When the Dallas Mavericks traded Luka Doncic, the deal triggered shock and outrage around the NBA. A year later, the Boston Celtics' trade of Jaylen Brown to the Philadelphia 76ers produced a more complicated reaction: anger from Celtics fans, certainly, but also a fierce debate about Brown's true value and about how players should be evaluated in the league's data-driven era. Brown is set to play against his old team for the first time on Tuesday night (8 p.m. ET, NBC/Peacock), opening a new chapter in what has become the NBA's best current rivalry.
On paper, Brown's credentials are elite. He is a five-time All-Star, a former Finals MVP and finished sixth in regular-season MVP voting in 2026 after averaging a career-best 28.7 points per game. Yet advanced statistics have long been more skeptical. According to Cleaning the Glass, which filters out garbage time, the Celtics posted a better net rating with Brown off the court than on it in seven of his 10 seasons in Boston, including each of his final four.
On-off data comes with serious caveats. One Eastern Conference analyst said a player needs roughly 10,000 possessions, about two full seasons for a starter, before the numbers become meaningful. Houston's Amen Thompson illustrates the danger of small samples: through nine games this season, the Rockets have been 47.3 points per 100 possessions worse with him on the floor, the worst mark of any rotation player in the league, yet evaluators have not lowered their opinion of him. "The public should not have access to on-off this early in the season," the East analyst said. "It should be banned. I think it's a disservice to the discourse."
Raw on-off figures are also vulnerable to all kinds of built-in biases. "I call it astrology for NBA professionals," a Western Conference executive said, "because whatever story you want to tell, you can go and find confirmation." To fix that, analysts rely on regularized adjusted plus-minus, or RAPM, which controls for context such as the other nine players on the court, home-court advantage and opponent shooting luck. Taylor Snarr, a former Jazz analyst who created the estimated plus-minus metric, said the process "does a much better job of understanding context and isolating the player's actual impact." The example often cited is Christian Braun, who two years ago posted the third-best on-off swing among rotation players, but shared the floor almost exclusively with Nikola Jokic, who finished first.
Models then blend in box score statistics to credit the right players, a step that shortens the evaluation window dramatically. "Instead of taking 10 years for on-off to be convincing, you can reduce it down to maybe two years with RAPM," said Kostya Medvedovsky, creator of the DARKO projection system. The harder question is how heavily to weight traditional stats against on-off data, and that is where Brown sits at the center of a genuine disagreement. Models favoring the box score see an elite scorer; models favoring on-off impact see a very good player, but not an MVP-caliber centerpiece, partly because of his modest efficiency and his turnover-to-assist ratio for a high-usage star.
Executives point to the Celtics and Thunder as the teams most committed to that approach, sometimes called RAPM maxxing. Players such as Derrick White, Isaiah Hartenstein and Alex Caruso have always graded well in adjusted on-off metrics; Caruso's teams have been better with him on the floor in all nine of his NBA seasons, and both DARKO and EPM rate him as a top-30 player per possession even though he has never averaged more than 10.1 points per game. The method is not foolproof, as Oklahoma City's misses on Patrick Patterson and Gordon Hayward showed, but the Thunder kept trusting their models and won a title a year after adding Hartenstein and Caruso. Paul George, meanwhile, has produced positive on-off splits in every season of his career, which helps explain why Boston preferred him, along with draft picks, despite his age, injury history and salary.
Many around the league were still surprised by the modest return for an MVP candidate in his prime. But Boston's willingness to move Brown was hardly a secret, and no rival team topped Philadelphia's offer. "In some ways, the Jaylen Brown case is the purest proof of this," Medvedovsky said. "It seemed like: You're trading a Finals MVP, so you're going to get a ton of value for him. But then we saw you just didn't."
Now Brown is trying to prove the market wrong in Philadelphia. He is averaging 21.6 points per game while often serving as the third or fourth option for the 76ers, who lead the East at 8-2. The early splits are striking: Philadelphia has outscored opponents by 4.3 points per 100 possessions with Brown on the floor and sits at minus-1.9 without him, a 6.2-point swing that would rank as the second-best single-season mark of his career. The sample, though, is only 678 possessions, less than 10% of the way to a reliable conclusion, which is precisely the cautionary lesson his own trade was built on.
