Is Kevin Garnett a defensive anchor?

[Geaux_Hawks]

G35 are you really posting team Drtg again? That's been covered in this thread. All your doing is giving Howard more credit for having better defensive teammates and playing in a better system. By APM and on/off numbers, KG's impact is clearly greater, yet you keep ignoring that...

What he really isn't paying attention to is how the Wolves had better dRtg than the Magic despite where they were ranked in multiple years.

[colts18]

G35 are you really posting team Drtg again? That's been covered in this thread. All your doing is giving Howard more credit for having better defensive teammates and playing in a better system. By APM and on/off numbers, KG's impact is clearly greater, yet you keep ignoring that...

What he really isn't paying attention to is how the Wolves had better dRtg than the Magic despite where they were ranked in multiple years.

What part of era differences don't you understand? By raw team Drating the 2004 TWolves would have been #1 this year over the Celtics and Bulls and so would 5 other teams. By raw team D rating this year's Spurs who were 11th in 2011, would be 25th in 2004. Do you not understand that those things like handchecking and illegal defense actually go against your premise because they made it easier to play defense hence the lower team D ratings.

[Rapcity_11]

Difference to league average is more important that raw numbers, so I can't agree with you there. Era differences exist and shouldn't be ignored.

[Geaux_Hawks]

Hand checking does not make a huge difference. If you are willing to say it does, then it helps prove KG never had a supporting cast with any defensive value.

[colts18]

Difference to league average is more important that raw numbers, so I can't agree with you there. Era differences exist and shouldn't be ignored.

Here are there team d rating differentials from league average from their 3rd year onward:

98: +2.1 (negative is good)
99: -0.7
00: -0.7
01: +0.7
02: +0.8
03: +0.2
04: -3.2 (only great year)
05: +0.5
06: -1.7
07: +1.4
Average: -0.06

Howard:
07: -2.4
08: -2.0
09: -6.4
10: -4.2
11: -5.5
Average: -4.1

So Howard's average is better than any of KG's years and all of his 5 years are better than any KG year except 1.

[Rapcity_11]

colts - What you're telling us there is that the Magic are better than the Wolves defensively.

When it's broken down further, it's clear that KG has the bigger impact on his team. This has been beaten to death by myself, drza and Doc MJ in this thread and keeps getting ignored.

[Doctor MJ]

3 points is good, I'll even go so far to say I think Howard does a bit more than this. Obviously though, when the difference between the best and worst defenses is more than a 10 point difference, a 3-ish point impact player won't make scrubs elite.

Ummmm someone explain to me what the STANDARD is for an elite defensive rating? Dr. MJ was right in that THERE ISN'T a large between the best teams and the worst teams. This isn't college where Duke gets to play Western Michigan. Three points is the difference between being rated 4th defensively and being rated 11th. That's the small window you are working with in the NBA. People don't realize every point is huge in a basketball game. Where do you think the idea of "You only have to watch the last two minutes of a game." came from?

The difference between the 1st and 2nd rated defenses was .2 points. The difference between the Bulls #1 rated defense and the Magic was less than 3 points. So exactly what more will it take to make the Magic elite.....

Huh. Not sure I understand what you're asking about. You appear to be interpreting my statement as saying that the Magic aren't really elite. I'm not saying anything about the Magic specifically.

What I'm saying is simply that the gap between the best & worst teams is quite a bit bigger than what one player's impact is. Hence all players' team success is contingent on the rest of the team, and if a player has some elite team success then some credit needs to go to teammates.

Hence, while team success is certainly a good thing for a player in a player comparison, it must be factored in at a reasonable level. The mere fact that another player had some seasons of weak team success is not sufficient to decide the comparison.

[WhateverBro]

lol @ relative to his teammates... meaning he needs a very good defensive supporting cast to make the kind of impact Dwight Howard makes with inferior talent.

and to drza pushing the KG hasn't had anybody remotely competent defensively agenda. He had more than acceptable defensive talent throughout his prime years in Minnesota. To name a few -- Chauncey Billups, Trenton Hassell, Latrell Sprewell, Eddie Griffin, Ervin Johnson, Rasho Nesterovic; it didn't translate to an elite team defense because KG isn't a true defensive anchor.

Reaching. You just listed players based on reputation? I assume.

Because Sprewell and Nesterovic wasn't more than acceptable defensively. Sprewell wasn't the defensive player he once was when he played with the Wolves. He gambled a lot defensively and was pretty lazy on the defensive end.

Eddie Griffin was a great help defender but other than that.. He was very slow defensively and didn't play the pick and roll well. Aside from blocking, he wasn't as intersted in defense as you may think. Ervin Johnson, sure he was solid but didn't play enough minutes to make a bigger impact.

The only one you listed that stands out is Hassell; who always was a great overall defender with the Wolves.

I think the Wolves would've gotten away with having OK defensive players during those years because of Garnetts ridiculous impact defensively. The problem was; they didn't have OK defensive players overall, they were liabilties. Hudson, Szczerbiak, Blount, Ricky Davis, Jaric, Olowokandi etc etc. Those are just awful defenders who would make a mess defensively on any team. It's just too bad Garnett had to deal with a lot of these players at the same time.

[WhateverBro]

Difference to league average is more important that raw numbers, so I can't agree with you there. Era differences exist and shouldn't be ignored.

Here are there team d rating differentials from league average from their 3rd year onward:

98: +2.1 (negative is good)
99: -0.7
00: -0.7
01: +0.7
02: +0.8
03: +0.2
04: -3.2 (only great year)
05: +0.5
06: -1.7
07: +1.4
Average: -0.06

Howard:
07: -2.4
08: -2.0
09: -6.4
10: -4.2
11: -5.5
Average: -4.1

So Howard's average is better than any of KG's years and all of his 5 years are better than any KG year except 1.

.......So Magic are better than Wolves defensively? Really? That's a shock.

Going by this logic; giving the best defensive player all the credit for a teams defense would result in Garnett somehow becoming a better defensive player during his Celtics years. Which of course anyone who has actually watched Garnett on the Wolves knows isn't true...

[NO-KG-AI]

You can't argue about KG on RealGM. Too many people made too many assumptions and jumped to too many conclusions about a player they rarely watched, and when tons of data came out showing that KG was overwhelmingly dominant on defense, they can't just swallow their pride and admit they are wrong.

[colts18]

Since when is APM, RAPM the end all be all for statistical analysis? The error levels and sample sizes problems are so huge that you have to take the data with a grain of salt or else you start believing crazy stuff like Nick Collison being more valuable than Durant, Kobe, LeBron, Wade, Howard. At least with stats like PER, you never got a discrepancy like that. Those guys will rise to the top. Sure you might get a guy like Ginobili who looks good by PER and might be ahead of a Kobe, but you never get the Nick Collison.

By RAPM, Garnett has been a top 5 in every season from 08-11 including 3 top 3 finishes. RAPM said that Garnett and Ginobili have had the most impact in the past 4 years. Do you really believe a stat that says that past his prime Garnett is a top 3 player in the NBA ahead of CP3, Kobe, Dwight, Durant, etc.?

[HeatRing2012]

Since when is APM, RAPM the end all be all for statistical analysis? The error levels and sample sizes problems are so huge that you have to take the data with a grain of salt or else you start believing crazy stuff like Nick Collison being more valuable than Durant, Kobe, LeBron, Wade, Howard. At least with stats like PER, you never got a discrepancy like that. Those guys will rise to the top. Sure you might get a guy like Ginobili who looks good by PER and might be ahead of a Kobe, but you never get the Nick Collison.

By RAPM, Garnett has been a top 5 in every season from 08-11 including 3 top 3 finishes. RAPM said that Garnett and Ginobili have had the most impact in the past 4 years. Do you really believe a stat that says that past his prime Garnett is a top 3 player in the NBA ahead of CP3, Kobe, Dwight, Durant, etc.?

well, I've seen with my own eyes that Ginobili has more impact than CP3 when he's on the court, you can't deny that. I don't know if its due to the better supporting cast but as soon as he subs in the Spurs go on a run.
Imho It's his age (34) that keeps him off the court for a long time. so while Manu is limited to play 30 minutes due to condition issues or health, CP3 can play 36 or up to 42 minutes (PS) for that matter.
in those extra 12 minutes he's on the court he'll produce more stats.
if we adjust fpr p36 numbers Gino is crushing CP3 in points, rebounds and blocks. he's also the better shooter (TS%). so theres that.
is Gino a better player than CP3? no, he lacks the passing (CP3 has twice as much assists) and mainly the age is a big concern. however he is impacting the game more in his minutes than CP3.

same with KG, he's doomed to be 30 minutes per game guy like Gino due to his 35 years of age. but as soon as he's on the court he's controlliong the defense.

don't get me wrong, guys like CP3 can reach higher ceilings and more highlights when they get their occasional superstar burst, but those RAPM heros like KG and Gino are doing elite stuff in their limited on court time too

[G35]

G35 are you really posting team Drtg again? That's been covered in this thread. All your doing is giving Howard more credit for having better defensive teammates and playing in a better system. By APM and on/off numbers, KG's impact is clearly greater, yet you keep ignoring that...

Tell me the better defensive teammates Dwight has had because I don't see it.

His impact is clearly greater? Where? In APM? On/Off numbers? And what does that say? That in 12 years KG was only able to get out of the first round once? I guess I am not "deep" enough to look at these advanced numbers that people pull out to make cases for players. But I look at results.

Performance + Results = Impact

A lot of players look good statistically. But not so good in the bottom line categories.

Wilt, Oscar, Barkley, Nash, Wilkins, Ewing, DRob and KG.

I see people trying to rate KG better than DRob and Barkley based off of these stats and completely change perception.

Barkley was seen as the second best player behind Jordan. This is when Magic, Bird, Hakeem, Ewing, Malone, Stockton, Pippen, Drexler were playing. I don't remember KG ever being seen as the best or even 2nd best player in the NBA even when he won the MVP. Shaq, Kobe, and Duncan were all seen as clearly better. But now people start pulling these stats out and it seems certain players are given these incredible passes.

KG is not the only player to have to play with scrub teammates. But he along with a few others seem to get more excuses and failures explained away than others.

If any other star had missed the playoffs 4 times, lost in the first round 7 times they would be no doubt they would not be seen in the same light as KG. You can't join 2 other potential HoF's, one of the premier defensive coaches of this era, a PG that has made the All defensive team the last 3 years and then be the singular reason why "it works".

And finally you can't cherry pick the stats you want to use in these comparisons....it makes ALL the stats suspect then......

[Doctor MJ]

I don't remember KG ever being seen as the best or even 2nd best player in the NBA even when he won the MVP. Shaq, Kobe, and Duncan were all seen as clearly better. But now people start pulling these stats out and it seems certain players are given these incredible passes.

Well, you remember incorrectly then. In '03-04 it was generally accepted that Shaq was no longer in his peak form, and Kobe certainly was not more established than KG. The debate was between Garnett & Duncan for the top spot.

I'll also say that after that year, video games gave Garnett the highest ratings. Obviously you can say "Who cares about video games?", but they do put a ton of effort into their ratings, and by no means does winning the MVP guarantee you the top rating. After Nash won his first MVP I believe he was still rated behind two of his teammates.

KG is not the only player to have to play with scrub teammates. But he along with a few others seem to get more excuses and failures explained away than others.

If any other star had missed the playoffs 4 times, lost in the first round 7 times they would be no doubt they would not be seen in the same light as KG. You can't join 2 other potential HoF's, one of the premier defensive coaches of this era, a PG that has made the All defensive team the last 3 years and then be the singular reason why "it works".

And finally you can't cherry pick the stats you want to use in these comparisons....it makes ALL the stats suspect then......

I think the key thing I want you to understand that there is a contingent of us that come to the pro-Garnett side of things as a result of considering everything, not because we have some weird KG-homer thing going. Just speaking for myself, by no means do I ignore the weak team success in my evaluation of KG. Indeed I used to hold it against him more strongly, but from my subsequent analysis it largely seems to me to that knocking KG too harshly for that teams stuff is just doing the classic backwards thinking that praises those fortunate enough to luck into good team situations.

Not saying you have to agree, but I'm really not cherry picking stuff to make my guy look good. I see the whole field and with great reluctance have come to believe it's easier to "explain away" the bad stuff than to "explain away" the good stuff.

[Rapcity_11]

G35 are you really posting team Drtg again? That's been covered in this thread. All your doing is giving Howard more credit for having better defensive teammates and playing in a better system. By APM and on/off numbers, KG's impact is clearly greater, yet you keep ignoring that...

Tell me the better defensive teammates Dwight has had because I don't see it.

His impact is clearly greater? Where? In APM? On/Off numbers? And what does that say? That in 12 years KG was only able to get out of the first round once? I guess I am not "deep" enough to look at these advanced numbers that people pull out to make cases for players. But I look at results.

Performance + Results = Impact

A lot of players look good statistically. But not so good in the bottom line categories.

Wilt, Oscar, Barkley, Nash, Wilkins, Ewing, DRob and KG.

I see people trying to rate KG better than DRob and Barkley based off of these stats and completely change perception.

Barkley was seen as the second best player behind Jordan. This is when Magic, Bird, Hakeem, Ewing, Malone, Stockton, Pippen, Drexler were playing. I don't remember KG ever being seen as the best or even 2nd best player in the NBA even when he won the MVP. Shaq, Kobe, and Duncan were all seen as clearly better. But now people start pulling these stats out and it seems certain players are given these incredible passes.

KG is not the only player to have to play with scrub teammates. But he along with a few others seem to get more excuses and failures explained away than others.

If any other star had missed the playoffs 4 times, lost in the first round 7 times they would be no doubt they would not be seen in the same light as KG. You can't join 2 other potential HoF's, one of the premier defensive coaches of this era, a PG that has made the All defensive team the last 3 years and then be the singular reason why "it works".

And finally you can't cherry pick the stats you want to use in these comparisons....it makes ALL the stats suspect then......

Everything always comes back to team success for you. It's unfortunate because it doesn't allow you to compare individuals fairly...

[WhateverBro]

Tell me the better defensive teammates Dwight has had because I don't see it.

His impact is clearly greater? Where? In APM? On/Off numbers? And what does that say? That in 12 years KG was only able to get out of the first round once? I guess I am not "deep" enough to look at these advanced numbers that people pull out to make cases for players. But I look at results.

Performance + Results = Impact

A lot of players look good statistically. But not so good in the bottom line categories.

Wilt, Oscar, Barkley, Nash, Wilkins, Ewing, DRob and KG.

I see people trying to rate KG better than DRob and Barkley based off of these stats and completely change perception.

Barkley was seen as the second best player behind Jordan. This is when Magic, Bird, Hakeem, Ewing, Malone, Stockton, Pippen, Drexler were playing. I don't remember KG ever being seen as the best or even 2nd best player in the NBA even when he won the MVP. Shaq, Kobe, and Duncan were all seen as clearly better. But now people start pulling these stats out and it seems certain players are given these incredible passes.

KG is not the only player to have to play with scrub teammates. But he along with a few others seem to get more excuses and failures explained away than others.

If any other star had missed the playoffs 4 times, lost in the first round 7 times they would be no doubt they would not be seen in the same light as KG. You can't join 2 other potential HoF's, one of the premier defensive coaches of this era, a PG that has made the All defensive team the last 3 years and then be the singular reason why "it works".

And finally you can't cherry pick the stats you want to use in these comparisons....it makes ALL the stats suspect then......

What? You're clearly remembering things wrong. His MVP year he was clearly the best player in the NBA and I don't remember anyone saying otherwise. Both Kobe and Shaq didn't have a great year and Garnett had Duncan beat across the board and led his team further in the playoffs.

Look at the MVP voting for that year: http://www.basketball-reference.com/awards/awards_2004.html

120 out of 123 first place votes? Come on now.

I've seen some ridiculous claims on Garnett on this board, even in this topic but this is a new one. How can anyone look back at the 03-04 season and say Garnett wasn't the best?

[Doctor MJ]

Since when is APM, RAPM the end all be all for statistical analysis? The error levels and sample sizes problems are so huge that you have to take the data with a grain of salt or else you start believing crazy stuff like Nick Collison being more valuable than Durant, Kobe, LeBron, Wade, Howard. At least with stats like PER, you never got a discrepancy like that. Those guys will rise to the top. Sure you might get a guy like Ginobili who looks good by PER and might be ahead of a Kobe, but you never get the Nick Collison.

By RAPM, Garnett has been a top 5 in every season from 08-11 including 3 top 3 finishes. RAPM said that Garnett and Ginobili have had the most impact in the past 4 years. Do you really believe a stat that says that past his prime Garnett is a top 3 player in the NBA ahead of CP3, Kobe, Dwight, Durant, etc.?

I'm totally with about the "grain of salt", but the question then is exactly what that entails. For me it's about about figuring out how much noise you think is involved, and behaving with confidence accordingly. For other people it's clearly about throwing the stat out altogether when it gives them results they don't like. Obviously, I don't think those people are being reasonable.

This is a list of the top big minute defenders by 3 different many-year APM samples across different periods.



http://asubstituteforwar.com/2011/04/23 ... o-garnett/

Look at how close the rest of the top 10 is, and far away Garnett is. My "grain of salt" is certainly big enough that I'm not going to use it to say the #5 guy did clearly better than the #6 guy, but it's not big enough to dismiss the lead Garnett has.

Understand also how I apply this knowledge: The situation is that Garnett has disappointing team success. We all know that that can happen due to a very weak supporting cast situation. Were that the case, what data would we expect to see? That the team actually did okay when Garnett played, but that the team fell apart without him. And of course this is exactly what happened.

So now we have a player that won the MVP, had his star fall due to weak team success that seems entirely explainable, and then spearheads the single greatest team turnaround in NBA history. What's the simplest explanation? That he really was the player people thought he was win he won the MVP, and that he then faced some remarkably bad luck.

Now, I realize the thread is technically talking about whether he's a defensive anchor, not whether he was a worthy MVP-level player, but to me the questioning of both largely revolves around the same issue.

Re: Collison. This is a good thing to talk about. Do I consider him a secret superstar? No I don't. +/- is not the be all end all. On what basis do seemingly accept +/- for other but not Collison? Well a few things:

1) Again, I don't rank guys strictly by +/-. I defend Garnett so fiercely because the +/- answers the one blot on his resume, but he's obviously got plenty of other stuff to go with it.

2) One issue with +/- is with using it to judge guys who are platooned. Understand, if I have a player who can really help the team in a certain situation, but is too limited to be used in others, then he ends up having his +/- distorted. Stars are out there in the best of times and the worst of times and thus are given a much more "honest" assessment by +/- in terms of their actual capabilities. So I look at someone like Collison and think he's a great role player, and question whether he should play more, but by no means would I try to build a team around him.

3) Note that with players like Ginobili and older Garnett you can ask similar questions. I'd be inclined to say they really are having huge impact when they play, but the fact that they play less limits their overall impact, and distorts things in the sense that if they were asked to play bigger minutes their lack of stamina would lower their quality of play.

4) I'm glad to have Engelmann's numbers here, but I'm also a little wary of them too when they show things I don't see elsewhere. Collison didn't appear super-high on the leader boards of previous multi-year samples (unlike Ginobili and Garnett), so I've an additional sense of caution with trusting Collison's placement than I do with other players.

5) And this also relates to the fact that we've never really seen role players consistently stay up on the +/- leaderboards unless you count Ginobili as a role player (which I don't). I look at someone like Shane Battier who probably exemplifies the modern role player and unlike Collison has typically been a big minute guy. Long term +/- studies tend to rank him as having an impact as a Top 40 player, which is quite good when you think about it. Good enough that it's easy to imagine him having a better career impact than the Randolphs and Arenases of the world but not good enough that you'd think that secret superstars really exist among role players. (Also worth noting that by career Win Shares, a box score based stat, Battier ranks 4th among his 2001 draftmates above such guys.)

[microfib4thewin]

I like how some still argue that Duncan is protected by great defensive teammates. Bowen is arguably the only constant in Duncan's career. Now, defensive metrics for basketball are still primitive, so how much impact Bowen has on defense is not fully figured out, but I have a hard time believing that the Spurs had a <100 Drtg for 10 straight years simply because Pop has a great scheme and Bowen is a GOAT perimeter defender. When Duncan's minutes started catching up to him in 09 the Spurs defense was no longer the same, with or without Bowen. If Pop is really this defensive genius then why couldn't he still maintain the defensive identity with his team?

I don't think APM is everything, but the statistical evidences that are present are at least strong enough to not fully conclude that KG is overrated as a defender. KG is a pretty odd case, raw numbers, defensive rating, and +/- are very favorable for him, but his team's Drtg has always been mediocre during his prime even when he was playing 40 minutes a night. In Boston, he is still having a similarly large +/- despite playing with a better team and with better defenders, and it looked like the only thing that has changed is the team's overall Drtg which had been top 1-2 in three of the last four years. I do think there needs to be more metrics available before we can judge what KG really means to a team's defense, but I don't think it's right to just write him off completely simply because the Wolves were a mediocre defensive team. For all we know 20 years from now when more data are available people will treat it the same way as baseball fans do with wins for a pitcher.

[Doctor MJ]

I like how some still argue that Duncan is protected by great defensive teammates. Bowen is arguably the only constant in Duncan's career. Now, defensive metrics for basketball are still primitive, so how much impact Bowen has on defense is not fully figured out, but I have a hard time believing that the Spurs had a <100 Drtg for 10 straight years simply because Pop has a great scheme and Bowen is a GOAT perimeter defender. When Duncan's minutes started catching up to him in 09 the Spurs defense was no longer the same, with or without Bowen. If Pop is really this defensive genius then why couldn't he still maintain the defensive identity with his team?

Duncan's great no doubt. He and Garnett were clearly the top 2 defensive players of the 00s. Like I say, I would only object to the idea that Duncan alone gives you an elite defense. He's the most important part, but he can't do it alone.