2020 SABR Analytics Conference research presentations

Schedule of events Featured speakers Diamond Dollars Case Competition
SABR Analytics Conference Research Awards Sponsors Renaissance Phoenix Downtown hotel information
Research Presentations Download the Digital Program Photo galleries


SABR and Baseball Info Solutions are pleased to announce the research presentations for the ninth annual SABR Analytics Conference, which was held March 13-15, 2020, at the Renaissance Phoenix Downtown in Phoenix, Arizona.

All baseball fans are welcome to watch our free YouTube livestream during the conference, so we hope you tune in!

Here is a list of presentation abstracts and bios. Click on a link below for the audio recording and PowerPoint slides, when available:


1:45-3:15 p.m., Friday, March 13
RP1-RP3 took place consecutively in a single session in the Grand Ballroom (3rd floor).

RP1: Swing Adjustments
Alex Caravan

The influx of more technical measurement of baseball performance has led to more availability and freedom to refine the definition of hot and cold zones beyond its vague place in sabermetrics. In this paper we use the full extent of 2008 to 2019 MLB (PITCHf/x; Statcast) data to devise three separate methods to interpret hot and cold zones and see what sort of reliability each method has in teaching us about what’s behind a hot and cold zone, and what type of hitters are especially prone to hot and cold zones. This paper demonstrates several controls and rigors to enact when studying zone discretization, and points to the idea of higher swing percentages and Isolated Power (ISO) being important variables in determining a hitter’s likelihood of running hot or cold.

Alex Caravan (acaravan@gmail.com) is a quantitative analyst in Driveline Baseball’s R&D department, where he has co-authored multiple peer-reviewed studies through publications like PeerJ, individual research pieces on the Driveline Baseball blog, and worked on various software features both internally and externally (Driveline EDGE). He graduated from University of California Berkeley and now lives in Seattle, where he splits his free time recording the Driveline Research & Drinks Podcast, training for the next big Spikeball tournament with his roommate Anthony Brady, and searching the near metropolitan district for the cheapest bulk sushi roll deals.


RP2: Strength of Players Faced — A WAR Analysis
Russell Eassom

In the last few seasons we have seen some of the biggest variance in team performance. Last season the Minnesota Twins had a strength of schedule (SOS) of -0.5, where as the Milwaukee Brewers had a 0.3 (SOS). Meaning there was a 0.8 run differential between the teams that the Twins played compared to the Brewers. I looked to see if this strength of teams faced translated down to the player level and it showed a growing variance in strength of player faced over the last couple of seasons. Given that both FanGraphs and Baseball-Reference WAR calculations for position players use a hitter’s wOBA performance and compare it against league average wOBA to generate batting runs, should we be accounting for strength of players faced like BRef does for its pitching WAR? This presentation will look into that and show the potential changes to players seasonal and career WAR values.

Russell Eassom is a Data Scientist for Algomi, a FinTech in the fixed income bond world. He is also a writer for Bat Flips & Nerds, a British take on baseball, where he uses skills learned as a data scientist to investigate the analytical side of baseball. His particular passion is defensive analysis of players and the shift.


RP3: Beating Boudreau: Measuring Success Against the Shift
Ken Kauffman

Note: This presentation was delivered by pinch-hitter Cameron Adams of SMT.

Infield shifts against pull hitters have become ubiquitous in today’s game. However, little quantitative information is available on what types of hits are must effective against it. Using MiLB FIELDf/x ball- and player-tracking data ([x, y, z] coordinates at 20 or 30Hz), we develop “Beats”—a hitting metric that incorporates exit velocity, spray angle, and launch angle to determine offensive success against the shift. We find a contiguous range—or Beat Box—in which shifted batting averages are higher than unshifted ones.  

Applying this new metric, we find that certain hit contact zones are more likely to produce Beats. We also look at Beats in the context of landing location and discover that launch angle plays a key role in the overall effectiveness of the shift.

Ken Kauffman is a Data Scientist for SportsMEDIA Technology (SMT), focused primarily on FIELDf/x and other baseball products. He received his B.S. in Industrial Engineering with minors in Statistics and Psychology from California Polytechnic State University (SLO), and he is currently pursuing his M.S. in Analytics at Georgia Institute of Technology. Prior to SMT, Ken worked at Zebra Technologies, where he worked on analytics projects for their RFID-based NFL player/ball tracking product.


9:35-11:05 a.m., Saturday, March 14
RP4-RP6 took place consecutively in a single session in the Grand Ballroom (3rd floor).

RP4: Using Clustering to Find Pitch Subtypes and Effective Pairings
Gregory Dvoroscik, Eno Sarris, and Joe Camp

Using Statcast data, it is now possible to compare individual pitches across baseball based on characteristics like movement, velocity, and spin rate, which are all measurements that become obvious and meaningful even in a single outing. Various research has used those physical characteristics to classify pitches and define optimality in isolation. However, even an elite pitch has to be mixed with less optimal ones, especially for starting pitchers. Therefore, it is imperative to study the interactions between pitches to fully understand the best shape a particular pitch should have — how a pitch is paired with others can be as important as its own characteristics. 

In this work, we find effective and ineffective subtype pitch pairings. To do so, we first attempt to understand how many different subtypes exist of a Statcast pitch classification by using k-means clustering of vertical and horizontal movement, velocity, and spin rate data for the entire 2016 and 2017 seasons. For both left-handed and right-handed pitchers, we find 30 subtypes within the 9 prominent Statcast pitch types. Using these subtype clusters, we consider resulting performance based on swinging strike rate, exit velocity, and extreme launch angles (popup and ground ball percentage). We then consider the effectiveness of each subtype (which we refer to as the reference subtype) when paired with each of the other 29 subtypes. Next, we consider the gain or loss for all pitchers who include both the reference subtype and paired subtype in their pitch arsenal from the average performance of the reference subtype. 

As a result, we find that the average gain across all subtypes by the most effective pairing increases swinging strike rate by almost 2 percent, raises extreme launch angle outcomes by over 4 percent, and reduces exit velocity by more than 1 MPH — all amounts that are similarly lost by the worst pairing. We also present specific pitcher examples of effective and ineffective pairings. Our work has potential impact on pitch design, player development, and scouting. For the former, teams could focus efforts on teaching young pitchers new pitch subtypes that have specific shapes according to the characteristics of the best pitches already thrown by that pitcher. For the latter, with very little in-game data, teams could seek to add pitchers that already possess effective pairings or avoid pitchers with ineffective pairings.

Gregory Dvorocsik is a junior at Wake Forest University majoring in mathematical business. He has worked with the school’s baseball analytics department creating scouting reports and other deliverables for the coaching staff to improve game preparation and player development. He is seeking an analytics internship or apprenticeship with a MLB team this summer.

Eno Sarris is a staff writer for The Athletic, where he specializes in pitching analytics. He takes the best public analytics findings to the players in the clubhouse to get their reactions. He has also been a contributor to FanGraphs, ESPN, MLB.com, Fox Sports, SB Nation, The Hardball Times, and others.

Joe Camp is an Associate Professor in Electrical and Computer Engineering (ECE) at Southern Methodist University (SMU). While his research efforts focus on wireless systems and drone communications, his hobby is baseball analytics and has collaborated with the Texas Rangers in recent years to give a SABR 101 talk to fans before games.


RP5: Measuring the Impact of Pitch Location on a Player’s Performance
Patrick Brennan

Pitch location plays a massive role in the performance of both pitchers and hitters. This is obvious. The issue is isolating and measuring the skill/luck/impact of pitch location and converting it into a single number is tough. For pitchers, the skill of locating the ball is defined by the term “command.” It’s an extremely integral part of scouting, but other than just using our eyes and making assumptions, there’s no mainstream quantifiable number for the important skill. In this research, quantifying command will be done by breaking the strike zone (and the space outside the strike zone) to detailed areas and evaluating league-average offensive production (wOBA) in each of those areas. Pitch type and pitcher/batter handedness will also be taken into account. With this information, each pitcher will be assigned a value and subsequently averaged out to create a final number.

Patrick Brennan is an analytical assistant for the baseball team at Kansas State University. In addition, he has published research at various sites, such as The Hardball Times, Beyond the Box Score, and Royals Review.


RP6: Scoring Reversals Revisited: The Mutability of MVP Voting
Colton Cronin

Each fall, the Baseball Writers’ Association of America announces Most Valuable Players for the American League and National League. To determine the MVP, a group of voters individually rank their top ten players, and then each player is assigned points based on each ranking; the player with the most points is named MVP. However, this point scheme has an interesting feature: The 10th-ranked player receives 1 point, the ninth player receives 2 points, and so on up to the second-ranked player who receives 9 points, but the first ranked player receives 14 points. Consequently, this method greatly rewards receiving first-place votes. But what if a different point scheme was used? In this presentation, I illustrate that MVP results are greatly impacted by the points scheme used. Using ballot results from all 58 NL and AL awards from 1990 to 2018, I apply 10 alternate point schemes to each election and find that not only would the rankings of one or more players change in all 58 elections, but also that in six of these elections, a different MVP would have been named. The results also suggest that, depending on the scheme, appearing on multiple voters’ ballots in lower rankings can be as important as receiving a few first-place votes. My goal of this presentation is not only to demonstrate an interesting facet of MVP voting but also to illustrate how, player performance and voter preferences aside, the construction of analytical systems can impact outcomes. It is perhaps unsurprising that the results of a subjective award like MVP are somewhat mutable, but the designers of any form of metric should be aware of how seemingly minor structural choices — like point schemes — can reflect the values of the metric’s designers and can influence who we perceive as the best.

Colton Cronin is a junior at Vanderbilt University (home of the 2019 College Baseball World Series champs!) He is majoring in mathematics and economics, and he is interested in how economic theory can examine what we consider “best” and “fair” in various competitive scenarios, including baseball.


1:45-3:15 p.m., Saturday, March 14
RP7-RP9 took place consecutively in a single session in the Grand Ballroom (3rd floor).

RP7: The Fielding Bible: Repositioning Defensive Runs Saved
Mark Simon and Brian Reiff

Baseball Info Solutions (BIS) has recently announced significant upgrades to the Defensive Runs Saved methodology. In particular, the Range & Positioning component of DRS (formerly known as the Plus/Minus System) is being replaced by the PART System. This presentation will highlight the new data and methodologies that allow us to explore some of the biggest trends and storylines arising from the new numbers.

PART stands for Positioning, Air balls, Range and Throwing. At its core, the system’s goal is to split a fielder’s contributions into its individual components. The PART System offers numerous additional advantages, the most notable being the consideration of positioning when assigning credit or debit to fielders and the ability to evaluate all plays, including ones in which the defense was shifted, which were previously excluded from DRS. The result is a better understanding of a player’s defensive performance.

Mark Simon is a Senior Research Analyst for Baseball Info Solutions. He previously worked as a researcher at ESPN, including nine years on “Baseball Tonight.” He now writes regularly about defense and other topics for The Athletic.

Brian Reiff is a Research Analyst at Baseball Info Solutions. He initially joined the R&D group as an intern during his senior year at Lehigh University and became a full-time member of the staff after graduating in May 2017.


RP8: Modeling and Projecting Offensive Value Using Combined Hit-Tracking and Speed Measurements
Glenn Healey

Outcome-based statistics for representing batter and pitcher skill have been shown to have a low degree of repeatability due to the effects of multiple confounding variables such as the defense, weather, and ballpark. Statistics derived from pitch and hit-tracking data acquired by the Statcast system have been shown to provide greater repeatability and predictive value than outcome-based statistics. The wOBA cube representation uses three-dimensional hit-tracking data to compute intrinsic batted ball statistics for batters and pitchers. While providing more reliable measures than outcome-based statistics, this representation also revealed that running speed is an important determinant of batter success. We address this issue by building a four-dimensional model for a batted ball’s value as a function of its physical contact parameters and the batter’s time-to-first speed. The model uses a Bayesian framework that employs a kernel method to generate nonparametric probability density estimates using a large set of sensor data provided by MLB. A cross-validation scheme allows the algorithm to adapt to the size and structure of the data. The result is a learning algorithm that generates a continuous mapping from batted-ball and time-to-first sensor measurements to run value defined using linear weights. Separate mappings are built to accommodate the effects of batter handedness. We show that the four-dimensional model provides more accurate predictions than the three-dimensional wOBA cube. Visualizations illustrate the dependence of batter performance on the 4-D measurement vector. We also present leaderboards showing the batters with the most significant gains and losses in offensive performance due to time-to-first speed. We show that the new model improves our ability to analyze other factors that affect batter performance including susceptibility to shifts, the ballpark, and the weather.

Glenn Healey is a professor of electrical engineering and computer science at the University of California, Irvine where he is director of the computer vision laboratory. He received the B.S.E. degree in Computer Engineering from the University of Michigan and the M.S. degree in computer science, the M.S. degree in mathematics, and the Ph.D. degree in computer science from Stanford University. Dr. Healey’s professional life is dedicated to combining physics, statistical signal processing, and machine learning methods for the development of algorithms that extract information from large sets of data.


RP9: Using Pitching Mechanics Analytics to Predict Injury and Improve Performance
Bill Leisenring, Josh Myers, Justin Orenduff

Baseball pitching imposes significant stress on the upper extremity and can lead to injury. In 2019, MLB teams spent over $390 million on pitchers while on the injured list for a total of 22,700 days missed. Many studies have attempted to predict injury through pitching mechanics, most of which have used laboratory setups that are often not practical for population-based analysis. This presentation will detail the findings of a study that sought to predict injury risk in professional baseball pitchers using a statistical model based on video analysis evaluating delivery mechanics in a large population. This model can be used to assess injury risk of professional pitchers and may be expanded to pitchers at other levels.

The United Shore Professional Baseball League (USPBL) is an independent professional baseball league in metro Detroit, Michigan, and provides development opportunities for former college and minor league players to sign with an MLB organization. A systematized pitching development program derived from this analysis of pitching delivery mechanics has been implemented for all teams in the USPBL. The league also uses pitching mechanics analytics as one factor in making personnel decisions. This presentation will also discuss research findings from the USPBL including increased arm health, more innings pitched, increased velocity, and over 35 USPBL players signing with MLB teams since the league’s inception in 2016.

Bill Leisenring is the Chief Technology Officer at DVS Baseball. Bill has a M.S. in Electrical Engineering from The Ohio State University where he was a Center for Automotive Research Fellow. In 2009, he founded Control-Tec, an automotive data analytics company, that was later sold to Delphi in 2015. A spin-off from Control-Tec, Novation Analytics, was sold to IHS Markit in 2019. Leisenring is now responsible for product development and technology at DVS Baseball. His passion for baseball analytics started with studying baseball cards and The Sporting News during his youth and winning a regional high school essay contest with research on the physics of pitching.

Josh Myers is a co-founder of DVS Baseball and serves as the Director of Statistics. Josh handles all data analytics and statistical analysis for DVS, including the pitching injury risk model. He also runs a successful statistical consulting business in the field of property tax assessment. Josh graduated from the University of Virginia in May 2005 with a B.S. in Physics and Mathematics, all while pitching for the varsity baseball team. He then earned a Master of Science degree in Statistics in January 2007, also from the University of Virginia.

Justin Orenduff is a co-founder of DVS Baseball and creator of the DVS Scoring System. The scoring system was built as a product of Justin’s pitching career, research into throwing patterns among pitchers, and hours of testing and validating the scoring system over a five-year period. Justin also serves as the Director of Baseball Operations for the United Shore Professional Baseball League. Before DVS, Justin co-founded and authored The Baseball Pitching Rebellion and was the Head Pitching Instructor at I.T.S. Baseball. Justin was an All-American pitcher at both George Washington University and Virginia Commonwealth University. He was a member of the 2003 USA Collegiate National Team and a 2004 first round draft pick of the Los Angeles Dodgers. After his playing career, Justin returned to Virginia Commonwealth University in 2009 and earned his B.S. in Business Management.


8:30-10:00 a.m., Sunday, March 15
RP10-RP12 took place consecutively in a single session in the Grand Ballroom (3rd floor).

RP10: Examining Neural Activity to Pitches and Feedback at the Plate: Psychological and Performance Implications
Jason Themanson

In an effort to gain a competitive edge, teams have begun to examine hitters’ neural activity while they classify different pitch types. That research is useful in learning the time course and neural structures utilized in pitch recognition processes, but it does not account for the influences of pitch decisions and performance feedback on the psychology and behavior of hitters throughout a plate appearance or beyond. We can obtain valuable insight into hitters’ behavior and their psychology, including their expectations, attentional focus, and self-regulatory learning processes, by examining the dynamic distribution of pitch-by-pitch neural activity in conjunction with neural activity to the performance feedback hitters receive between pitches. To examine this topic, collegiate baseball players and non-players (novices) completed a computerized video task assessing whether pitches were balls or strikes. Following each pitch, hitters received feedback on the accuracy of their decision before the next pitch was thrown. Participants’ neural activity was recorded throughout the task. Results indicated relationships were present between college players’ neural activity to feedback, their neural activity to the pitches, and their performance in the task. These relationships were not present in the novices. This finding suggests that players associated the information received in their performance feedback to their processing of the task and ultimately their task performance. Neural activity can index many psychological processes that underlie task performance and self-regulatory efforts to improve behavior, including information evaluation, focused attention, and the monitoring and control of motor performance. Implications and uses for this research include assisting in scouting evaluations and player development procedures. Obtaining this level of psychological data, when combined with advanced analytic data and physiological data, could provide new insights into performance modeling and player development plans and evaluations.

Jason Themanson is a Professor in the Department of Psychology and the Neuroscience Program at Illinois Wesleyan University. He received his B.S. in Psychology from the University of Illinois, his M.A. in Social Psychology from the University of Connecticut, and his Ph.D. in Kinesiology (with an emphasis in exercise psychology) from the University of Illinois. Dr. Themanson’s research utilizes both neuroelectric and behavioral measures to examine self- regulation and cognitive control processes during task execution.


RP11: What’s Hanging? An Empirical Definition And Defining Attributes For The Hanging Pitch
Jeremy Losak

The “hanging pitch”’ is a colloquial phrase used by media pundits and fans alike, although a precise definition does not seem to exist. Presently, hangers are mostly defined via the eye test, creating opportunities for subjective classification differences. Using Statcast pitching data from 2015-2018, this paper applies empirical techniques to propose an objective definition for a hanging pitch, and then uses that definition to analyze pitchers based on their hanging pitch tendencies. We start by modeling the probability that any given slider or curveball is barreled, given the pitch is put in play, and based on characteristics of the pitch. Based on the density of expected barrel rates returned by the model, we identify an expected barrel rate threshold, approximately 8%, and set that as our hanging rate criteria. We then use our classification to test if there are any internal or external factors that impact the probability of throwing a hanging pitch. We identify that pitcher usage plays a role; pitchers are less likely to throw hanging pitches the more they throw in a season, game, or at bat. We also identify that there is some season-to-season pitcher-specific correlation for giving up hanging pitches, although the driver of that correlation has yet to be determined.

Jeremy Losak is an Assistant Professor in the Department of Sport Management at Syracuse University, serving as a faculty member for the sport analytics bachelor’s degree program. He earned a Ph.D. in economics from Clemson University, where his research focused on the economics of sports, particularly baseball labor markets and daily fantasy betting markets. Jeremy is assisted by Gareth Jobling, originally from Cleveland, Ohio, who is a senior Sport Analytics and Economics double major with an interest in business analytics.


RP12: Fundamentals of Projecting Defensive Performance
Alex Vigderman

Thanks to the interest that comes from fantasy sports, yearly projections of hitting and pitching statistics are quite easy to find. Less attention is paid to the prospect of projecting defensive performance, especially considering the perception that defensive metrics are unreliable. This presentation will dig into the considerations involved in projecting defensive performance. 

This presentation will discuss the stability of defensive metrics compared to offensive and defensive statistics, particularly focusing on the new version of Defensive Runs Saved featuring the PART System (featured in another presentation at this conference). Additional factors like player aging, the relative difficulty of different positions, and minor league performance will also be discussed, with some specific examples of how they come into play when evaluating a player’s future prospects defensively.

Alex Vigderman is a Senior Research Analyst at Baseball Info Solutions, where he takes proprietary baseball and football charting data and bakes it into exciting analysis. He was previously an intern with the Boston Red Sox in their analytics department after graduating from the University of Pennsylvania with a degree in Psychology and working in the healthcare software industry.


11:15 a.m.-12:45 p.m., Sunday, March 15
RP13-RP15 took place consecutively in a single session in the Grand Ballroom (3rd floor).

RP13: The 2019 Baseball, and the Unanticipated Consequences of Change
Meredith Wills

2019 holds a unique place in baseball history; never before has the ball undergone two dramatic changes in one calendar year. The regular-season ball showed an unexpected decrease in drag—lower even than 2017—leading to a season in which any number of home run records were shattered. Then, in the postseason, the ball changed again, demonstrating drag that was both higher and more erratic.

Having previously looked at the baseball’s construction and how it may have contributed to the 2017 Home Run Surge, I performed similar studies on balls from the 2019 regular season and postseason. The regular-season ball showed a number of physical differences—including seam height, leather smoothness, and spherical symmetry—some of which were consistent with improved aerodynamics. As with the 2017 ball, the changes were consistent with standard manufacturing improvements. Meanwhile, the balls used in the postseason appear to have come from both 2018 and 2019 regular-season populations—a mix that could account for unpredictable drag variability.

The introduction of 2018 balls in the postseason may be connected to higher regular-season ball usage rates. In its first season with the Major League baseball, Triple-A exceeded projected usage by roughly 30%. In addition, MLB went through >25% more regular-season balls than in previous years. These increases appear to have depleted the 2019 surplus to the point that balls from previous seasons were required to supply the postseason.

Dr. Meredith Wills is a Sports Data Scientist for SportsMEDIA Technology (SMT) and a contributing writer for The Athletic, whose independent research on the composition of the baseballs helped shed new light on the home run surge in recent seasons. In her spare time, she is also a knitting designer, working in partnership with both the Baseball Hall of Fame and the Negro Leagues Baseball Museum to create reproductions of vintage baseball sweaters. She received her B.A. in Astronomy and Astrophysics from Harvard University, and her M.S. in Physics and Ph.D. in Physics from Montana State University-Bozeman.


RP14: Introducing SRC and OSWC: Using Game Theory to Assign Credit for Offensive Outcomes
Michael McBride

How should individual baseball players be credited for the collaborative outcomes of their teams? My work proposes a new framework for assigning individual credit that is theoretically grounded and easy to interpret and use by baseball specialists and everyday fans. I utilize the Shapley Value concept from coalitional game theory to create two new offensive credit measures: Shapley Run Credits (SRC) and Offensive Shapley Win Credits (OSWC).

The Shapley Value is a mathematically-defined solution concept that calculates a fair allocation of credit for the gains realized from team collaboration. When applied to runs, SRC partitions the credit for each run scored among the players who contributed to the production of that run in proportion to each player’s importance in the scoring. When applied to offensive wins, OSWC splits a win credit among the players in proportion to how much each helped the team outscore the opponent. Calculating SRC and OSWC is not trivial because it requires the play-by-play computation of counterfactual innings for each possible subset of the team members, and this in turn requires the explicit programming of human judgments about how each player’s offensive event would hypothetically impact all base-out states, not just the base-out state in which the event actually occurred. However, because the total SRC and OSWC will equal the number of actual runs scored and wins achieved, these measures have straightforward interpretations and require no technical sophistication to use once calculated.

This presentation introduces SRC and OSWC, illustrates their calculation, compares them with existing measures, and demonstrates one application, namely, the evaluation of MVP candidates. I calculate SRC and OSWC for all players in the 1990-2018 World Series, ALCS, and NLCS and identify which MVP awards went to players with the most (or not the most) impactful offensive performance. SRC and OSWC have other potential applications, including the estimation of how a player’s offensive contributions are hindered or enhanced by teammate quality.

Michael McBride is Professor of Economics at the University of California, Irvine and Founding Director of the Experimental Social Science Laboratory. He received B.A. and M.A. degrees from the University of Southern California and M.Phil. and Ph.D. degrees from Yale University. His research uses game theory and experimental methods to study conflict, cooperation, and strategic interaction.


RP15: Can an Across-the-Board Increase in Minor League Pay Reduce PED Use?
Scott A. Brave

In March 2019, the Toronto Blue Jays unexpectedly increased minor league pay across-the-board by 50 percent. We provide evidence that a potentially unintended byproduct of this policy change may be a reduction in the incentives for players to use performance-enhancing drugs (PEDs). Using the universe of minor league PED suspensions and a statistical model of minor league level assignments that accounts for endogenous player performance and PED suspensions, we show that a 50 percent increase in minor league-wide pay could reduce PED use by a significant amount. Our model suggests that much of this impact will occur among players near promotion and demotion thresholds, where a small boost in performance can have a significant impact on level assignment and consequently on pay as well. Moreover, much of the impact arises at the lower levels, where salaries are especially low and PED use is more common. A large across-the-board pay increase for these players reduces the relative return of minor league level progression thereby resulting in lower PED use. However, if MLB were to announce an across-the-board increase in minor league pay concurrently with the elimination of a large number of minor league affiliates, as has been recently reported, our results also suggest that any positive effect on PED use may not immediately materialize. In this case, a decline in roster spots at the lower levels where contraction is expected will likely only amplify the benefits to PED use for those players now at risk of falling into the independent leagues.

Scott A. Brave is a senior policy economist in the economic research department of the Federal Reserve Bank of Chicago, where his responsibilities include the releases for the Chicago Fed’s National Activity, National Financial Conditions, Brave-Butters-Kelley, Midwest Economy, and Detroit Economic Activity Indexes. Brave received a B.A. in economics with honors from the University of Chicago and an M.B.A. with concentrations in economics, statistics and finance from the University of Chicago Booth School of Business. His research on the competitive effects of performance-enhancing drugs and team synergy in major league baseball has been published in the Journal of Sports Economics and the Journal of Sports Analytics.


For more coverage of the 2020 SABR Analytics Conference, visit SABR.org/analytics.

Originally published: January 24, 2020. Last Updated: January 24, 2020.