The Decline of the Mid-Range Jump Shot in Basketball: A Study of the Impact of Data Analytics on Shooting Habits in the NBA

The Decline of the Mid-Range Jump Shot

in Basketball: A Study of the Impact of Data

Analytics on Shooting Habits in the NBA

BY Shawn Kilcoyne

ADVISOR • Son Nguyen

EDITORIAL REVIEWER • Judith McDonnell

_________________________________________________________________________________________

Submitted in partial fulfillment of the requirements for graduation

with honors in the Bryant University Honors Program

NOVEMBER 2020

Table of Contents

Abstract ..................................................................................................................................... 1

Introduction ............................................................................................................................... 2

Analytics in the NBA ............................................................................................................ 4

The Evolution of the Three-Point Shot ................................................................................. 5

The Decline of the Mid-Range Shot ..................................................................................... 7

Literature Review ...................................................................................................................... 9

Utilizing Analytics to Evaluate Individual Player Performance ........................................... 9

Utilizing Analytics to Evaluate Shot Selection ................................................................... 12

Utilizing Analytics to Evaluate Shot Quality by Location.................................................. 14

Data Analysis .......................................................................................................................... 17

Data ..................................................................................................................................... 17

Methodology ....................................................................................................................... 18

Visualization Findings ........................................................................................................ 19

Linear Regression Model Findings ..................................................................................... 25

Discussion ............................................................................................................................... 26

Conclusion .............................................................................................................................. 29

Appendices .............................................................................................................................. 30

Appendix A – NBA Database Snippet – 2005-06 Season .................................................. 30

Appendix B – Visual Illustration of 5 Primary Shot Locations .......................................... 31

Appendix C – Simplified NBA Database Snippet – 2005-06 Season ................................ 32

Appendix D – Pace Correlations ......................................................................................... 33

Works Cited ............................................................................................................................ 34

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 1 -

ABSTRACT

The purpose of this thesis paper is to investigate the strategic shift away from the mid-range

jump shot in basketball over the past decade. This paper will cover the rationale for the

decline of the mid-range, as well as the general impact of data analytics on the way the game

of basketball is played at the professional level. Following a review of the existing literature

relating to the use of analytics in the NBA, this paper will analyze the differences in shooting

habits between two seven-season periods. Data visualization tools, including boxplots,

statistical trends, and distribution plots, will be used to illustrate the changes in shooting

habits from the 2005-06 season through the 2018-19 season. Additionally, a predictive

statistical model will be used to identify the variables that are most important to winning in

the NBA, including shot locations, defensive rating, and pace of play.

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 2 -

INTRODUCTION

The game of basketball is constantly evolving. Data analytics has overhauled the way players

and teams are evaluated, focusing more on statistics and quantitative measures rather than

evaluating players purely on subjective opinion. The impact of data analytics is most

profound on the shooting habits of both individual players and teams. In the NBA, there has

been a drastic shift in shooting habits over the past decade as a result of the three-point

revolution. The way the game is played in the NBA today is completely different than the way

it was played in the early 2000s. The main difference? The absence of the mid-range shot.

The “mid-range” is the area of the court between the three-point line and the area immediately

surrounding the basket, commonly referred to as “the paint”. The charts above depict the shift

in shot locations over the past 20 years in the National Basketball Association (NBA). The

chart on the left shows the 200 most frequent shot locations over the course of the 2001-02

NBA season, while the chart on the right depicts the top 200 shot locations for the 2019-20

NBA season. The mid-range jump shot is now a rarity in the NBA, and league offensive

strategies have changed accordingly. Nowadays, players are primarily encouraged to do one

of two things on offense: drive all the way to the basket or shoot a three-pointer. This has

resulted in the death of the mid-range jump shot.

Figure 1: Shot Locations 2001-02 vs 2019-20 (Goldsberry)

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 3 -

This phenomenon is not only evident in the strategies of teams. The shot makeup of a current-

day NBA superstar greatly differs from the superstars of the past. For example, the jump shot

habits of Michael Jordan (1996-1998) and James Harden (2018-2020) are essentially inverted,

despite each of them being arguably the greatest scorer of their era.

This paper will explore the rationale behind the decline in the mid-range jump shot,

explaining the inefficiency of the shot using NBA data. The paper will provide a brief history

of the use of data analytics in the NBA and how it led to a shift in league-wide shooting

habits. Using datasets of two seven-year periods, shooting trends will be analyzed using

statistical visualization tools, including boxplots, statistical trends, and distribution plots.

Additionally, a statistical model will be used to analyze the impact of numerous variables on

winning in the NBA, including shooting locations, defensive rating, and pace.

Figure 2: Jordan vs Harden Jump Shot Locations (Goldsberry)

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 4 -

Analytics in the NBA

The National Basketball Association has embraced analytics unlike any other professional

sports league. Whether it is the tracking of player movement, analysis of specific shot

locations on the court, or lineup analysis to identify a team’s most effective combination of

players, analytics are a major driver behind almost every basketball-related decision an NBA

organization makes.

Matt McLaughlin captures the rise of data analytics in the NBA in his BizTech article How

Data Analytics Is Revolutionizing Sports: “The NBA has embraced data analytics in a way

that surpasses most other major U.S. sports leagues… Nearly every team in the NBA has

hired data analysts as full-time staff members to work with coaches and front office staff.

These analysts help teams identify trends that may improve” (McLaughlin 1). Data analytics

is crucial for success in today’s NBA, and teams such as the Golden State Warriors and

Houston Rockets have taken the lead in the NBA’s recent data analytics revolution.

NBA Commissioner Adam Silver says it himself: “Analytics are part and parcel of virtually

everything we do now… I think it is part of the result” (Wharton School). No league in the

world has embraced the recent data analytics movement like the NBA. From analyzing ticket

sales in order to increase team revenue, to using wearables in order to monitor players’ sleep

schedules, the NBA has incorporated analytics into nearly every aspect of its product.

NBA front offices, coaches, and fans have always used statistics to supplement their

evaluation of players. That said, until the mid-2010s, those statistics were predominantly basic

box score numbers: points, rebounds, assists, steals, blocks, field goal percentage, etc. One of

the major breakthroughs that helped start the current data analytics revolution is the league’s

investment in video tracking technology. At the start of the 2013-2014 season, the NBA

implemented SportsVu software into every NBA arena, a brand specializing in video tracking

tools (Mudric 1). Since the initial investment into video tracking, the NBA has only expanded

its partnerships with data-focused companies, including SecondSpectrum and Sportradar. The

league’s six-year partnership with Second Spectrum and Sportradar are “worth in excess of

$250 million to the NBA” according to Forbes (Heitner 1).

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 5 -

The leading pioneer for the NBA’s data revolution is former Houston Rocket’s General

Manager Daryl Morey. Coming from an analytical background, Morey was hired as GM of

the Rockets in 2007. Following three consecutive losing seasons, Morey upheaved the

Rocket’s offensive strategy prior to the 2012-2013 season, installing a system centered around

taking highly efficient shot attempts, predominantly three-pointers and layups. Morey also

traded for budding star James Harden and reconstructed the Rocket’s roster to complement his

star guard. In his development of the Rocket’s groundbreaking strategy, Morey took

advantage a simple reality: three is more than two. With the rationale that the 50% increase in

points for a three-pointer outweighs the increased difficulty of a three-point attempt, nearly all

jump shots that the Houston Rockets attempted moving forward have come from behind the

three-point line. Morey’s Rockets are the first example of a team abandoning the mid-range

jumper, the main topic that this paper focuses on. Houston’s strategy has resulted in sustained

success; to date, the team has the third-best record in the NBA since 2012, with only the

Golden State Warriors and San Antonio Spurs posting more wins. The Rocket’s strategy has

been dubbed Moreyball, a reference to the Oakland Athletics infamous Moneyball strategy.

The Evolution of the Three-Point Shot

Out of all the strategic adjustments that have been made since the beginning of the analytics

revolution, the increasing reliance on the three-point shot is the most staggering. In the

modern NBA, teams such as the Houston Rockets and Golden State Warriors structure their

entire offensive attack around the three-point shot. The current reality of the NBA would have

been unimaginable 40 years ago, when the three-point line was first introduced.

The use of the three-point shot increased only marginally over the first decade and a half of its

existence following its introduction prior to the 1979-1980 season. The league average for

attempts per game remained under 10 attempts per game through the 1993-1994 season. The

early 1990s was also a time of decreased scoring across the league. For this reason, the NBA

decided to move the three-point line closer to the basket, from 23 feet, 9 inches and 22 feet at

the corners to a uniform 22 feet around the arch, in an attempt to increase scoring across the

league. The first season after this change, the 1994-1995 season, was subsequently the first

major jump in three-point attempts in league history, to 15.3 attempts.

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 6 -

As it turns out, the change in the three-point line did not stick, as the move did not result in

the increased scoring that the league anticipated:

Though the average number of 3-point attempts per game increased by over 50

percent, the line was moved back to the original distance after the 1996-97 season

because the change had actually lowered the average score of games. In the three

seasons leading up to the new rule, teams averaged 105.6 points per game. In the three

seasons with the shorter 3-point line, that average fell to 100.8. (Pierson 1)

Although the league moved the three-point line back, this three-year experiment did have

some lasting influence on team shot selection. Coaches had begun to incorporate the three-

point shot into their offensive strategies, while players who specifically excelled in three-point

shooting, dubbed three-point specialists, became more commonplace across the league.

Following the move of the three-point line back to its original distance, three-point attempts

across the league increased slightly each year from 1997 through 2008, but then leveled out at

roughly 18 attempts per team per game for each season through 2012. The 2012-2013 season

represents the real beginning of the three-point revolution, as it was the first season in NBA

history in which the average number of three-point attempts across the league eclipsed 20

attempts per game. That figure rose exponentially over the next 7 seasons, and during the

2018-2019 season, teams attempted 32 three-pointers per game on average (Basketball

Reference).

If one player can be credited with spurring this three-point

barrage, it is Golden State Warriors’ superstar Stephen Curry.

Drafted out of Davidson College in 2009, Curry is largely

considered the greatest three-point shooter of all time. At the

time of this writing, he ranks 6

in NBA history in three-point

percentage, despite the absurd level of difficulty of many of his

attempts. Curry has broken the record for most three-pointers in

a single season three times, setting the current record of 402 in

the 2015-2016 season. Prior to that season, no player had even

eclipsed 300 three-pointers in a season. The adjacent shot chart

Figure 3: Curry Jump Shot Locations

(Goldsberry)

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 7 -

depicts Curry’s jump shot activity during his 2015-2016 MVP campaign, which features

almost exclusively three-point attempts (Goldsberry, 2020).

Curry’s impact on the game of basketball is historic. As the Ringer’s Kevin O’Conner

highlighted in his 2019 article It’s More Than Just the Shot, Curry’s influence on younger

generations is the greatest since Michael Jordan, arguably the best player of all time:

Kids used to lower their hoops or use trampolines to dunk like Michael Jordan. Now

they shoot from deep like Stephen Curry. Whether it’s good or bad for the game is

moot; it’s happening no matter what. Players of all shapes and sizes are entering the

league with shooting skill. This season, 20 teams attempt over one-third of their shots

from 3. (O’Conner 1)

Curry’s shooting prowess has helped him capture two NBA MVP awards, as well as three

NBA championship rings, while inspiring an entire generation and effectively changing the

game.

Curry’s coach, Steve Kerr, who was hired by the Warriors in 2014, ushered in a system of fast

pace and three-point shooting that maximized Curry’s skillset. In fact, Kerr, a former player

himself, shot 45.4% from behind the three-point line over the course of his career, the best

percentage of any player in NBA history. Kerr, who won 5 championships as a player, has

coached the Warriors to three championships in the 2010s behind the shooting of Curry. The

Warriors broke the typical prototype of a championship team; the team played at a fast pace,

shot 30+ three-pointers per game, and utilized a lineup of undersized players. Golden State

proved that the three-point shot is not a gimmick, but instead an efficient tool that, if utilized

correctly, can be the driving force behind a dynasty.

The Decline of the Mid-Range Shot

Given the dramatic increase in three-point shots attempted, shots from other locations on the

court have declined as a result. The primary victim of the rise of the three-pointer has been the

mid-range shot. The mid-range is the area of the court between the paint and the three-point

line; the shot is too close to the basket to be a three, but is far enough way that it is not

considered a layup or floater. As the graph to the right indicates, since 2003, teams have

steadily moved away from the mid-range jump shot in favor of taking more 3s. As one of the

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 8 -

leading teams behind the three-point revolution, the Warriors 2015 championship coincides

with when the three-point shot overtook the mid-range shot in terms of percentage of league

shot attempts.

The decline of the mid-range is rooted in analytical research. As Wade McCagh synopsizes in

his article How Spatial Analytics Killed The Mid-Range Jump Shot: “It took the most

advanced spatial tracking technology we've seen in sports to reveal a simple truth: 3 points are

more than 2” (McCagh 1). The decline of the mid-range is truly that simple; if a player gets

three points for making a 23.8 foot jump shot, but only two points for a 23.7 foot jump shot

(assuming they are shooting from straightaway), the three-point shot is undeniably the more

efficient shot.

To put it another way, if a player manages to shoot 50% from mid-range in a season, they are

considered an extremely good shooter. Despite that, over the course of that season, that

shooter will only average 1 point per shot from the mid-range. If this shooter were to aim to

score 1 point per shot while attempting three-pointers, they would only have to make 33.33%

of their shots to reach this level of efficiency. The difference between these percentages is

huge; if the shooter can shoot within 16.67% of their mid-range percentage on three-pointers,

they will be a more efficient scorer of the basketball. The additional point for a three-point

shot more than compensates for the increased difficulty of the deeper shot attempt, therefore

Figure 4: The Rise of the Three & Decline of the Mid-Range

(FlowingData)

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 9 -

teams have almost completely abandoned the mid-range attempt as part of their offensive

strategy.

All in all, there has been a staggering shift in strategy in basketball over the past few decades,

and it is rooted in data analytics. Statistical analysis has helped reveal the inefficiency of the

mid-range jump shot and has spurred the three-point revolution that has overtaken basketball

at all levels.

LITERATURE REVIEW

With the rise of data analytics and the three-point revolution in the NBA, a multitude of

scholarly research has been conducted examining player and team performance, as well as

shot selection and spatial analysis. This section of the paper will examine these past studies to

provide an overview of the prior research completed in this area of study.

Utilizing Analytics to Evaluate Individual Player Performance

The first area in which several studies have been completed is using data analysis to evaluate

individual player performance. Points, assists, rebounds, and other basic counting stats are the

typical metrics used to evaluate players, yet they do not encapsulate all the potential impact,

positive or negative, a player can have on the court. For this reason, researchers have

developed different, more advanced metrics for evaluating individual performance that

consider more than just basic statistics.

A 2011 study, Evaluating Basketball Player Performance via Statistical Network Modeling,

introduces new metrics to evaluate player performance that considers the interaction effects

by teammates. The study was presented at the 2011 MIT Sloan Sports Conference by James

Piette, Lisa Pham, and Sathyanarayan Anand. The goal of the study is to evaluate how

effective certain combinations of players are while measuring the contributions of each

individual player using statistics. The study acknowledges that the commonly used plus-

minus statistic inherently has biases, and the metric the researchers use attempts to correct for

this.

The study evaluates players based on three areas, offense, defense, and total efficiency, and

computes player centrality scores to measure the importance of each player to their respective

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 10 -

teams. Using those scores, the researchers identified several players that overperform on

offense, but underperform on defense, as well as those who underperform on offense, but

overperform on offense. One of the advantages to the study is that it identifies players who are

currently under-utilized by their teams; they overperform in one or multiple areas of the game,

yet their centrality scores are low. Generally, the study contributes an algorithm to the

growing body of metrics used to evaluate individual player performance, taking the

contribution of teammates into account. (Piette, et al. 1)

In another study, Cervone et al. (2014) incorporate other factors into individual player

evaluation. While most analytical models focus on the results of the end of a possession, such

as points, turnovers, or assists, this model considers players’ decisions to pass, dribble, or

shoot over the course of the possession. By using the metric of expected possession value

(EPV), this source offers a different type of analysis of offensive play in basketball.

The graphic above visually conceptualizes EPV. At this given point in the San Antonio Spurs’

possession, Kawhi Leonard, the player with the ball, must decide whether to shoot or pass,

and if he passes, which player to pass to. The EPV metric evaluates the probable points that

Figure 5: EPV of Spur's Possession (Cervone et al.)

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 11 -

result from each of his options, highlighting that the smartest play is to pass the ball to Danny

Green for a corner 3. Typical statistical analysis would only acknowledge Leonard’s decision

if he either scored or assisted a teammate’s made shot. The EPV metric instead evaluates

Leonard’s decision making; if he passes the ball to Danny Green, regardless of whether or not

Green makes the shot, Leonard is credited with making the best play. This study, and the

introduction of EPV as a metric, further contributes to the tools used to evaluate individual

performance, and “overcomes many shortcomings of the conventional boxscore approaches to

analyzing the game” (Cervone, et al. 1).

In evaluating player performance in basketball, most studies and metrics focus on a player’s

contributions on the offensive side of the game. The Dwight Effect: A New Ensemble of

Interior Defense Analytics for the NBA, a 2013 study by Kirk Goldsberry and Eric Weiss,

instead introduces spatial and visual analytics that evaluate multiple aspects of defensive play.

The study investigates “The Dwight Effect”, named after dominant defender Dwight Howard,

which is defined as “the ability of an interior defender to reduce the efficiency of an

opponent’s shooting behavior” (Goldsberry & Weiss 3). Common statistics, such as steals and

block shots, do not always accurately correlate to a player’s defensive effectiveness; this

study accounts for both the efficiency and frequency of shot attempts of opposing players.

The incorporation of frequency as a key variable

is relatively unprecedented, as it attempts to

account for the fear factor of great defenders.

Some interior defenders, such as Dwight Howard,

have a presence that deters opposing players from

attempting shots around them, which the

frequency variable attempts to account for.

The study seeks to accomplish two goals: to present new metrics to evaluate defensive

effectiveness in the NBA, and to acknowledge the challenges of measuring defensive

performance. By comparing the efficiency and frequencies of opposing offenses when a

certain player defends the interior, the researchers were able to identify the best, as well as the

worst, interior defenders in the NBA over the course of the 2011-2012 and 2012-2013

seasons.

Figure 6: Opponent Field Goal Percentage by Individual Defender

(Goldsberry & Weiss)

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 12 -

Utilizing Analytics to Evaluate Shot Selection

Another major area in which data analytics has had a distinct impact is shot selection. Recent

studies have used advanced metrics to evaluate the efficiency of shots from certain positions

on the court and have brought about increased emphasis on the importance of shot selection in

basketball. Other studies have focused on spatial analysis, which takes into account both the

location of shots taken and the distance of defenders.

A 2012 study, The Problem of Shot Selection in Basketball, by Brian Skinner, offers

interesting insight into shot selection in the NBA. Skinner investigates the decision making of

players, particularly their decision of whether to take a shot or pass the ball. The analysis

explores what shots are worth taking and which shots should be passed up and takes

extraneous factors such as time left on the shot clock and probability of a turnover into the

analysis. In the NBA, teams have 24 seconds to attempt a shot and if the team does not

attempt a shot within the shot clock, it expires and the team loses possession of the ball.

Skinner uses a theoretical model of the shot selection process in hopes of answering the

question “how likely must the shot be to go in before the player

should take it?” (Skinner 1).

As Skinner finds, the shot clock has a large impact on shooting

habits. As the charts to the right depict, as the shot clock gets

closer to expiring, teams attempt shots at a higher rate (out of

necessity), but the quality of those shots decreases. For the

games in his dataset, Skinner’s analysis finds that: “For NBA

teams, the expected number of points per possession is 0.86, or

0.83 if one considers only possessions lasting past the first

seven seconds of the shot clock” (Skinner, 6). If teams were to

employ the optimal shooting strategy identified by Skinner’s model, that expected number of

points per possession would rise to 0.91, or 0.88 if one only considers possessions that last

past the first seven seconds of the shot clock. The rise from 0.86 to 0.91 is hugely impactful:

“This improvement of 0:05 points/possession translates to roughly 4.5 points per game…

such an improvement could be expected to produce more than 10 additional wins for a team

Figure 7: Shooting Rate by Shot

Clock Time (Skinner)

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 13 -

during an 82-game season” (Skinner 6). In short, Skinner’s study highlights the importance of

shot selection in basketball and how it can have a profound impact on winning.

Another study, Quantifying Shot Quality in the NBA, evaluates shot quality in a different way.

This Second Spectrum study introduces two new metrics in effective shot quality (ESQ) and

EFG+, effective field goal percentage minus effective shot quality, which computes shooting

ability above expectation. As the study highlights, the current advanced metric to evaluate

shooting efficiency is effective field goal percentage (EFG%), which accounts for the reality

that three-pointers are worth 1 more point than two-point shots. The study introduces ESQ

and EFG+ to offset the fact that EFG% confounds the quality of any shot attempted and the

ability of a player to make that shot.

The charts above depict EFG% for different locations on the court. The mid-range area is by

far the least efficient area on the court according to the effective field goal percentage metric.

The issue with EFG% is that it does not account for shot quality. If Player A attempts

significantly harder shots than Player B, and as a result Player B makes more shots than

Player A, Player B will have the higher EFG%. That is not necessarily a fair evaluation of

Player A, because if Player B was able to take the same quality shots that Player B attempts,

Player A may very well be a better shooter.

Figure 8: ESQ and EFG+ (Chang et al.)

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 14 -

This is where ESQ and EFG+ come in. The following graphic depicts ESQ and EFG+ applied

to NBA players:

Figure 9: Player Specific ESQ and EFG+ (Chang et al.)

This graphic proves that some players must take much more difficult shots than others. For

three-pointers for example, Spencer Hawes has an effective field goal percentage of 68.7%,

while Kevin Durant has an EFG% of 62.1%. That said, Kevin Durant, a former NBA league

MVP, faces much tougher defense every game than Spencer Hawes, a career role player. The

ESQ metric takes this into account and shows that Hawes takes significantly easier three-

pointers than Durant (Chang et al. 6). All in all, this study introduces two new metrics useful

for evaluating shot quality and efficiency at the NBA level.

Utilizing Analytics to Evaluate Shot Quality by Location

Another study by Kirk Goldsberry, CourtVision: New Visual and Spatial Analytics for the

NBA, introduces a “new ensemble of analytical techniques designed to quantify, visualize, and

communicate spatial aspects of NBA performance with unprecedented precision and clarity”

(Goldsberry, 2012). In this study, Goldsberry also introduces new metrics in an attempt to

make up for the oversimplifying nature of commonly used statistics. Goldsberry’s

CourtVision metrics account for spatiality, clarifying that “although conventional metrics are

simple ways to summarize the probability of a shot attempt resulting in a made basket, they

fail to expose true differences in shooting ability across the league” (Goldsberry, 2012). In

this particular study, Goldsberry uses the CourtVision model to identify the best shooters in

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 15 -

the NBA during the 2010-2011 NBA season, pinpointing Ray Allen and Steve Nash as the

league’s best shooters.

Goldsberry also contributed to Andrew Miller, Luke Bornn, and Ryan Adams’ research on

spatial analysis of professional basketball. In that study, the researchers conducted a spatial

pattern analysis to identify to the best

shooters from different locations on

the court. Based on the results, Miller

et al. discovered great variety in shot

selection, as well as effectiveness on

those shots, amongst NBA players. In

summary, “some shooters specialize in certain types of shots, whereas others will shoot from

many locations on the court” (Miller et. al 5).

Another study, Live by the Three, Die by the Three? The Price of Risk in the NBA by Matthew

Goldman and Justin M. Rao, looks into how a team determines the right proportion of two-

point and three-point shots to take over the course of a game. In determining the best

combination of shot attempts, a trade-off exists between risk and reward. Goldman and Rao

highlight that early in a game, teams should maximize points per possession, regardless of

risk, and therefore should attempt more three-pointers. That said, late in a game, the winning

team should pursue low risk, “predictable scoring opportunities”, while the trailing team must

shoot more threes in order to tighten the score (Goldman & Rao 2).

One of the main takeaways from this study is that the leading team takes more three-pointers

when the best play would be to take more high percentage shots. This trend continues as the

winning team’s lead decreases: “As a lead decreases, the leading team should become more

risk-neutral, but teams in this circumstance actually tighten up and become more risk averse,

contrary to what their risk preferences ought to be to maximize the chance of winning the

game” (Goldman & Rao 8). This is an important finding; the statistics show that teams that

are trying to solidify a win do not shoot the shots that would give the team the best chance of

winning. Further research needs to be done to see if the mid-range jump shot could be an

effective shot for the winning team late in games.

Figure 10: Factorized Point Process (Goldsberry, Miller, Bornn, Adams)

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 16 -

Paul Zuccolotto, Marica Manisera, and Marco Sandri also conducted a study on shooting

performance of players in high-pressure game situations. Using data from the Italian Serie A2

Championship, the researchers identified that the situation “most impacting the scoring

probability is when the shot clock is about to expire and, for free throws, when the player has

missed the previous shot” (Zuccolotta et al. 587). The study investigates players’ personal

reactions in the last two seconds before the shot clock expires, as well as when the score is

within a reasonable range within the last five minutes of a game. The findings of this study

contribute to the growing body of analytics research related to basketball that are increasingly

used to measure and evaluate player and team performance.

The existing body of research highlights the impact the rise of data analytics has had on

player evaluation in the modern NBA. While the current research covers how analytics are

used to evaluate individual player performance, shot selection, and shot quality by location,

the research focusing specifically on the mid-range is not extensive. This paper aims to add to

the existing body of research while identifying the key contributing factors to the decline of

the mid-range jump shot in the modern NBA.

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 17 -

DATA ANALYSIS

Data:

The data analysis in this paper is based on a dataset of statistics gathered from fourteen

seasons worth of NBA data. The larger dataset is split into two equally sized data sets of

seven seasons. The first dataset is made up of NBA statistics from the 2005-2006 season

through the 2011-12 season. The second dataset is made up of NBA statistics from the 2012-

2013 season through the 2018-2019 season. As identified in the literature review, the 2012-

2013 NBA season represents the beginning of the three-point revolution, making it a logical

split for evaluating the shift in shooting habits in the NBA.

The data was retrieved from NBA.com/statistics. Microsoft Excel was used to organize the

data. See Appendix A for a snippet of the dataset. Most of the values in the dataset are based

on season totals, such as the total amount of shot attempts from designated areas on the court

over the course of a full season. For this study, shot attempts are specified as coming from one

of five areas on the court: the restricted area, the paint, the mid-range, corner three-pointers,

and above the break three-pointers; see Appendix B for a visual illustration of these shot

locations.

It is important to note that there are a few outlier data points that deviate from the typical 82

game NBA season. The 2011-2012 NBA season was only 66 games due to the lockout

shortened season following disputes between players and owners regarding the league’s

collective bargaining agreement. All data points for the 2011-2012 season are based on 66

games, therefore the 2011-2012 season represents an outlier datapoint for some charts used in

the analysis. Also, the Boston Celtics and Indiana Pacers only played 81 games each in the

2012-2013 season due to the tragic events of the Boston Marathon bombing leading to the

teams’ April 15, 2013 game being cancelled.

The dataset includes the following 30 variables:

Variable:

Description:

Year

NBA Season

Games Played

RAFGM

Restricted Area Field Goals Made

RAFGA

Restricted Area Field Goals Attempted

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 18 -

PFGM

Paint Field Goals Made (Non-Restricted Area)

PFGA

Paint Field Goals Attempted (Non-Restricted Area)

MRFGM

Mid-Range Field Goals Made

MRFGA

Mid-Range Field Goals Attempted

C3FGM

Corner Three-Point Field Goals Made

C3FGA

Corner Three-Point Field Goals Attempted

ATB3FGM

Above the Break Three-Point Field Goals Made

ATB3FGA

Above the Break Three-Point Field Goals Attempted

FTM

Free Throws Made

FTA

Free Throws Attempted

Personal Fouls

PFD

Personal Fouls Drawn

OFFRTG

Offensive Rating

DEFRTG

Defensive Rating

AST%

Assist Percentage

AST/TO

Assist to Turnover Ratio

PACE

Pace of Play

OREB

Offensive Rebounds

DREB

Defensive Rebounds

TREB

Total Rebounds

AST

Assists

TOV

Turnovers

STL

Steals

BLK

Blocks

Wins

Games Won

Rank

League Standing (based on Wins)

Methodology:

The two datasets are analyzed by two primary methods. First, visualization tools will be used

to illustrate the statistic trends in shooting habits over the past 14 years. Graphs highlighting

the trend over the years, boxplots based on variables such as offensive strength and league

rank, and distribution plots based on league rank are used to visualize the data. Secondly, a

linear regression model will be used to measure the effect selected variables have on an NBA

team’s chance of winning. Python is the primary software with which this analysis was

completed.

For the visualization tools, each type of chart displays different measures of the data. The

statistics trending graph simply showcases the totals for a selected variable each year. When

graphed together, the trend of the data can be seen visually. Boxplots are used to identify the

interquartile range, which is the 25

to the 75

percentile of the data. The middle line

highlights the median of the selected data set. Scatterplots are used to observe the relationship

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 19 -

between two variables; the dots represent individual datapoints, while the accompanying line

graphs highlight the strength of the relationship between the two variables. The distribution

plots identify the most common count of the data, as well as the shape of the sample

distribution.

The linear regression model is used to predict the number of wins a team will have based on

selected variables from the dataset. The model is computed using ten variables, as opposed to

the 30 variables that influence the visualization charts. The use of fewer variables simplifies

the model and gives a general idea of the variables that impact winning. See Appendix C for a

snippet of the simplified database. The ten variables are:

Variable:

Description:

RAFGA

Restricted Area Field Goals Attempted

PFGA

Paint Field Goals Attempted (Non Restricted Area)

MRFGA

Mid-Range Field Goals Attempted

C3FGA

Corner Three-Point Field Goals Attempted

ATB3FGA

Above the Break Three-Point Field Goals Attempted

DEFRTG

Defensive Rating

PACE

Pace of Play

TREB

Total Rebounds

AST

Assists

TOV

Turnovers

Visualization Findings:

The initial findings focus on the change in shooting distributions for teams over the 14-year

period analyzed. A statistics trending analysis was completed for the field goals attempted by

teams on average over the course of each season from the 2005-06 season through the 2018-

19 season. The number of field goal attempts from each of the major areas on the court was

analyzed: the restricted area, the paint outside the restricted area, the mid-range, the corner

three pointer, and the above the break three pointer. Again, it is crucial to note that only 66

games were played in the 2011-12 season, so this represents an outlier data point on each

graph.

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 20 -

Over the period analyzed, there is a slight

positive trend in number of restricted area field

goals attempted. In the 2005-06 season, teams

averaged 2100.6 restricted area field goal

attempts over the course of the season, compared

to 2414.2 attempts in the 2018-19 season, a

313.7 shot increase. That change represents a

14.9% increase from the restricted area attempts

in 2005-06.

There is a clear positive trend in the number of

paint field goal attempts (non-restricted area). In

the 2005-06 season, teams averaged 836.9 paint

field goal attempts, compared to 1170.7 attempts

in the 2018-19 season, a 333.8 shot increase.

That change represents a 39.9% increase from

the paint attempts in 2005-06.

There is a clear negative trend in the number of

mid-range jump shot attempts. In the 2005-06

season, teams averaged 2230.9 mid-range

attempts over the course of the season, compared

to 1109.2 attempts in the 2018-19 season, a

1121.7 shot decrease. That change represents a

50.3% decrease from the mid-range field goal

attempts in 2005-06.

Figure 11: Restricted Area Field Goal Attempts by Year

Figure 12: Paint Field Goal Attempts by Year

Figure 13: Mid-Range Field Goal Attempts by Year

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 21 -

There is a clear positive trend in corner three-

point attempts. In the 2005-06 season, teams

averaged 369.2 corner three-point attempts,

compared to 596.1 in the 2018-19 season, a

226.9 shot increase. This change represents a

61.4% increase in corner three-point attempts

since the 2005-06 season.

There is a very clear positive trend in above the

break three-point attempts. In the 2005-06

season, teams attempted 927.2 above the break

three-point shots, compared to 2009.5 attempts

in the 2018-19 season, a 1082.3 shot increase.

This change represents a 116.7% increase from

the average number of above the break three-

point attempts in 2005-06. Furthermore, the

average number of attempts in the 2012-13

season was 1170.0, a 242.8 shot increase from 2005-06. The average number of attempts then

increased by 839.5 attempts from 2012-13 to 2018-19.

The next stage of the analysis focuses on looking at how these shooting habits differ based on

the strength of a team’s offense, as well as how these shooting habits translate to winning.

Boxplots and distribution plots were used in this part of the analysis. For many of the

boxplots, the data is dispersed based on a team’s offense being neutral, weak, or strong; in this

analysis, offense strength is quantified using Offensive Rating, a metric which measures a

team’s offensive points produced per 100 possessions. For the boxplots that deal with rank,

the rank is based on where a team finished in terms of number of wins relative to all 30 teams.

Figure 14: Corner Three-Point Field Goal Attempts by Year

Figure 15: Above the Break Three-Point Field Goal Attempts by Year

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 22 -

As the boxplots to the left depict, the frequency

of mid-range jump shot attempts clearly

decreased in the 2012-2019 data compared to

the 2005-2012 data. Interestingly, there was no

major difference in mid-range field goals

attempted for teams with strong or weak

offenses. On the other hand, there are clear

differences in the offensive strength of teams in

the 2012-2019 dataset based on mid-range field

goal attempts. Strong offensive teams attempted

the least amount of mid-range shots, while the

weakest offenses utilized the highest number of

mid-range attempts.

The use of the corner three-point shot has

changed over the past 15 years. As depicted by

the boxplots to the left, teams in the 2012-2019

dataset attempted more threes across the board

than the 2005-2012 teams, regardless of the

strength of a team’s offense. That said, the

strongest offenses attempted the most corner

three-pointers on average, although it was close

to the level of league average (neutral) offenses.

The weak offenses lagged farther behind, clearly

attempting the least corner three-point shots over

the course of a season.

Figure 18: Corner Three-Point Attempts by Strength of Offense

Figure 19: Corner Three-Point Attempts by Team Rank

Figure 17: Mid-Range Attempts by Team Rank

Figure 16: Mid-Range Attempts by Strength of Offense

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 23 -

The frequency of corner three-point shooting

directly correlates with winning. As the rank

boxplot and distribution plot show, regardless of

the dataset, top ranked teams utilized the corner

three-pointer much more than lower ranked

teams. The corner three-point shot is a staple of

winning teams.

The use of the above the break three-pointer has

changed. Teams with strong offenses attempt the

most above the break three pointers, while weak

offenses shoot the least amount. The number of

above the break three-point attempts is

significantly higher in the 2012-2019 dataset

than the 2005-2012, regardless of offensive

strength. Although it is not a large difference,

teams that attempt more above the break three-

pointers rank higher in league standings on

average.

In looking for variables that help explain the shift in shooting habits in the NBA over the 14-

year period analyzed, a surprising variable presented itself as a factor: pace. Pace, a metric

that simply measures the number of possessions a team has over the course of a game, is

typically used to measure how often a team gets out in transition and pushes the ball down

court. After looking at the pace variable in comparison with the shooting statistics in the

datasets, specifically the mid-range field goal attempts, it seems that there is a relationship

between pace of play and a team’s shooting habits.

Figure 20: Corner Three-Point Attempts Distribution Plot by Rank

Figure 21: Above the Break Three-Point Attempts by Strength of Offense

Figure 22: Above the Break 3-Point Attempts by Team Rank

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 24 -

First, there seems to be a relationship between mid-range jump shooting and team pace of

play. For the 2005-2012 data, the teams are heavily concentrated in the 1500-2500 MRFGA,

89.0-95.0 PACE area. On the other hand, the 2012-2019 teams are concentrated in the 1000-

2250 MRFGA, 92.5-100.0 PACE area. Teams in the 2012-2019 data set are evidently

shooting less mid-range jump shots and playing at a faster pace of play.

After the realization that pace of play is a variable that may influence a team’s shooting

habits, the relationship between pace and the strength of team’s offense and defense,

respectively, was looked at. As the boxplots below show, teams that have a strong offense

play at the highest pace on average. Also, teams with a weak defense seem to play at the

highest pace, while strong defenses play at a slower pace. Across the board, teams in the

2012-2019 data play at a higher pace than teams from the 2005-2012 dataset. That said, pace

itself does not necessarily relate to winning. As the distribution plot depicts, pace increased

across the league from the 2005-2012 period to the 2012-2019 period, but an even distribution

of top and bottom ranked teams played at all different levels of pace.

Figure 25: Pace by Strength of Defense

Figure 23: Correlation Between Mid-Range Field Goal Attempts and Pace

Figure 24: Pace by Strength of Offense

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 25 -

Linear Regression Model Findings:

The second focus of the data analysis is to develop a model that predicts winning. A linear

regression model using the nine variables identified in the methodology section produced the

following results:

Figure 28: Output for 2012-2019 Data

The coefficient measures the impact of the variable on the expected number of wins in a

season. For example, a one-point increase in defensive rating (DEFRTG) would result in

1.174 less expected wins in a season for the 2005-2012 data. It is important to note that

Figure 26: Pace Distribution Plot

Figure 27: Output for 2005-2012 Data

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 26 -

defensive rating measures the number of points a team allows over 100 possessions. By this

logic, a one-point increase in defensive rating signifies a weaker defensive team.

The two variables that impact winning the most, based on the model, are defensive rating and

pace. These variables are measured differently than the other seven, as defensive rating and

pace are measured on a possessions basis, as opposed to season totals. For this reason, it

makes sense that an increase in one of these variables would have a more profound impact on

winning than the same increase in one of the season-total based variables. Defensive rating

had about the same impact on winning for both datasets. Pace had a stronger impact on

winning for teams in the 2005-2012 dataset than teams in the 2012-2019 dataset.

Turnovers had the strongest impact on winning out of the season-total variables, with the

impact being negative for both datasets. Mid-range field goal attempts had a nearly identical

negative impact on winning for both datasets. Corner three-point attempts had a negative

impact on winning for the 2005-2012 dataset, and a positive impact on winning for the 2012-

2019 dataset.

DISCUSSION

The initial trending analysis highlights the major focus of this paper. Shot attempts from the

mid-range have consistently declined across the NBA over the past 14 years, with a

significant plummet in mid-range shot attempts from the 2012-2013 season on. This analysis

confirms the statistical decline of the mid-range, as well as the types of shot attempts that

have replaced those mid-range shots. Above the break and corner three-pointers have both

increased exponentially since 2012. On top of that, paint attempts outside the restricted area,

which are typically floaters and short pull-ups, have increased over this time period, with a

significant spike between 2016 and 2019. Restricted area field goal attempts, which include

essentially only layups and dunks, have also increased marginally over the time period.

These trends are indicative of the league-wide change in offensive strategy between the two

time periods. Numerous teams in the mid-2000s focused on playing through star players that

thrived in the mid-range, including league leading teams like Tim Duncan’s Spurs, Kobe

Bryant’s Lakers, and Paul Pierce and Kevin Garnett’s Celtics. From 2012 on, teams evolved

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 27 -

to create pace-and-space offenses that focused on shooting almost exclusively three-pointers

and layups. Teams in this new era surrounded backcourt and/or wing stars with shooters, such

as LeBron James’ Heat and Cavalier teams, James Harden’s Rockets, and Steph Curry’s

Warriors.

That said, not all teams enacted this new focus on three-pointers and layups to the same

degree. Some teams, such as the Houston Rockets, fully embraced this analytics-based

strategy, and almost completely removed the mid-range jump shot from their team offensive

attack. The Rockets only attempted an astonishingly low 5.5% of their shots from the mid-

range during the 2018-19 season (the league average was 15.2%). Other teams, such as the

San Antonio Spurs, still heavily depended on the mid-range jump shot as an integral part of

their team’s offensive strategy. With Kawhi Leonard and Tim Duncan leading the team in the

early part of the 2010s, followed by LaMarcus Aldridge and DeMar DeRozen in the latter part

of the decade, the Spurs have structured their attack around their star players, who all excel as

individual mid-range shooters. The Spurs attempted a league-leading 28.2% of their shots

from the mid-range in the 2018-19 season. Despite their conflicting offensive strategies, the

Spurs and Rockets are both top three in the NBA in wins since 2012, suggesting that there are

ways to win with and without the mid-range shot.

The boxplots in the data findings echo this sentiment, as there were no major differences in

mid-range frequency between teams with strong offenses and those with weak offenses. The

highest ranked teams did shoot less mid-range attempts than low ranked teams, but the

difference was not as profound as it was with other shot attempt types. This suggests that the

decline of the mid-range is a league-wide trend regardless of team strength.

Corner three-point attempt frequency did vary by team strength. Top ranked teams attempted

significantly more corner three-pointers than low ranked teams. The corner three-pointer,

which is a shorter distance from the hoop than above the break three-pointers, is an asset of

strong teams. As mentioned, many teams in the new era have focused on surrounding star

players with shooters, and the corner is an essential spot on the court to place a shooter. When

a player drives to the basket, the kickout to the corner is one of the easiest passes that player

can make, so it makes sense that strong teams heavily rely on corner three-point attempts.

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 28 -

Additionally, over 90% of corner three-pointers are assisted. Teams in this era highly value

three-point specialists, such as the Rocket’s PJ Tucker, whose sole purpose on the offensive

end of the floor is to stretch the floor by spacing to the corner.

Above the break three-point shooting also did vary by team. Teams with strong offenses

heavily incorporate the above the break three into their offensive strategy, while teams with

weak offenses do not attempt nearly as many of these threes. The above the break three is not

strongly related to team rank, but higher ranked teams did shoot slightly more above the break

threes than low ranked teams.

The league-wide shift to play at a faster pace certainly impacts the changes in shot

distribution. Although pace by itself was not related to team rank, strong offenses played at

the fastest pace when compared to weak offenses. Weak defensive teams also played at a

higher pace than strong defensive teams. Logically this makes sense, as teams that allow a lot

of points need to play at a faster pace so that they get more possessions and have a greater

chance of outscoring their opponent.

The league-wide increase in pace and league-wide decrease in mid-range shooting are not

coincidental. As teams have changed to play at a faster pace, they get out in transition more

often, and have less possessions in the half-court. When out in transition, a player is more

likely to attempt to drive all the way to the basket, pull-up for an above the break three, or

kick out to a corner 3-point shooter. At the same time, players are less likely to take a mid-

range shot when in transition, as that shot is more typical in a slowed down, half-court style

offense. For the 2012-2019 data, pace was positively correlated with restricted area field goal

attempts, above the break three-point attempts, and corner three-point attempts, but negatively

correlated with mid-range attempts. See Appendix D for the correlations between pace and

the five types of shot attempts.

The linear regression model gives an interesting look into the variables that impact winning in

the NBA. For this model, defensive rank and pace were the two most important variables for

predicting winning. This finding follows classic logic in NBA history: offense wins games,

defense wins championships. Teams that play at a high pace attempt a lot of shots and

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 29 -

therefore score more points, which translates positively to winning. That said, the most

important variable to winning in the model is the strength of a team’s defense. This model

used defensive rating as a blanket metric to measure team defense, so additional research

would need to be done to identify the specific factors that contribute to strong defense. It is

important to note that the model simply suggests that defensive rating and pace are the biggest

predictors of winning of only the nine variables tested. It showcases their importance relative

to the other variables but is not meant to claim that they are the end-all-be-all predictors for

team success in the NBA.

CONCLUSION

This paper sought to investigate the strategic shift away from the mid-range jump shot in the

National Basketball Association in the last decade. The findings of the data analysis

statistically confirmed the decline in mid-range jump shot attempts from the 2005-06 season

through the 2018-19 season. The visualization tools also highlighted the rise in paint field

goal attempts, corner three-point attempts, and above the break threes over the time period.

Another main takeaway of the study is the relevance of pace of play in the shift in shooting

behavior in the NBA in the past decade. Pace, which measures the number of possessions a

team has a game, has increased league-wide over the time period, regardless of the strength of

the team. The increase in pace signifies an increase in transition opportunities, in which teams

focus on getting all the way to the basket for a layup or finding a shooter for an open three-

pointer. The style of play of this new pace-and-space era differs significantly from that of the

mid-2000s and heavily contributes to the decline of the mid-range jump shot.

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 30 -

APPENDICES

Appendix A – NBA Database Snippet – 2005-06 Season:

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 31 -

Appendix B – Visual Illustration of 5 Primary Shot Locations

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 32 -

Appendix C – Simplified NBA Database Snippet – 2005-06 Season:

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 33 -

Appendix D – Pace Correlations:

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 34 -

Works Cited

Cervone, Dan, Alexander D’Amour, Luke Bornn, and Kirk Goldsberry. “POINTWISE:

Predicting Points and Valuing Decisions in Real Time with NBA Optical Tracking

Data.” Harvard University, MIT Sloan Sports Analytics Conference, 28 February

2014, Boston, MA, USA.

Chang, Yu-Han Chang, Rajiv Maheswaran, Jeff Su, Sheldon Kwok, Tal Levy, Adam Wexler,

and Kevin Squire. “Quantifying Shot Quality in the NBA.” Second Spectrum, Inc.,

MIT Sloan Sports Analytics Conference, 28 February 2014, Boston, MA, USA.

Goldman, Matthew, & Justin M. Rao. “Live by the Three, Die by the Three? The Price of

Risk in the NBA.” 42 Analytics, MIT Sloan Sports Analytics Conference, 1 March

2013, Boston, MA, USA.

Goldsberry, Kirk & Eric Weiss. “The Dwight Effect: A New Ensemble of Interior Defense

Analytics for the NBA.” Harvard University, MIT Sloan Sports Analytics Conference,

1 March 2013, Boston, MA, USA.

Goldsberry, Kirk. “CourtVision: New Visual and Spatial Analytics for the NBA.” Harvard

University, MIT Sloan Sports Analytics Conference, 2 March 2012, Boston, MA,

USA.

Goldsberry, Kirk. “How Mapping Shots In The NBA Changed It Forever.” FiveThirtyEight,

FiveThirtyEight, 2 May 2019, fivethirtyeight.com/features/how-mapping-shots-in-the-

nba-changed-it-forever/.

Goldsberry, Kirk. “Kirk Goldsberry on Instagram: ‘Curry 15-16 vs Iverson 00-

01.’” Instagram, 26 Apr. 2020, www.instagram.com/p/B_aTSEqFw9E/.

Goldsberry, Kirk. “The NBA Is Obsessed with 3s, so Let's Finally Fix the Thing.” ESPN,

ESPN Internet Ventures, 30 Apr. 2019, www.espn.com/nba/story/_/id/26633540/the-

nba-obsessed-3s-let-fix-thing.

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 35 -

Heitner, Darren. “The NBA's Six Year, $250 Million Data Deal.” Forbes, Forbes Magazine,

27 Sept. 2016, www.forbes.com/sites/darrenheitner/2016/09/22/the-nbas-six-year-250-

million-data-deal/#1f6c6481481d.

McCagh, Wade. “How Spatial Analytics Killed The Mid-Range Jump Shot.” Vice, 26 Sept.

2016, www.vice.com/en_au/article/pgj338/numbers-game-how-spatial-analytics-

killed-the-mid-range-jump-shot.

McLaughlin, Matt. “How Data Analytics Is Revolutionizing Sports.” BizTech: Technology

Solutions That Drive Business, 13 Dec. 2018,

biztechmagazine.com/article/2018/12/how-data-analytics-revolutionizing-sports.

Miller, Andrew, Luke Bornn, Ryan Adams, and Kirk Goldsberry. “Factorized Point Process

Intensities: A Spatial Analysis of Professional Basketball.” Harvard University, 8

January 2014.

Mudric, Milos. “How The NBA Data And Analytics Revolution Has Changed The Game.”

SmartData Collective, 26 July 2019, www.smartdatacollective.com/how-nba-data-

analytics-revolution-has-changed-game/.

“NBA & ABA Year-by-Year Top 10 Leaders and Records for Effective Field Goal Pct.”

Basketball Reference, www.basketball-reference.com/leaders/efg_pct_top_10.html.

“NBA League Averages - Per Game.” Basketball Reference, www.basketball-

reference.com/leagues/NBA_stats_per_game.html.

O'Connor, Kevin. “It's More Than Just the Shot.” The Ringer, The Ringer, 12 Feb. 2019,

www.theringer.com/nba/2019/2/12/18221183/steph-curry-warriors-2019-all-star-

game-charlotte.

Pierson, Justin. “That Brief Time in the 90s When the 3-Point Line Was Shorter.”

ClutchPoints, 7 June 2016, clutchpoints.com/brief-time-90s-3-point-line-shorter/.

Piette, James, Lisa Pham, and Sathyanarayan Anand. “Evaluating Basketball Player

Performance via Statistical Network Modeling.” University of Pennsylvania, MIT

Sloan Sports Analytics Conference, 4 March 2011, Boston, MA, USA.

The Decline of the Mid-Range Jump Shot in Basketball

Honors Thesis by Shawn Kilcoyne

- 36 -

Skinner, Brian. “The Problem of Shot Selection in Basketball.” Fine Theoretical Physics

Institute, University of Minnesota, PLoS ONE, 25 January 2012, Minneapolis,

Minnesota, USA.

Wharton School. “Keynote Conversation with Adam Silver and Dan Pink | 2017 Wharton

People Analytics Conference.” YouTube.com, 17 May 2017,

www.youtube.com/watch?v=mUSFAYslACQ.

Zuccolotto, Paola, Marica Manisera, Marco Sandri. “Big data analytics for modeling scoring

probability in basketball: The effect of shooting under high-pressure conditions.”

International Journal of Sports Science & Coaching, 6 November 2017.