protector.

saying kobe is done is just premature, stupid and shows that you're trying to flat out hate on the guy. that other measure of how well a player is doing that stat nerds love to throw out is PER, where kobe's currently at 19. that indicates a well above-average player. so which is it?

[gts]

if you want to educate, you would break this down, maybe explain in depth why he is a negative impact overall. Not just google "kobe, sucks, rpm, nba" and copy and paste paragraphs. That's called plagiarizing.

He can't do that because he's using the stat wrong


There was quite a discussion on this on one of the analytic sites and their biggest fear is that people would use this stat to compare players against other players without proper context...

look at how many average role players you have with positive +- numbers on great teams and that shows you right there the problem of using it like the OP has just done

[RoundMoundOfReb]

Problem is, they're not standardized.

97-00: ascreamingcomesacrossthecourt.blogspot.com
08-14: gotbuckets.com

For 03-07, there's nothing super reliable, but J.E. (maker of RPM) posted these awhile back:

https://sites.google.com/site/rapmstats/

NPI doesn't use a prior, standard does.

I compute my own RAPM as well, when I have some new results (trying to clean the data up to produce prior-informed 01 and 02, which don't exist yet...J.E. computed 02, but used in-sample numbers as a prior) I'll PM them for you.

In one place though, this is the best link I think:

https://docs.google.com/spreadsheet/ccc?key=0Am56xyn_resAdHV2M185TGhoazBTcm5WMFV0eXZqa FE#gid=0

The owner of the sheet normalized them to 1.0 SD. Normally, RAPM corresponds to raw impact (i.e. how much better or worse a guy's impact is per 100 possessions).

Hope that helps.

LeBron James 2010 :bowdown: :bowdown: :bowdown: ****ing GOAT

Thanks btw. Do you know when rapm for this season will be available?

[Artillery]

People try and put him on CP3's or Nash's level. :oldlol: :oldlol: :oldlol:

It's always hilarious to hear people say he's better than Paul because Parker's team beats CP's team in the playoffs(despite the fact that Paul's head-to-head numbers are so much better than his). Same goes for Nash. No doubt the Spurs offense would look all-time great tier with those two at point instead of TP.

[fpliii]

fair enough. i don't understand this statistic enough to really comment on it, but you're crazy if you think kobe is actually a below average player. he's still easily an above average player and a borderline all-star in terms of how well he's playing. he's definitely not done. he just can't lead a team of scrubs to the playoffs or even to be really competitive. offensively, he's still a capable first option but would thrive more in a second option role at this point. he just needs a team around him. he'd be PERFECT for a team like the bulls, for example, where he has teammates that can create looks for each other, cover for him on defense, but still has an opportunity to iso and get buckets.

on top of that, you have to consider that kobe's coming off an achilles injury and he's still adjusting to the game. it's still early in the season and his shooting percentages will go up. his defense is definitely his biggest problem at this point, and it's particularly bad because there's also no interior defense at all so he can't lead offensive players to bynum or pau back when he was a decent rim protector.

saying kobe is flat out done is just premature, stupid and just shows that you're trying to flat out hate on the guy. that other measure of how well a player is doing that stat nerds love to throw out is PER, where kobe's currently at 19. that indicates a well above-average player. so which is it?
RPM (and xRAPM, RAPM, etc., all variants) only compares players in the roles they are playing. It just means that Kobe in his current role is having a negative impact, when adjusting for teammates and opponents. Doesn't mean in a different role he wouldn't be productive, and isn't an indictment of his earlier career (during his peak years from 06-08, he rates as the best offensive player in the league).

But yeah, the people championing PER aren't supporting RPM. PER is a garbage, it's a formula of linear weights of box score stats. RPM (again, and xRAPM, RAPM, etc.) requires legitimate mathematical and programming knowledge/ability to produce. It's a peer reviewed mathematical procedure to emulate ordinary least squares, when the number of equations isn't exactly equal to the number of unknown variables. Teams have been using it since 01 (the Mavericks were the first to use APM), and it's still the best metric out there. There's no question about that. Again, it doesn't dislike Kobe. It's just saying that right now, in his present role, he is very far removed from his impact in his prime.

[Quizno]

RPM (and xRAPM, RAPM, etc., all variants) only compares players in the roles they are playing. It just means that Kobe in his current role is having a negative impact, when adjusting for teammates and opponents. Doesn't mean in a different role he wouldn't be productive, and isn't an indictment of his earlier career (during his peak years from 06-08, he rates as the best offensive player in the league).

But yeah, the people championing PER aren't supporting RPM. PER is a garbage, it's a formula of linear weights of box score stats. RPM (again, and xRAPM, RAPM, etc.) requires legitimate mathematical and programming knowledge/ability to produce. It's a peer reviewed mathematical procedure to emulate ordinary least squares, when the number of equations isn't exactly equal to the number of unknown variables. Teams have been using it since 01 (the Mavericks were the first to use APM), and it's still the best metric out there. There's no question about that. Again, it doesn't dislike Kobe. It's just saying that right now, in his present role, he is very far removed from his impact in his prime.
yeah i'm not arguing that at all, i'm just arguing against the "kobe is done" narrative. being done indicates that a player can no longer contribute in the NBA, and anybody who's seen kobe play this season knows that's not true. it's sillly when people (such as OP) try to use objective statistics to support their subjective (biased) opinions.

is kobe done being able to carry a squad of bad players to the playoffs? clearly. is he no longer capable of playing a big role on a competitive team? absolutely not

[fpliii]

LeBron James 2010 :bowdown: :bowdown: :bowdown: ****ing GOAT

Thanks btw. Do you know when rapm for this season will be available?
Well you have to normalize for standard deviations to compare between seasons. But I think other than 04 KG I think 10 LeBron is the highest on record (since 96-97) when adjusted.

[RoundMoundOfReb]

is kobe done being able to carry a squad of bad players to the playoffs? clearly. is he no longer capable of playing a big role on a competitive team? absolutely not
This is the problem with Kobe though. If he was in a Paul Pierce type of role ( a mentor who picks his spots on offense) he could still be an effective player, however, there is no way he would accept that role or any other role other than 1st option taking 20 fgas/game - and thus he is done.

[fpliii]

yeah i'm not arguing that at all, i'm just arguing against the "kobe is done" narrative. being done indicates that a player can no longer contribute in the NBA, and anybody who's seen kobe play this season knows that's not true. it's sillly when people (such as OP) try to use objective statistics to support their subjective (biased) opinions.

is kobe done being able to carry a squad of bad players to the playoffs? clearly. is he no longer capable of playing a big role on a competitive team? absolutely not
Oh, obviously Kobe isn't finished as an NBA player, or even as an All Star caliber player perhaps. He can still contribute positively on a winning team.

I've watched every game this season (painfully). But obviously he's a negative in this role. Doesn't matter since there's no reason to win, and the goal is to tank.

[RoundMoundOfReb]

Well you have to normalize for standard deviations to compare between seasons. But I think other than 04 KG I think 10 LeBron is the highest on record (since 96-97) when adjusted.
According to the links you posted LeBron 2010 is the highest ever. Not surprising. LeBron has a top 5 peak all time.

[Quizno]

This is the problem with Kobe though. If he was in a Paul Pierce type of role ( a mentor who picks his spots on offense) he could still be an effective player, however, there is no way he would accept that role or any other role other than 1st option taking 20 fgas/game - and thus he is done.
do you really think that's what the lakers want him to do? do you think that's what upper management wants him to do? what the fans want him to do? there's really 2 options here:

1) take about 10 fga/game, pass the ball a lot, don't really create for your teammates, lose almost every game

2) take 20+ fga/game, create for teammates (but also turn the ball over and miss a lot of shots), give the fans something at least interesting to watch, lose almost every game

kobe isn't taking jeremy lin, wesley johnson and carlos boozer to the playoffs in this conference. it's just not happening. if he were in the east that might actually be a different story, but he's not. this season was a lost cause before it started, kobe clearly has one goal for this season and it's to pass MJ in points. the fans know this team can't win and the if the lakers management knows one thing, it's that people in LA love kobe. shoot away

[chazzy]

LeBron James 2010 :bowdown: :bowdown: :bowdown: ****ing GOAT

The year he mailed it in against the Celtics? Damn, our league has low standards for its GOAT :oldlol:

[RoundMoundOfReb]

1 bad game = mail it in...:oldlol:

Guess Kobe mailed it in vs them in game 7 as well/

[pastis]

i wonder why some members are thinking that roundmoundofreb has a kobe agenda.
he is just right. he admitted, that kobe had a good prime and that kobe won due to goat coach, goat peak player and very stacked teams. that istn against kobe. that are just facts. the west is tough, so you need super stack team and goat coach for being relevant over a longer stretch.
right now kobe is playing shit. so what:confusedshrug:
roundmoundofreb also sees kobe top15. agenda where:confusedshrug:

roundmoundofreb is just an objective unbiased basketball poster

[Quizno]

According to the links you posted LeBron 2010 is the highest ever. Not surprising. LeBron has a top 5 peak all time.
he's saying that if you want to compare seasons, you have to standardize them into z-scores because you can't compare two statistics with different scales (means and standard deviations)

[chazzy]

1 bad game = mail it in...:oldlol:

Guess Kobe mailed it in vs them in game 7 as well/
Try the last 3 games after being up 2-1. F**ing GOAT, doe

[riseagainst]

1 bad game = mail it in...:oldlol:


he's was up 2:1 with home court advantage. Then lost the next 3. That's gotta hurt.

[fpliii]

According to the links you posted LeBron 2010 is the highest ever. Not surprising. LeBron has a top 5 peak all time.
Guess he's normalizing differently than I am (I use a cutoff of guys who played 24 equivalent minutes per game over a season, so my SDs might be different), but the owner of that spreadsheet is legit, I don't have a problem with that.

Not sure if LeBron is top 5, but he might be. Top 2 peaks since I started watching are 92-93 Hakeem and 99-00 Shaq. I missed MJ's peak (which, from going back and watching tape, seems to be one of 89-90, 90-91, 91-92), but obviously he's in the mix, as are LeBron, Kareem, etc.

[J Shuttlesworth]

he's was up 2:1 with home court advantage. Then lost the next 3. That's gotta hurt.
Yeah totally his fault. 27/9/7 isn't enough when you're on the Cavs

[fpliii]

he's saying that if you want to compare seasons, you have to standardize them into z-scores because you can't compare two statistics with different scales (means and standard deviations)
The Google Docs link does that I believe. It just depends on what the cutoff is for calculating SD (he used everybody in the league I believe, while I use the equivalent of 24mpg over a season; some others use 375 minutes in a season, because that works well).

I'm really just nitpicking though. It's probably going to be #1 or at worst #2 no matter what criteria are used. Actually, I think that spreadsheet is using J.E.'s numbers. gotbuckets.com's numbers are newer (meaning, they used a cleaner datasource, improved priors), but they still have 10 LeBron about the same.

[I<3NBA]

this league is 1 on 1 now? cause i remember it being a 5 on 5 league

give kobe a few above average team mates. then talk sh*t
i find this so funny on so many levels, not least because Lebron fans used to be the ones to say this :lol :lol :lol and Kobe stans gave them so much grief for it, and now the shoe is on the other foot... :lol :lol :lol

[Quizno]

The Google Docs link does that I believe. It just depends on what the cutoff is for calculating SD (he used everybody in the league I believe, while I use the equivalent of 24mpg over a season; some others use 375 minutes in a season, because that works well).

I'm really just nitpicking though. It's probably going to be #1 or at worst #2 no matter what criteria are used. Actually, I think that spreadsheet is using J.E.'s numbers. gotbuckets.com's numbers are newer (meaning, they used a cleaner datasource, improved priors), but they still have 10 LeBron about the same.
word. yeah i'm taking my last statistics class at my university right now and actually have my final in a week so this stuff is kind of interesting to me right now and i can actually understand wtf you're talking about, but i don't have time to really delve into RPM right now and see how it's calculated. did you study stats in college or something? you know your shit

[fpliii]

word. yeah i'm taking my last statistics class at my university right now and actually have my final in a week so this stuff is kind of interesting to me right now and i can actually understand wtf you're talking about, but i don't have time to really delve into RPM right now and see how it's calculated. did you study stats in college or something? you know your shit
Majored in math, did my grad work in stats. I actually didn't take a dedicated regression class (studied more theoretical topics, focussed on Bayesian coursework; of course early applied stats classes teach OLS), so I learned a lot of this stuff on the fly.

On some level it's just basic matrix algebra, with some optimization. Once you understand that, you can proceed. Calculating isn't terribly difficult, as long as you know your way around Excel and a program like R or Stata.

[riseagainst]

Yeah totally his fault. 27/9/7 isn't enough when you're on the Cavs

21-11-8 on 34% and 6.3 turnovers is definitely not enough. Sure his rebounds and assists went up compared to his first 3 games, but that's only because he couldnt score now so now he's passing and rebounding, which is great. But he's also turning it over 4 more times for 1 more assist.

But dipping from 32 points on 54%FG for the first 3 games down to 21 points on 34% in the last 3. That's a Wilt Chamberlain-esque drop off.

[Quizno]

Majored in math, did my grad work in stats. I actually didn't take a dedicated regression class (studied more theoretical topics, focussed on Bayesian coursework; of course early applied stats classes teach OLS), so I learned a lot of this stuff on the fly.

On some level it's just basic matrix algebra, with some optimization. Once you understand that, you can proceed. Calculating isn't terribly difficult, as long as you know your way around Excel and a program like R or Stata.
that's cool man, good on you for putting all that work in. lmk if you ever develop a stat that confirms that gerald wallace is the GOAT, i love that guy. thanks.

[fpliii]

that's cool man, good on you for putting all that work in. lmk if you ever develop a stat that confirms that gerald wallace is the GOAT, i love that guy. thanks.
:lol

Good luck to you with your exams bro. :cheers: Good talk.

[DMAVS41]

RPM does though. The initial seed is 2001, and each year uses the RPM from the previous season in constructing the prior.

As I said though, RPM includes box score stats, and isn't pure RAPM (just based on scoring margin from play-by-play). LeBron will rate better in standard RAPM.

Actually...I read that this years RPM does not use any seeding and it's only based on the games played so far this season.

Which would make sense given the results actually.

[AirFederer]

So you're saying Kobe doesn't excel in RAP'M like in his prime?

[NoGunzJustSkillz]

if he is done, why do you constantly do research on him and make threads on him?
RoundMoundOfReb is a certified weirdo.

[Yao Ming's Foot]

i find this so funny on so many levels, not least because Lebron fans used to be the ones to say this :lol :lol :lol and Kobe stans gave them so much grief for it, and now the shoe is on the other foot... :lol :lol :lol

:facepalm

2007 Cavs Def Rtg: 101.3 (4th of 30)
2014 Lakers Def Rtg: 115.2 (30th of 30)

[J Shuttlesworth]

21-11-8 on 34% and 6.3 turnovers is definitely not enough. Sure his rebounds and assists went up compared to his first 3 games, but that's only because he couldnt score now so now he's passing and rebounding, which is great. But he's also turning it over 4 more times for 1 more assist.

But dipping from 32 points on 54%FG for the first 3 games down to 21 points on 34% in the last 3. That's a Wilt Chamberlain-esque drop off.
So you're saying LeBron couldn't go 6-24 like Kobe and still get bailed out by teammates?

[Yao Ming's Foot]

So you're saying LeBron couldn't go 6-24 like Kobe and still get bailed out by teammates?

Lebron shot 41.6% in the playoffs surrounded by "nothing" and made the Finals. :confusedshrug: