Advanced Metrics Hate Jaylen Brown, Why?
Jaylen Brown made the All-NBA Second team, but his metrics tell a different story. What do they see that we don’t? And what do we see that they can’t?
By some measures, Jaylen Brown is one of the league’s best players. He was voted an All-Star this season, finishing third in Eastern Conference backcourt voting among fans and players and second in the media vote, and just earned a Second Team All-NBA selection. According to the fans, players, and media, Brown is one of the best guards in the NBA, and he has backed it up.
His 26.6 points per game was the ninth-best figure in the league, and he even garnered an All-Defensive team vote. Top-ten scorers who are strong defenders don’t grow on trees and are usually one of the best players in the association. While most reasonable people wouldn’t place Brown on their top-10 players list, they also wouldn’t place him where many all-in-one advanced metrics place him.
Basketball Reference’s Box Plus/Minus (BPM) has Brown pegged as the 66th-best player, his Win Shares per 48 minutes (WS/48) is 104th, FiveThirtyEigth’s Overall Raptor has him 111th, and Dunks and Threes’ Estimated Plus Minus (EPM) views him as the 46th best player this season. While the individual rankings are somewhat meaningless, the overall message is clear– Jaylen Brown is not a top-30 player and, at best, a fringe top-50 player.
People’s eyes and a rudimentary understanding of box score-based statistics make Brown out to be one of the best players in the NBA, which is not unexpected. Brown finished the season averaging 26.6 points, 6.9 rebounds, and 3.5 assists per game on an effective field goal percentage (eFG%) of 55% (league average was 54.5%). Only eight players finished the season averaging 25 points, 6.0 rebounds, and 3.0 assists on an eFG% of 54.5% or greater, and they were Joel Embiid (9.2 BPM), Luka Doncic (9.0), Giannis Antetokounmpo 8.5), Stephen Curry (7.6), Kevin Durant (7.1), LeBron James (6.1), Zion Williamson (5.8), and Jaylen Brown. While Brown’s basic box score stats get him in elite company, his BPM of 1.3 is nowhere near their level, which is interesting because the box in box plus/minus is referencing the box score.
One of the things I find most frustrating about the eye test vs. analytics debate is when a single case study is used to dismiss an entire avenue to measure players. BPM, WS/48, Raptor, and EPM aren’t useless because they are less glowing in their assessment of Jaylen Brown, and the fans, media, and players aren’t idiots because they think Brown is great. For the most part, advanced metrics, fans, media, and players agree. These are the top-25 players in BPM.
Quibble over this list all you want, but 13 of these players ended up on an All-NBA team, and most, if not all, received votes. What is it about Jaylen Brown that makes him a media darling but an analytics dud?
Why Jalen Brown is a star
Let’s start with why Jaylen Brown is viewed as a great player. The most obvious reason is good old points per game. Brown is a strong volume scorer and ranked ninth in points per game at 26.6. He also isn’t a chucker who trades in efficiency for sheer volume. His effective field goal percentage (eFG%) this season of 55% was above the league average of 54.5%, and so was his true shooting percentage (TS%). The easiest way to get noticed is to score points, and Brown is a proven commodity.
Another important factor is Brown is viewed as a plus defender. There aren’t any basic box score stats that suggest this, but there are a host of good perimeter defenders who don’t rack up statistics, and the eye test, particularly on the perimeter, still has a strong hold on defensive reputation.
It’s hard to fault people for viewing Brown as an excellent two-way player. He simply looks the part. He’s strong and explosive, and when he blows by a defender or locks someone down, the optics make sense. Not only does he physically fit the mold of a great player, he has the style of play to back it up. He can create his own look and has become a reliable midrange jump shooter. When he gets it going on offense, he feels unstoppable. Superstars win one-on-one matchups, and Brown wins his fair share of duels.
The final piece in the puzzle concerns who Brown plays for. The Boston Celtics are a well-supported historic organization that has consistently won and won big with Brown in town. In his seven seasons, the Celtics have won 63% of their games, made the playoffs every season, reached the Eastern Conference Finals four times, and the NBA Finals once. The NBA markets the playoffs as their everything and the Celtics have afforded Brown the platform to perform on center stage.
When you add it all up, Brown’s sterling reputation among the fans, media members, and players makes sense. He can score the ball, has a strong defensive reputation, and every summer he’s on the NBA’s biggest stage having success. When a player has a platform, looks the part, and backs it up, they’re usually a star. So why do the metrics vehemently disagree?
Why the metrics don’t love Jaylen Brown
The problem, or perhaps the beauty, with metrics is they are unswayed by the aesthetics and platform a certain player possesses. They’re solely interested in effectiveness relative to one’s peers.
Jaylen Brown is an excellent scorer, and fans, media members, and players have long placed a premium on high-level scorers. Metrics also love excellent scorers, but does his 26.6 points per game tell the whole story?
Brown’s scoring, while still valuable, lags behind the absolute best players in the league. Because I’m an insane person, I hate how shooting fouls don’t count as a field goal attempt. The result is that it’s difficult to know who is actually taking the most shots per game in the league. So I did it manually by adding field goal attempts to shooting fouls drawn and subtracting and-ones. From there, I had true shots per game, and I could calculate points per true shot.
Of the top-101 players in field goal attempts, Brown’s 1.181 points per true shot was below the average of 1.205 and ranked 65th. At his shot volume, that’s still tremendous and slightly below Jamal Murray and Jordan Poole’s per shot production but at a higher volume. However, Brown’s scoring production alone won’t make him a star if there are deficiencies in his game.
(If you’re wondering who was first, Nikola Jokic came in at 1.478, Kevin Durant was second at 1.433, and Domantas Sabonis was third with 1.386)
Brown is clearly a talented scorer, even if his points per game flatter his offensive impact, and metrics are sure to give him a sizable boost because of it. However, metrics also love passing, offensive rebounds, free throw attempts, 3-point attempts, and limited turnovers. And it’s in these areas where Brown lags behind many of the top players in the NBA.
A quick look at where Jaylen Brown stacks up in the basic box score stats outside of points per game presents us with a player with significant holes in their game. Brown’s 3.5 assists per game rank 67th, his 1.2 offensive rebounds 75th, his 5.1 free throw attempts 30th, his 7.3 3-point attempts 21st, and his 2.9 turnovers per game are 17th.
Brown is a high-volume scorer but is comparably a low-volume free-throw generator, and his sizeable 3-point volume is hurt by inefficiency (33.5%). Free throws and 3-pointers aren’t everything, but they’re the modern-day foundation for just about every perimeter scorer. Free throws necessitate a foul, and foul trouble is the best way to neutralize an interior presence. They’re also a good proxy for rim pressure, and defensive efficiency dramatically improves after a made or missed free throw. 3-pointers are also valuable beyond their points, although those points are plenty valuable too. Spacing is the name of the modern game, and nothing has unlocked spacing quite like 3-point volume.
The two best metrics to capture Brown’s shortcomings at the line and beyond the arc are free throw rate (FTr) and 3-point attempt rate (3PAr). Of the qualified top-25 points per game leaders, which doesn’t include LeBron James, Devin Booker, Kevin Durant, Anthony Davis, or Bradley Beal due to games played, only one had a lower FTr than Brown’s .249, Kyrie Irving at .227. When we flip it to 3PAr, Brown does better, ranking 13th, but he paired that with the 14th-best 3P% of the group. If you were wondering, James, Booker, Durant, Beal, and Davis all have better FTr than Brown.
Brown’s inability to generate free throws and his mediocre 3-point volume and efficiency for a perimeter player compared to other volume scorers significantly weigh down his offensive impact. He was the only player in the top-25 in points per game to be below league average in FTr and 3PAr, and six were above average in both.
Brown’s combination of slightly above-league-average shooting efficiency at tremendous volume is still impressive, but to become a truly elite scorer, he’ll need to either ramp up his free throw generation or 3-point efficiency. The good news is his tracking data suggests he is close to being a much more dangerous threat from behind the arc and at the line.
He was 15th in tightly contested threes (defender two to four feet away) per game at 1.8 and 15th in open threes (four to six feet away) at 4.1. His efficiency wasn’t elite, 29.1% on contested threes and 34.1% on open threes, but those figures are acceptable and in line with other elite 3-point shooters. What doomed Brown was he was dragged down by only hitting 39.1% of his wide-open threes. A slight uptick in his 3-point volume and efficiency on open threes would go a long way in boosting his overall line, and he’s already good enough at the hard part.
Brown’s inability to generate free throws is perplexing. His 7.2 restricted area (less than five feet from the rim) attempts per game were tied for tenth with Nikola Jokic and Anthony Edwards and are more than Joel Embiid, the patron saint of free throws. While some players garner free throws through sheer force, for many perimeter players, it is a skill. Brown does the hard part, getting to the rim, but he needs to add another layer to his offensive game where he gets the whistle.
The fact is, Jaylen Brown’s actual scoring lags behind the prestige of his points per game. That’s not a death sentence. Plenty of elite offensive players have had more middling scoring production than you would assume. They just make up for it by generating points for their teammates.
Metrics love passing, and unfortunately for Brown, his passing is well below the norm of an offensive star who isn’t a point-per-shot god. His 2.9 turnovers per game were the 17th most in the league, which is neither good nor bad on its own. The best passers and offensive players rack up turnovers because turnovers are the price of point production. The problem is Brown didn’t generate enough assists to get his passing into the black.
Turnovers are the worst possible offensive outcome. When you shoot, you can score or draw a foul, and a missed shot can become an offensive rebound. But a turnover is always zero points and is more likely to lead to a highly efficient transition opportunity for your opponent. While playmaking isn’t perceived as beneficial as shot-making, it’s still plenty valuable. Players are more efficient on assisted shots as opposed to unassisted ones. When the difference between the worst and best shooting team this season was only 5.7% in eFG%, each percentage point a pass adds to a shot is massive and assists are a fantastic proxy for high-quality passing.
Brown’s 3.5 assists per game were 67th in the league, and the players with the most comparable assist and turnovers per game to Brown are Kyle Kuzma and Paolo Banchero. Kuzma is a fine player, but the only people that ever mentioned him in the All-Star conversation are weirdo Wizards fans (which I can say because I am a Wizards fan), and Banchero was a 20-year-old rookie.
Expanding the peer group further doesn’t help Brown either. The only other players to average over 2.5 turnovers and under 4.0 assists per game were the aforementioned Kuzma and Banchero and Jalen Green, Alperen Sengun, Kelly Olynyk, and Bam Adebayo. Green and Sengun were second-year players on a bad team, Olynyk is a journeyman floor spacing big, and Adebayo is a solid offensive player that has made his name on the defensive end. Brown led that group in offensive BPM at 1.5, which was the 53rd-best figure this season, but it shows how capped his offensive production is if he continues to turn the ball over without assists.
The reason why metrics aren’t in love with the league’s ninth-leading scorer feels relatively simple. His offensive impact is significantly weighed down by turnovers and poor passing. Plus, his scoring impact wasn’t as impressive as his points per game would suggest. Jaylen Brown was still a plus-offensive player, but he wasn’t elite. However, offense is just one part of the equation, and Brown has a strong defensive reputation for a reason.
Defensive metrics are admittedly imprecise, but over a large sample, they do offer insights. Since 2017-18, the best single-season defensive BPM (DBPM) for a shooting guard or small forward is Matisse Thybulle’s +4.5 in 2020-21. Simply reaching +2.0 gets you into the top 20. Brown has posted a -0.2 DBPM in that time frame, with seasons ranging from 0.3 to -0.7. According to DBPM, Brown is a good to solid but not great perimeter defender.
FiveThirtyEight’s Raptor has the best defensive season from a shooting guard since 2017-18 at +6.1 (Alex Caruso this season) and +5.0 for a small forward (Matisse Thybulle in 2020-21). Jalen Brown’s best season came in at +1.2 in 2017-18, and his worst, outside of his rookie season, was 2020-21, where he came in at 0.0. Once again, Raptor says Brown is solid to good defender but not a great one.
Dunks and Threes’ defensive component to EPM tells a similar story. The best defensive shooting guard this season was Alex Caruson at +4.5, and the best small forward was Matisse Thybulle at +2.6. Brown’s 2017-18 season came in at +2.1, but since then, he has been between +1.0 and +0.1.
Honestly, the metrics and the public seem to agree about Jaylen Brown’s defense. He’s a good defender but isn’t one of the league’s best. As a one-on-one defender, he is highly effective, but he isn’t a great team defender and doesn’t rack up a ton of steals. As a player that handles a sizable offensive load, that’s still impressive. The only wings who grade out as elite defensively and offensively are Paul George, Kawhi Leonard, and Giannis Antetokounmpo, who is really a frontcourt player.
The final piece of the Jaylen Brown puzzle is his on-off ratings. At the end of the day, catch-all advanced metrics try to pinpoint how many points per 100 possessions a player adds over an average player. While on-off ratings are team dependent, the best players in the league see their teams prosper when they’re on the court and suffer when they sit.
Once again, Brown doesn’t pop like many star players in on-off. Since 2017-18, his on-off net rating is only +1.0. When we shrink the search to after Brown’s breakout in 2019-20 when he averaged over 20 points per game for the first time, his on-off net rating declines to +0.3. According to PBP Stats, over that period, the Celtics' offense would improve from 114.49 points per 100 possessions to 116.09 with Brown on the court, but their defensive rating would decline from 109.29 to 110.41.
Plus/minus and on-off ratings can be noisy over a single season, but over a large sample, they usually spot the best players in the league. For instance, since Kawhi Leonard joined the Clippers (2019-20), he has a +11.2 on-off net rating, Jimmy Butler’s is +5.6 with the Heat, Paul George’s is +6.5 with the Clippers, which includes a full season without Kawhi, and Jayson Tatum, Brown’s own teammate, has an on-off net rating of +9.0 since 2019-20.
The metrics don’t love Jaylen Brown because he doesn’t impact winning in a measurable way as much as other elite players. His scoring is a plus but not as impressive as his points per game suggest, and he adds so little as a playmaker that he’s a subtraction. While his defense is good, it’s not elite, and perimeter players' defensive impact is far less substantial than their interior brethren.
I’m a sucker for metrics, and even I find it hard to believe that Jaylen Brown isn’t a surefire top-50 player. Averaging 25 points per game ain’t what it used to be, but it’s still incredibly difficult, and I believe that’s where the massive disconnect between the analytics and eye test lies. It’s hard to be as good at basketball as Jaylen Brown. In fact, there probably aren’t 30 people on Earth that can do what he does, but that doesn’t mean he helps his team win basketball games better than 30 others.
Jaylen Brown is an enigma. For some, he’s one of the best basketball players in the world, and for others, he’s a second-tier second option. Paradoxically, he exposes the fans, media, players, and metrics’ blind spots. The ability to do incredibly difficult acts on a basketball court is valuable beyond what many metrics are capable of capturing. Is a contested midrange jumper good? No, but sometimes the shot clock is winding down, and it’s all you’re going to get, and Brown is a good bet to make the best of it. Conversely, passing and the environmental effects of scoring are picked up by metrics but often overlooked by fans, media, and players. Is a contested midrange jumper good? No, and someone was open in the corner, and there were 14 seconds left on the shot clock.
The Celtics are a great team, and Jaylen Brown is a major reason why. More credit should go to their excellent supporting cast of Derrick White, Marcus Smart, Robert Williams III, Al Horford, Malcolm Brogdon, and Grant Williams, but Brown allows each of them to be stars in their roles, something no metric can fully capture. While some players make a living doing dirty work, Brown makes a living doing a superstar’s work, even if he isn’t quite at that level.
Jaylen Brown is really good at basketball. He’s just not one of the absolute best players in the league. He has many of the tools necessary to be a star, but he needs to refine their deployment. The fans, media, and players are right about his ceiling, and the metrics are right about his impact. He’s definitely not a top-15 player, but he’s also not well out of the top-50. Let’s split the difference and say he’s a borderline top-25 player.
Excellent analysis
Time will tell as we watch his career on the 76rs
Celtics got some old timers with George and Conley