Going to Extremes: Building a Winner in the Modern NBA

By Evan Barone and Kees Van Hemmen

There’s no other way to say it: the Chicago Bulls are better off than half the league. The Phoenix Suns are better off than half the league. Even the Nets – the Brooklyn Nets – are better off than half the teams in the NBA. Let’s get a little more specific: what we mean to say is that, if a team’s goal is to win the championship, the worst thing that team can possibly be is average. This is not because winning on its own is a negative; it is because the NBA’s talent acquisition system–the draft lottery–is set up to reward teams that perform horribly. The best players on championship teams are disproportionately taken at the top end of the draft.

Based on VORP1 (value over replacement player), of the top three players on teams with the top four records during the 2017-2018 season, 75% were drafted in the top ten, 50% were drafted in the top five, and 8.3% were drafted number one overall. Of the top three players on the last four different championship winning teams, 66.7% were drafted in the top ten, 50% were drafted in the top five, and 33.3% were drafted number one overall. Moreover, of the twelve all time VORP leaders, 83.3% were drafted in the top ten, 75% were drafted in the top five, and 41.7% were drafted number one overall. In short, tanking and drafting are the two controllable steps to building a contender in the modern NBA.

Most NBA general managers recognize that the draft incentivizes losing. As a result, the majority of executives have shown a willingness to tank2 for a season or two. That is, before either fan outrage, media pressure, or an overestimation of their current assets causes them to attempt to win again. However, both the need for multiple star caliber players in the modern NBA and the uncertainty of the draft prove this strategy to be insufficient if the ultimate goal is to win a championship.

Over the past 10 years (2007 to 2017), NBA championship winning teams have averaged 2.2 All-Stars3 and 1.8 All NBA4 players on their roster. The only way to draft All-Stars with any measure of consistency depends heavily on selecting at the top of the draft. The odds of acquiring an all-star player are, on average, 47% if selecting with the number one pick and 37.5% if selecting in the top five, even though only 10.3% of all NBA players ever become All-Stars in their respective careers.5


The analytical data validates these numbers and supports a drastic skew in favor of picking in the top five to ten draft positions. The average WOR6 is 4.73 for number one draft picks, 3.77 for top five draft picks, and 3.13 for top ten draft picks, compared to 0.98 for the draft as a whole.7 Furthermore, in terms of Win Shares,8 the data demonstrates a nonlinear downward trend wherein the additional value of having a high draft pick is significantly greater than would be expected from a linear relationship.9


With enough data to confirm the viability of using All-Star games as a metric for player success, and knowing the average amount of All-Stars needed to win an NBA championship, we can project the number of years a team needs to tank to have a legitimate chance to win the championship, assuming one first and second round pick per year. Finishing with a bottom five record gives you an average draft position of 4.2 in the first round under the new lottery format,10 and drafting at an average position of 4.2 correlates to 31% of players becoming all-stars.11 From the time that the NBA went to the two round format in 1989, 1.9% of second round picks have become all-stars.12 Therefore, with the mean number All-Stars on a championship team being 2.2, it would take an average of 6.7 years of finishing in the bottom five to acquire this amount of talent. Most importantly, this method of team building can be feasibly executed within the framework of the league’s rules, based on the financial structure of player employment. With the presence of economical rookie scale contracts and the salary cap veteran exception, known as Bird Rights,13 the NBA has set up a system where teams are initially afforded cheap labor, and then are able to retain those players as they become veterans and sign more lucrative contracts, even if the sum of those contracts exceeds the salary cap. It is true that rookie contracts expire after a maximum of four years, and the players then have the opportunity to sign with another team, which might become problematic to our idea of a six or seven year rebuild. However, NBA All-Star caliber players rarely sign their second contract with a team other than the one that drafted them due to restricted free agency. Of the seven All-Stars that have signed their second NBA contract in the previous three seasons (2015-2017), all of them have re-signed with the team that had drafted them, and none of them signed a contract for less than four years. Based on this data, teams can be fairly certain that if they manage to draft an All-Star, they will have that player on their roster for at least the next eight years. Another concern for such a long rebuild is that players drafted at the beginning of the process will not be in their primes at the same time as the players drafted at the end of the rebuild. However, the average age of players drafted in the last four drafts was 20.3, and the average peak of player performance is age twenty-nine, according to calculated efficiency per thirty-six minutes.14 Therefore, even in an extended rebuild lasting seven years, there will be a reasonable window of opportunity to compete before the players selected in earlier drafts begin to decline.   

It becomes difficult to compare this theory to historical results, as few teams have gone to such an extreme, but the Philadelphia 76ers, led by Sam Hinkie, provide a decent case study to analyze. They have had six lottery picks in the last five years, accounting for an average draft position of 4.8. Despite draft busts such as Jahlil Okafor, Nerlens Noel, and Michael Carter Williams, the Sixers still managed to draft a plethora of promising young talent due to the high quantity of top end picks they had. They have selected one current All-Star in Joel Embiid, and one almost certain future All-Star in phenom Ben Simmons. They have also traded the rights for Elfrid Payton at pick number ten for potential future All-Star Dario Saric at pick number twelve, who is currently averaged fifteen points and seven rebounds at age twenty-three. As a result, the Philadelphia 76ers ranked sixth in the ESPN NBA future power rankings for 2017-2018.15 In contrast, perennially mediocre teams such as Detroit and Charlotte come in at only twenty and twenty-one, despite both of those teams making the playoffs just two years ago (2015). In the last decade, multiple finals participants have been built their team around drafting in the lottery multiple times. This includes the 2011-2012 Thunder with Russell Westbrook, Kevin Durant, and James Harden, as well as the 2014-2015 Warriors with Klay Thompson, Steph Curry, and Harrison Barnes, proving the viability of building through the draft. Fortunately for these teams, they were able to capitalize on a disproportionate amount of their draft picks, allowing them to cut their rebuild short by a number of years. While this is an ideal scenario for any team, the numbers support the notion that this amount of drafting fortune can not be reliably expected.  

We acknowledge that waiting nearly seven years on average before even beginning to attempt to win games is less than ideal in a world of ravenous fan bases and unforgiving media, but for small and mid-size market teams that do not historically attract premiere free agents, it is the most viable path to a championship. However, there is hope that this procedure can be expedited through the implementation of prudent drafting measures. General manager success evaluations16 show that the top general manager drafts 58.3% more effectively than the average general manager. While it will still be necessary to finish in the bottom of the standings for an extended period of time, prudent drafting habits will shave valuable years off of what will surely be a grueling rebuilding process.

To New Orleans, Memphis, Milwaukee, and the rest of you teams that the Lebrons of the world refuse to even return a phone call from, there is hope. There will be many hardships ahead, there will be people calling for your jobs, there may even be hordes of protesters in the streets. But when you hold that Larry O’Brien trophy, when you are championed as a visionary, a hero, it will all be worth it. Sam Hinkie was praised by analytics junkies for coining the phrase “trust the process.” While we appreciate the sentiment, it came under particular scrutiny for the amount of blind faith it necessitates. Instead, we ask fans to “trust the numbers,” and have confidence that through aggressive tanking, their team will have the maximum probability of championship success. And, to all the GMs out there, once again we beg of you: go to extremes.

  1. Value Over Replacement Player (available since the 1973-74 season in the NBA); a box score estimate of the points per 100 TEAM possessions that a player contributed above a replacement-level (-2.0) player, translated to an average team and prorated to an 82-game season. Multiply by 2.70 to convert to wins over replacement (WOR). The average WOR for all qualified players from 2007-16 is 2.45 (basketball-reference).
  2. A strategy that NBA general managers use in order to finish at the bottom of the standings and secure a high draft pick. It is generally executed by putting together a roster composed primarily of players that were undrafted or were unsigned free agents, often paired with whatever high potential players a team can accumulate by taking on bad contracts via trade.
  3. 12 best players from each conference as selected by the fans, media, players, and coaches.
  4. The 15 best players in the NBA as selected by broadcasters and sportswriters, split into All NBA first, second, and third teams.
  5. he last 50 years the data was available (1964-2014), 317 out of 3017 NBA players have become all-stars (10.3%). https://en.wikipedia.org/wiki/List_of_NBA_All-Stars http://www.red94.net/3071-nba-players-past-50-years-guess-many-better-start-james-harden/13856/#
  6. VORP multiplied by 2.7 (see footnote 1)
  7. http://www.ballandone.com/2016/08/24/expected-return-draft-position/
  8. Win Shares is a player statistic which attempts to divide credit for team success amongst individuals on the team. It is calculated using player, team and league-wide statistics. The sum of player win shares on a given team will be roughly equal to that team’s win total for the season https://www.basketball-reference.com/about/ws.html
  9. http://berkeleysportsanalytics.org/feed/
  10. https://www.theringer.com/nba/2017/9/28/16381426/nba-draft-lottery-reform-passes
  11. See footnote 4
  12. 16 second round All-Stars divided by 843 second round selections equals 1.9% (since 1989)
  13. https://basketball.realgm.com/analysis/239564/CBA-Encyclopedia-Bird-Rights
  14. PTS+(FGM-FGA)+(FTM-FTA)+REB+AST+STL+BLK-TO http://www.nbaminer.com/golden-ages-of-basketball-players/
  15. ESPN Insider’s projection of the on-court success expected for each team in the 2017-18, 2018-19, and 2019-20 seasons.
  16. Composite Scoring: Each GM’s selected player’s efficiency endurance points are multiplied by draft picks values and added together, then divided by total games played.
    Player Efficiency Endurance Points: PTS+(FGM-FGA)+(FTM-FTA)+REB+AST+STL+BLK-TO multiplied by % of games played

Leave a Reply