Reviewing and previewing the deal: How long can we expect James Harden to be elite?

A few days before the trade deadline, I posted a piece about what was likely to happen (essentially, Harden for Simmons, Curry and draft capital) and why it was so tremendous for the Sixers. On the plus side, I nailed not only the basic shape of the deal but also the basic shape of the negotiation. As best we can glean from media reports, Brooklyn asked for Maxey, we said no, it’s Danny. They said how about Matisse? We said no, but maybe we could talk ourselves into sacrificing Curry. They said no, really it’s gotta be Matisse. We said we’d rather give an extra pick to make it Curry, and they said OK. [I’m leaving out of the narrative the seven heart attacks I had while all these rumors were flying.] Additionally, I had some analysis of why Harden was worth all we gave and far more besides that is as valid now as it was way back then! The bad news is, new articles were posting with updated rumors so fast that my piece got pushed down the stack and I fear many people didn’t see it. I’d like to simply repost it, but that’s not the way we roll at Liberty Ballers. So instead I have prevailed upon the editors to allow the following: I’m posting some new thoughts and evidence on the trade, and urging you again to read my thoughts from before the trade went down.

Alright, my throat is properly cleared. The primary objection many fans and media figures had to the Sixers’ actions was that James Harden is, to use the phrase most commonly employed, a fat, old man. Sometimes kinder words like “past his prime” or “late in his prime” are used. Father Time is undefeated and it’s certainly possible that Harden’s time of elite play has passed, or soon will pass. As in my previous piece, I’m going to use adjusted plus-minus statistics as a starting point for player evaluation, and 538’s RAPTOR as my adjusted +/- statistic of choice. The former is because only adjusted plus-minus statistics take account of everything a player does to help his team win; while flawed in that they can be noisy over anything less than very long periods, when they deliver a consistent message over multiple seasons they are extremely unlikely to mislead. As to RAPTOR, I’m not arguing it’s superior to the alphabet soup of different stats of this type (RAPM, RPM, EPM, DPM, etc. etc.). It’s just the one I’ve been using for a while, so people will know I didn’t look around for a stat that comported with the argument I wanted to make; plus their website is reasonably blind-friendly.

What does RAPTOR say about the players in the deal? This year it says Harden has been good but not special, Curry has been solid-backup level but not good enough to start, Simmons hasn’t taken the court, and Drummond and Millsap are eerily similar per minute with Drummond deserving credit for having played a lot more. But again, these methods aren’t great with just half a season of data to work with. Over a longer timespan, RAPTOR says:

Harden: superb; perhaps the greatest player of his era (recognizing that RAPTOR only evaluates regular-season play; I won’t delve into analyzing small-sample playoff stuff in this piece)

Simmons: very good, not deserving of All-Star appearances but probably top 50

Curry: poor starter or very good backup

Drummond: OK starter or excellent backup

Millsap: good-starter level per minute (but probably can’t handle too many minutes)

Let’s pause for a moment on Harden. If we take the three-season run from four years ago through the year before last, Harden’s excellence and consistency are as impressive as anything we’ve ever seen; only CP3 was similarly-year-in-year-outeffective in his prime among players from this era. Beard was over +10 on RAPTOR every year; for comparison this year there is only one +10 player in the entire league (the Joker). Prior to those three years he’d been spectacular for a long time; the first year RAPTOR is available for is 2013-2014, about which more below. That season, a full eight years before this one, RAPTOR rated Harden the sixth-best player per minute in basketball.

What about last year? This seems to have slipped down the memory hole, but from when he arrived in Brooklyn early last season to when he injured his hamstring late in the campaign, Harden was probably the best player in the league. There was quite a bit of discussion about how he was playing better than anyone, but it just wasn’t right to deem him the most valuable player of the year when he’d played for two teams and had been super-valuable to one but kinda sorta totally destroyed the other (Houston). I consider that logic sound, by the way. Harden could have scored 80 a game last year for the Nets and I’d have argued he shouldn’t be the season-long MVP; forcing a trade the way he did, and has now done twice, is the antithesis of “valuable.” But the point I want to make is this: literally a hundred times in the past week I’ve heard phrases like “if we could have gotten the Harden of 4 or 5 years ago....” when the guy they’re talking about was the best player in the game two years ago and arguably one year ago. The whole “four or five” thing is somewhere between off-base and flat-out lie.

Honestly, you just have to try this yourself and you’ll see what I mean: it’s not normal for a player to be so far up the adjusted plus-minus rankings year after year like this. All-timers like Lebron and Durant always do well, of course, but you don’t see them finishing so high so consistently, or anything close to it. Either Harden is one of the very best players ever, at least in the regular season, or there’s a weird glitch in the way these stats are calculated that just happens to favor Harden. And the latter seems pretty much impossible given that adjusted plus-minus stats do not account for points and assists and rebounds (all of which the Beard racks up amazingly well) but simply ask how your team does when you’re playing, and who you played with and against so that the positives and negatives can be allocated among the participants.

Now, this year Harden has been a quite good player by normal-human standards, but crappy by his own standards. So what’s going on? Here are the main theories, more than one of which may be valid:

1) He’s old and washed

2) He’s too fat to be a top player over the next few years

3) He’s injured, in particular his hamstring(s)

4) He’s been malingering to show he’s angry at the Nets’ situation

5) His game has been negatively and severely affected by the rules changes

All of these are possible, and none of them are good news for fans of the team where James Harden now plays. Hey, there’s a reason we got him for a guy who refuses to play and who couldn’t shoot the ball when he was willing to play! It’s my belief that it’s mostly 4), malingering. He’ll probably do it to us someday and it will suck! But hopefully not for at least two or three years. Anyway, my particular belief is that 1) is very implausible. The idea that Harden, at 32, is in age-related decline and is now a +1 or +2 or +3 player when healthy, rather than a +6 or +8 or +10 player as he’s been over the past decade, does not fit with what’s been going on in the NBA in the 21st century.

In last week’s piece I pointed to the longevity of fellow superstars like LeBron and CP3 as evidence that the NBA has changed, the days when players should be expected to play far worse at 35 than they did at 30 — conditional on being total dominators at 30 — really existed but now are gone. A smart friend noted the serious cherry-picking risk of saying things like “LeBron did it, so we should assume Harden can do it!” LeBron might be the best player in history, period. And he’s a known fitness fanatic. And plus he’s just one data point. You need to do a study! I’m a lazy, lazy man. So I did the tiniest possible study. But the results are awfully compelling. As I mentioned, RAPTOR goes back to 2013-2014, this season is the ninth for which we have RAPTOR data. Let’s look at who the superstars of 2013 were, and how long they have lasted or can be expected to last. Here are the top 10 in RAPTOR from its first year:

1) Chris Paul

2) Stephen Curry

3) Kevin Durant

4) Kawhi Leonard

5) Kevin Love

6) James Harden

7) Paul George

8) Manu Ginobili

9) Joakim Noah

10) Kyle Lowry

For you young folks whose memories don’t go back to 2013, I’ll mention that Chris Paul was a superb classical point guard of the era and Steph Curry was... what’s that you say? You know who they are? Indeed this pretty much appears to be a list of the current best players in basketball? Well, then you see my point!

One of these guys, Manu, is an all-time great who was already 36 when this season started, so obviously he didn’t fall apart in his mid-30s. I guess I should note that I know not everyone realizes Manu was as good as he was, but I am 100 percent confident it’s true. I have adjusted +/- data from the years prior to RAPTOR getting started and the top 10 are all the guys you’d expect, Dirk and LeBron and Duncan and all, and Manu is right in there with them — he was unbelievably valuable.

But the real point is this: Over and above the retired Manu, seven more of them are and have consistently been among the best players in basketball now, almost a decade later, even though some, like CP3, are as old as 37. And that’s over and above the left-for-dead-due-to-injuries Kevin Love, having a great year at 34 leading a surprise Cleveland team. The only guy with a short career is Noah who a) is a big man and it does seem possible the rules are different for those guys, b) had horrible injuries, and c) although a fine player was obviously not a dominator at the level of Durant or Kawhi, he just had a terrific year that year and maybe got a little lucky on his adjusted plus-minus.

If you look at the next 10 for guys who were known at the time to be legit superstars, you’ll find LeBron (still as good as anyone at 37), Dirk Nowitzki (already 35 in 2013-2014 and still had plenty in the tank) and that’s about it. Every player who consistently put up +5 or better around that time has either been terrific into their mid-to-late 30s or, like Kawhi and PG-13, is expected to. If you squint hard enough you can argue that Boogie Cousins was a legit superstar, and of course he was derailed by injuries, as I say big men probably really don’t last as consistently as perimeter players. And there’s a group of point guards who are not anywhere close to the CP3-Harden level but who were very good, and they have lasted OK but not great, just as you’d expect from terrific-but-relatively-speaking-lesser players. That group is Lowry (still excellent), Mike Conley (better than ever lately), Kemba Walker (not good anymore), Russell Westbrook (not good anymore), and Kyrie Irving (was never anything close to as good as his rep but still about as good as he was).

To summarize: anyone looking at on-off stats in 2014 or so would have seen clearly that the most valuable players in the NBA in recent seasons had been:

KD

CP3

Steph

PG-13

LeBron

Harden

Kawhi

Manu

Dirk

All the non-Harden players on the list either lasted forever or seem poised to. Maybe Harden will be the exception, but I don’t see any reason to expect that. Of course I am blind, if you’re looking at the Beard and saying “he’s too fat to last, all those other guys are ripped,” well, I accept I may be underestimating the importance of that. Same with the rules changes, my guess is it’s a minor effect. After all James was great in 2013 which was long before we started hearing about him jumping into defenders. And then there are injuries, which can strike anyone. My view is that if he’s healthy and happy, he’ll be awesome.

This turned out longer than expected, so I’ll just add the following:

-See below for why I am not at all bothered to lose Seth Curry, who is not really a good NBA player and indeed is only good-not-great even just counting offense.

-Hey, you know who else was top-15 in RAPTOR in 2013-2014? Danny Green! That’s why he’s still a +2 player now despite his advanced age, because he was really terrific in his prime and that’s what you expect, is that great players at 27 become very good at 34, while good players at 27 become backups or out-of-league at 34! I like to think Daryl knows he’s more valuable than Seth and just offered him first in a (successful!) attempt to fool Brooklyn in demanding Seth. “Don’t throw me into that Seth Curry briar patch, Mr. Marks!”

-I didn’t forecast Drummond-for-Millsap, sorry for not seeing that coming! I may write more about Millsap in another piece, but for now let’s list his RAPTOR the past four years:

+4

+1

+2

+1

He’s 36 now so I’m not weighing the +4 much except to say that The Other Bball Paul was a hell of a player in his prime so him being good at 36 is not impossible. As far as I can tell he’s never had a bad year. His defense this year is over +3, just as Andre’s has been, and his traditional defensive stats — steals, blocks, rebounds — are in line with that so I’d guess it’s real. As to his offense, the big ugly is that he’s shooting 22 percent from distance, which is probably what got him benched in Brooklyn. But that 22 percent is 6 for 27, not much of a sample for a guy with plenty of evidence he’s a 33 percent+ three-point shooter. If he’d hit 33 percent, nine of his threes instead of six, he’d have added nine additional points for his team in around 7.5 full-games worth of minutes. I.e. his RAPTOR would likely be over +2 this year. I suspect he’s as good as Drummond and a better fit and that we’ll be loving him soon; It may be that he was benched because Griffin and Aldridge have been so good more than due to any failing of Millsap’s. But if I’d seen him play perhaps I’d agree with Steve Nash that he belongs on the pine.

Millsap or no Millsap, we have James F. Harden — that is to say, James Freaking Harden — on our team. Sky’s the limit for this team, and I hope to write more on that subject soon.