Written by John McCool (@Desertrose28)
Shot selection, which includes any field goal attempt from a layup to a three pointer, is a vital component to the game of basketball. NBA teams that consistently take open or high percentage shots around the basket will likely have an advantage over their competition. Teams should also play to their strengths. For example, the Golden State Warriors run their offense around the three-pointer while the San Antonio Spurs favor a more balanced attack from the paint and mid-range.
Three-point field goal attempts and shots in the paint are expected to give more points than mid-range jumpers. For the 2015-2016 season, we examined the differences between team shot selection by grouping all 30 teams into three groups using k-means clustering, which is a easy way to classify and organize data based on similarities within the data. The clustering is based on similar field goal attempt frequencies across six different shot distance ranges: 0-5 feet, 5-9 feet, 10-14 feet, 15-19 feet, 20-24 feet, and 25-29 feet. It is important to note that the NBA three-point line deviates from 22 feet to 23.75 feet, so not all shots attempted from 20-24 feet are actually three-pointers. 
Cluster 1: 76ers, Cavaliers, Warriors, Hornets, Pistons
Cluster 2: Wizards, Spurs, Knicks, Pelicans, Timberwolves, Heat, Lakers, Clippers, Pacers, Mavericks, Bulls, Nets
Cluster 3: Jazz, Raptors, Kings, Trail Blazers, Suns, Thunder, Bucks, Grizzlies, Rockets, Nuggets, Hawks, Celtics
Cluster I: Teams That Frequently Shoot Three-Pointers
(Field goal attempt frequencies compared to the NBA league average from these distances. The decimal value 0.02 means to 2% above league average).
Cluster one consists of the Warriors, Hornets, Pistons, Cavaliers, and 76ers. Each team in this group finished among the NBA league leaders in three-point attempts per game (3PA) last season. Around 17.8% of this group’s shot attempts came from 25-29 feet, about 5.5% above the NBA league average. For reference, the next closest cluster averaged 11.2% of their field goal attempts from this range.
A moderate 0.41 correlation between three-point field goal percentage and field goal attempt (FGA) frequency from 25-29 feet suggests that teams that shot more threes also made a higher percentage of them. However, in this case, cluster one shot just 35.5% from 25-29 feet, which was only 0.64% higher than the next closest cluster of teams.
The Cavaliers, Warriors, and Hornets took advantage of their three-point opportunities, shooting a combined 38% from beyond the arc, or 3% better than the rest of the league. On the other hand, the Pistons and 76ers shot 34.5% and 33.9% respectively from the three-point range, falling into the bottom tier in this category. The 76ers led this group with 48% shot attempts from less than nine feet, 7.7% higher than the Hornets.
Unsurprisingly, the star-studded Warriors were easily the most prolific three-point shooting team in this group (and in the NBA), shooting 42.7% on catch and shoot and 39.2% on pull-up three attempts. Opposing defenses had no answer for Stephen Curry and Klay Thompson’s combined 44% three-point average. Curry alone attempted over 55% of his shots from the three-point line.
Unlike the other teams in cluster one, the Cavaliers favored taking fewer three-pointers, averaging 19.3% of their attempts from 22-24 feet while attempting just 15.9% of threes from 25-29 feet. All of the other teams averaged more 3PA from 25-29 feet.
Cluster II: Teams That Frequently Attack the Paint
Cluster two, which contains 13 teams, is a collection of teams with more varied shot attempt patterns. For instance, the Mavericks and Wizards favored shots from beyond 20 feet while the Heat clustered their shot attempts from within nine feet of the hoop. As a whole, however, this group averaged 2.8% mid-range field goal attempts above the league average from 15-19 feet.
We can see that cluster two teams (in green), grouped in the upper region of the scatterplot, led the rest of the NBA in 15-19 foot shot attempts. These teams also tended to take 10-14 foot jump shots more frequently than the other two clusters.
All together, however, cluster two only shot marginally better than cluster three from 15-19 feet and slightly worse than cluster one from 10-14 feet at a respective 40.7% and 40.4%. Teams including the Spurs, Pacers, Timberwolves, and Knicks favored the 15-19 jump shot, shooting at least 18% of their shots from this range. For reference, the Spurs’ David West and LaMarcus Aldridge scored over a third of their points from mid-range.
The Mavericks and the Clippers were outliers in this group, shooting an above average number of mid-range jumpers, but still shooting three-pointers at a fairly high volume, finishing 5th and 9th in 3PA last season. For example, the Mavericks relied on Dirk Nowitzki’s effective mid-range game, but balanced their attack with Wesley Matthews and Chandler Parson’s three-pointers. Both of these teams, however, averaged only a combined 38.1% of their points in the paint, falling just ahead of the New York Knicks, who finished dead last in this category. Following this trend, cluster two finished 2.1% below the NBA average on attempted 0-5 foot shots. 
Cluster III: Teams With a Balanced Attack
The third cluster had the best balance between three point and paint (interior) shot attempts last season. These 12 teams averaged 37.8% FGA within 5 feet of the hoop (2.3% above NBA average) and 36.7% from 20-24 feet, which was 1.1% above the league average.
The Bucks were the most aggressive in terms of getting to the paint. Nearly 44% of their total shot attempts fell within 0-5 feet, thanks to the efforts of power forwards Greg Monroe, Miles Plumlee, and John Henson. The Kings and Grizzlies followed a similar footprint, feeding the ball to DeMarcus Cousins and Marc Gasol, respectively, in the post.
We can see that a handful of teams from cluster three and the Mavericks (in the plot above) took a high number of shots from 20-24 feet. Of these teams, only the Jazz finished with outside of the top ten teams in three-point attempts per game last season. The Trailblazers, however, were the only team in this subgroup to consistently make three-pointers at a high clip, making 37.1% of them last season.
Also within this cluster were the Rockets, which took the fewest number of mid-range jumpers (10-19 feet) last season. The Rockets attempted just 14.8% mid-range jumpers instead taking 77.4% of their shots within 0-5 feet and 20-29 feet. Similarly, the Kings and Nuggets also followed part this blueprint, attempting only 16.4 and 16.8% of their shots from mid-range ranking 28th and 29th respectively in the NBA.
We ultimately know that shot attempt frequencies will change from year to year for every team. Players switch teams or are benched for draft picks because coaches want their young players to play more and improve for the long term. These factors alter the offensive structure and inevitably shot selection patterns.
Understanding the differences in shot patterns among NBA teams has important practical consequences. First, it gives fans and organizations a snapshot into how different teams run their offenses. For example, if we clustered teams on a individual level, we would see that the Grizzlies and Thunder favored similar types of shots despite a 3.7 PACE differential while the Heat and Timberwolves had very close shot tendencies from 0-19 feet to varying levels of success. Second, NBA teams can use this information to build a loose defensive strategy against their opponents based on shot frequencies. This research overall provides a benchmark that allows fans to track if shot patterns evolve or remain unchanged across the NBA this season.
 Statistics in this article come from NBA.com, ESPN, and Basketball Reference.
 The Wizards, Spurs, Magic, Knicks, Pelicans, Timberwolves, Heat, Lakers, Clippers, Pacers, Mavericks, Bulls, and Nets make up cluster two.
 In cluster two, the Dallas Mavericks had the lowest shot attempt frequency from 15-19 feet at 14.3%.
 Cluster three consists of the Jazz, Raptors, Kings, Trail Blazers, Suns, Thunder, Bucks, Grizzlies, Rockets, Nuggets, Hawks, and Celtics.
 Over 99% of Monroe, Plumlee, and Henson’s shot attempts were two-pointers last season