Nylon Calculus: Playoffs teams are living comfortably by the 3-pointer
The Houston Rockets set a record for 3-point attempts this season with 3306 shots from distance, leapfrogging their own record from 2014-15 by 626 attempts. They went deep in to unknown territory by attempting 46.2 percent of all their shots from behind the 3-point line, also an NBA record.
Their offense has been well chronicled. But a part of the conversation around the Houston Rockets of 2016-17 is also whether their reliance on the 3-pointer will come back to bite them. Their game plan has in-built potential for offensive explosions, but at the same time a bad shooting night might leave them with very little on the offensive end.
But does Houston’s offensive output actually have that much more variance compared to other offenses?
Looking at the top three offenses of this season, Houston does not at stand out in terms of variance. This season, both Golden State and Cleveland have had quite a few games below Houston’s worst offensive game.
To compare the variance of different offenses, I have calculated the median absolute deviation for every team this season. It is basically an attempt to measure the spread of performance for an offense. I used median absolute deviation over standard deviation to not give too much weight to outliers.
Houston is among the offenses with a higher spread on their offensive output. But plotting median absolute deviation versus 3-point attempt rate for the last five seasons does not give a strong indication that teams relying on threes are more variable:
There’s no significant correlation between the 3-point attempt rate and the spread in offensive performances. Teams shooting more 3s do not seem to have a more variable offensive output.
Cleveland’s position in the top of the graph above did interest me and it looks like it is partially caused by LeBron James’ resting. The team has had absolutely dreadful offensive performances the last three seasons, when LeBron has not played. But LeBron sitting does not account for all of it and this high spread in performance has created a lot of questions about the team during the regular season in the last three years.
It might seem odd that shooting more 3s, do not lead to a higher spread in offensive output. But shooting more 3s also means that there’s a higher chance that the resulting 3-pointer percentage will be closer to average. There’s a 10 percent chance for a league-average shooter to shoot 6 for 10 or better, but only a 0.2 percent chance for him to shoot 24 for 40 or better. Same shooting percentage, just more shots.
Houston’s best shooting night was against Dallas on the Dec. 27 when the Rockets hit 53.1 percent of their threes. This season, 27 teams had better shooting nights than that. Only Brooklyn and Philadelphia have not had better shooting performances this season. Houston’s best night was only the 97th-best night in the league all season, when it came to 3-point percentage. On the other hand, their worst shooting night was the 38th-worst night overall.
Looking back again at the last three seasons, this relationship between 3-point attempt rate and spread of 3-pointer percentage is also quite clear:
Helping to lower the spread in offensive output for the Houston Rockets is all the other things that they do well. They were second in free throw rate and fourth in fastbreak point per 100 possessions. Their shooters still had to be respected even on bad shooting nights and that gravity opened up the defenses.
Next: Dion Waiters might be the quirkiest free throw shooter in the NBA
In a single game, 3-point percentage is a big factor in a team’s offensive rating, but not any more so for teams that tend to shoot a lot of 3-pointers.