ElGee wrote:drza wrote:Re: Point guard impact (Kidd, Nash, Payton and Frazier)
...
Well, Englemann's single season RAPM stats give me some interesting food for thought.
I feel like there's a little too much stock being put into Englemann's numbers. They aren't gospel.
It's nice to see Kidd do well in them, but I have serious concerns about him. I love seeing people acknowledge how good he is defensively, even late in his career. So savvy, so smart, and like Payton and Dennis Johnson, boy does size to seem help. But Kidd is a guy who has never really run an elite offense, and that's a major reservation since he very clearly is a Quarterback style PG looking to distribute.
Here are his team's ORtg rankings in his prime:
96 19th
97 TRADE
98 12th
99 4th
00 16th*
01 22nd
02 17th
03 18th
04 25th*
05 26th*
06 25th
07 17th
That gives me serious pause. Then you dig deeper, and you notice that in 2000, a year after leading his ONLY top-10 offense, Phoenix had an offensive rating of 104.7 with Kidd starting and 104.1 in 15 games he missed with Randy Livingston "replacing" him (Penny was still offensively savvy then). That's sort of like an anti-Steve Nash thing...
Then he goes to New Jersey and receives a lot of MVP love in the weakest conference basically ever, but no one notices the team is defensively driven. Why? Because they never do. The Suns offense goes from 100.3 (19th) and 2.7 points below league average to 103.3 (19th) and 1.2 points below league average. Relatively speaking,
the Suns improved by a 1.5 points there. It was nice to see a non-scorer receive so much love, but the narrative there was way off.
Meanwhile, in New Jersey, the Nets go from -3.0 relative to the league and ranked 24th to -0.5, ranked 17th, and still below league average. Only the 01 season was marred by injury (Kittles for the season, 15g for Marbury, 33 games for Van Horn, 14 games for Martin) and the 02 season a healthy one that also saw the addition of a dynamic 6th-man in rookie Richard Jefferson. Hard to give Kidd credit for all of that
2.5 point relative shift on offense. In 2004, the team was +7.0 with Kidd on vs. off on offense. That's about what I'd expect from him. And in 2005, +14.1 (which is where his RAPM figure must come from). And another monster number in 06. Those figures give me pause too, because while I expect Kidd to help offenses and be good, those numbers are suggesting that he was on horrendous offensive teams. I think this is a quintessential case of APM models having no way to account for the fact that it's easier to take a 90 ORtg team and make them a 104 team than a 104 team and make them 118. Those are very different achievements.
Jason Kidd is a good offensive player, but he's not a great offensive player. He won't pressure defenses like someone who can score well too, doesn't use his scoring as a weapon, and for most of his career was basically a bad outside shooter. He, unlike Nash, clearly was better in transition/the open court, especially off of his own (amazing) defensive rebounding. I believe Kidd's strength in the halfcourt comes from being quick with his decisions, and he helps mediocre to bad offenses -- ones that had no backup PG at times in NJ --with those decisions and his ability to push in transition and get easy baskets but this is not someone who was ever an elite offensive player.
I respect your points, especially the one about RAPM not being gospel. But I would suggest that you re-read what I wrote (or perhaps I didn't do a good job of explaining it the first time), because I didn't start from the RAPM angle. Instead, I started from the angle that I've watched Kidd and considered him to be in the offensive quarterback mold...then, I noted that offensive quarterbacks tend to make larger impacts than the more lead-guard types historically speaking...then, I used RAPM as a check for whether my impression and that historical trend held up to the best (not perfect, but the current state of the art) single-season +/- stat we have. And it did. As such, I don't feel like I'm just using RAPM as an out-of-the-blue measure...it seems to confirm trends seen elsewhere.
I do find it very interesting that Kidd's offenses have never been at the top of the league, but I don't find that to be compelling in-and-of itself. Same with the part-time on/off numbers. Essentially, and I've spoken of this elsewhere, I see those things as earlier iterations of the APM families that we have now. If we have no +/- numbers, as in previous generations, then team rankings and injury absences are all that we have to estimate impacts and they are better than nothing. But I consider raw +/- to be a step up from them...and complete on-court/off-court to be a step up from that...and finally APM to be a step up from that as well, because each iteration accounts for the info in the previous evolutionary step but with more info added and more confounds accounted for. I actually like to look at all of the above, but when a less refined method is in conflict with a more refined method, I tend to believe the latter more than the former unless there's a compelling reason why.
And when we get out of the stats and into the reasoning portion of your rebuttal, I don't find it compelling. At its heart, I understand your argument to be essentially that Kidd doesn't do it the way that it's usually done so he can't really be an elite offensive player. You note that he isn't an elite scorer and doesn't have a great jumper, but to me those aren't a comprehensive list of skill sets nor a condemnation of being an elite offensive player. Kidd is also large for his position, was extremely fast for his position, has excellent court vision, is a quick and excellent decision-maker, and is extremely intelligent about the strengths/weaknesses of his teammates and where they need the ball to be effective. You say that he doesn't put pressure on defenses like a scorer would, well I'd counter by saying that scorers don't pressure defenses the way that Kidd would either. The methods are different, but that doesn't of itself make one better than the other.
Plus, though we're focusing as much as we can on offense, as Dr. Mufasa likes to point out there is a connection between offense and defense that isn't easy to separate. Having a point guard that is excellent at recovering possessions (crashing the defensive boards, steals) and facile at using that to quick-start the offense is its own kind of pressure. The personnel on those Nets teams were such that they usually weren't built to excel in the half-court offense anyway...the secondary players were finishers like Martin and Jefferson, not go-to scorers of their own right. And while if Nash led those units they might end up with higher offensive ratings, that would also come with catering the offense to suit his strengths/weaknesses more. Styling the Nets as a defensive team that sparked fast-break opportunities doesn't work as well with Nash as it does with Kidd, and not all of that is due to pure offense vs defense capabilities of the individual PG. Kidd's ability to make a strong individual offensive impact on those teams while allowing them to play a style that catered overall to the skillsets of the team shouldn't just be written off, just because the way he did it isn't the way that others might have.
Plus, I've seen you say this several times in other threads, but I need to see more support on your theory that it's easier to take a bad unit to average than an average unit to great. That sounds like something that we say because it's been said before, but I'm not sure I see the basis or proof of it. Yes, it's easier to take a poor unit to average than from great to great-er, because of diminishing returns. But to me, an individual can only maximize the team he is given. If you're given scraps and make it respectable, that is not necessarily an easier thing to do than taking prime cuts and making a feast out of it.
Anyway, I've gone a bit afield trying to respond to all of your points, but returning to the crux, I would say that my pro-Kidd arguments don't rely on RAPM as a standalone measure. As DocMJ has pointed out, his accolades (independent of RAPM) would have him right there with Payton and ahead of Frazier as is. And his box score advanced stats (also independent of RAPM) like PER are also very competitive with both Payton and Frazier. Even the complete non-statheads would agree that Kidd is one of the better point guards in history. From there, I don't see where it's that much of a stretch for the +/- stats (both net on/off court and APM) to agree that he is among the best as well. As I said above, if anything it seems to me that the RAPM results I site just support the trends that I was already seeing from a lot of different angles.
