richboy wrote:Here some issues I have.
Before KG left Minnesota was already falling. They were 21 in the league in defense. Forget the fact I can't imagine a team with Dwight Howard on it being at the botton of the league in defense. Fact is its very possible Minny was going to be a bad defense with or without KG. Also would point out that they replaced KG with a horrible defensive big in Al Jefferson.
One reason I like having these kinds of debates is that I often learn something I didn't know. See, I didn't come to my conclusions by looking at the stats...I used to watch the Wolves play on a regular basis and it was blatant that KG's impact was huge, but it wasn't until later on with the advent of the advanced stats that there were tools to quantitatively show it. I was reminded of this just now because, in order to rebut this point I had to go back and look again at that "21st" ranked Wolves defense in 2007. And in doing so, I learned something yet again.
KG sat out 6 games in the 2006-07 season. During those 6 absences, the Wolves gave up 117.5 points/game on 51% shooting from the field. I point this out here not as another argument for how bad the cast was (though seriously, look at those numbers again), but instead because those 6 games were SO bad that they actually skewed the overall results for the year. After spending some time navigating the complex way B-R calculates team defensive rating, I was able to quantify just how much those 6 games changed things:
overall season: Wolves DRtg = 107.9 (21st in NBA)
6 games w/o KG: Wolves DRTG = 121.6 (on pace for worst in NBA by FAR)
76 games w/ KG: Wolves DRTG = 106.5 (tied for 14th in NBA)
Again, the point isn't to pile on for how bad KG's teammates were (should be self evident by now). In fact, you can actually ignore the 6 games KG didn't play if you want to, because the more germane point is that when KG DID play the Wolves weren't one of the worst defenses in the league. In fact, they were again right there at league average.
Which means that most of your points in this paragraph no longer have any basis. KG didn't lead a bottom-feeding defense in 2007, despite having a putrid cast. Which also means that you can't really support the notion that the Wolves were likely to have a terrible defense in '08 with or without KG. Which, again, makes their immediate fall to arguably THE worst defense in the league since KG left more support for just how much heavy lifting KG was doing in Minnesota.
richboy wrote:My next issue is KG pulled Minny up to mediocre defense. That very well may be possible. I just don't see how that means KG has greater impact defensively than other dominate defensive big men. Most years of Hakeem's career Houston was an elite defense. Same with Dwight, Zo, Duncan.
No question KG has impact defensively. The question is the impact as great as adjusted plus minus suggest. People are saying Dwight Howard is no KG. KG has bigger impact than Tim Duncan. I'm just looking and wondering what is this based on. Looks like by some KG is playing with a group of HS kids in Minnesota. Because the reality is the only way you can suggest KG is comparable in defensive impact to those top bigs is say well Minnesota may have had the worst starting 5 of all-time. (snip)
I snipped here because I want to keep the arguments that you're making clear. You say that you disagree with the APM results (from 2003-04 thru 2010-11), but the rest of your post talks about the Wolves during time outside of that time range. While I have absolutely no problem discussing KG's impact on the 2000 - 2003 Wolves, I want to keep this as focused as possible on the 8 year time-window where we have APM results. 8 years is plenty of time, so if you want to argue against a stat's conclusions you should be able to find sufficient support from SOME time within that window.
You say that for KG's defensive impact to have been that large in Minnesota he would have had to have been playing with the worst starting 5 of all time. While earlier in the thread I referenced a post that made that exact argument for the '07 team (
viewtopic.php?f=64&t=1075089 ), more importantly on page 11 of this thread (7th post down) I went into a reasonable amount of detail describing the casts of all of his teams from '04 - '07. And not just names, actually pointing out the strengths and weaknesses of the personnel at that particular point in their careers and how that then played out on the court. As I pointed out, the vast majority of the defenders on that team brought little if anything to the table defensively. The majority of the perimeter players were unable to prevent penetration (often without need for pick), they were unable to disrupt opponents dribble/pass lanes, they were poor on rotations, and they were poor rebounders for their position. The majority of the bigs were unable to maintain post position, were unable to offer help defense, and were also unable to rebound their positions. The one year there was even a mild change in that pattern (in '04 Erv and Madsen were able to maintain their post position...not much else, and Hassell/Spree were able to provide some degree of perimeter resistance) KG was able to take a still subpar defensive cast to among the leagues best. Other than that year, defensively, he WAS leading some of the worst defensive casts in the NBA.
Conclusions: Here is where it's important that we come to some kind of consensus on what the +/- stats are saying. Because subjectively, I can say that the Minnesota casts were awful and the Boston cast has frankly become overrated...and you can say that I'm wrong, that the Minnesota cast wasn't that bad and the Boston defensive cast is "elite". And we essentially stale-mate at that point, because it's opinion against opinion.
But if I point out that the '07 Wolves had an average defensive rating when KG played and an abysmal 122 B-R DRtg in the 6 games he missed, that's not opinion. It's fact. It actually happened. For this particular data point you might argue that the sample size is too small for definite conclusions, and I'd even agree, but it is still an evidence point that actually happened.
And if I then point out that from KG's arrival in Boston in '07 until the present he has played in 268 games and missed 60 (large sample sizes each way), and that when KG has played with Rondo/Ray/Pierce/Perkins the unit's real defensive rating is a sparkling 97.3 but when that same 4 guys have played with any other big man their real defensive rating is an awful 112.1...again, this is fact. It's not my opinion. And here we're talking huge sample sizes, so there's no ambiguity.
And these are just snapshots of the data that a multi-year APM calculation measures. The job of APM is to note the differences in circumstances, and what happens in those circumstances, and over multiple years to quantify the trends. And as Doc MJ has pointed out, there are now THREE mutli-year APM calculations, spanning different sets of years, with many different KG supporting casts, calculated by different analysts, that all agree unanimously that Garnett has been having the largest defensive impact in the NBA over the last 8 years by a clear margin.
Again, my life's mission isn't to convince you that APM is perfect, or that you should even rely on it. But in this example it just seems abundantly clear. We have a player that already was one of the most decorated defensive players of the last 8 years in terms of accolades. We have a very reasonable subjective argument that he really was that good and his teammates really were that bad. Then, if you look in-depth quantitatively it is clear that his defensive impact was massive even if you try to avoid APM. Then, APM just further confirms it. If you disagree with all of that evidence, you really have to bring more to the table from within that 8-year time window than just "he can't have been that good, because if he was he'd have been as good as Larry Bird or David Robinson". That's not a counter-argument. And if that's the best one that you've got, maybe it should make you reconsider a bit where KG might deserve to slot in your personal rankings.
