By Andy Feng '24
Introduction
Twenty-five years ago, Michael Jordan led the NBA in scoring by averaging 29.6 points per game. Last season, Steph Curry led the league with 32.0 points per game. However, despite both Jordan and Curry playing the guard position, the ways in which each player achieved their respective scoring titles were vastly different: Jordan dominated with his deadly mid-range jumper and fadeaway, while Curry barraged defenders with his hot shooting from beyond the arc with over five made threes per game. This difference in shot selection isn’t simply a difference between Jordan and Curry’s game, however. In fact, the drastic increase in three-point attempts is a league-wide phenomenon: since the introduction of the three-point line to the NBA in the 1979-80 season, the three-pointer has increased in popularity nearly every year. In this paper, I aim to explore how shot selection in the NBA has changed in the last twenty-five years and how this varies among the five basketball positions.
Methods
To collect data for field goal attempts by zone, I scraped regular season and playoff data for each year from NBA.com since the 1996-97 season, which is the first year the NBA started tracking player field goal attempts by zone. I did not include the current 2021-22 season for the sake of having a complete data set for each season. The zones provided by the NBA are restricted area, paint (non-restricted area), mid-range, corner three (left and right), and above the break three. However, I combined the corner and above the break three zones into one three-point zone. Furthermore, the NBA filters out backcourt heaves, so this eliminates field goal attempts that don’t accurately reflect shot selection. Since the data provided is at the player level, I summed each player’s data to obtain season-level data for field goal attempts by zone. In addition, I combined regular season and playoff data for each season into a one data set. Furthermore, I scraped data from Basketball Reference to collect each player’s position. In some cases where a player had two positions listed, I used the first one.
Results
What once was the most popular shot in the NBA, is no longer. In twenty-five years, the mid-range jumper has fallen from the most attempted field goal to the least attempted one. In direct contrast, the three-point shot has risen from the least to most attempted field goal with nearly 40% of all shots being launched from three this past season. This increase is especially apparent in the last decade in which three-point attempts became the most popular shot, catapulting past mid-range and restricted area shots. Since the 2014-15 season, players have attempted more three-pointers than mid-range shots, and the gap between the two continues to grow every season. Meanwhile, shots in the restricted area and in the paint have stayed roughly consistent over the last twenty-five seasons.
By comparing just the 1996-97 and 2020-21 season, we see similar dramatic change. In the 1997 season, the mid-range jumper was the most common shot in the NBA with 40% of all field goal attempts from that zone, barely edging out the restricted area. Now in 2021, that number has fallen drastically to a mere 13%. Compare that to the increase of the three-pointer, which has skyrocketed from just 11% to 39%.
When examining the changes within each position, we see similar trends. Across all five positions, the mid-range shot has fallen while the three-pointer has risen. Save for the center position, all other positions are attempting the three more than all other shots. Notably, shooting guards are attempting the highest percentage of three-point field goals with nearly 50% of all their shots coming from beyond the arc. In addition, centers and power forwards seem to be the most recent neglectors of the mid-range shot in exchange for the three; only in the last half-decade did both positions start to take more threes than mid-range jumpers.
The center position is especially interesting to consider. Traditionally played by big men who dominated the inside, such as Shaquille O’Neal, Hakeem Olajuwon, and Wilt Chamberlain, modern-day centers are still taking roughly 50% of their shots from inside the restricted area, which is significantly more than shot attempts from the other zones and comparable to centers twenty-five years ago. This suggests that despite the recent trend of players opting for the three, centers have still mostly stuck to their traditional role of playing close to the basket. However, the center position is still evolving and being affected by the three-point revolution: while shots from the restricted area and paint have remained consistent for the last quarter-century, centers have decreased their mid-range shots while increasing their three-point attempts from nearly zero a game to comprising 20% of their shots just within the last decade.
Discussion
The data illustrates a clear story: over the last twenty-five years, NBA players have gradually exchanged the mid-range jumper for the three-pointer. Now, players are launching three-point field goals at a historically high rate with almost 40% of all field goals attempted from beyond the arc. Thus, it is logical why 3-and-D players are so sought after in today’s NBA: the ability to shoot the three and play defense is more valuable than ever.
But why are players shooting so many three-pointers now? Have they gotten more accurate? When comparing three-point field goal percentage from twenty-years ago to today, we see that players have not necessarily gotten much better, if at all, at shooting the three. The expected value of a field goal is simply calculated by multiplying the number of points a field goal is worth and the league-wide average field goal percentage from that zone. In 1997, the expected value of a three-point field goal was roughly 1.06 points; in 2021, the expected value of the same shot was 1.10. This marginal difference is surprising given how much three-point shooting has increased. In fact, shooting percentages across all zones have stayed very consistent for the last twenty-five years. Thus, it is interesting to see that players have only recently drastically increased their three-point field goal attempts.
Still, it is no surprise that the three-point shot is so popular given that the expected value of a three is roughly .3 points higher than that of a mid-range or paint shot. For a player deciding between the mid-range and three-point shot, this means that they either make 50% of their mid-range shots or 33% of their threes. Since the league-average from mid-range is well below 50% while the average for threes is roughly 35%, it makes sense that players are giving preference to the three. Of course, this also depends on different and changing in-game scenarios. The graph above also provides insight on why players, especially centers, are still taking many shots from inside the restricted area: it has the highest expected value since players shoot significantly more accurately closer to the basket.
Conclusion
Over the last twenty-five seasons, the NBA has undergone a major change in shot selection: the mid-range jumper has fallen to the least attempted field goal while the three-pointer has risen to the most popular. By taking player field goal attempts by zone since the 1996-97 season, I could visualize trends in total shot selection as well as for each of the five positions. In both cases, the trend of attempting more threes is consistent.
Since this paper only explores the trends in NBA shot selection from the last twenty-five years, it would be interesting to examine the applications of this. For example, if teams attempting more three-pointers are in fact winning more games, or if shooting more mid-range jumpers has a negative impact on winning. Likewise, it would be useful to analyze if this change is largely due to superstars who are inherently attempting the most field goals, or if it is due to role players settling for three-pointers.