BY HENRY ABBOTT and STEVE ILARDI

Having just led the Warriors into the rarefied air of a fourth NBA title, Stephen Curry’s value to the Warriors is nearly incalculable. 

By one respected measure, Estimated Wins (eWins) from Dunks and Threes, Curry accounted for 15.28 of the Warriors’ 53 wins. No other Warrior approached half of that total. 

Whatever the cost, Steph is the proverbial “bargain at twice the price.” But he’s also the highest-paid player in the history of the NBA. At $45.8 million in 2021-2022, the cost is pretty steep.

And … we’ll get into the weeds below … NBA regular-season wins last year cost about $3.35 million each. If you do the math, the Warriors came out a little ahead. 

Curry also stands out among All-Star guards. The Grizzlies’ budding superstar, Ja Morant, clocked in with a respectable 8.29 eWins, about half of Curry’s total. So it makes sense that Morant should make less money. 

But Morant makes a lot less money—like a fifth as much. Near the end of his rookie deal, Morant earned just $9.6 million last season. 

Morant’s salary is a fraction of what he deserves, which is bad for him and, in the harsh reality of the competitive NBA, magical for his team. To contend for the NBA title, you typically need to win north of 50 games. The Warriors’ math is that if Curry generates 15 wins, then there’s some tap dancing—and cap dancing—to figure out how to afford the other 35-40. After paying Steph, the Warriors had only $67 million left to spend under the salary cap. That would only buy 20 more wins at league-average prices, and who wants to win a measly 35 games? So to leap back into title contention, the Warriors shattered the salary cap and set all-time spending records.

Over the broad sweep of NBA history, we see the occasional Joe Lacob, James Dolan, Mikhail Prokhorov, or Paul Allen—billionaires determined to build dynasties with gold bricks. It usually doesn’t work. Even when it does, the league stacks on such punitive luxury-tax bills that everyone eventually loses their appetite for overspending. 

Sooner or later, every team sees great value in great value. 

Which makes the up-and-coming Grizzlies, and not the champion Warriors, a more sensible team-building model. Memphis won five more games than Steph’s team, and they did it with the league’s second-lowest payroll. They paid just two-thirds of the Warriors’ salary outlay, and that’s before taxes. Only the flagrantly tanking Thunder paid less.

Morant is one of the key reasons why the Grizzlies are now the NBA’s reigning Moneyball team. Even after paying their franchise guy, the Grizzlies have room under the cap to spend over $100 million more on Ja’s teammates—and that’s enough for some pretty amazing teammates. 

But who was a better value for their team last year: Ja or Steph? Is it even possible to do an apples-to-apples comparison? We believe it is. That’s why we’ve invented an intuitive new stat to assess the value of every NBA contract, combining highly respected, battle-tested on-court analytics with commonsense financial analysis: Bonus Wins. 

It’s our new metric of Moneyball in the NBA, and it has a hell of a story to tell … about a world beneath the surface of the NBA, where victories are created.

Trae Young led the NBA by creating 11.74 Bonus Wins last season. At a cost of just $8.36 million, he was arguably the NBA’s most underpaid player. MVP Nikola Jokic also did something incredible, creating almost as many Bonus Wins despite making nearly four times Trae’s salary. 

There’s a lot to notice from the first look at Bonus Wins:

• Bonus Wins are rare. In the whole NBA, only Jokic and Trae delivered more than 10 Bonus Wins. Only 16 players, including Joel Embiid, Jayson Tatum, Luka Doncic, and Darius Garland, delivered as many as six.

• It’s a lot easier for low-salaried players to create Bonus Wins. Most of the league leaders in Bonus Wins made under $10 million. 

• While the list of winning players includes guys of various ages, players who deliver lots of Bonus Wins are overwhelmingly young. Trae is 23. The average age of the league leaders in Bonus Wins is 26.

Steve Ilardi, TrueHoop’s partner in creating Bonus Wins, has spent nearly two decades as part of the league’s analytics movement, designing, tweaking, and refining tools to measure player performance for NBA teams and ESPN.com. 

HENRY: Steve, the basic assumption here is that we can put a number on the impact that a player has on his team? Do we feel good about that?

STEVE: Yes. It’s at least a great start. The core methodology has been around for nearly 20 years. It’s all built around the simple insight that box-score counting stats like points and rebounds—and especially the eye test—are often misleading. We can get a better gauge of player impact if we focus instead on how some players consistently make their teams better (in terms of plus-minus scoring margin), while others simply do not. In fact, we can take a team like the Lakers and use advanced math to sort out how much of their performance in a given season is due to the play of LeBron; how much to the five opposing players; and how much to LeBron’s teammates. With a few caveats that we can get into, over the course of an entire NBA season we can put a reliable number—a solid estimate—on the true plus-minus impact of each player. 

So that’s what all the advanced plus-minus metrics are about. Jerry Engelmann and I worked together to popularize one for ESPN that many fans probably recognize: Real Plus-Minus (RPM). But analytics whizzes like Taylor Snarr of Dunks and Threes have now used fancy player-tracking data to make plus-minus metrics even better. I’m a huge fan of his Estimated Plus-Minus (EPM), which has proved in one recent analysis to be the best single-number metric ever developed. 

Through the years, when similar plus-minus statistics have returned puzzling results, it has typically been with players in limited roles. Matt Bonner is an interesting case of a guy who once put up impressive plus-minus numbers in short minutes. He struggled to guard many of the league’s centers, but against a limited set of opponents, he was the Spurs’ secret weapon because he could make corner 3s. In his case, his plus-minus impact likely would have plummeted had he played twice as much and had his shortcomings exposed.

Remember that when you see Gary Payton II’s jaw-dropping EPM impact for the Warriors. He saw limited burn in the regular season, and then played only 16 minutes a game in the Finals. Although Payton’s EPM was considerably higher than Klay’s, it matters—it matters a lot—that Thompson carried a much bigger workload, especially in the postseason. The best ability, as they say, is availability. 

Remember: EPM measures efficiency, how you help your team turn possessions into points. And sometimes efficiency is the thing you most want to know. But if we’re talking about salaries, that’s when we want to gauge a player’s full contribution to his team’s success or failure, so we need to factor in playing time. That’s where Estimated Wins (eWins) comes in. It’s simply the player’s total EPM impact across all possessions played. So eWins rewards the players who are the most available, on the court, and succeeding against all kinds of opponents. It truly reflects what teams pay for.

Of course, even sophisticated measures like EPM and eWins can obscure the potential of players who are simply in the wrong role. Early in his career, Kevin Durant played a lot of minutes at shooting guard, with a predictably horrid plus-minus impact. It worked out so badly that one of the pioneers of plus-minus stats famously said (and later retracted) that he wouldn’t even want Durant on his team! 

Plus-minus metrics like EPM and eWins can send some wonky signals here and there, but overall they’ve proven highly useful in telling us about each player’s bottom-line impact every season. And we know the best predictor of future behavior is past behavior. (As my Psych students all hear about 100 times before they graduate!.)

So how do we get from Estimated Plus-Minus to Bonus Wins?

EPM measures the player’s on-court plus-minus impact per 100 possessions. So we multiply each player’s EPM by his total season possessions to get an estimate of his contribution to the team’s aggregate point differential; we can use that number, in turn, to estimate the number of wins he helped generate.

For example:

• The Celtics outscored opponents by an estimated 5.84 points per 100 possessions with Jayson Tatum on the court (adjusting for the effects of his generally superb on-court teammates and all opponents). 

• He played about 70 percent of Boston’s possessions, so Tatum single-handedly added about +4.0 (5.84 x 0.70) to Boston’s team point differential on the season.

• We can now use a formula similar to “Pythagorean wins” to calculate that Tatum accounted for about 16 of the Celtics’ 51 wins. 

Once we have a sense of how many wins a player has generated, we can factor in the money component:

• The entire NBA had a collective payroll last year of $3,905,190,611 across all 30 teams.

• Teams spent that money to win as many of the league’s 1,230 regular season games as possible. 

• So it’s just simple math: We can divide the total NBA payroll by 1,230 available wins to find that last season each win cost about $3.35 million. 

Players who can deliver wins at lower costs are generating Bonus Wins. Tatum contributed his 16 wins on a salary of $28.1 million. That comes to only $1.76 million per win—a bargain price that reflects a 48-percent discount on the league average! Think of it like this: The Celts only paid Tatum to deliver about eight wins ($28.1 million salary with wins costing an average of $3.35 million per win). Instead, he gave them a whopping 16. And Tatum’s high-value contract is a key reason the Celtics were able to create such a killer roster, with room under the salary cap to pay many other excellent players. 

Likewise, the Grizzlies had 56 wins with a team payroll of just $98 million last year. They spent a paltry $1.75 million per win, which means they averaged Tatum-esque value across their entire roster. For the production they provided, the Grizzlies’ players were wildly underpaid. It’s no wonder Zach Kleiman was named NBA Executive of the Year.

That’s how you crush the GM game. 

Players who deliver exceptional value are essential to team success. The ultimate goal of a general manager is constructing an entire roster of players who produce far more than they’re being paid for. In other words, a GM aims to maximize Bonus Wins.

How did the Grizzlies stand out? They had three elite players on cheap rookie contracts (Desmond Bane, Ja Morant, and Jaren Jackson Jr.) who created around 10 wins each. Bane, who made just over $2 million, accounted for an estimated 10.18 wins and delivered 9.57 Bonus Wins, third in the entire league. Such value contracts are the key to building an NBA title contender; the Bonus Wins metric is a Rosetta Stone that helps us solve the otherwise impenetrable mystery of how to accurately identify them.

We are almost finished building a database of EPM, eWins, Bonus Wins, and salary for every player in the league over the last nine years. Before that’s totally complete, we already have a steady stream of interesting findings coming from our analyses as we examine the league through the powerful lens of this new metric. 

Some of what we have been discussing:

• Spotlighting the sneaky-efficient contributors

• The GMs who excel at finding value

• The economic anatomy of a champion 

• The characteristics of a bad deal

• The Bonus Wins career trajectory of LeBron, Giannis, Kawhi, Steph, Jokic, and all the NBA’s biggest names

There’s a ton more to come. We welcome your suggestions. 

But after only a week of exploring all this information, our most obvious takeaway (and we hinted at it earlier): Superstar salaries are out of hand.

The early days of NBA analytics revealed two groups that stood out as massive bargains: superstars and players on rookie contracts. During the 2011 lockout, the highest-paid players in the world were making a bit over $20 million (capped by league rules), even as analytics gurus estimated peak LeBron James could be worth three or four times that much. Teams’ search for surplus value became largely a search for superstars. 

At his peak, James Harden contributed 25 wins for the 2018-19 Rockets. That’s an absurd number, way up there with the apex seasons of LeBron and Steph and the biggest contributors in league history. A player like peak Harden could make $75 million—far more than anyone has ever been paid—and still deliver solid value! (This fact may help explain why Daryl Morey, the Pied Piper of NBA analytics and Steve’s former boss on a consulting gig, has been so persistently smitten with the Beard.)

That same season, when the Lakers signed LeBron, it seemed like there would be near-infinite ways to construct a winning roster around him. After all, he had won a title in Cleveland with the likes of Tristan Thompson and J.R. Smith. 

But at the same time, the league’s top players were getting more expensive. LeBron’s move came a year after Kevin Durant abandoned small-market Oklahoma City for super-rich Golden State. The stunned league reacted by cooking up a new way for teams to keep their homegrown superstars: Pay them more. The lucrative supermax contract, available only to a short list of elite players who’d been with the same team since their rookie deal, peaks at a staggering 35 percent of the team’s salary cap in the first year (with huge annual raises). Before long, many of the league’s biggest names—Stephen Curry, Damian Lillard, John Wall, Russell Westbrook, James Harden, Giannis Antetokounmpo, Joel Embiid, Luka Dončić, Rudy Gobert—qualified. 

Superstar salaries skyrocketed.

We will explore this salary explosion in the weeks ahead. But here’s a quick spoiler: Only the two or three players at the very top of the league—we’re talking MVP candidates—represent world-class bargains on annual salaries in the $30-million range. Those are typically superstars in their prime years of productivity, usually in their mid-to-late 20s.

But the league’s highest salaries (those priced at $40 million or more) always go to players in their 30s—aging superstars with slipping production per minute, to go with reduced minutes and missed games. In terms of return on investment, most of these huge-salaried players aren’t Steph, some of them are John Wall. Simply put: The players making the biggest salaries are almost all on the downslopes of their careers. 

As a result, most of the league’s brightest stars—its highest-paid marquee veterans—can’t match the league leaders in Bonus Wins. LeBron, adding 12.78 wins last season, was in the league’s top ten, and by far the Lakers’ best player. Yet, with a massive $41.2 million salary, his wins cost very close to the league average. Clearly, the Lakers had a hard time crafting a winning roster around him.

Part of the reason for the Lakers’ demise is their bet on another supermax player: Russell Westbrook. He played perfectly well for a player his age; his 3.4 eWins made him one of the Lakers’ more productive players. But Russ made $44.2 million. Remember when we said a win cost $3.35 million, on average? You’d count on a star player with Westbrook’s salary to deliver over 13 wins—not 3.4! 

This is probably a good time to mention that Bonus Wins can be negative. Westbrook’s Bonus Wins total clocked in at -9.8 last year. If the Lakers had instead spent his $44 million salary on average-value NBA players, the team would have won ten more games. 

There’s a lesson in team building here: Role players, once seen as worse value than superstars, deserve a fresh look. The players the Lakers ditched to make room for Westbrook—Kyle Kuzma, Alex Caruso, Kentavius Caldwell-Pope, and Montrezl Harrell—collectively earned $44.4 million last year and accounted for 14 more eWins than Westbrook.

And another surprise finding: The Lakers’ team leader in Bonus Wins last season was one of the least famous players on the roster: Malik Monk. On the court, he offered roughly the same productivity as Westbrook. But by contributing more than four eWins for under $2 million, Monk was a legit difference-maker, generating a tidy 3.86 Bonus Wins for the Lakers. He’s a model for the kind of player who can help stars like LeBron win titles, and the reason we created Bonus Wins.

Thank you for reading TrueHoop!