Calculating NBA Player Net Value

Refreshing a favorite front-office metric

Nov 14, 2024

NBA discussion often focuses on an asset (player performance) while forgoing the liability (player contract). From a front-office perspective, considering the liability is vital. In this analysis, I propose Player Net Value, a metric that considers both the current production and the cost of an NBA player.

Player Net Value

To assess the net value of a player, consider (i) how to calculate a player’s value and (ii) what the cost of that player is. For both the value and the cost, I use salary, as a percentage of the salary cap. A player’s actual salary is treated as their ‘cost’, and their predicted salary using current stats as their ‘value’.

To calculate player value, using current stats, I build upon Predicted Market Value1, a metric introduced in 2018. This article aimed at predicting free-agent salaries using 7 key advanced metrics as well as age, position, and experience controls.

I build upon this approach by adding additional controls and I also use Multivariate Adaptive Regression Splines (MARS). This allows for a model that explains roughly 10% more of the variance in salary (from 60% to 70% for my sample).

Another contribution is adjusting season-end metrics to be rates, rather than their season-end raw values. This allows for in-season evaluation. Injuries or strength of schedule may skew these values early in the season, but even with a 10-game sample, the results pass the eye test.

Player Net Value is calculated as a player’s predicted market value minus their current salary:

Key Examples (at time of writing)

Shai Gilgeous-Alexander (OKC):

Shai Gilgeous-Alexander is a great example of someone who would be worth more than the max contract if allowed. At the time of writing, Gilgeous-Alexander, Tatum, and Jokic all would yield more than 40% of the cap if they could. These players all have positive net value, even though they’re already paid handsomely.

Evan Mobley (CLE):

Mobley represents arguably the best deal in basketball: rookie contracts. Players on rookie contracts are consistently high in net value, especially after they’ve found their way into teams’ rotations. The Oklahoma City Thunder’s current roster is another great example of players out-performing their relatively small rookie deals.

Lonnie Walker IV (no team, not in data):

The Celtics cut Walker before the current season. Is he better than at least some of the league? Undoubtedly. Unfortunately for him though, the salary he’d require as an NBA veteran is more than the value he is projected to add. My career trends analysis also supports this, where we see that younger players don’t have to be as good as their older counterparts to stay in the league.

Aggregate Team Ratings

One benefit of player-value statistics is that they can often be aggregated to the team level. You can do that here, with a couple of caveats2.

Large, positive values in the ‘difference’ column of the above table correlate highly with flexibility in assets. Players are worth more than their current salary, and in trades, this surplus in value can be balanced with draft picks. Teams underperforming their salary cap likely have the opposite situation: to get rid of bad contracts, they’ll have to pair a draft pick with an outgoing player.

I also think these value percentages are the first step toward creating franchise power ratings. There are other factors to consider too, including draft picks, talent development, coaching quality, and ownership support. But it’s a starting point!

Key Takeaways & Next Steps

Although net value isn’t a new idea by any means, this analysis is a starting point from the basketball operations perspective. Understanding optimal salary allocation, and where efficiencies can be found, is a key aspect of building a championship roster. The best teams take a patient approach while rewarding deserving players.

From the “entertainment” perspective, net value remains an interesting concept. One subsequent study could quantify the premium that small-market teams have to pay to land free agents. Another fascinating approach would be to use all-star votes, Twitter/X followers, or some other popularity metric to see if player brand recognition increases a player’s expected salary.

There are countless benefits of this approach and even more avenues to explore. This metric is just the tip of the iceberg, but one that is both effective in communicating a message and simple to understand.

Thanks for giving this article a read. For any methodology questions or to continue the conversation, reach out at vaughnhajra@gmail.com or on X/Twitter @vaughnhajra.

This article proposes “Estimated Player Profit” which I’m calling “Player Net Value”

The biggest caveat worth mentioning is if a player hasn’t appeared in a game yet (for example, Kawhi Leonard), they aren’t included in this. That’s why the Clippers have a “Team Total” under 80%. On the flip side, though, you can get a good idea of which teams are outperforming their current salary.

Another thing worth considering is that because older players require more salary, an equivalent old player would garner a higher predicted contract than their younger counterpart. Additionally, there is no “max” contract in these projections, so teams with superstars may be inflated by one player’s contributions.

All of this said these shortcomings can be seen as positives when assessing long-term franchise ratings