If you mention the 2016 offseason to a Los Angeles Lakers fan, don’t be surprised to be met with a deep sigh followed with what sounds like “dang” being whispered under their breath. In reality, the word they uttered distressingly was actually a name: Deng. That dreadful summer, the Lakers not only dealt with their adored legend Kobe Bryant stepping away from the game, but they also watched as the front office signed Timofey Mozgov and Luol Deng to long-term deals that paid $16 million and $18 million per year, respectively.

Since then, the Lake Show has given up 2015’s No. 2 overall pick D’Angelo Russell to dump Mozgov to the Brooklyn Nets. And as the trade deadline approaches this year, Deng—who has only played in one game this season—appears to be immovable without potentially giving up another young talent.

A multitude of managerial decisions befuddle us every year, and whether trades or free-agent signings, they are often so perplexing because we lack a dependable understanding of how front offices value their athletes. Sure, the sports analytics boom has yielded a plethora of excellent tools and metrics to evaluate production, but translating our opinions from those numbers to the viewpoint of a general manager trying to offer a competitive contract to a free agent in the available market is exceedingly difficult.

Placing a dollar value on players has become even more arduous in recent years, due to the league’s sky-rocketing salary cap.

For example, Andre Iguodala signed a contract worth $12 million per year in 2013. Just three years later, Solomon Hill took a similar $12 million annual deal for the New Orleans Pelicans. Though the pair of wings were offered the same amount of money nominally, how teams actually rated these players at the time of their signings was far from similar.

In 2013, the salary cap was just $58,679,000, making Igoudala’s contract worth 20.45 percent of his team’s cap. Meanwhile, with the cap up to $94,143,000 in 2016 when Hill agreed to his terms, his indenture only dented the Pelicans’ cap by 12.75 percent. The spending limit has enlarged to such a degree that by pure dollar amount, nine of the 10 largest contracts in NBA history have been signed within the past two years.

Introducing Predicted Market Value (PMV)

Using the concept that a team considers the value of a player as the percentage of salary cap space he encompasses, I performed a multiple regression to predict that figure, which we will call Predicted Market Value (PMV). Referencing stats from a player’s contract season, as well as their age, games played and playing time in that year as independent variables, the model yielded an R-squared value of 0.75 and identified these seven statistics as significant predictors of PMV: player efficiency rating (PER), minutes per game, games played, win shares (WS), defensive win shares (DWS), steal percentage (STL%) and true shooting percentage (TS%).

At a glance, most of these variables make perfect sense as important factors for predicting a player’s value as a percentage of the cap.

PER takes into account the players’ box-score stats, and DWS provides balance for that offense-heavy metric. Additionally, WS takes into account an individual’s contribution to team success, and TS% incorporates offensive efficiency. Meanwhile, games played and minutes per game consider health and playing time. The one component I was stunned to discover as having an impact is STL%, which is defined as an estimate of the percentage of opponent possessions that end with a steal by a player while they are on the floor.

In his 2014 article on FiveThirtyEight titled “The Hidden Value of the NBA Steal,” Benjamin Morris breaks down how irreplaceable steals are and explains how they are one of the most informative statistics in basketball. The entire piece is absolutely worth reading, but one particular segment that illustrates the importance of takeaways stands out. It comes when Morris used box-score stats to predict how much teams would suffer when someone couldn’t play. His model revealed that “A marginal steal is weighted nine times more heavily when predicting a player’s impact than a marginal point.”

In other words, a steal is worth as much as nine points.

Estimated Player Profit (EPP) and other PMV insights

To more easily consume the data, here is a visualization comparing each player’s PMV, based on their current season statistics, to the actual cap hit percentage their team incurs:

After adding up each team’s individual player scores, these are the top five in total PMV this season:

1. Golden State Warriors: 157.84%
2. Houston Rockets: 151.82%
3. Philadelphia Sixers: 149.15%
4. San Antonio Spurs: 146.15%
5. Miami Heat: 141.15%

Of those teams, Golden State and Houston were a fairly predictable top two, boasting several superstar athletes. But Miami sneaking into the top five is eye-opening. Despite not having any single contributor with a PMV of at least 21 percent, Heat head coach Erik Spoelstra has inspired a complete effort from his squad with a league-high nine players boasting a PMV greater than 10.

The primary goal of a business is to make a profit, and by taking the difference between PMV and the player’s real cap hit percentage, we can come up with the estimated player profit (EPP). According to EPP, here are the league’s 10 most profitable contracts:

1. Karl-Anthony Towns: +25.58%
2. Nikola Jokic: +22.49%
3. Clint Capela: +21.97%
4. Tyreke Evans: +18.31%
5. Joel Embiid: +18.31%
6. Gary Harris: +17.39%
7. Kristaps Porzingis: 17.30%
8. Lou Williams: 17.17%
9. Devin Booker: 15.81%
10. Spencer Dinwiddie: 15.72%

As expected, the new class of elite big men is front and center on this list, producing at levels well above the pay grades established by their rookie contracts. Only two players on this list, Tyreke Evans and Lou Williams, are above the age of 25. In fact, they’re the only two guys that old among the top 20. The major shocker on this list is Dinwiddie, who came into this season still fairly unknown, but made the most of an opportunity to start for an injured D’Angelo Russell.

Evans, who has now been in the league for eight years, is having his best season since he won Rookie of the Year back in 2010. As somewhat of an enigma throughout his career, one marred by inconsistency and injuries, the multi-positional talent has found himself in elite company as one of the league’s most balanced players. Evans is one of eight players averaging at least 19 points, five rebounds and five assists per game, joining LeBron James, Stephen Curry, Russell Westbrook, Kevin Durant,  Chris Paul, DeMarcus Cousins and Blake Griffin.

Williams has been known as a walking bucket throughout his NBA life, but he has elevated to new heights in his 13th year and is being considered by many an All-Star snub. The former Sixth Man of the Year is averaging over 20 points per game for the first time in his career (23.5) and doing it more efficiently than ever. Among players with at least 1.5 isolation possessions per game, Williams ranks first in efficiency, scoring 1.29 points per possession.

That said, the contracts that are the least profitable actually reveal the most, highlighting potential flaws in the way front offices evaluate talent. Excluding players who have missed significant time due to injury, these are the least profitable contracts according to EPP.

1. Luol Deng: -18.71%
2. Joakim Noah: -16.61%
3. Ian Mahinmi: -15.62%
4. Chandler Parsons: -15.15%
5. Darrell Arthur: -14.78%
6. Timofey Mozgov: -13.71%
7. Brook Lopez: -13.01%
8. Cole Aldrich: -11.44%
9. Omer Asik: -10.77%
10. Corey Brewer: -10.13%

Opposite to the most profitable contracts, this group consists primarily of guys who are past their primes, with nobody on the list under the age of 29.

The model justifies the payment of some of these players at the time of their acquisition, which leads me to believe that maybe NBA front offices are not putting as much weight on a player’s age before signing them as they should.

For example, Deng, who agreed to his four-year deal at the age of 30 for 19.12 percent of the cap, had a PMV of 18.88 percent. Obviously, hindsight is 20/20, and his numbers plummeted almost immediately after agreeing to the terms. But his offer was near his estimated market rate at the time.

Another key flaw in how some general managers judge players emerges just based on the accuracy of the model itself. GMs around the league are falling victim to recency bias, placing too much importance on an individual’s contract year while disregarding their career up to that point. A prime example of this is Mahinmi, who in 2016 signed a four-year, $64 million contract.

At the time of his acquisition by the Washington Wizards, the now overpaid big man had a respectable PMV of 15.72 percent, which is just slightly lower than the amount of cap space the Wizards allocated for him (17 percent). However, they completely failed to acknowledge that although Mahinmi had a solid year and started 71 games on the Indiana Pacers during his contract season,  the French native started just 21 games combined in his seven seasons prior, averaging 4.3 points and 3.8 rebounds per game over his career. And, of course, some decisions are still complete headscratchers,  including when the Lakers offered Mozgov 17 percent of their cap even though his market value according to PMV reveals he was worth just over 7 percent.

We may disagree with many of the choices NBA front offices make with regard to salaries, but with PMV, we can now see where they’re coming from. To see the full list of PMV and EPP scores across the league, along with breakdowns by team, continue here.

Follow Faizan on Twitter @Faizan_Has.

Follow NBA Math on Twitter @NBA_Math, Facebook, Instagram and YouTube.

Unless otherwise indicated, all stats are from NBA Math, Basketball Reference or NBA.com and are accurate heading into games on January 30.