While developing our Change Your Game exhibition, we have explored several non-technical sports innovations. These innovations include novel coaching strategies, training regimens, or any techniques that enable athletes and teams to be more competitive. In an earlier blog, I documented how Dick Fosbury transformed high jumping by introducing the “Fosbury flop” in the 1960s. In this post, I describe some of the early pioneers of sports analytics.
Sports analytics encompasses a set of data management technologies and computational techniques for gathering and interpreting observable statistical data about athletes and game play. Team executives use statistical analysis to evaluate players for drafting, trades, and contract negotiations. Coaches use analytics to understand competitors’ tendencies, to develop in-game strategies, and to identify areas for player improvement.
Many sports fans first became aware of sports analytics from Michael Lewis’s bestselling book Moneyball (2003), and the 2011 film adaptation starring Brad Pitt as Oakland Athletics general manager Billy Beane. The book and film describe how Beane deployed unorthodox mathematical techniques to find high-performing but undervalued baseball players, which helped his low-payroll baseball team compete successfully against wealthier clubs.
Moneyball: The Art of Winning an Unfair Game, by Michael Lewis, 2003. Courtesy of W. W. Norton
But analytics did not begin with Moneyball. Rather, the book’s publication in 2003 signaled a growing enthusiasm for sports analytics that had been building since the 1950s. The sports analytics community emerged from three distinct subcommunities: operations researchers, freelance sports journalists, and internet hobbyists.
“Operations research” (OR) originated during World War II as a new science of warfare, in which statisticians analyzed their military’s operational data and suggested tactical adjustments. For example, by accounting for the weather or time of day, operations researchers could help their air forces optimize the lethal efficiency of aerial bombing runs. As a diversion from their military work, postwar OR practitioners applied their analytical tools to various sports. In a 1954 paper, for example, US defense analyst Charles M. Mottley observed that football resembled “ground combat” and suggested that coaches, like generals, could use OR methods to make significant improvements in team performance. In an analysis of 400 running plays, Mottley argued that football teams could maximize yardage gained by balancing end runs and runs into the line of scrimmage.
In a 1954 article, operations researcher Charles Mottley argued that football teams could maximize yardage gained by balancing end runs and runs into the line of scrimmage. From the Journal of the Operations Research Society of America
Likewise, in a 1959 Operations Research article, the Royal Canadian Air Force’s George R. Lindsey empirically demonstrated that right-handed batters had a higher batting average against left-handed pitchers (and vice versa) because they could see the baseball better. He suggested that managers should substitute different players into the lineup depending on the “handedness” of the opposing pitcher, a now common practice known as “platooning.”
In a 1959 article, operations researcher George R. Lindsey empirically demonstrated that right-handed batters had a higher batting average against left-handed pitchers (and vice versa) because they could see the baseball better. “Same” handed batters hit for a 0.231 average, while “opposite” handed batters averaged 0.263. Courtesy of Operations Research
Complementing the OR work, an eclectic community of amateur hobbyists and freelance sports journalists developed new statistical metrics to animate their research and writing. During the 1960s, freelancer Bob Davids published statistically informed baseball columns for The Sporting News, Washington Post, and Chicago Tribune. In the spring of 1971, Davids sent a letter to thirty-two like-minded “statistorians” and proposed forming an association. Sixteen charter members founded the Society for Baseball Research (SABR) in August 1971.
Led by freelance journalist Bob Davids, sixteen “statistorians” founded the Society for Baseball Research (SABR) on August 10, 1971 at the Baseball Hall of Fame’s National Baseball Library in Cooperstown, New York. Courtesy of SABR
Bill James was not one of SABR’s founders, but he eventually became its most influential and prominent member. In his freelance articles for Baseball Digest and The Sporting News, James used statistics to challenge received wisdoms and showed a knack for asking probing questions. For example, how did a pitcher’s run support impact his win-loss record? How did hitters fare in day versus night games? Between 1977 and 1988, James published a preseason subscription newsletter, Baseball Abstracts, and introduced several new advanced metrics. For example, James’s formula for Runs Created added up the positive offensive impact of walks, extra base hits, and stolen bases, while subtracting a negative effect when baserunners were caught stealing. James eventually became the poster boy for “sabermetrics”—the use of statistical methods to glean new insights into baseball—a term he coined in 1980.
Baseball executives ignored James’s thinking until the 1990s, but he inspired countless statisticians working across several sports. Bob Carroll, a Pennsylvania high school teacher and freelance author, gathered five fellow hobbyists at the Pro Football Hall of Fame in Canton, Ohio, and founded the Professional Football Researchers Association in June 1979. In The Hidden Game of Football (1988), Carroll and his co-authors developed new strategic metrics, such as “expected points,” that forecast scoring probabilities for different field positions and game situations. That same year, Bob Bellotti published Basketball’s Hidden Game (1988) and introduced “points created,” the basketball analog to James’s runs created. In another nod to James’s Baseball Abstracts, Dave Heeren published several Basketball Abstracts (1988-1994) and introduced TENDEX, a linear weights model that added positive events (e.g., points, rebounds) and subtracted negative events (e.g., turnovers) to rate a player’s total performance.
During the 1990s and early 2000s, the sports analytics community began to flourish among hobbyists and bloggers on the nascent World Wide Web. Dean Oliver, an engineering student and varsity basketball player at Caltech, was among the first statisticians to transition from traditional books and subscription newsletters to personal web pages and user groups as a cheaper and more accessible publishing platform. In 1990-1991, Oliver began posting his basketball analyses on a personal website he called the Journal of Basketball Studies, and later, on the Usenet group rec.sports.basketball. His 1991 paper, “New Measurement Techniques and a Binomial Model of the Game of Basketball,” introduced the so-called “Possession Scoring System,” which recognized each team’s possessions as basketball’s fundamental unit of analysis. Oliver used a team’s total possessions as a normalizing metric to account for pace of play and defined a team’s offensive and defensive rating as the number of points scored or allowed per possession, a now-standard measure of efficiency.
It is important to note that these operations researchers, freelance journalists, and internet bloggers were quirky iconoclasts and outsiders. Generally, professional scouts, coaches, and team executives derided the sports analysts since most of them had never played the games. “This guy has never played baseball,” Detroit Tigers manager Sparky Anderson said dismissively of Bill James. “I don’t think he knows very much about it.”
The tide began to turn in the early 2000s. When author Michael Lewis described James’s influence on the Oakland Athletics and published Moneyball in 2003, it signaled a growing enthusiasm for analytics and inspired a hiring spree for statisticians among professional teams. In 2004, the Los Angeles Dodgers recruited Oakland’s 31-year-old assistant general manager Paul DePodesta to be its new general manager, even though he had never played or coached professional baseball. That same year, the Seattle Supersonics hired Dean Oliver as the first full-time basketball analyst in the NBA. By 2012, twenty-six of thirty Major League Baseball teams and twenty-four of thirty NBA basketball teams employed at least one person dedicated to analytics.
MIT graduates Daryl Morey (left) and Jessica Gelman (right) founded the MIT Sloan Sports Analytics Conference in 2006; it annually attracts 2,000 participants. Morey is president of the Philadelphia 76ers; Gelman is CEO of the Kraft Analytics Group. Courtesy of SportTechie
Over the next several years, sports analytics blossomed into a full-blown academic subdiscipline. In 2006, Boston-based analysts Daryl Morey (then with the Boston Celtics) and Jessica Gelman (Kraft Analytics Group) co-founded the MIT Sloan Sports Analytics Conference; it annually attracts 2,000 participants. The two main statistical societies—the Institute for Operations Research and Management Sciences (INFORMS) and the American Statistical Society—each added sections devoted to sports analytics. Sports analysts also founded three peer-reviewed journals to share their non-proprietary methods, including the Journal of Quantitative Analysis in Sports (2005), the International Journal of Sports Science and Engineering (2007), and the Journal of Sports Analytics (2015). In 2016, Syracuse University launched the first bachelor’s degree program in the US in sports analytics to prepare students to fill the growing demand for analysts in professional sports management.
“If we win, on our budget, with this team,” Brad Pitt announced in his film portrayal of the A’s Billy Beane, “we’ll have changed the game.” His prediction was prophetic. Sports analytics emerged after World War II as a quirky pastime practiced by operations researchers, freelance journalists, and internet bloggers. By the time the Moneyball movie hit theaters in 2011, analytics had evolved into a rigorous academic subdiscipline and an innovative technique practiced by executives across professional sports.
SOURCES:
“About – Sloan MIT Sports Analytics Conference,” http://www.sloansportsconference.com/about/, accessed 9 September 2020.
Alamar, Benjamin C. Sports Analytics: A Guide for Coaches, Managers, and Other Decision Makers. New York: Columbia University Press, 2013.
Alamar, Benjamin and Vijay Mehrotra, “Beyond ‘Moneyball’: Rapidly Evolving World of Sports Analytics, Part I,” Analytics, 3 October 2011, https://pubsonline.informs.org/do/10.1287/LYTX.2011.05.05/full/.
Anderson, Sparky. Quoted in Alan Schwarz, The Numbers Game: Baseball's Lifelong Fascination with Statistics. New York: Thomas Dunne, 2004, p. 131.
Ballard, Chris. “Measure of Success,” Sports Illustrated, 24 October 2005, 87-91.
Baumer, Benjamin and Andrew Zimbalist. The Sabermetric Revolution: Assessing the Growth of Analytics in Baseball Philadelphia, PA: University of Pennsylvania Press, 2014, 1-22, 26, 87-91.
Bellotti, Robert S. Basketball’s Hidden Game: Points Created, Boxscore Defense and Other Revelations. Night Work Publishing, 1988.
Belth, Alex. “Bronx Banter Interview: Rob Fleder Bronx Banter,” accessed July 27, 2022, http://www.bronxbanterblog.com/2012/03/19/bronx-banter-interview-rob-fleder/.
Carroll, Bob, Pete Palmer, and John Thorn, The Hidden Game of Football. New York: Grand Central Publishing, 1988.
Cochran, James J. “The Emergence of Sports Analytics,” Analytics, 1 February 2010, https://pubsonline.informs.org/do/10.1287/LYTX.2010.01.06/full/.
Fortun, M. and S. S. Schweber. “Scientists and the Legacy of World War II: The Case of Operations Research (OR),” Social Studies of Science 23, no. 4 (November 1993): 595-642.
Glockner, Andy. Chasing Perfection: A Behind-the-Scenes Look at the High-Stakes Game of Creating an NBA Champion. Da Capo Press, 2016, pp. 18-23.
Gray, Scott. The Mind of Bill James: How a Complete Outsider Changed Baseball. New York: Doubleday, 2006.
Heeren, Dave. The Basketball Abstract. New York: Prentice Hall, 1988.
Kirby, Maurice W. Operational Research in War and Peace: The British Experience from the 1930s to 1970. London: Imperial College Press, 2003.
Lewis, Michael. Moneyball: The Art of Winning an Unfair Game. New York: W. W. Norton, 2003.
Lindsey, George R. “Statistical Data Useful for the Operation of a Baseball Team,” Operations Research 7, no. 2 (March-April 1959): 197-207.
Mottley, Charles M. “The Application of Operations Research Methods to Athletic Games,” Journal of the Operations Research Society of America 2 no. 3 (1954): 335-38.
Okrent, Dan. “He Does It By the Numbers,” Sports Illustrated, 24 May 1981, 43-51 [re: Bill James].
Oliver, Dean. “About The Journal of Basketball Studies,” http://www.rawbw.com/~deano/about.html, accessed 6 September 2020.
Oliver, Dean. “New Measurement Techniques and a Binomial Model of the Game of Basketball” (1991) at http://www.rawbw.com/~deano/articles/bbalpyth.html, accessed 6 September 2020.
Rau, Erik P. “Combat Scientists: The Emergence of Operations Research in the United States during World War II,” PhD diss., University of Pennsylvania, 1999.
Reimer, Alex. “Syracuse University Will Launch First Sports Analytics Degree in the U.S.,” Forbes, 11 May 2016, https://www.forbes.com/sites/alexreimer/2016/05/11/syracuse-university-will-launch-first-sports-analytics-degree-in-the-u-s/#1f1f146b73b4, accessed 9 September 2020.
Thompson, Dick and Tom Hufford. “A History of SABR,” Society for American Baseball Research, https://sabr.org/history, accessed 28 July 2022.
“Tribute to PFRA Founder Bob Carroll (1936-2009),” Professional Football Researchers Association, 2009, http://www.profootballresearchers.org/bob-carroll-tribute.html, viewed 17 December 2019.
Waring, Stephen P. “Cold Calculus: The Cold War and Operations Research,” Radical History Review 63 (Fall 1995): 29-51.