Eli Pattison '22
Abstract
This project aims to measure the value a basketball player contributes to his team relative to his financial cost to them. This was quantified by dividing a player’s 2017-18 win shares by their salary for that year. Values of win shares per million dollars (WS/$Million) were calculated for all 276 NBA players that played more than 1000 minutes over the course of the 2017-2018 season. Rankings of overall win shares, which were similar to MVP voting patterns, were then compared with player rankings for WS/$Million. Graphs of win shares vs salary and WS/$Million vs salary were also analyzed.
Introduction
When discussing the “most valuable” player in the NBA, names are often debated without considering cost. The NBA operates with a soft salary cap, which was set at $99 million for the 2017-2018 season. This means that if a team’s payroll exceeds the cap they face certain penalties, such as the reduced ability to sign free agents and having to pay a luxury tax. Therefore, in order to build a competitive team, it is crucial that front offices are able to sign productive players to cheap contracts. For example, five time all-star Blake Griffin looks much less appealing from a team-building standpoint when his $29,727,900 salary for 2017-18 is considered. This project seeks to derive a more accurate measurement of true value by comparing a player’s contributions to his team’s success to his cost. It also attempts to find which types of players’ contracts are the most undervalued in today’s NBA.
Methods
Win shares are an advanced metric used in basketball analytics circles to attempt to quantify the number of wins a player adds to his team over the course of an 82 game regular season. More information about how win share totals are generated can be found here. Win shares were gathered from Basketball Reference and divided by 2017-18 total salaries according to Basketball Reference.
Results
Win Shares Leaders:
| Player | Age | Team | WS | Salary ($Million) | WS/$Million | |
| 1 | James Harden | 28 | Houston Rockets | 15.4 | 28.30 | 0.54 |
| 2 | Lebron James | 33 | Los Angeles Lakers | 14 | 33.29 | 0.42 |
| 3 | Karl-Anthony Towns | 22 | Minnesota Timberwolves | 14 | 6.22 | 2.25 |
| 4 | Anthony Davis | 24 | New Orleans Pelicans | 13.7 | 23.78 | 0.58 |
| 5 | Damian Lillard | 27 | Portland Trail Blazers | 12.6 | 26.15 | 0.48 |
| 6 | Giannis Antetokounmpo | 23 | Milwaukee Bucks | 11.9 | 22.47 | 0.53 |
| 7 | LaMarcus Aldridge | 32 | San Anto nio Supurs | 10.9 | 21.46 | 0.51 |
| 8 | Nikola Jokic | 22 | Denver Nuggets | 10.7 | 1.47 | 7.27 |
| 9 | Kevin Durant | 29 | Golden State Warriors | 10.4 | 25.00 | 0.42 |
| 10 | Andre Drummond | 24 | Detroit Pistons | 10.3 | 23.78 | 0.43 |
The rankings for overall win shares for the 2017-18 season are unsurprising. As would be expected, the entire All-NBA first team finished in the top ten. James Harden, the league MVP finished first in win shares, and 6 of the top 7 MVP vote getters were in the top ten in total win shares.
WS/$Million Leaders:
| Player | Age | Team | WS | Salary ($Million) | WS/$Million | |
| 1 | Nikola Jokic | 22 | Denver Nuggets | 10.7 | 1.47 | 7.27 |
| 2 | Clint Capela | 23 | Houston Rockets | 10.2 | 2.33 | 4.37 |
| 3 | Larry Nance | 25 | Los Angeles Lakers/Cleveland Cavaliers | 5.8 | 1.47 | 3.94 |
| 4 | Spencer Dinwiddie | 24 | Brooklyn Nets | 5.7 | 1.52 | 3.74 |
| 5 | Montrezl Harrell | 24 | Los Angeles Clippers | 5.5 | 1.47 | 3.74 |
| 6 | Josh Richardson | 24 | Miami Heat | 5.4 | 1.47 | 3.67 |
| 7 | Pascal Siakam | 23 | Toronto Rapters | 4.7 | 1.31 | 3.58 |
| 8 | Fred VanVleet | 23 | Toronto Raptors | 4.7 | 1.31 | 3.58 |
| 9 | Jerami Grant | 23 | Oklahoma City Thunder | 5.4 | 1.52 | 3.54 |
| 10 | Maxi Kleber | 26 | Dallas Mavericks | 2.8 | 0.82 | 3.43 |
The league leaders in win shares per million dollars are not similar to the league leaders in overall win shares and do not correlate with All-NBA teams, all-star rosters, or MVP voting. Nikola Jokic is the only player in the top ten in both categories, and league MVP James Harden ranks 122nd in win shares per million dollars. Essentially, the list is dominated by role players who made decent contributions on small contracts.
Analysis
This graph compares player win share totals and salaries. Predictably, the chart shows that player salary and wins generated are positively correlated.
This chart compares player win share totals with win shares per million dollars. The graph shows that generally the more expensive contracts are, the less valuable they become. The league’s most valuable contracts tend to be in the 1-5 million dollar range. Beyond that range the value of contracts seem to level out with no noticeable difference in WS/$Million as salary increases.
Conclusion
Win shares per million dollars give a good indication of which NBA players have the most and least valuable contracts. However, it is an imperfect statistic. It is a measure of regular season value and does not factor in postseason performance. Players like LeBron James and Stephen Curry are not payed more than $30 million a year just to put together good regular seasons. Rather, they are expected to win championships. Their actual value will be higher than WS/$Million indicates because of their postseason success.
The statistic also doesn’t account for the value of draft picks. For example, Ben Simmons’s 1.49 WS/$Million seems slightly less valuable when one considers that the 76ers used a valued commodity in the number one overall draft pick to acquire Simmons.
Additionally, the NBA is a superstar driven league. While this statistic would indicate that the best approach to team building is to forego players worthy of max contracts in favor of slightly less productive players on cheaper contracts, this is not the case. Superstars are needed to compete for championships, as the winner of the finals every year shows us. The best team building strategy might instead be to land a superstar on a max deal and surround them with cheap, productive players.
There are several intriguing future applications with this statistic. I’m interested in tracking overall money earned for retired players and comparing it with their total win shares to find the most undervalued and overvalued players over the courses of their careers. Basketball Reference also breaks down win shares into offensive and defensive win shares. I’m interested in tracking offensive win shares versus salary. I would expect these metrics to be more positively correlated than total win shares and salary. Teams like to pay players who create points and often undervalue those who shine on the defensive end.