A glossary of statistical terms has been provided at the conclusion of the article. All statistical data is sourced from Basketball-Reference unless otherwise noted.
This article statistically analyzes two NBA players - to keep this process completely objective we will call them “Player A” and “Player B.” To give you some context, these players are on the same team, both play a very similar position, have been playing under seven seasons, and have been All Stars. With the NBA trade deadline approaching on Thursday, February 10, 2022, there have been discussions in the media as to whether these two can coexist on a winning team, and if one of them should be traded. If you’re a fan of this team you probably know which players I’m talking about - if not I will reveal them at the end of the article.
Career Statistics
If you read ALLCAP: What Will The Pelicans Do With Zion Williamson you know that to gain a full appreciation for both players you are comparing, career statistics is a good place to start as it helps us attempt to answer a baseline question - who has been (objectively) better during their entire tenure in the NBA? Note that this is not even close to the same as who is better now, or who has more value (now or throughout their career) but it at least gives us a large enough population to understand the two players.
I separated the statistics into four categories: Standard Per Game, Standard Per 36 Minutes (meaning what the players’ statistics would be if they played 36 minutes per game), Rating Per 100 Possessions, and Advanced. The glossary at the bottom of this article will help you understand any of the statistics that we use here.
As previously mentioned, let’s first dive into these buckets from a career overview for these two players:
When looking from a career perspective, you can see Player B wins in almost every standard statistical category (per game or per 36 minutes). Although having less time in the NBA, Player B has started in more games (he has started all of his career games), gets more minutes, and when looking at the traditional box score categories (Points, Rebounds, Assists, Steals, Blocks and Turnovers), Player B is better in almost all of these as well. This flows into his offensive and defensive rating and advanced statistics, showing his win shares (WS), box plus / minus (BPM), and value over replacement player (VORP) statistics all significantly higher than Player A.
What Player A has is more efficient shooting percentages. Player B has a slight edge in 3-Point field goal percentage, but as far as effective field goal percentage (which takes into account a combination of two and three point percentages) and true shooting percentage (which adds in free throw percentage) Player A is above Player B.
This got me thinking - if Player A has higher shooting percentages but on less volume, maybe there is something about the way he is utilized that has caused him to have less impact in some of these statistical categories. I partially handled that by looking at the per 36 minute Standard Statistics, but as you can see, Player B was still superior in most of these categories, although if you look at points, for example, the gap between the players decreases significantly.
Another method I considered to normalize the data was to normalize usage rate, as this shows us how much a player is involved while on the court. I found that the mean for usage rate was 25.2%, so I looked at both players’ statistics again, modifying the numbers based on a 25.2% usage rate. Surprisingly, this had virtually no effect on the results. Obviously, as with the per 36 numbers, this decreased the gap but did not change the overarching results of who won each category. Also, I decided that utilization should not necessarily be something that is normalized because to some degree, it can be controlled by the player like any other statistics. Obviously, coaches will decide which player should bring the ball up, which player should be taking most of the shots, etc., to make the team successful, but there is also an element of a player being able to get open, command the ball, and be able to make a play. So I decided to scrap this now and going forward and did not normalize any of the data in this way.
Comparing Progress
The next thing we need to look at is comparing the statistical progress of these two players throughout their career. I looked at each player’s statistics (the same as above) from Year 1 - 3 of their career and then the years since then (for Player A, he has played three seasons since, and for Player B, he has played two seasons since, including the current season). The reason we compare the first three seasons the seasons after is just based on contract structures in the NBA. As we learned in the last ALLCAP article on Zion Williamson, a player is eligible to receive an extension from their Rookie Scale Contract after their third season. Obviously there are different circumstances for every player, but generally speaking this tells us that the league and teams feel that they have had enough of a sample size of games for players after three seasons to determine the size of their next contract, and ultimately the role they will have on their respective teams and in the league.
Clearly, we can see both these players improved in almost every statistical category since their first three seasons. It’s not surprising to see the jump in the traditional box score statistics as usage rate and minutes increase, as well as just having a better feel for playing in the NBA. It’s also not surprising that, when comparing the two players, Player A increased significantly in offensive and defensive rating while Player B stayed pretty much the same. Offensive and defensive rating tells us how many points a team scores / allows per 100 possessions when a player is on the court. As your offensive rating gets higher, it is harder to improve, and for defensive rating, as it gets lower, it is harder to improve. Additionally, since Player A and Player B are on the same team, they are playing with a lot of the same players (including each other) so it is not surprising to see their offensive and defensive ratings eventually become nearly identical, as these statistics are just as much affected by the other players on the court.
What really stood out here is the improvement in WS and VORP by Player A, especially when compared to the improvement (or not) by Player B. Obviously, Player B started off much higher in these categories, but these are important in showing who is contributing more to wins - especially when looking at two players on the same team. Even though Player B still has a higher VORP (although not by much) this is trending down while Player A is trending way up. This could say something about their respective replacement players, or just the fact that Player A did not start all of his games during his first three seasons, but his increase in this category comes with an increase in WS, while for Player B, it has been the opposite. Below represents the percentage increases in each category for Player A and Player B:
Note that VORP needed a separate graph due to the difference - Player A’s VORP improved 2400% while Player B’s VORP actually declined by 16.92%.
Conclusion - Analysis of Player A v Player B
Now that we understand the data I’ll pose the question - who do you think is the better player going forward? Personally, I always felt Player A was better than Player B. This opinion flip flopped over time, to the point where last season I actually bet on Player B to win MVP (which he did not). When I looked at the data my initial thought was that I was wrong - Player B is far superior to Player A. But when digging deeper, based on the improvement of both players, especially when you look at Player A’s increase in WS, I actually personally would rather have Player A.
If you haven’t figured it out yet - Player A is Jaylen Brown, and Player B is Jayson Tatum. Both players are starters for the Boston Celtics, and as their statistics may indicate, they play very similar positions. Both have been All Stars in the NBA and both have a very high perceived value. What’s interesting is that in trade discussions, most of what we have heard in the media is that Boston would likely only consider trading Brown because of Tatum’s upside. One of these examples is Brown for Ben Simmons which we discussed in an ALLCAP Trade Rumor article back in November, 2021. It’s true that if you look at “upside” from the standpoint of who is more likely to have a 40+ point performance (Tatum has 10 and Brown has 5). However, based on what the data shows it seems like Tatum is leveling out as far as his improvement is concerned, whereas Brown is still getting better. So as far as upside, it is arguable that Brown’s is still to be determined while we have a better idea of where Tatum’s is.
We won’t get into it too much here, because this analysis is already a lot to process, but if we were to decide who we would keep between the two, it is also vital to look at contract situations. Including this year, Brown has three years until hitting free agency including this year getting paid around 22-23% of the cap, whereas Tatum has five years on a max extension (minimum 25% of the salary cap). Both will likely be demanding a veteran max contract (30% or higher of the salary cap) assuming they remain on the same trajectory as far as level of play, so it would be crucial to assess the team’s cap situation along with statistical analysis in order to determine which player better fits the bill.
Glossary - Standard Statistics
G (Games Played)
GS (Games Started)
MP (Minutes Played)
FG (Field Goals Made)
FG% (Field Goal Percentage) - Percentage of made shots per attempt.
3P (3 Point Field Goals Made)
3P% (3 Point Field Goal Percentage) - Percentage of 3 point shots made per 3 point attempt.
2P (2 Point Field Goals Made)
2P% (2 Point Field Goal Percentage) - Percentage of 2 point shots made per 2 point attempt.
eFG% (Effective FG%) - Percentage of field goals made accounting for the fact that a three point field goal is worth more than a two point field goal.
FT% (Free Throw Percentage) - Percentage of free throws made per free throw attempt.
TRB (Total Rebounds)
AST (Assists)
STL (Steals)
BLK (Blocks)
TOV (Turnovers)
PTS (Points)
Glossary - Per 100 Possessions Statistics
Ortg (Offensive Rating) - Number of points scored per 100 possessions.
Drtg (Defensive Rating) - Number of points allowed per 100 possessions.
Glossary - Advanced Statistics
PER (Player Efficiency Rating) - A measure of per-minute production standardized such that the league average is 15.
TS% (True Shooting %) - Shooting percentage taking free throws into account.
USG% (Usage Rate) - Percentage of utilization for a player while they are on the court.
OWS (Offensive Win Shares) - How many wins a player contributes per season based on their offensive output.
DWS (Defensive Win Shares) - How many wins a player contributes per season based on their defensive output.
WS (Win Shares) - Combination of Offensive and Defensive Win Shares.
WS/48 (Win Shares / 48 Minutes) - How many total wins a player contributes to per 48 minutes based on both offensive and defensive output (league average is .100).
OBPM (Offensive Box Plus / Minus) - A box score estimate of the offensive points per 100 possessions contributed over a league-average player, translated to an average team.
BPM (Box Plus / Minus) - Combination of Offensive and Defensive Box Plus / Minus.
DBPM (Defensive Box Plus / Minus) - A box score estimate of the defensive points per 100 possessions contributed over a league-average player, translated to an average team.
VORP (Value Over Replacement Player) - A box score estimate of points per 100 TEAM possessions that a player contributed over a replacement-level player, translated to an average team and prorated for an 82 game season. This number multiplied by 2.7 equals wins over replacement.