You Are Here > Basketball-Reference.com > BBR Blog > NBA and College Basketball Analysis

SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all Basketball-Reference content, so you can quickly and easily find the content you want.

Also, our existing Basketball-Reference blog rss feed will be redirected to the new site's feed.

Basketball-Reference.com // Sports Reference

For more from Neil, check out his new work at BasketballProspectus.com.

Player Audit: Adrian Dantley

Posted by Neil Paine on April 8, 2010

As his prize for winning the 2010 Basketball-Reference NCAA Tournament Pool contest, reader Ian was able to request a post on a basketball-related subject of his choosing. The topic he went with:

"As for a subject of the blog, I'll go with my childhood hero, who I feel has long been unheralded by the masses, although I'll admit my bias. I'd love to see a blog dedicated to one Adrian Delano Dantley."

Great choice, Ian! Let's get our Player Audit on...

Adrian Dantley

Position: SF/SG
Height: 6-5    Weight: 208 lbs.
Born: February 28, 1956 in Washington, District of Columbia
High School: DeMatha Catholic in Hyattsville, Maryland
College: University of Notre Dame
Draft: Selected by the Buffalo Braves in the 1st round (6th pick, 6th overall) of the 1976 NBA draft.
Honors: NCAA All-American (2x), NBA All-Star (6x), 1977 NBA ROY, NBA Scoring Leader (2x). #4 Retired by Utah Jazz. Inducted into Hall of Fame as Player in 2008.

To this day, Dantley remains one of the most unique and underrated players in NBA history. Zander Hollander once said Dantley was "listed at 6-5, but more like 6-3"; he also said Dantley "never met a foul line he didn't own". In The Book of Basketball, Bill Simmons wrote that we'll never see another Dantley-type player again, and with good reason: Dantley was a 6'3½" small forward with a ridiculously effective low-post game, a guy who could basically get fouled whenever he felt like it (he shot 11 free throws per 36 minutes at his peak), had good midrange touch but no 3-point range, poured in nearly 31 PPG at his peak as strictly a halfcourt scorer, and didn't defend a soul. In short, Dantley was the absolute antithesis of everything you think of a small forward being today (aside from the foul-drawing, which LeBron James has gotten pretty good at for reasons we can all disagree on). And yet he shares one thing in common with almost every great small forward in league history -- he could fill it up like nobody's business.

Dantley was always pegged for stardom -- he was one of Morgan Wootten's all-time best players at DeMatha High School, and he had an impressive NCAA career at Notre Dame, helping the Irish snap UCLA's 88-game winning streak as a freshman and garnering All-American honors during his sophomore and junior seasons. Drafted 6th overall by the Buffalo Braves, the 20-year-old Dantley promptly averaged 20 PPG on .520 shooting and won the Rookie of the Year award in 1977.

But in a move that foreshadowed so many others, Dantley was traded by Buffalo to Indiana after winning ROY honors -- making him the only NBA player to hold this distinction -- in exchange for high-scoring Pacer Billy Knight (whom the Braves/Clippers would end up shipping to Boston just a year later anyway). After a brief layover in Indiana, it was off to the Lakers for Dantley, where he dropped below 20 PPG due to a redundant role with Jamaal Wilkes. When the Lakers dealt Dantley to Utah for former All-Star Spencer Haywood in September 1979, Dantley finally found some measure of stability in his fourth city in three pro seasons.

The Lakers would go on to win the NBA title without Dantley (thanks to some guy named Magic), but the drug-addled Haywood was a total non-factor, and Dantley averaged 28 PPG as the Jazz's star player. Over the next 7 seasons, the undersized low-post machine would post truly ridiculous efficiency marks, leading all players in Offensive Rating from 1980-86 despite one of the highest Usage %'s in the league. He also shot 53 foul shots for every 100 field goal attempts, one of the best marks in the game (and certainly the best by a player 6'5" or shorter). As a result, Dantley had more Offensive Win Shares than any other player during his Jazz days -- more than Bird, Magic, Kareem, Moses, you name it. You want to be shocked? Look at Dantley's Offensive Rating and Possession% numbers next to Larry Bird's over that span:

Dantley Bird
Year Age ORtg %Pos Age ORtg %Pos
1980 23 118.5 26.6 23 109.2 24.9
1981 24 118.5 27.5 24 106.9 24.1
1982 25 120.6 27.5 25 113.6 25.2
1983 26 123.7 25.4 26 114.7 24.8
1984 27 125.6 27.3 27 115.5 26.6
1985 28 117.0 26.8 28 119.1 27.8
1986 29 120.7 28.5 29 117.3 27.6

Unbelievable... From 1980-86, at the same ages, Dantley was without question a more effective offensive player than Larry Bird was for six of the seven seasons... Wrap your brain around that.

Unfortunately for Dantley, though, defense is the other half of the game -- and as good as A.D. was offensively, that's almost how bad he was on D. Frank Layden locked horns with his star often, and it's not hard to see why: with Dantley as their best player, Utah ranked dead last in points allowed per possession in 1980, and 4th-to-last in '81 & '82. Surprisingly, the Jazz made a quantum leap on defense in 1983 and were the NBA's best by 1985, but that had little to do with Dantley and everything to do with the addition of 2-time DPOY Mark Eaton at center. According to Defensive Points Added, Dantley didn't show much (if any) improvement despite the Jazz's meteoric rise on D:

Year Team DPA
1980 UTA -2.15
1981 UTA -2.43
1982 UTA -2.80
1983 UTA -2.66
1984 UTA -2.81
1985 UTA -1.83
1986 UTA -2.92

The other knock on Dantley was his poor reputation as a teammate -- allegedly he was not a team player, unwilling to sacrifice shots & touches for the greater good, and prone to a bad attitude when his scoring numbers were threatened. After the 1986 season, Layden had grown so weary of his battles with Dantley that Utah shipped him to the rising Pistons for Kent Benson and Kelly Tripucka. In 1987, Dantley seemed to get the last laugh over Layden and every coach who ripped him (Cotton Fitzsimmons once said that no team could ever win with Dantley) when Tripucka flopped badly in Utah, and A.D. led the Pistons in scoring en route to the Conference Finals, losing to Boston in no small part because Dantley was knocked out of Game 7 with a concussion (in addition to the obvious: Isiah Thomas' infamous turnover to Bird in Game 5).

In 1988, Detroit moved closer to the brink of a title with Dantley as their leading scorer, but they lost the 7th and deciding game of the Finals after spotting the Lakers a 15-point lead in the 4th quarter, a deficit that proved too big to overcome despite a furious rally (really, you should watch the whole thing play out, it was like the 2000 Blazers-Lakers Game 7 if the Blazers had managed to hold on). After that disappointing defeat, the Pistons prepared for another deep postseason run, but Dantley and Thomas clashed over who was the team's Alpha Dog, and Dantley had issues with Detroit coach Chuck Daly as well. In the middle of the 1989 season, "Trader" Jack McCloskey swapped Dantley with Dallas' Mark Aguirre, a similar (if less productive) all-offense, no-D forward who was friends with Thomas and more willing to accept a lesser role as Thomas and Joe Dumars grew into more offensive responsibility. As if to validate Fitzsimmons' old critique, the Pistons went on to win the first of 2 consecutive titles within months of Dantley's departure to Big D.

In Dallas, Dantley was markedly less effective than he'd been in Detroit, even before a broken leg in 1990 essentially ended his career. But when he retired in 1991, Dantley was 13th all-time in career PPG with 24.3 and had taken the 6th-most free throw attempts of any player in league history (he was 4th in career FTM). He was also the 3rd-most prolific scorer of the 1980s, scoring more points in fewer games and with a higher FG% than Larry Bird.

Any assessment of Dantley's career has to acknowledge that he didn't win a championship, that his teams typically got better after he left, that he was a poor defensive player, and that he generally had a one-dimensional game (he didn't really rebound, pass, handle the ball, shoot from long range, get out in transition, block shots, defend, or do much of anything anything beyond scoring massive amounts of points around the basket and at the line). You also have to judge him on the fact that his coaches disliked him, and he was traded 5 times because of personality/ego conflicts. But you can't overstate Dantley's brilliance as a pure scorer. Hollander once wrote, "The Sun rises in the East and Adrian Dantley averages 30 PPG," which is what they should put on his headstone someday. In the pantheon of guys who could simply fill up the basket with volume and efficiency, the list of players ahead of Dantley is short. Every player has weaknesses, and Dantley certainly had his share, but few have had strengths as remarkable as A.D.'s.

ShareThis

29 Responses to “Player Audit: Adrian Dantley”

  1. Ian Says:

    Thanks for that, Neil! I love the Bird comparison, and it really illustrates just how amazingly efficient Dantley was. I think it could be argued that Adrian Dantley is one of the top 10 offensive players ever, and certainly presented the most unique combination of volume and efficiency of scoring ever, given his size and the fact that one leg was 2 inches shorter than the other! Watching him play on offense was a true thing of beauty for me growing up. Those Piston heartbreaking postseason losses proved to be tough early lessons for me, and I bawled my eyes out when he was traded to Dallas.

    I think it's worth noting that in the years since he played, a lot of his former adversaries have come around to acknowledge him and put past grievances in perspective, most notably Frank Layden. And while it is true that several of his teams improved after his departure, his teams also tended to improve during his tenure, as Utah evolved from a perennial playoff doormat to a solid playoff team and Detroit went from a solid playoff team to a genuine championship contender. He also ranked a conspicuous 4th on your list of the greatest disparity between offensive win shares per minute and that of his teammates.

    Again, thank you so much for this!

  2. mrparker Says:

    Feel like I'm reading about Reggie Miller minus the trades. He's another all time offensive force. Will we be saying the same about Durant years from now?

  3. Mike Says:

    Reggie Miller couldn't touch AD offensively

  4. Gabe Says:

    "he didn't really rebound, pass, handle the ball, shoot from long range, get out in transition, block shots, defend"

    So, in summary, he was not familiar with the rules of the game of basketball. :P

  5. Jason J Says:

    I'm a Dantley fan mostly because I grew up hearing stories about what a great guy he was from my uncle who attended basketball camps that AD taught. I thought he got a little hosed as far as his legacy went by the trade in '89 because quite frankly he and Aguirre did the exact same thing, and Dantley did it better (and if Isiah isn't playing on a shot ankle in game 7 of the '88 finals, he likely wins a ring right there). In defense of his one-dimensionality, that was a pretty common expectation of small forwards of the day. Dantley, Aguirre, King, Tripuka, English, Person, and to a lesser extent Wilkins were all pure scorers who didn't do much else. I think Pippen's success helped change the position to a more well-rounded one.

    I do take a bit of offense to the Bird comparison though, only in that we're comparing a do-everything exceptionally well while making sure everybody else gets theirs team leader to a do-one-thing savant while making sure everybody else gets you the damn ball as often as possible one man team. If Bird stopped rebounding and defending and told McHale and Parish to get out of the post so that he could use his every-game size advantage (he was 6' 9" in an era of 6' 7" small forwards), there's a pretty good chance that his scoring efficiency would jump up while his team wins descended because of course taking post opportunities away form McHale and Parish is stupid.

    Really if you look at Bird's numbers, the one glaring "weakness" is that he didn't get to the line as often as you'd like, but it's pretty clear that the reason for that is because his team needed him to be a perimeter playmaker rather than a post scorer. He certainly had the chops to score in the post.

  6. Anon Says:

    "If Bird stopped rebounding and defending and told McHale and Parish to get out of the post so that he could use his every-game size advantage (he was 6' 9" in an era of 6' 7" small forwards), there's a pretty good chance that his scoring efficiency would jump up while his team wins descended because of course taking post opportunities away form McHale and Parish is stupid."

    Ehhhhhh...not so sure about that for two reasons that come to mind. Keep in mind that playing someone more in the post would take away some of his 3-point opportunities, which is a big part of efficiency especially for guys like Bird who can shoot it well from downtown. It's a tradeoff that can play a part in someone's shooting. Plus, I can't imagine Bird becoming more efficient without McHale and Parish doing what they're doing. By virtue of being reliable forces on offense in the post, they were able to stretch opposing defenses and as a result other teams wouldn't be able to cheat as much towards Bird. Perhaps if you took those players out of the lineup Bird would have more chances down low (and everywhere else for that matter), but then defenses would also be able to key in more on his every move and make things more difficult for him.

  7. Jason J Says:

    Take a look at Bird's 3 point attempts. He was great at it, but the team didn't call on him to do it much. For his career he took about 2 per game. That's an era difference. In today's game he'd probably average 5 or 6.

    Your point is well-taken, and you may be right, that removing efficient scorers who demanded attention could have made Bird's life tougher. But when we look at similar scorers like Bryant, who takes a lot of tough shots and hits them, we see that his TS% dips when Gasol, Bynum, and Odom start taking more of the load. Kobe's not giving up his difficult shots. These other guys aren't offering to take over the end of the shot clock heroics. He's giving up his forays to the paint which lead to layups or free throws, because that's where the bigs need to be. It's a good thing for the team. It's not a good thing for his TS%.

    I think, generally speaking, with the top tier players, you see them sacrificing the easy stuff and doing more of the fill in work in order to make their teams better. Having better players doesn't necessarily make their lives easier statistically. Jordan's efficiency peaked before Scottie developed. Shaq's peaked before Kobe developed, but both had their success as their teammates got better and took some of their control of the game, away. Spreading the wealth I guess.

  8. Neil Paine Says:

    That's an interesting comment, Jason, because it directly contradicts Dean Oliver's concept of skill curves -- that all players start with the basic high-percentage shots and add more difficult ones as the team needs them to (to the limit of the ability, that is). Under Oliver's paradigm, when a player reduces his Poss % he's giving up harder shots, not easier ones. Because the basic 5-10% of possessions you get "for free" are all easy shots and everyone gets those, they remain constant. It's the shots above and beyond that which become increasingly difficult, and giving up more of those makes your ORtg go up. Oliver confirmed this (and Eli Witus re-confirmed this) when he found an inverse relationship between usage and efficiency. Without this as an assumption for all players (the effects being felt in varying degrees, but absolute in its general impact), our basic model of the game of basketball breaks down.

  9. Jason J Says:

    Pre-apology for the length of this!

    I realize that Neil, and yet it does not necessarily seem to be true for great players, and it's something I have trouble reconciling in my mind. Unfortunately, I don't have the mathematical chops to present a proof here, just instances and my own logic for what it's worth.

    Kobe's best TS% came in the season with his second highest Usage. When he played with Shaq his usage rate was at its lowest and his TS% was worse than after O'Neal left. Universally. The game wasn't easier for him as an individual scorer because Shaq was drawing attention. Shaq created easy scoring opportunities for specialized role players who needed him to - Fox, Horry, Fisher, Shaw - but a slashing creator who loses opportunities to get to the rim because those possessions are designated to a more efficient interior scorer is not going to fare better statistically speaking. That's fewer layups and free throws (and Kobe's jumper is not in line with a Fisher or Horry's anyway).

    With Jordan you don't see it so much in his TS% which is always pretty high (though maybe it's worth noting that after having a TS% over .600 for 4 consecutive years, the first time (1992) that his usage dropped below 32, his TS% dropped to .579). Still you see it in his PER more clearly. As Pippen took more responsibility in '92 and '93, Jordan's per minute production dropped. Now professional sports observer Bill Simmons says Jordan peaked in '92. He certainly did not seem significantly slower or less explosive at 28 than he was at 25, and he had more skills in his toolset. So why the drop off in efficiency?

    Magic Johnson's shooting efficiency doesn't seem to coincide with his usage at all. When his usage bottomed out at 19% in 1983, his TS% was .603, and when his usage peaked at 27% in 1987, his TS% was .602.

    Larry Bird's highest Usage was 30%, and he had his second highest TS% that year.

    LeBron's Usage has been between 33.5 and 33.8 every year for the last 5 except for one, and each of those years except one his TS% was .568 or above. In 2007, when his usage fell to 31, his TS% dropped to .552.

    I'm not saying there's some direct link between increased usage and better true shooting percentage, but for the truly exceptional players, it seems pretty clear that increased usage does not decrease efficiency as an absolute - and I think my interpretation above does a decent job of accounting for the difference. I know for a fact that you can't blame these instances on health or age because there are too many of them. Here's a more concrete example of where I think the logic of it lies:

    In 1989 Doug Collins called a ton of screen and roll action between Jordan and Grant or Cartwright (Collins was notorious for micromanaging as a coach). This led to a lot of Jordan penetration. He shot 9.8 free throws per game and had his career high TS% that year. Phil Jackson came in, instituted the triangle, and gradually moved Pippen to the primary ball-handler, giving virtually all of the pick and roll action to Scottie (probably because Jordan was a superior outside shooter so it made more sense for Pip to create jump shots for Jordan than vice versa). By 1992 MJ's free throws had dropped to 7.4 per game though his minutes had only dropped by 1 per game. His usage was basically static (it dropped from 32.1 to to 31.7), but his TS% fell from .614 to .579. Yet the Bulls as a team shot much better in 1992. The Bulls team eFG% in 1989 = .507; in 1992 = .518; is this DESPITE the fact that the player on the team with the highest Usage had his shooting efficiency drop dramatically (conventional statistical wisdom), or BECAUSE of it (my very poorly supported hypothesis), or a total coincidence?

    And I think therein lies the notion of sacrifice that Bill Simmons termed "the secret" in his Book of Basketball.

    Looking at a few individuals - in 1989 the Bulls top 4 by win share had the following Usages / TS%:

    Jordan 32.1 / .614
    Grant 15.2 / .547
    Pippen 21.3 / .524
    Paxson 14.8 / .536

    The same list in 1992 (happen to be the same folks at the top of the WS column so it works for this unscientific BSing - again Usage / TS%):

    Jordan 31.7 / .579
    Grant 15.5 / .618
    Pippen 24.6 / .555
    Paxson 12.1 / .551

    Here's what I see:

    Jordan, who according to the folks who were there was pretty much peaking, did less and did it less efficiently in 1992 than in 1989.

    Grant's usage stayed pretty even, but his TS% skyrocketed. This is either a result of his personal improvement with age, the team getting him easier opportunities (which one might suggest used to belong to Jordan), or some other explanation I haven't thought of.

    Pippen's Usage went up and so did his TS% (like Mike, this is the opposite of what the whole "more work = less efficient" concept tells us), but like Grant this is likely due to getting older / better - and for the record Scottie's best TS% came in 1991 when his usage was relatively low which fits the Basketball on Paper expectation.

    Paxson's Usage drops and his TS% goes up, which is again precisely what Dean Oliver's research tells us will happen.

    Lord I just realized I should have been using ORtg instead of TS%, but a spot check shows basically the same thing. Jordan's Ortg was 2 points higher in 1989 than 1992 despite a small drop in Usage. Magic's highest Usage accompanies one of his best ORtgs. His lowest Usage accompanies one of his worst. Same with Kobe. Same with LeBron. Same with Bird. For role players like John Paxson as the Usage goes down, the ORtg goes up, as we would expect - less responsibility = more efficiency for him, but not for the greats. Maybe I'm looking at it the wrong way, but I think what I'm seeing is true, and I think it makes sense when you consider how teams have to distribute shots / touches.

  10. Neil Paine Says:

    The problem is, you can't look at it that way, because you can't tell whether the reduced touches were because he was less efficient or if he was less efficient because of the reduced touches (and vice-versa)... It's a classic chicken and egg problem to look at it on the seasonal level. However, Eli looked at it on the possession-by-possession level using lineup data, which skirts the problem:

    http://www.countthebasket.com/blog/2008/03/06/diminishing-returns-for-scoring-usage-vs-efficiency/

    He found a significant trade-off as a general rule. Perhaps some players can actually avoid this, and perhaps the rule doesn't apply to star players, but you can't escape the "chicken-and-egg" conundrum without doing what Eli did and looking at it on a lineup-based level.

    So I think we're safe in assuming the general rule applies to everyone, just not as much to some as to others. Or as Dean Oliver said, "Low usage players (I used 18% as a cut off) are more sensitive to increases than high usage players (23%). But both are still very significant. This implies that increasing Hoiberg's possessions 5% causes a bigger decline (about twice the size) than a similar increase in, say, Kevin Garnett. Or, from an optimization perspective, taking Garnett's possessions (who increases in efficiency only a little) and giving them to Hoiberg (who declines in efficiency a lot) has a pretty big cost."

    That's why I can safely say that, ignoring stylistic differences, peak Adrian Dantley would help your offense slightly more than peak Larry Bird. It's just that Bird was so much better in the other facets of the game, he was able to make up the difference and then some. That said, Dantley is still underrated.

  11. Anon Says:

    "But when we look at similar scorers like Bryant, who takes a lot of tough shots and hits them, we see that his TS% dips when Gasol, Bynum, and Odom start taking more of the load."

    I think there are some examples of skill curves for Kobe Bryant floating around on the web. You should take a look, because it demonstrates Dean Oliver's concept of "high possession rate/tough shots = lower offensive efficiencies" pretty well.

    Anyway, to your point about Kobe's efficiency going down when other players take more of the load -- it's not something that I agree with. Neil already beat me to it with the points he made above, but let me just ask you this question: how many times have you heard Lakers fans defend Kobe's low shooting %s when he has jacked up lots of shots in games over his career with something along the lines of "His fg% is low because he HAS to shoot a ton and no one else is doing anything"? Personally speaking, I heard this argument too many times too count -- and I'm NO Lakers fan, but they have a valid point. Kobe is a great tough-shot shooter, but I think that even he would say that it makes scoring alot easier when the opposing team has someone else to paying attention to, and he doesn't have to exert himself trying to shoot over double-teams.

  12. Jason J Says:

    Neil & Anon - It occurred to me as I was going to bed after writing that novel up there at 2AM last night, that what we're talking about here are skill-curves, which I'm assuming have something of a line of best fit associated with them and would even-out over time while allowing for a few seasons here or there to not line up right, and as such my notion and Dean's notion are not necessarily mutually exclusive.

    For the chicken / egg conundrum, that had certainly crossed my mind. What if these guys are getting their touches reduced specifically because their efficiency is slipping? Probably pretty likely (wish Doug Collins would have done that for Jordan in 2002). Again, looking at individual seasons, I'm pretty sure Jordan in 1992 wasn't significantly worse than Jordan in 1991 but his usage dropped 1.2 points and ORtg dropped 4 points while his team wins jumped up from 61 to 67. Jordan played almost 2 minutes more in 1992 than 1991 yet had a lower usage rate, took fewer free throw attempts per game, and shot 2% worse from the field. Certainly you'd think if the play style didn't change, the personnel didn't change, and the other teammates improved with age, Jordan's burden would ease and his efficiency would increase, but that's not what happened here. I don't have the resources to watch every possession, but again, in this instance, I would suggest that the young guns on the team were getting better, and Phil purposefully spread the ball around more, reducing Jordan's driving opportunities.

    I think the LeBron example is another where it's tough to say that he got worse when he went from 21 to 22. His usage dropped 2.6 and his ORtg dropped 3 points, and his PER dropped 3.5 points. Now that was Larry Hughes' first full season as a Cav. Could Hughes, also primarily a slashing wing player, have been taking some of LeBron's easy opportunities? I'm not sure in this case. LeBron's minutes dropped, his FGAs and FG% dropped. His FTA / min and 3PTA / min both dropped, so it does look like he got to drive less per minute on the court but not because he was shooting more from deep... So maybe, maybe not.

    Anon - When I first typed Kobe's name I realized I was opening a can of worms. Let me just say I'm a Celtics fan and a Jordan fan and not a Kobe apologist by any means. As you can see from the whole thrust of my argument here, I tend to disagree with whiny notion that having to take all the shots means you're going to play less efficiently. For Kobe in particular that's very hard to back-up. When his Usage peaked at a staggering 38.7 in 2006, when his fans claim the team was too terrible for him to play efficiently, his ORtg was 114, just one off from his career best. When he was at his best with Shaq (what I would consider the most help he had), his ORtg never surpassed 112. It's pretty clear (to me anyway) that Kobe is as efficient as he is. I do need to give a mea culpa regarding the Gasol years. Kobe's ORtg has stayed constant since Gasol arrived. His TS% fell, but the ORtg has been 115 every year since 2007, the year before Gasol arrived.

    Now I just noticed something on LeBron's page that seems very much inline with Neil's prognosis. LeBron's usage rate peaked last year at 33.8, and his ORtg peaked at 124 at the same time, AND it was his team's best season, AND I think it was pretty clearly the strongest team he was ever a part of. So in this case it looks like Brown is getting LeBron more opportunities as his efficiency is increasing or vice versa. Maybe having superior teammates is helping to create driving lanes this season, but having watched a ton of Cavs games, my opinion is that LeBron is just better than he has been at reading defenses and taking opportunities. He's also shooting an unreal 5 threes and 10.5 free throws per game at the same time. That'll keep ye ol' scoring efficiency nice and high!

    For what it's worth this has been a very interesting topic to discuss.

    Out of curiosity, Neil, what do you think of this http://arxiv.org/abs/0908.1801 paper by Brian Skinner titled "The Price of Anarchy", which suggests that to optimize team efficiency plays sometimes need to go away from the most efficient scorer - which I think can be taken to support both Dean's mentality (less work = more efficiency for the individual) and my notion (sometimes a star surrenders some of the easy stuff to lesser players, because he can do the hard stuff, and they can't).

  13. Jason J Says:

    Also - if we look at Usage as a continuum between 0-100, then logic dictates that, yes, at some point on that line increased usage will lead to lower efficiency. LeBron cannot use 100% of his team's possessions. He would become exhausted, and the defense could literally put 5 guys on him. On the other hand if he was to only use 1 possession per game, he would never develop rhythm or be able to set the defense up for counter moves, and his efficiency may suffer for not getting enough of a chance, unless we assume that LeBron always takes an easy first shot. There must be a balance point. Is it possible though that for elite scoring stars, the balance of usage to efficiency may be a on a higher scale than for the vast, vast majority of players?

  14. Anon Says:

    Jason J,

    I typed up a long post to discuss some of the points you brought up only to accidentally lose it all by closing my browser. **sighs** I need some more coffee. I'll type up my more expansive points at another time.

    Right now, to quickly paraphrase the things I was typing up -- I think that the Kobe Bryant you're talking about before 06 was mainly a player who was still not in his prime offensively. Watching him throughout his career, I can definitely tell you that during the Shaq-Kobe days, he was not doing some of the things on a skill level (particular form the post and the midrange) that he was doing when he reached his mid to late 20s (and still does right now). That's just one of the reasons why it's hard to look at the usage/efficiency paradigm season-by-season.

    Also, I think this particular point you made is actually something that is already part of the model:

    "Is it possible though that for elite scoring stars, the balance of usage to efficiency may be a on a higher scale than for the vast, vast majority of players?"

    The tradeoff isn't a nice, constant continuum for all players. For those elite players, taking on a small additional amount of offense isn't going to have any real noticeable effect compared to say, the "garbage cleaners on even your borderline stars who don't have their ability to create and make tough shots. But the effect is still there, especially when they dominate more and more of their team's offense and strain their skill sets.

  15. Neil Paine Says:

    I wish I had attended Brian's presentation at Sloan, to be honest. All of the paper presentations sounded a lot more interesting than listening to some (not all, but some) team executives make tight-lipped statements about things we already knew, trying to avoid giving away too much to the enemy. I think the traffic analogy is really fascinating, because it does seem to support skill curves -- but it also implies that a team's optimal mix is closer to giving 20% of possessions to everybody, from Eric Snow to LeBron James, which I don't anybody would have expected to be the optimal solution for an offense. That said, I definitely need to read the paper instead of hearing secondhand summaries to be more well-versed in what his research concluded.

    Also, with regard to the anecdotal examples of Jordan, Kobe, etc., it's worth noting that just because Jordan had a 125 ORtg one year and a 121 the next, it doesn't necessarily mean his "true ORtg skill levels" were 125 & 121... We can never know his true ORtg skill in any given season, we can just estimate it, and those estimates are prone to random variance. How much variance is tough to say, though, because almost any results you get are fundamentally tainted by the skill curve phenomenon. You're probably seeing why it's really tough to nail down a conclusion on this topic, to the point that some analysts (Dave Berri, Bob Chaikin) don't even believe there is a "skill curve" phenomenon for any players. At the other end of the spectrum, I'm content to assume the general rule of thumb for all players because what little evidence we have points in that direction, but I'm standing on admittedly shaky ground there, too.

  16. Jason J Says:

    Anon - Good point re: Kobe. Bryant's skillset certainly expanded over the years (though his athleticism has dipped pretty sharply as well), though I still stand on the notion that playing with Gasol isn't making things any easier as Kobe's ORtg has remained steady over that span while his Usage decreased, and we should have seen on increase in ORtg with a decrease in Usage, even if minimal, if I understand things right. To be accurate Kobe's ORtg is off this year, but I'm pretty sure his many injuries account for that.

    I'm trying to think of another reason to explain these season to season aberrations for Jordan and LeBron though, who seemingly should have been as efficient or better with the drop in Usage in 1992 and 2007 respectively but clearly weren't. Is there any reason to believe that a decrease in usage for an elite player (or any player I guess) SHOULD result in a decrease in efficiency?

    To eliminate the possibility of skills appreciating or depreciating as a player gets older, let's check out regular season vs. playoffs instead of season to season.

    Jordan Reg Season Career (only up to 1998 so his non-playoff Wiz years don't drag the numbers down): Usage = 33.5 ORtg = 120

    Jordan Playoff Career: Usage = 35.6 ORtg = 118

    Which is what is expected. Usage increase = minor efficiency decrease in a superstar. Check! His Usage and ORtg both increased in 1986, 1993 and 1998, but those are the exceptions to the rule. I think the main thing these numbers demonstrate is that Michael Jordan was ridiculously good.

  17. Jason J Says:

    "Also, with regard to the anecdotal examples of Jordan, Kobe, etc., it's worth noting that just because Jordan had a 125 ORtg one year and a 121 the next, it doesn't necessarily mean his "true ORtg skill levels" were 125 & 121... We can never know his true ORtg skill in any given season, we can just estimate it, and those estimates are prone to random variance."

    I hadn't been thinking that way at all. That makes a lot of sense. Given the small relation we're seeing in MJ, it's probably all fairly negligible. I still enjoyed the conversation.

  18. Neil Paine Says:

    Me too. It's nice when we can have a civil, thoughtful conversation together. As opposed to when certain other elements join the thread (not to name any names)...

  19. Anon Says:

    "Bryant's skillset certainly expanded over the years (though his athleticism has dipped pretty sharply as well), though I still stand on the notion that playing with Gasol isn't making things any easier as Kobe's ORtg has remained steady over that span while his Usage decreased, and we should have seen on increase in ORtg with a decrease in Usage, even if minimal, if I understand things right."

    I think his case is pretty interesting actually, because if you look at the seasons after 06 in which he takes a step back in his offense, you see that he had his best shooting seasons from the field -- his effective field goal and true shooting percentages have come in the 07 and 08 seasons. So I think that at least THAT part of his game benefited from not having to outshoot the other team one on five like he did during 06. But here's the interesting thing: what actually hurt his efficiency in 07 and 08 (compared to his 06 season) was that he didn't take care of the ball as well! During those seasons his turnover rates went up, which can lower your ORtg. You also have to remember that Kobe injured his shooting hand in 08 and has re-aggravated the injury over the years. So I think these are all little things that can be taken into consideration, at least if you're looking to explain why he hasn't made a quantum leap in ORtg (along with the other reasons that have already been brought up).

  20. Mike G Says:

    When a player truly believes he should get 20 shots per game, and the coach or the other players don't get him that many opportunities, he may well start taking shots he normally wouldn't take -- and his efficiency will drop.

    This is true for Dantley (career .617 TS%) and for Walker (.484). Not everyone needs 20 shots or any particular number. But maybe everyone has his own optimum.

    When Dantley went from Utah to Detroit, his Usg% went from 30 to 24, but his TS% remained right around .620 . He was already 30 YO, so that faded gradually.

    Jordan in 1991 averaged a career low 37 minutes. For the first time in his career, he had decent help in the backcourt: BJ and Pax filling the other slot, Hodges and Hopson backing MJ himself. And Pippen in a pinch.

    In what would be a recurring theme, Phil Jackson always wanted MJ to stay under 38 MPG, and he never could. In '92, Hopson's career went belly-up, Hodges was on his last legs, and Bobby Hansen was brought in well past his prime.

    So, while Jordan's Usage rate may not have been taxing, his minutes were. He didn't exactly coast through many minutes, at either end of the court. His only recourse was to 'retire' every few years.

  21. Jason J Says:

    Good point, Anon. Maybe Kobe's playmaking for others, rather than launching so many shots, contributed the TOs, which hurt the ORtg despite the fact that his Usage dropped notably. I wonder if Asst% could be factored in as well (decreased playmaking should decrease turnovers).

  22. Anon Says:

    @ Jason J

    Knew I'd finally find this somewhere on the web: an example of a skill curve for Kobe. This is the one from 2008 http://www.basketballgeek.com/wp-content/uploads/2009/02/picture-4.png

    The beauty of this example is that you only look at the usage vs. efficiency from one season so you can avoid some of the problems of looking at this phenomenon from season to season. Of course variance always plays a part in these things (Kobe actually improved his efficiency a little when he went from around 20% to 23% of the offense, so it could be showing what Neil said earlier about elite players not being really affected by a small workload increase.I think it also partly might be due to he usually doesn't have many games working under 24% of the offense as he does for 25%-33%, so it could skew the regression curve), but overall you can see that his efficiency does take a gradual hit as he increases his load especially as he hits the 30% + range.

  23. Jason J Says:

    good find, anon. i'm surprised kobe's optimal workload seems to be less than 25% usage (and you may be right that the hit he takes below 24% is just that he's not used to it - though it might be that he needs a certain number of reps in a game to be at his best), but then again not that surprised. and it kind of jibes with the even split Brian Skinner mentioned too...

  24. Romain Says:

    Bill Simmons thinks that Jordan peaked in 1992, I'd say he peaked in 1991:

    PER = 31.6 in the regular season (2nd only to his 1988 season) & 32.0 the playoffs (his best ever)
    WS/48 = 0.321 in the regular season (his best ever) & 0.333 the playoffs (also his best ever)
    ORTG = 125 in the regular season (his best ever) & 127 the playoffs (also his best ever)

    All of this while his USG% is among the lowest of his careers.

    In other words 1991 is the year where Jordan relied the most on his teammates to win.
    You can notice that with his outstanding average of 11.4 assists per game in the NBA Finals. This is also the year when the Bulls had their easiest title run, losing only 2 games in the playoffs.

    By 1993 he had returned to a higher USG% and lower efficiency (by his standards of course), as illustrated by his 41.0 ppg average the NBA Finals.

    The question of course is, did Jordan share the ball less because his team was weaker, or was his team weaker because Jordan shared the ball less?

  25. Anon Says:

    "The question of course is, did Jordan share the ball less because his team was weaker, or was his team weaker because Jordan shared the ball less?"

    I know I'm not a pro basketball player by ANY means, but after playing a pick-up game yesterday where the players on my team couldn't make a single shot and I had to keep taking the ball on offense to make some something happen, I'm going to say that players tend to share the ball less when your don't have a great team around you (or your team is struggling) rather than the other way around.

    Now I know how Allen Iverson sometimes feels on offense.

  26. Jason J Says:

    Anon - I remember a Jordan quote from way back when people always criticized him for not making his teammates better and considered Magic and Bird the superior players ('87 / '88 or thereabouts). He said something to the effect of, "Give me Worthy and Kareem to make better, and we'll see." Great for building team unity, huh? Of course once Pippen and Grant and BJ all developed and they got some bench depth nobody questioned Jordan's ability to make his teammates better. Chicken or egg again. Did Jordan start deferring to his teammates and therefore the players improved with more reps, or did the players get better therefore Jordan gave them more touches?

    I remember Chris Webber and Gary Payton (I think it was GP) saying that they played the same on good teams and bad teams and results had everything to do with the quality of their teammates - this was an argument about whether or not Al Jefferson deserved to be an All-Star last year when he was putting up great numbers on a losing team.

    I don't know if I really buy that though. Are Peja and Bibby really better than Juwan and Strickland in the late 90s, or are they just a better fit with Webber allowing him to play his game?

    It gets me thinking about player roles and the whole shot distribution question we touched on above. If the optimal split of shot attempts for any 5 man unit is 20% of the shots each instead of a higher proportion going to the better scorer, how does that work from a logistical standpoint? If the plays aren't going to the star who demands a double team, how does the spot shooter get open? If the slashing scoring specialist or post scorer isn't drawing help defenders at the rim, how does the offensive rebounding specialist get his tip ins? Sure it might make sense for the ball to be spread evenly, but how do you accomplish it?

  27. Anon Says:

    Jason J,

    It's indeed an interesting question. I think that in an ideal sports world where all basketball players are around the same talent level, having each player take on around 20% of the offense WOULD be the best way to play the game. It would certainly be intuitive too -- having all five players on the floor carry a equal brunt of the load and maximize their efficiencies in a team game I think would be better than a team that has one player shouldering the entire load and and bunch of non-impact players around him. The only problem is as you suggested, that not everyone is the same in skill and talent level. As a Cavs fan, I heard the criticisms that LeBron dominates the ball too much and does too much for his team, and would be better off taking a step back in the offense. I actually think that LeBron himself would agree with that one...the only thing is, where would those shots go? Delonte West might be a good candidate -- he can handle the ball well (the Cavs best ball handler outside of Lebron), set others up, go to the basket and has a nice stroke especially off the dribble (the dribble-drive and stepback jumper on the wing is his specialty), but he can be a notoriously streaky shooter. Mo Wiliams is the best shooter on the team and is also a good playmaker, but he's prone to turning over the ball and doesn't have great ability to get by his man with the first-step and get to the rim. Andy Varejao has been really efficient on the offensive end and has improved his ability to finish around the rim. He's even surprised me this season with some moves in the paint, but I still wouldn't be calling plays for him in the post yet on a consistent basis, where you have Mo or LeBron take the ball and just dump it down to him on the block. Shaq can still post up and score, but he has been injury his game is not what it once was, and you can neutralize him by sending him to the foul line (and Dwight's not a good free throw shooter either, but unlike Shaq now he's both strong AND freakishly athletic, so he can punish you down-low and also with the pick-and-roll even if his post repertoire is still a work in progress). I think that picking up Jamison would be LeBron's best bet for shouldering that load offensively, but it's taken him a while to find his way in the offense and he has also been banged up this season. And NONE of them has Lebron's ability to hit tough shots especially at the basket. So I don't think it's bad at all to allow your star player (who is also the league's best player) to take the shots that you would probably struggle with your skill level. Perhaps if Lebron had someone like Bosh or Wade around him, he would certainly take a step back and give up the ball more often on offense (and you certainly saw that in the '08 Olympic games. LeBron setting up his superstar teammates in Kobe, Wade, Bosh, etc. was great for the game). And in several Cavs games I've watched this season, LeBron has always looked for his teammates when they are working well offensively moving without the ball and off of screens, cutting towards the basket, and knocking down open shots. But when they don't do that and the team needs offense or if the game is close in the 4th, I don't see how Lebron distributing his workload to the Delontes and Andys of the world will help you win more games. I know it's a team game, but SOMEONE has to shoot the ball, and I'd rather have my star player who has the best ability to put the ball into the rim (and also create looks for others by doing so) handle that responsibility more often than not.

  28. Ryan Says:

    I don't mean to take a step back in time, but this discussion, in contrast to the early-late 80's (minus a few specific high-scoring super stars), may be entirely different.

    The perimeter orientated, like-Mike focus of the league has certainly changed the way a team on the opposite end would approach this situation.

  29. Scott Says:

    You might want to clarify this statement - "making him the only NBA player to hold this distinction" - since many ROTY have been traded, including at least one (Elton Brand) after his second season. Dantley's only possibly unique in being traded *immediately* after winning ROTY.