🧊 Measuring Young's impact on his teammates' statistics when they play with Young versus others
In 2018, Atlanta Hawks fans collectively groaned when the team decided to trade its first-round, third overall pick, Luka Dončić, and another future first-round pick for the skinny, short point guard, Trae Young.
But by the start of the season, Young put on 15 pounds of muscle, indicating to fans he was ready to prove something — and did, by way of skill, swagger and, most important, teamwork.
Now, in the ongoing 2021-22 season, the Hawks have not yet reached the pace they had last spring. As of Feb. 20, 2022, the team has a 28-30 record, but is ranked No. 2 in ESPN's John Hollinger's measurement for team efficiency. Together, these circumstances provide an environment ripe for data analysis.
Meanwhile, Young continues to break records for the Hawks, clearly contributing to the team’s success when they find it, but by how much?
NBA.com keeps collective stats for every team’s lineup history. For five-, four-, three- and two-player combinations, the site logs the number of games these combinations have played in together, their collective offensive and defensive stats, and more.
Using this data, I wanted to measure Young’s impact on the player combinations he is a part of, demonstrating just how much his presence on the floor increases a group’s performance, when compared against group’s he is not a part of.
To do this, I used the Hawk’s two-player lineups to assess partnerships between a select group of Hawks, including:
I chose to start with just three measures of success: average points, assists and rebounds per game. In all three, Young improved the performance of the pairs of which he was a part by over 40 percent. Of all these measures, he improved points per game the most.
To some degree, the idea that Young would improve the performance of his partnerships, compared to partnerships without him, is not surprising — he is the team’s best player.
But by comparing collective as opposed to individual stats, fans have a unique window into something different: team performance and player chemistry rather than individual strengths (and to read more about my methodology, you can find my code here).
Furthermore, while Young might be the team’s best player, he is not its top performer by every measure. When it comes to rebounds, for example, Young averages 3.9 per game, far below Clint Capela’s 10.9 or John Collins’s 7.9.
Nevertheless, Young still has a substantial positive impact on the collective rebounds among his partnerships, compared to partnerships without him. While Young averages less than four rebounds per game himself, when he is paired with the players above, the pairs average roughly 20 rebounds per game, nearly doubling the average rebounds per game for these players pairings without Young.
When it comes to assists, these Young’s bread and butter. He is ranked third overall in the NBA for assists per game. But just how much does his gameplay influence his teammates’ assists? Could Young’s tendency to ball movement rub off on his teammates?
While Young adds roughly 9 more assists per game to collective performance when paired with these players, he averages 9.3 assists per game himself.
If anything, it seems Young has the greatest impact on John Collins, who averages less than two assists per game on his own. When Collins and Young play together, they average 16 assists per game, showing these players improve each others’ average performance when paired together.
When I started this project, my goal was to not only demonstrate Young’s impact on his teammates’ pace, but to measure it. There are many more statistics I could have chosen, too many confounding variables to count and infinite ways to display this data (for alternate versions of this article with different data visualizations, click here). For now, I’m content with some data-backed evidence that Young lifts his teammates up. That is the Hawks Way after all.