How free agency changed NBA team building
This summer, the NBA free agency period was more frenzied than ever before. Dudes were getting crazy deals! Four guys who have never been to an All-Star game — Mike Conley, Bradley Beal, Nic Batum, and Hassan Whiteside — combined to bank about a half-a-BILLION dollars on the open market. I already told you that the Green Falcon, Harrison Barnes, was getting maxed out this summer, and he did. And those were some of the good decisions! I mean, the Portland Trail Blazers felt compelled to pay $184M to three guys — Allen Crabbe, Evan Turner, and Mo Harkless — who all play the same position!
It wasn’t just the big paychecks that were startling, either; there was some dramatic player movement, too.
After nine years in Atlanta, Al Horford left the Hawks for the Boston Celtics. Dwyane Wade — twelve years a Heat player — took his talents out of South Beach and headed back to Chicago. Big names like Dwight Howard, Pau Gasol, and Derrick Rose changed teams, too. But no free agent made a bigger splash than Kevin Durant. After nearly a decade as the marquee player of the Thunder franchise, KD headed to the Golden State to join Stephen Curry and the Warriors.
There has been a fair amount of digital hand-wringing on NBA-Twitter over the recent accumulation of star players on a small number of super-teams. It’s easy enough to draw the parallels between the new Durant-Curry partnership and the one formed by LeBron James and D-Wade in Miami and then blame LeBron’s Decision as the dark genesis of some new anti-competitive NBA landscape. However, the construction of super-teams through free agency and trades is nothing new. To find out where it all started, we need to go back to the summer of 1988, all the way back to Tom Chambers.
The first summer of free agency
There’s a reason Oscar Robertson spent the first ten years of his career toiling away for the Cincinnati Royals (a basketball team). He couldn’t leave! There was no such thing as a free agent back then. In fact, it wasn’t until the summer of 1988 that a new collective bargaining agreement created the concept of unrestricted free agency. Prior to that time, a player in his prime would only part ways with his team if — like the Big O — he was traded away (or, in rarer cases, via a funky mechanism akin to the modern sign-and-trade agreements). Likewise, NBA teams had basically two ways to add new players: draft a rookie or make a trade with another team. As such, success on draft day was absolutely paramount to having success on the court.
All that changed in the summer of 1988 when a 6-10 All-Star power forward out of the University of Utah became the first ever unrestricted free agent. The 29-year-old, seven-year NBA vet, Thomas Doane Chambers, left a crowded Seattle SuperSonics front court for Phoenix. He teamed up with an impressive young collection of talent — Kevin Johnson, Jeff Hornacek, and Tyrone Corbin — acquired by the Suns via the draft. The pairing was an immediate success as Chambers was more productive the following season than ever before — averaging 26 points per game — and with him the Suns went 55-27 and reached the 1989 Western Conference Finals.
Chambers started a revolution. More free agents followed after him and the process of team building was never the same. You can see in the chart below how successful teams, championship teams, were dependent on drafted players during the pre-free agency era, whereas players acquired through other means (free agency, trade) became more critical to championship teams during the post-free agency era.
To help visualize the paradigm shift in team building, I’m using VORP — the Value Over Replacement Player — as a summary measure of a player’s contribution to his team. The plot shows the percentage of positive team VORP contributed by players acquired through the draft for each championship team since 1980. Some minutiae: Scottie Pippen wasn’t technically drafted by the Bulls; he was acquired by Chicago via a draft-day trade, but I treat him as if he was native Chicagoan, anyways.
I did analogously with any other draft-day trades. LeBron James was drafted by Cleveland, but he was acquired by the 2016 Cavs via a sign-and-trade agreement, so I did not consider him to be a drafted player for that team. I did the same with any other player who was drafted, left, and came back. In the plot, key contributors — players with individual VORP of at least 2.5 — are highlighted within the bars and the year they were drafted is shown. I’ve divided the teams into two groups — those constructed around a franchise player who was drafted before 1989 (pre-free agency, blue bars) and those teams constructed around a franchise player who was drafted in 1989 or after (post-free agency, red bars).
Percentage of positive team VORP contributed from players acquired by the draft for each NBA champion, 1980-2016.
*Ben Wallace joined the league as an undrafted free agent in 1996.
The two dotted lines demonstrate the dramatic shift in the role of drafted players on championship teams before and after the introduction of NBA free agency. Championship teams constructed before the free-agency era relied more on the players they drafted — with those “lifers” contributing 65 percent of positive team VORP, on average. In contrast, championship teams constructed after the free-agency era have relied more on other means of team building — as drafted players accounted for only 41 percent of positive team VORP, on average.
More anecdotally, you can see how the old-school champs required a pair, if not a trio of great players drafted at short intervals: Larry Bird (‘78) – Kevin McHale (‘80) for the Celtics, Magic Johnson (‘79) – James Worthy (‘82) for the Lakers, Isiah Thomas (81’) – Joe Dumars (‘85) – Dennis Rodman (‘86) for the Pistons, and Michael Jordan (‘84) – Scottie Pippen (‘87) – Horace Grant (‘87) for the Bulls. Since the onset of free agency, several teams have succeeded in winning a title with just one foundational star acquired from the draft: Kobe Byrant (‘96) with the Lakers, Dwyane Wade (‘03) with the Heat, Paul Pierce (‘98) with the Celtics (with an assist from Rajon Rondo), and Dirk Nowitzki (‘98) with the Mavs. The Spurs are special — they overlapped the 20-year career of their No. 1 draft pick, Tim Duncan, with two other franchise players at the beginning (David Robinson) and the end (Kawhi Leonard) of his career.
In the year 38 A.C. (After Chambers), the 2015-Warriors were a throwback in championship team construction. Like the dominant teams of the 1980’s and 1990’s, Golden State’s focus on continuity and player development resulted in a powerful nucleus of draftees in Stephen Curry (‘09), Draymond Green (‘12), and Klay Thompson (‘11).
But, were these Warriors the best-drafted team in the free agent era? And, if not, who has benefited the most from the draft since the start of free agency?
The best-drafted team of the free-agency era
If we start with the 1989-90 season and sum the individual VORP scores for each player making a positive contribution, we find 50 teams that had a total VORP of at least 19.6. That list comprising the best regular-season teams includes 17 (of 37) eventual champions plus 33 other ultimately unsuccessful championship contenders.
The top-10 is populated almost entirely by three dynasties: the Jordan-Pippen Bulls, the Curry-Green Warriors, and the Durant-Russell Westbrook Thunder.
We already touched on the 2015 Warriors above. With positive contributions from Curry, Green, Thompson, and Barnes (plus 0.1 VORP tossed in by garbage time-Warrior, Ognjen Kuzmic, just for laughs), Golden State’s drafted players tallied a total VORP of 17.1, which was good enough for tenth place in the free-agency era. In 2016, the drafted Warriors — Curry, Green, Thompson, Barnes, and, this time around, Festus Ezeli– were even better. They accounted for a VORP contribution of 18.4. Of course the Warriors had some other key players — Andre Iguodala, Andrew Bogut, and Shaun Livingston — that joined the team through other means. This fact is borne out in the table above, as you see that the two Warriors teams each had a relatively modest proportion of total team VORP contributed by drafted players (75 percent and 79 percent in 2015 and 2016, respectively) compared to other teams in the top-10.
By contrast, the 2016 Thunder received the largest percentage of their positive VORP contributions from players acquired via the draft — very nearly 100%, in fact. Seven Thunder draftees — Westbrook, Durant, Stevie Adams, Andre Roberson, Serge Ibaka, Nick Collison, and Cameron Payne — pitched in for 19.5 of the team’s 19.6 positive VORP (99 percent). That was the third-best squad acquired through the draft in the free-agency era, even a bit better than the more starry Thunder quartet of 2012 that featured Durant, Westbrook, Harden, and Ibaka.
The honor of best-drafted team goes to 1990s Bulls, specifically to the 1992 Bulls that trotted out Jordan, Pippen, and Grant. This all-time-great team racked up a huge total VORP: a sum of 23.9 from players making positive contributions. The four players drafted by Chicago — Jordan, Pippen, Grant, and Perdue — represented 93 percent of that total, a VORP of 22.3.
In case you were wondering, Chambers and his fledgling 1989 super-Suns fell just short of making the cut, with positive VORP contributions totaling 18.1, 65 percent of which came from Phoenix draft picks: KJ, Hornacek, Thunder-Dan Majerle et al.
What were the chances?
OK, so we’ve established that the two 2016 Western Conference Finalists were outliers in the era of modern team building — benefiting from the draft more than any other non-Michael Jordan-led team. But, does their success stem from effective pre-draft player evaluation? An emphasis on continuity? Or did they just get lucky?
We can evaluate the recent hot streaks of these two franchises in the draft relative to league-wide expectations. First, we just need to create a simple model of expected career success based on a player’s position in the draft.
For our data, we’ll consider all the players that have been selected since the 1989 NBA Draft. This is a convenient cutoff point, because the 1989 Draft was the first to take place after the initiation of unrestricted free agency with Chambers in 1988. Moreover, the draft in 1989 was the first to use the modern, two-round format.
Below, we visualize the career success of each player in each draft since 1989. Each row represents a single draft with picks No. 1 thru (up to) 60 displayed from left to right. Each column represents a single draft-pick slot, with the years moving from 1989 (top) to 2015 (bottom). Each player selected is color-coded to represent his current level of career achievement: dark blue = an MVP Award winner, light blue = an All-Star, orange = a starter (i.e., 100+ career games started or 50+ percent of career games started), red = an NBA pro (i.e., 50+ career games played), and gray = a washout, bust, hack, what-have-you (i.e., < 50 games played)
Career success for NBA Players drafted since 1989, pick-by-pick.
As one might expect, the blue-blooded players (MVPs and All-Stars) tend to be concentrated in the lottery picks, with a smattering of pleasant surprises like Manu Ginobili found further to the right. However, even the No. 1 draft pick isn’t necessarily guaranteed career success, with many achieving no more than generic starter status and one yet to find his way off the bench (I’m looking at you, Anthony Bennett). On the other end of the spectrum, the gray squares (washouts), tend to be concentrated in the second round, but there are plenty of busts from the first round, too, with Royce White being one of the most recent casualties. Of course, the draft classes of 2013, 2014, and 2015 still have a disproportionate amount of red and gray squares, because those younger cohorts haven’t really made their mark on the league, yet.
It’s easier to visualize career success by draft-pick slot if we reorganize the plot to look more like a histogram. That is, rather than plotting the draft year on the vertical axis, we shuffle the data so that each draft-pick slot is shown sorted by the players’ career achievements. Each column still represents a single draft-pick slot from No. 1 thru 60 (left to right). But, now, the bottom row represents the best player selected at each draft position (or at least, the one with the highest career achievement, to date) and the top row represents the worst player selected at each draft position.
Career success for NBA players drafted Since 1989, a histogram
Using the histogram, you get a better sense for the probability of, say, pocketing an All-Star with a lottery pick. Take, for example, the No. 1 draft picks. To date, the 27 No, 1 picks since 1989 have produced five league MVP winners: Shaq O’Neal (‘92), Allen Iverson (‘96), Duncan (‘97), LeBron (‘03), and Derrick Rose (‘08), 12 All-Stars, nine starters, and one replacement-level pro. We can smooth out the pick-by-pick fluctuations in the histogram and quantify the probability of each career achievement for a given draft position using a series of logistic regression models, as shown in the plot below.
Career success for NBA players drafted since 1989, modeled
This model is very simple — accounting for only one potentially predictive factor, draft position — but, it will allow us to estimate the expected probability of each level of career achievement for each of the players drafted by the Warriors and the Thunder (previously the SuperSonics) during the relevant team-building periods. For the Warriors, this included nine players drafted from 2009 to 2012 and for the Thunder/Sonics this included nine players drafted from 2007 to 2009.
Expected-versus-actual career achievements for recent draft picks of the Golden State Warriors (2009-2012) and the Oklahoma City Thunder (2007-2009, previously the Seattle SuperSonics)
From 2009-2012, the Warriors drafted a few players that have yet to make much of an impact in the league: Udoh (‘10), Tyler (‘11), Jenkins (‘11), Ezeli (‘12), and Kuzmic (‘12). But, as you can see, this was pretty much to be expected given their positions in the draft pecking order. As such, the Warriors were still able to accumulate 1.38 more pros and 0.11 more starters than expected. But, the Warriors really made their hay with three very special draft picks. Thompson (16 percent probability) and Green (2 percent probability) were long-shots to achieve All-Star status based on their draft positions of No. 11 and No. 35, respectively, and Curry was a real MVP dark horse (1 percent probability) with his starting position of seventh in the draft. Overall, the Warriors brought together three All-Stars with those nine picks, when most teams would have been expected to find only one (#lightyears).
From 2007-2009, the Thunder did a great job drafting too. Of nine selections, seven made it in the league, five of which became starters, and three of those were ultimately All-stars. Of course, Sam Presti wasn’t perfect either — selecting not only James Harden, but also his less successful pseudo-cousin Devon Hardin, too — but, a haul of Durant, Westbrook, Harden, Ibaka, and Jeff Green is pretty darn impressive. Relative to the Warriors, though, the Thunder/Sonics had better draft picks with which to work, including a pick each from the No. 2, 3, 4, 5 draft slots. As such, it was slightly less surprising that they turned their nine draft picks into Durant – Westbrook – Harden — three eventual All-Stars — as they could have been expected to find 1.6 All-Stars with those picks, on average.
The continuity challenge
Of course, drafting great players is only half the battle in the modern NBA and a smart draft pick will pay the biggest dividends only if you can keep him from leaving your team. Maintaining roster continuity is infinitely more challenging in the free-agency era, as exemplified by the Thunder’s own experience with Harden in 2012. Rather than risk having The Bearded One leave Oklahoma the following summer as a free agent without getting anything in return or, worse, commit to a bidding war in order to keep him around, they opted to trade the future superstar instead. While this oft-criticized personnel decision of the summer of 2012 chipped away at a young nucleus acquired through the draft, it pales in comparison to what went down this summer.
First, the Thunder swapped Ibaka for Orlando’s Victor Oladipo, Ersan Ilyasova, and Domantas Sabonis. Then, they were blindsided when KD spurned them for Golden State. As a result, the Thunder, a team that was modeled on growing through the draft and maintaining continuity suddenly features lots of unfamiliar faces with uncertain prospects for success next season..
Durant’s move to Golden State didn’t just break up the nucleus that the Thunder built through the draft, it also harshed the home-grown mellow of the Warriors. Golden State was forced to part ways with their former lottery pick, Barnes, to make room for Durant. More subtly, one might expect slightly less productive seasons (and thus, smaller VORP) for two other career-Warriors, Thompson and Curry, as the Splash Brothers will likely shift some of their scoring burden onto the shoulders of their new teammate, Durant. In 2017, a few less splashes might result in a smaller team VORP contributed from players acquired by the Warriors via the draft. Thus, in this freest of free agent periods, both of the greatest drafted teams in recent NBA history have been dismantled in one fell swoop.
I’ll leave it to you, dear reader, to decide if it’s worth being upset about the breakup of two great drafted teams and the formation of the most recent super-team. But, if you are mad, don’t blame KD, don’t blame LeBron. Blame Tom Chambers!
The data for this article was collected from Basketball-Reference