The trade of Jaylen Brown from the Boston Celtics to the Philadelphia 76ers created a mix of shock and intense debate across the NBA. Fans felt outraged, but front offices focused on a heated debate over Brown’s true value on the court. This debate centered on how modern data analysis of players in a league flooded with information works. Brown is a five-time All-Star and a former Finals MVP who finished sixth in regular-season MVP voting in 2026 after averaging a career-best 28.7 points per game, according to ESPN.
Despite these impressive traditional numbers, advanced statistics were more skeptical of his stardom. In seven of Brown’s 10 Boston seasons, the Celtics achieved a superior net rating when he sat rather than played, covering each of the final four. Cleaning the Glass, which removes garbage time, is the source for all on/off figures in this article. With Brown set to face his former team for the first time Tuesday at 8 p.m. ET on NBC/Peacock, it’s the right time to examine the divide between his reputation and the data.
The Basics of On and Off Data
A modern NBA front office uses vast amounts of data to evaluate every player in the league. Teams track points, rebounds, assists, shot quality, pick-and-roll frequency, and isolation efficiency. While individual data points are nearly endless, the objective remains the same. A front office should identify players who help the team outscore opponents rather than players who just collect strong statistics in a vacuum.
Players boasting top individual stats frequently deliver the greatest winning impact, though this is not always the case. These mismatches can lead to different, sometimes counterintuitive, evaluations of players. When examining Orlando’s frontcourt pair of highly drafted players, former No. 1 pick Paolo Banchero records higher career averages per game in points, rebounds, and assists than his teammate Franz Wagner.
In every season of his career, Banchero has posted a negative or neutral on/off split, which indicates that the Magic always perform better when he sits. Conversely, Wagner has posted a strongly positive on/off split in every single season. While the Magic’s defensive rating fell by 5.4 points when Wagner played, it rose by 3.8 points with Banchero on the floor during his tenure, meaning defense accounts for the majority of this gap.
Hidden Value on Defense
Steals and blocks are scarce for Wagner, but on/off data suggests he clearly executes correctly on that end even if box scores miss it. Inquire with league insiders about a player profile more valuable than basic stats imply, and they frequently cite defensive standouts. The past two champions relied heavily on Alex Caruso and OG Anunoby, both of whom were part of that group.
Shane Battier, famously labeled the “No-Stats All-Star” in a 2009 piece by Michael Lewis, set the path for these players. Exemplary on/off numbers throughout their careers were no surprise to those role players. Public perception of a defensive impact reputation often requires a very long time to stabilize.
An Eastern Conference team analyst stated that 10,000 possessions, roughly two seasons of starter data, represent the minimum sample size before on/off data becomes useful. This caution runs counter to the modern media cycle’s incessant demand for information and immediate answers. The public often wants to see results right away without waiting for a large sample to form.
The Problem with Small Samples
Because of the pressure to find answers quickly, many people jump to conclusions based on limited data. According to the East analyst, the public should not access on/off data this early in the season. It should be banned. I think it’s a disservice to the discourse. This statement highlights the frustration many feel when raw numbers are used to judge players too soon.
Transforming on/off data from a raw metric into a useful tool requires patience and context. A Western Conference team executive observed that the issue involves inherent biases of all sorts, including the sample size problem. Whatever narrative you wish to construct, you can find confirmation, which is why an executive for a Western Conference team called it astrology for NBA professionals.
This bias allows people to pick the data that supports their pre-existing opinion about a player. For this reason, professionals often look for more sophisticated methods that adjust for these variables. The goal is to isolate the player’s actual impact from the noise of small sample sizes and team context.
Advanced Metrics That Fix the Data
Taylor Snarr, a former model builder, noted that some metrics understand context and isolate actual player impact much better. This describes stats like RAPM, which incorporate various adjustments to refine raw data. This process helps teams make better decisions by looking deeper than simple on/off splits.
Two years ago, Snarr noted that the key factor is how frequently and consistently players play together. This consistency makes the data more reliable and accurate over time. Kostya Medvedovsky, a model builder, stated that RAPM could reduce the convincing timeline for on/off from 10 years to maybe two years.
By isolating impact, these adjusted metrics enable much quicker, more accurate player analysis. This shift in methodology has changed how teams evaluate talent and build their rosters. The ability to get a clearer picture faster helps front offices make trades and sign players with more confidence.
How Data Influenced the Brown Trade
The debate over how heavily you prefer the box score statistical components versus the on/off components remains a tricky question for model builders. However, the process can lean more toward art than science, making the trickier question how to weigh these different types of information. The Western executive remarked that legitimate debate and variance amongst the nerds emerge there, in his view.
For a player like Jaylen Brown, this variance in interpretation played a major role in his departure from Boston. The Celtics’ internal view of his value likely differed from the public perception of his stardom. Medvedovsky said the Jaylen Brown case is, in some ways, the purest proof of this. Lessons from the adjusted on/off metrics appear to have permeated most front-office thinking.
The trade was not just about scoring numbers but about how the team performed with him on the floor. The data suggested that the Celtics could be more effective with different lineup combinations. This analytical approach led to a blockbuster move that surprised many fans but made sense to those who trusted the adjusted stats.
Changing Minds in the League
According to the Western Conference executive, acceptance among nontechnical people is what has changed the most. This shift has altered the conversation around player value across the entire league.
It helps that many of the most successful teams in recent years have relied on these advanced metrics to build their rosters. The public is starting to see that traditional stats do not tell the whole story. This acceptance has made it easier for teams to make controversial moves based on data that the average fan might not fully understand.
The Jaylen Brown trade is a clear example of how this new way of thinking has taken hold.
