As Devin Booker and Trae Young lead their respective teams in the 2021 NBA conference finals, it's time to retire the phrase "good stats, bad team."
Both Booker and Young were tagged with the unflattering label early in their careers, when they accumulated impressive per-game statistics for lottery-bound teams. The notion might even have helped keep Young off this year's All-Star team when the Atlanta Hawks started the season 13-18 while dealing with injuries to key starters prior to the announcement of reserves. (The Hawks went a mere 28-13 the rest of the way.)
Now, with Young powering Atlanta's pair of upset series wins as a lower seed, and Booker carrying the Phoenix Suns to a lead in the Western Conference finals -- without star guard Chris Paul -- heading into Tuesday's Game 2 (9 p.m. ET on ESPN), we should learn from their examples and forget about a concept that always made more sense in theory than practice.
The clumsy relationship between individual and team success
I can understand where the concept of "good stats, bad team" originated. Back when players were evaluated primarily based on their per-game statistics, and specifically points per game, it was easy to confuse volume scoring for performance that translated into winning.
Without the evaluation framework provided by advanced stats, team record was a shortcut to telling the imposters from the genuine article.
Consider the group of nearly 1,000 players since individual turnovers were first tracked in 1977-78 to average at least 20 points per game while playing at least two-thirds of their team's games. Of these, a little more than a third came from teams with below-.500 records. The conventional wisdom is right, to an extent. There is a relationship between the record of a high scorer's team and his own performance, as measured by my player net rating on a per-possession basis.
On average, my net rating for players on teams that finished .500 or better (plus-4.5 points per 100 possessions) is more than twice as good as for those on below-.500 teams (plus-2.1). Yet individual performance for a high scorer still explains only about 20% of the variation in team record.
Essentially, judging a player's ability by his team's performance is a blunt measure. Modern statistical analysis gives us more surgical tools that can better separate individual performance. And though those tools indicate that most high scorers on lottery teams are in fact less valuable than those on playoff ones, that's not always the case.
After all, there are four other players on the court at all times, not counting minutes a player is on the bench. So it is that the lowest-rated player by my metric in this group of 20-point scorers (Jeff Malone with the 1991-92 Utah Jazz) could play on a team that went 55-27 and reached the conference finals thanks to stars Karl Malone and John Stockton.
At the other extreme, Anthony Davis' 2018-19 performance ranked in the top 25 of this group, highest for any player on a below-.500 team. (Admittedly, Davis played sparingly after publicly requesting a trade, but that happened only after the New Orleans Pelicans fell out of the playoff race.) A year later, after Davis was traded to the Los Angeles Lakers, he helped them win the championship. Turns out he definitely wasn't an example of "good stats, bad team."
Better support surrounds Booker and Young
It didn't require requesting a trade for Booker and Young to be put in better position to succeed. Their own teams managed that by eventually parlaying a series of lottery picks into a core of young talent and making skillful additions to those pieces through trades and free agency.
Weighted by minutes played, the Hawks have the third-youngest rotation among playoff teams, while the Suns are sixth youngest. That stands in sharp contrast to their opponents in the conference finals, the Milwaukee Bucks (third oldest) and LA Clippers (oldest). In addition to Young, Atlanta's core includes recent first-round picks John Collins, Kevin Huerter and the injured De'Andre Hunter, while Phoenix boasts a pair of third-year starters in No. 1 overall pick Deandre Ayton and Mikal Bridges.
To those groups, both teams made key pickups last offseason. For the Suns, that was All-NBA point guard Chris Paul, whose veteran example has helped Booker and the other young Phoenix players translate an undefeated run in last year's seeding games into a full season of success, plus forward Jae Crowder. The Hawks weaponized their cap space to add starter Bogdan Bogdanovic and top reserve Danilo Gallinari, having already added Clint Capela in the trade at the 2020 trade deadline.
Because of the newcomers, both Booker and Young saw their scoring averages decline this season. Young's drop-off of 4.3 PPG is particularly notable. Young was one of 28 players in the league to see his scoring decline so dramatically at an age where most players' stats are on the rise. That happened in part because Atlanta's pace slowed dramatically, but Young's assist rate also increased 10% on a per-possession basis while his usage went down. It's no surprise Young was more willing to trust a better set of teammates.
Given Paul took over primary ballhandling duties, and the Suns typically had a point guard on the court after using Booker in that role at times in previous seasons, his assist rate actually went down while his usage went up. But the improved depth allowed Phoenix to rest Booker more, as his minutes per game dropped from 35.9 to 33.9. His scoring fell as a result.
How both stars have developed their games
Although the change in outlook for the Hawks and Suns has more to do with the quality of their rosters, they've also benefited from the inevitable improvement of Young (age 22) and Booker (24). That's been particularly evident on defense, a more legitimate critique of both players earlier in their careers.
During 2019-20, Young's minus-1.8 defensive RAPM (regularized adjusted plus-minus via NBAshotcharts.com) ranked in the league's bottom 10 when accounting for the teammates and opponents who were on the court with him. At minus-1.3, Booker was also in the bottom 25. While defense remains a weakness, both Booker (minus-0.7) and Young (minus-0.9) have improved out of that range.
We've also seen Booker reach a new level of shot making during this playoff run. Second Spectrum's qSI (quantified shooter impact) metric measures how much players outperform the expected effective field goal percentage for an average player on the same shot attempts based on location, type and distance of nearby defenders. Booker has also excelled in this category, ranking 59th in the league (minimum 100 field goal attempts) with a plus-7.1 qSI during the regular season. In the playoffs, that has jumped to plus-12.0, sixth best in the same group.
In Young's case, playmaking has stepped to the forefront. He's assisted on nearly half his teammates' field goals while on the court in the playoffs, per Basketball-Reference.com data, the highest estimated rate in a run of more than 10 games since Russell Westbrook in 2016. According to Stathead.com, Young is on track to join Westbrook (twice), James Harden, LeBron James (four times) and John Wall as the fifth player since 1977-78 with an assist rate higher than 40% and a usage rate higher than 30% in the same playoffs.
As Devin Booker and Trae Young lead their respective teams in the 2021 NBA conference finals, it's time to retire the phrase "good stats, bad team."
Both Booker and Young were tagged with the unflattering label early in their careers, when they accumulated impressive per-game statistics for lottery-bound teams. The notion might even have helped keep Young off this year's All-Star team when the Atlanta Hawks started the season 13-18 while dealing with injuries to key starters prior to the announcement of reserves. (The Hawks went a mere 28-13 the rest of the way.)
Now, with Young powering Atlanta's pair of upset series wins as a lower seed, and Booker carrying the Phoenix Suns to a lead in the Western Conference finals -- without star guard Chris Paul -- heading into Tuesday's Game 2 (9 p.m. ET on ESPN), we should learn from their examples and forget about a concept that always made more sense in theory than practice.
The clumsy relationship between individual and team success
I can understand where the concept of "good stats, bad team" originated. Back when players were evaluated primarily based on their per-game statistics, and specifically points per game, it was easy to confuse volume scoring for performance that translated into winning.
Without the evaluation framework provided by advanced stats, team record was a shortcut to telling the imposters from the genuine article.
Consider the group of nearly 1,000 players since individual turnovers were first tracked in 1977-78 to average at least 20 points per game while playing at least two-thirds of their team's games. Of these, a little more than a third came from teams with below-.500 records. The conventional wisdom is right, to an extent. There is a relationship between the record of a high scorer's team and his own performance, as measured by my player net rating on a per-possession basis.
On average, my net rating for players on teams that finished .500 or better (plus-4.5 points per 100 possessions) is more than twice as good as for those on below-.500 teams (plus-2.1). Yet individual performance for a high scorer still explains only about 20% of the variation in team record.
Essentially, judging a player's ability by his team's performance is a blunt measure. Modern statistical analysis gives us more surgical tools that can better separate individual performance. And though those tools indicate that most high scorers on lottery teams are in fact less valuable than those on playoff ones, that's not always the case.
After all, there are four other players on the court at all times, not counting minutes a player is on the bench. So it is that the lowest-rated player by my metric in this group of 20-point scorers (Jeff Malone with the 1991-92 Utah Jazz) could play on a team that went 55-27 and reached the conference finals thanks to stars Karl Malone and John Stockton.
At the other extreme, Anthony Davis' 2018-19 performance ranked in the top 25 of this group, highest for any player on a below-.500 team. (Admittedly, Davis played sparingly after publicly requesting a trade, but that happened only after the New Orleans Pelicans fell out of the playoff race.) A year later, after Davis was traded to the Los Angeles Lakers, he helped them win the championship. Turns out he definitely wasn't an example of "good stats, bad team."
Better support surrounds Booker and Young
It didn't require requesting a trade for Booker and Young to be put in better position to succeed. Their own teams managed that by eventually parlaying a series of lottery picks into a core of young talent and making skillful additions to those pieces through trades and free agency.
Weighted by minutes played, the Hawks have the third-youngest rotation among playoff teams, while the Suns are sixth youngest. That stands in sharp contrast to their opponents in the conference finals, the Milwaukee Bucks (third oldest) and LA Clippers (oldest). In addition to Young, Atlanta's core includes recent first-round picks John Collins, Kevin Huerter and the injured De'Andre Hunter, while Phoenix boasts a pair of third-year starters in No. 1 overall pick Deandre Ayton and Mikal Bridges.
To those groups, both teams made key pickups last offseason. For the Suns, that was All-NBA point guard Chris Paul, whose veteran example has helped Booker and the other young Phoenix players translate an undefeated run in last year's seeding games into a full season of success, plus forward Jae Crowder. The Hawks weaponized their cap space to add starter Bogdan Bogdanovic and top reserve Danilo Gallinari, having already added Clint Capela in the trade at the 2020 trade deadline.
Because of the newcomers, both Booker and Young saw their scoring averages decline this season. Young's drop-off of 4.3 PPG is particularly notable. Young was one of 28 players in the league to see his scoring decline so dramatically at an age where most players' stats are on the rise. That happened in part because Atlanta's pace slowed dramatically, but Young's assist rate also increased 10% on a per-possession basis while his usage went down. It's no surprise Young was more willing to trust a better set of teammates.
Given Paul took over primary ballhandling duties, and the Suns typically had a point guard on the court after using Booker in that role at times in previous seasons, his assist rate actually went down while his usage went up. But the improved depth allowed Phoenix to rest Booker more, as his minutes per game dropped from 35.9 to 33.9. His scoring fell as a result.
How both stars have developed their games
Although the change in outlook for the Hawks and Suns has more to do with the quality of their rosters, they've also benefited from the inevitable improvement of Young (age 22) and Booker (24). That's been particularly evident on defense, a more legitimate critique of both players earlier in their careers.
During 2019-20, Young's minus-1.8 defensive RAPM (regularized adjusted plus-minus via NBAshotcharts.com) ranked in the league's bottom 10 when accounting for the teammates and opponents who were on the court with him. At minus-1.3, Booker was also in the bottom 25. While defense remains a weakness, both Booker (minus-0.7) and Young (minus-0.9) have improved out of that range.
We've also seen Booker reach a new level of shot making during this playoff run. Second Spectrum's qSI (quantified shooter impact) metric measures how much players outperform the expected effective field goal percentage for an average player on the same shot attempts based on location, type and distance of nearby defenders. Booker has also excelled in this category, ranking 59th in the league (minimum 100 field goal attempts) with a plus-7.1 qSI during the regular season. In the playoffs, that has jumped to plus-12.0, sixth best in the same group.
In Young's case, playmaking has stepped to the forefront. He's assisted on nearly half his teammates' field goals while on the court in the playoffs, per Basketball-Reference.com data, the highest estimated rate in a run of more than 10 games since Russell Westbrook in 2016. According to Stathead.com, Young is on track to join Westbrook (twice), James Harden, LeBron James (four times) and John Wall as the fifth player since 1977-78 with an assist rate higher than 40% and a usage rate higher than 30% in the same playoffs.
