Nylon Calculus: Reactive teams in the NBA playoff bubble

Whether or not the hot hand is real, NBA players certainly loosen their shot selection when they’ve got it going.

NBA teams make thousands of decisions every game. Their decision-making strategies have been refined in the analytics era, but some approaches–such as those dictated by short-term successes and failures–remain flawed. We at Nylon Calculus present the teams in the NBA playoff bubble that employ such an approach.

Changing shot selection has been the Moreyball Era’s most obvious alteration to basketball strategy. Three is worth more than two, and NBA teams have come to realize (much like Mr. Incredible) that math is math. Math is math!

But while 3-point attempts are on the rise, the when and why of letting it fly may still be affected by old-school thinking. When some teams get hot from beyond the arc, they adjust to what is working at the moment and take 3s at a higher rate. Research tells us that 3-point success varies greatly over small sample sizes, suggesting that such a reactionary approach is unwise.

The degree of reactionary-ness can be measured by how much the number of 3s made in the prior quarter affects the number of 3s attempted in the current quarter (while controlling for the number of 3point attempts in the prior quarter).

Across the league, here is how many 3PA can be expected in a quarter given the number of made 3s in the prior quarter (all numbers are standardized to the NBA average FGA per quarter):

On average, teams took 0.9 more 3s following quarters with five or more made threes than they did following quarters where they didn’t make a 3 at all.

Which NBA teams are most reactive to a hot-shooting quarter?

Only three teams in the Orlando bubble had results that were reactionary at a statistically significant level:  Utah, Phoenix, and Houston. Here is how many 3-point attempts would be expected after a zero-3PM quarter vs. after a three-3PM quarter for each of the three teams. For context, a three-3PM quarter is an average quarter for an NBA team, so this is the change from a poor-shooting quarter to an average one.

Utah, Phoenix, and Houston could stand to re-evaluate their offensive tendencies based on this data. If they don’t adjust, opposing defenses can capitalize on the knowledge that when these offenses catch fire from 3, there is a good chance they are going to keep on chucking.

Here is how all 30 teams stack up (Note: p-values less than 0.05 are the only ones that are statistically significant/meaningful):