Kevin Durant, Alpha Dogs, and Supporting Casts
Posted by Neil Paine on November 21, 2008
Last week, Wages of Wins author David Berri posted about the Oklahoma City Thunder over at the WoW Journal, pointing out that Kevin Durant and his teammates are having really a terrible offensive season in 2008-09 (something which, frankly, cannot be debated -- OKC ranks dead last in the NBA in offensive efficiency, and by a pretty wide margin). In turn, Berri's post sparked a discussion at APBRmetrics regarding just how much of the Thunder's offensive woes are Durant's fault (his 94.5 ORtg is pretty abysmal) and how much of the blame falls to his lackluster supporting cast. The argument that ensued was fairly circular: is Durant playing inefficient basketball because his team is bad (and he is forced to take on a huge role he's perhaps not ready for yet), or is Durant's team bad because he's playing inefficient basketball? Or both?
Out of all this, our esteemed colleague Kevin Pelton (of Basketball Prospectus) posed the following question:
If we take out each team's leading scorer, who has played with supporting casts comparably bad to the one employed by the Thunder, especially so early in their career?
Well, that's a question we can at least attempt to answer objectively. First, let's classify the "leading scorer" of a team as the player who consumed the most possessions -- this is preferable to taking the leading scorer because straight-up points account for both usage and efficiency, and here we're interested in identifying the team's offensive "alpha dog" regardless of how well he actually produced points. Then let's calculate the cumulative offensive rating for his teammates, which should give us a rough estimate of his supporting cast's offensive ability.
This season, the Thunder have scored 1076 points on 1148 possessions (through Nov. 20); Durant has produced approximately 211 of those points and has used roughly 224 of those possessions by himself. That means his teammates have combined for an offensive rating of 93.6, which is really bad by the way. Last year, Durant's Seattle teammates produced an ORtg of 101.3 -- better, but still not good at all. So, the question is, which "alpha dogs" have had to put up with similarly inept teammates in the past? Have they been as inefficient as Durant? And has anyone so young (KD is 20 this season) had to carry such a terrible supporting cast?
Here's every "alpha dog" whose teammates posted an offensive rating roughly as low as Durant's fellow Thunderites in '08-09, sorted by the differential between their individual ORtg and that of their awful teammates:
Year Team Player Age ORtg t_ORtg Diff 2002 MIA Eddie Jones 30 109.2 97.1 12.2 2000 CHI Elton Brand 20 104.2 92.8 11.4 1993 DAL Derek Harper 31 109.2 97.9 11.3 2002 MEM Pau Gasol 21 107.6 96.6 11.0 2003 DEN Juwan Howard 29 101.6 91.0 10.6 1983 WSB Jeff Ruland 24 107.4 96.9 10.5 2001 GSW Antawn Jamison 24 104.4 96.1 8.2 1983 HOU Allen Leavell 25 102.8 95.6 7.2 2001 ATL Jason Terry 23 104.8 97.8 7.0 2001 CHI Elton Brand 21 103.3 96.7 6.7 1999 VAN S. Abdur-Rahim 22 104.2 97.5 6.6 1999 CHI Toni Kukoc 30 98.7 92.1 6.6 1988 LAC Mike Woodson 29 102.3 96.7 5.7 1981 DET Phil Hubbard 24 102.7 97.4 5.3 1979 NJN Bernard King 22 101.7 96.8 5.0 1978 NJN Bernard King 21 98.0 93.7 4.3 1998 GSW D. Marshall 24 99.6 95.7 3.9 2003 MIA Caron Butler 22 99.2 97.1 2.1 2004 CHI Jamal Crawford 23 99.0 96.9 2.1 2004 TOR Vince Carter 27 99.4 97.3 2.1
As you can see, despite their horrendous supporting casts, most of these players managed to rise above their teammates and post reasonable efficiency numbers. In fact, many of them did it at an age not significantly older than Durant (while one of the major points made in KD's defense is his youth). So it's not like it's unprecedented to expect Durant to perform more efficiently than he has so far this season.
However, the good news for Durant is that he's not your ordinary "alpha dog" -- he's more like a super-mega-alpha dog, taking on about 29% of OKC's possessions when he's on the court. By contrast, most of these guys, while leading their respective squads in possessions, were only expected to create on 25% of the team's possessions (or fewer). So Durant deserves a little leeway when we evaluate his efficiency numbers, because he's being asked to do more, at a younger age, than most of the guys on this list.
Even so, Durant's early returns aren't exactly those of a future megastar; no alpha dog in the past has had such horrible teammates and still played down to their level in terms of efficiency. Durant still may very well develop into a solid player, but if the history of similar players (alpha dogs saddled with poor teammates) is any indication, it's doubtful he'll live up to the considerable hype that surrounded him when he first came out of college.
November 21st, 2008 at 4:36 pm
Hey Neil,
I think this is a very good analysis. I wonder if you might have any opinions on the comments I left on Dave Berri's analysis (I came a bit late to the part; they're the third-to-last and last posts up right now.
Basically, I think you can give Durant even more of a pass because there exists an equilibrium at which player's additional shots are more harmful than helpful. It stands to reason (and empirical examples bear out) that shots have a diminishing chance of going in the more you take them. But someone has to shoot, even for terrible teams like the Thunder. So basically, the fact that the Thunder are bad dictates that Durant takes so many shots, not the other way around.
Also, I think that Dave Berri's system undervalues perimeter and unassisted shots. Assuming that shot attempts for paint and perimeter players have the same value seems naive, because it seems that a lot of the high-percentage shots that exist in the paint couldn't exist without what the perimeter players do.
November 21st, 2008 at 6:01 pm
Wonder where Kobe Bryant's Lakers circa 2006 fall in this list..
November 21st, 2008 at 6:23 pm
Not as high as you'd think. That year, Kobe's ORtg was 113.7 and his teammates' combined ORtg was 107.5, for a difference of only 6.2 pts/100. In fact, that season was only the 3rd-biggest differential of Kobe's career -- in 2006-07, Kobe's ORtg was 115.4 and the rest of the Lakers had 106.6 (a difference of 8.8), and in '03-04 Kobe had a 112.4 ORtg vs. his teammates' 104.2 (+8.2).
Unsurprisingly, this year represents the smallest differential of Kobe's career as an alpha dog; his teammates (110.1 ORtg) are actually outpacing him (109.5) for the first time. Last year, Kobe was still 1.9 pts/100 ahead of the rest of the Lakers, but a full season of Gasol, Bynum, Ariza, etc... That's scary depth.
November 21st, 2008 at 7:25 pm
Wow - thanks!
November 21st, 2008 at 7:46 pm
I just keep thinking that there is only so much losing and bad teams before the stink of losing stays on a player, regardless of obvious potential.
The Thunder miight not to be average before Durant's rookie contract ends. What is the effect of 4-5 years of that kind team production? I suppose I could look at some older Clipper teams to see waisted potential.
November 21st, 2008 at 8:00 pm
Andrew: I think you have a very good grasp of what the kind of analysis we do here is all about, as well as the weaknesses of Wins Produced. In your comments, you basically echoed everything I've ever written about the flaws in Prof. Berri's system -- what produces wins at the team level isn't necessarily as pertinent when applied to individuals. At the team level, shot-creation is assumed, so the only thing that matters is efficiency; at the individual level, the ability to create high-% looks is anything but a given. At the team level, defensive rebounds represent complete defensive stops; at the individual level, they only represent the final step in a series of defensive contributions by multiple players. In fact, we've had many debates on the topic over at APBRmetrics, if you're interested in reading some older threads...
As for Durant, I absolutely believe he would be more efficient on a team that didn't ask him to do too much. If he would use only, say, 22% of a team's possessions when on the court, it would make a world of difference to his ORtg. However, it would hurt the Thunder's offense (if you can call it that) if he were to take that approach & pick his spots more often. It's the reality of being one of the lone talented creators on a bad team.
That said, the list above shows that other players have had to put up with similarly poor teammates, and they managed to at least do better than KD is doing right now, even at such a young age in the cases of players like Brand and King. That tells me that Durant, while talented, will probably never become the "rich man's Kevin Garnett" we kept hearing about when he was drafted.
November 21st, 2008 at 9:01 pm
Does anyone else think that Kevin Durant is playing woefully out of position? Does it make sense that they're trying to turn a player that played mostly in the post at small forward in college and into a shooting guard? Am I the only one who has a problem with this?
November 21st, 2008 at 9:24 pm
Durant is younger than any of the those players when they came into the league and this is his second year. check out his numbers next year when he would be the same age as most of those players their rookie year.
November 21st, 2008 at 11:44 pm
It is true that most of those on the list were older and most of those close to his age were interior players who may have had it easier as exceptional talent warranting or getting high usage is rarer and hard to cover. Bernard King and Caron Butler are the best comparisons and Durant being younger does for me grant him a little slack.
November 21st, 2008 at 11:46 pm
I think it would be insightful to look at the type of shots that the various bad teams were asking the heavy-lifting players to take (or, if you like, what kind of shots they're capable of taking, given team context); if you ask me, it seems a bit like comparing apples and oranges to compare Durant's numbers with his outside-oriented game with Pau Gasol or Elton Brand, who operate more inside.
November 21st, 2008 at 11:49 pm
rarer "at those interior positions"
and perimeter efficiency is generally lower than inside
Butler and King did pretty well. Not above criticism or maybe ideal but they created / filled a certain type role that can sometimes work if you have the right pieces (efficient 3 pt shooting, post scoring or even just the offensive rebound game) around that type which the Thunder hasn't fully assembled yet at a minimum and may be way off on.
November 22nd, 2008 at 2:51 am
I'd be inclined to agree, except that offensive rating is not biased toward/against particular positions or sizes of players. In general, a perimeter-oriented player will actually have higher a offensive rating than a big man with the same Floor Percentage (% of possessions that score at least 1 point), for the simple fact that the outside guy will mix in a few 3-pointers. I'd say it'd be easier for Durant, who took 15% of his shots from beyond the arc as a rookie and is taking 7% from deep this year, to post a good offensive rating than Brand, who almost never took a 3. It's just that Brand's floor % was .523 as a 20-year-old, and Durant's is .455.
November 22nd, 2008 at 9:32 am
What about Lebron James and his Cavs team as a rookie. That team didn't have anyone that could put the ball in the basket and LeBron had to carry the team from day 1.
November 22nd, 2008 at 12:06 pm
Well Mark, you bring up a good point. Based on LeBron's performance, barring injury he is looking like the best player ever. That isn't hyperbole, he's been that good.
Neil, I think you've done a better job than Berri here, this seems like a much more realistic breakdown. But I think it is tough to compare Durant to Brand because of Durant's lack of physical maturity, and the unique challenges of being a perimeter player on a bad team. The physical part is obvious, Durant is very physically undeveloped and that makes a huge difference in a ton of different areas. As to being a perimeter guy on a bad team, it's normally easy to pick up a couple extra points per possession off of open threes, but Durant doesn't get many of those, and 3 and 2 point jumpers taken under overwhelming defensive pressure are percentage killers. I think it's actually kind of incredible that Durant is shooting .455, because the Thunder offense is just a ghastly sight to behold. Finally, the sample size of players that were as young as Durant and carried his kind of offensive responsibilities is so small that I think any conclusions are sort of premature. It's tough to project for him, especially because the only guys we can compare him to, King and Brand, were very different players.
November 22nd, 2008 at 1:00 pm
Needless to say, the Thunder aren't on national TV much, so last night I watched them for the first time this season. I was blown away at how bad they are offensively, just in an aesthetic sense. It's horrific, that's the word I used to describe it. This is a case where the #s definitely back up what you're seeing.
In 2004, LeBron's production was similar to Durant last season -- he had an offensive rating of 99.2 and used 28% of Cleveland's possessions when on the floor. And his teammates actually had a higher ORtg than he did (102.9). However, in year 2 LBJ made a huge improvement, to a 113.5 ORtg on 29.5% of Cleveland's possessions. Durant's having to take on a similarly huge % of OKC's offense, and his teammates are considerably worse than LeBron has ever had to deal with.
But LBJ still made that "leap" in year 2 (7.7 pts/100 better than his teammates), and Durant simply hasn't, at least not yet. If Durant's offensive rating was, say, 100 or 99, like last season, that would be encouraging (similar to James' +7 differential over his teammates). But at 94.5 through last night's game, he's right down there with the rest of the Thunder (92.9 ORtg). That's what bothers me most about his production this season: I'd be fine with below-average #s from KD if they were still far better than his teammates, given that this OKC offense is tracking to be one of the worst of all-time. But it's disturbing that he's playing down to their level -- it's admittedly a small sample, but you just don't see that with the other guys on that list.
November 22nd, 2008 at 3:15 pm
Give Durant a couple more years. Lebron was an anomaly physically so do not include him in any comparisons, it is not fair. Durant just needs a couple more years for his body to mature and to pack on some pounds. He is a very tall and lanky, young 20 year old. And quite frankly, comparing him to other players, such as Brand or Bernard King is not fair simply because there has not been a player similar to Durant. Ever. Dirk comes as close to him as I can think of off the top of my head. And dirk came along very slowly for similar reasons. As a young 7 footer it takes time to put on muscle and strengthen such a long frame.
Durants perimeter skills are so far ahead of any other relatively tall and young player that came before him. He shoots over 80 percent from the line and can bury any kind of jump shot or set shot imaginable. And he handles the ball extremely well. Once his physicality and strength catch up, which it surely will in a matter of time, he will be the most well rounded and dynamic player in the NBA.
November 22nd, 2008 at 4:26 pm
I have to agree with the poster who said that Durant is playing out of his position. Hopefully the P.J. firing will knock some sense into the coaching staff, and they shift him to the SF position where he belongs. I don't know if there's anything to back this up, but I feel like his numbers would more accurately reflect his potential if he were playing SF as opposed to SG.
November 22nd, 2008 at 6:09 pm
"Once his physicality and strength catch up, which it surely will in a matter of time, he will be the most well rounded and dynamic player in the NBA."
I really wonder about this part. It was said when he was coming out of college that this guy was an absolute gym rat. He played basketball constantly and never wanted to do anything else, and this was always considered to be a good thing. On the other hand, maybe he's not very committed to getting in the weight room. That could be an absolute killer to his potential, and I absolutely think he could become the most dynamic player in the league, but he has to commit to doing the heavy lifting before he can ever become the player we all think he can be.
November 22nd, 2008 at 7:13 pm
Durant IS a 'super-mega-alpha' dog, and that's the problem. He's already been pigeon-holed as a franchise player, which has completely destroyed the Thunder's future, because no franchise will ever win with a limited perimeter player as its centerpiece, especially one playing outside of his natural position and body. When was the last time a player like that carried a team to a Championship?
You can blame his teammates and his coach all you want, but a supporting cast won't matter if the guy they're supporting is fundamentally incapable of leading a winning team. Durant is a glorified role player who puts up enough shots to have All-Star numbers, thus we view him as a Star, and thus he's prevented from being relegated as the third option of an efficient team, which is the only way he'll ever win anything.
Yes, Durant is an unprecedented player. But he still has to play defense. He still has to play efficiently. He still has to learn how to win.
Judging by his history, and by the 'eyeball test', and by the past performance of inordinately skinny players, and by his remarkable inability to gain strength (I was stronger than he currently is when I was 12 years old, literally), I see no reason to believe he'll be anything but a rich man's Rashard Lewis.
November 22nd, 2008 at 11:09 pm
I completely agree with Brian. Lebron was fully developed physically by the time he was 15, while Kevin Durant still has an adolescent body. Keeping this in mind, Durant appears to be farther away from his prime than any of the players on that list. Durant may still be a superstar, but it will take a few years for him to get there.
November 23rd, 2008 at 11:16 am
Hi,
I was wondering about how D-Wade is doing this season. He is stuffing the stat sheet and shooting at a 49-50% FG clip. It seems to me he's miles away from his teamates production? Any toghts on this one? Thanks.
November 23rd, 2008 at 12:42 pm
As of last night, Wade's offensive rating is 115.4, and he's using a staggering 37% of Miami's possessions when on the floor. His teammates have a combined offensive rating of 106.9 (league average is 105.8), so Wade's differential is +8.5 over his teammates. That's really good from an historical perspective -- it puts him in the top 20% of all "alpha dogs" since 1973, and with that 37% usage rate, he's doing it with a far bigger workload than just about anyone else on the list. The record for single-season usage rate is 36.5%, set by Kobe Bryant in 2005-06, so if D-Wade keeps this up we would actually be witnessing the biggest one-man show in the league since '73.
November 23rd, 2008 at 1:19 pm
Well, good news if your a Thunder fan. The change of coach to Brooks from PJ made a big difference last night. Neil you said you saw the friday game against the Hornets, and it was horrific. But it was typical of the PJ team thus far this season.
Last night I watched very closely as they played the same team back to back. The team had 112.8 OR, and worked repeatedly to get better shots instead of launching jumpers.
Durant seemed to thrive, he went 11/16, 30 points. His night would have been better if he had rebounded better (5), and not had 5 turns. But, the new coach Brooks put Durant at the three, which was well past due.
FWIW, fellow Sophomore Jeff Green was more efficient last night.
November 23rd, 2008 at 1:21 pm
I forgot to mention in the comment above, that ten of the first seventeen FGA's (the whole first quarter), were shots close in to the basket. Much better than the first dozen games.
Joe-Thunderguru.com
November 23rd, 2008 at 1:48 pm
Re: KD playing out of position for his body type.
Kevin reminds me of a young Clyde Drexler. At 6'7" Clyde seemed tall and skinny in his first seasons at guard, but his ball handling and three point range made him a good fit. I think KD could become a great SG also, with the flexibility to drop down as a small forward. Also look at the change in average height in the NBA. How many 7' centers are there? I think if he can bulk up he can create some great match-ups taking a guard in the post.
November 24th, 2008 at 6:38 pm
Durant seems to have a variety of problems:
1) Too thin
2) In college he was a great rebounder, but now they're asking him to be only a jump-shooter. This could be a bit jarring to his confidence.
3) Jeff Green duplicates his game to some extent. It's difficult to develop two young players who do the same thing.
November 24th, 2008 at 8:27 pm
Nick, it's one thing to say that Durant is playing poorly now, and might not be all he's hyped up to be. I would disagree, but until he produces like a star it's not an unreasonable thing to say. But glorified role-player? I hope you remember that as his career continues, because that is about the dumbest thing I've ever heard.
December 5th, 2008 at 9:12 pm
If Neil still reads comments on older posts, do you have a take on Durant since Brooks took over? I believe his usage% has increased, yet he's seen an increase in TS% and the differential between his ORating (102) and the team's (97.6) has also increased (these numbers were taken from bbr).
I don't know what this could be attributable to, as there are a number of factors that changed (new coach, starting Westbrook and Durant's position switch). Thoughts?
December 7th, 2008 at 9:20 pm
Glad to hear Durant's finally getting some burn at the 3 spot. Putting him at the 2 for development seemed like the stupidest thing I'd ever heard. He's 6-10 with long arms!!! Why put him in a position to guard guys who can routinely beat him off the dribble? Why put him as far from the bucket as possible?