Sports analytics

Sports analytics is a relatively modern field of study that uses statistical data to maximize the performance of athletes and teams. Analysts examine data from past performance—either of individual players, teams, or the league as a whole—and try to use that data to give athletes and teams an advantage. Analytics can be used as a guide for in-game strategy, to suggest personnel moves for players, or to manage a player’s workload.

The basic framework for sports analytics developed over the course of the twentieth century, as athletes, coaches, and front office personnel searched for ways to gain a competitive advantage. However, the practice did not fully catch on until Major League Baseball’s Oakland Athletics and their general manager Billy Beane built a successful team with statistical analysis in the early 2000s. Beane’s strategy was detailed in the 2003 book Moneyball: The Art of Winning an Unfair Game, which also inspired a 2011 film. Beane’s strategy was soon copied by other baseball teams and adapted into other sports. As technology improved and became more widespread, the use of sports analytics further evolved. By the 2020s, it had become an integral part of both on-field play and off-field decisions.

Overview

While sports have been around for thousands of years, it was not until the nineteenth century that a wide variety of organized sports began to flourish across the United States and Great Britain. Societal changes brought about by the Industrial Revolution enabled people to have more leisure time, and some began to fill that time by playing sports. Baseball developed in the United States in the mid-nineteenth century. The game was likely inspired by several British games brought to the American colonies in the eighteenth century.

In 1859, sportswriter Henry Chadwick was credited with creating the box score, one of the first statistical tools used to analyze baseball games. The box score was a printed record of statistics such as at bats, hits, runs scored, runs allowed, etc. A person could read a box score and see that a player went to bat four times, got two hits, and drove in two runs. It would also show that a pitcher pitched for five innings, gave up six hits, and three runs. The box score also showed other stats, such as which players hit home runs or if any errors were made.

The box score allowed the growth of other baseball statistics, such as batting average and earned run average. A player’s batting average is the number of hits divided by the number of at bats. An earned run average (ERA) is the number of earned runs a pitcher gives up divided by innings pitched and multiplied by nine—the number of innings in a standard game. This number shows how many runs a pitcher gives up on average over the course of a nine-inning game.

Box scores became a must-read feature for baseball fans and later became common in almost all professional sports. However, box scores had some basic problems that illustrated their shortcomings as an analytical tool. For example, a baseball box score may say that a player hit a double, but it does not say how the ball was hit. It may have been a hard-hit ball to the outfield or a lucky “bloop” hit that just happened to fall between two outfielders.

In the twentieth century, coaches, team officials, and sportswriters began to look at statistics to see if they could find some way to predict performance in the numbers. F.C. Lane, a writer for Baseball Magazine, published a 1925 book in which he analyzed the ability of Major League Baseball hitters. He determined that a high batting average did not necessarily make players better hitters. He placed more value on hits that went for multiple bases—such as doubles, triples, and home runs—than singles, which went for one base.

In the 1940s, Brooklyn Dodgers General Manager Branch Rickey hired a statistician named Allan Roth to analyze his team’s performance. Among his innovations was the use of a platoon system where managers insert different players in the lineup each game based on the matchup against the opposing pitcher.

These early forms of analytics were best-suited for baseball because the sport is more statistically driven and comes down to the matchup between a batter and a pitcher. However, coaches and team officials in other sports also looked for an advantage in statistics. Paul Brown, the coach of his namesake Cleveland Browns from 1946 to 1962, was the first to analyze game film to scout opponents and developed a system for evaluating college players in preparation for the NFL Draft. In the 1950s, Bob Spear, basketball coach at the US Air Force Academy, and his assistant Dean Smith, created an evaluation system to grade their team’s offensive and defensive performances. They felt that total points scored was not as important as points per possession. Smith would go on to have a hall of fame career at the University of North Carolina.

Sabermetrics and Moneyball

In the 1970s, writer and baseball fan Bill James began scanning box scores for unique statistical information and publishing his findings in the Bill James Baseball Abstract. In 1980, he called his research sabermetrics, a name honoring the Society for American Baseball Research (SABR), a group to which he belonged. Sabermetrics examined baseball statistics though an objective lens that looked at different data than commonly accepted statistics such as home runs, wins, and runs batted in.

Among James’ innovations was the statistic of runs created, a measure of a player’s value to his offense. Each offensive stat, such as hits, walks, stolen bases, etc., was assigned a weighed value—a home run would be worth more than a single—to determine a final score. He also predicted a team’s “true” winning percentage by examining the relationship between its runs scored and runs allowed. In addition, James examined a player’s minor league statistics to determine which Major League player his skill set would best match.

James’ work took a while to catch on, but by the mid-1980s, his books were selling thousands of copies, and he had caught the eye of several baseball general managers. One of those was Sandy Alderson, the general manager of the Oakland Athletics from 1981 to 1997. Alderson was a fan of James’ Baseball Abstract and read it as he built the A’s roster in the late 1980s. The team would go on to play in three consecutive World Series and win one.

Billy Beane was a former-Major League player working as an assistant to Alderson. When Alderson left the team in 1997, Beane took over as general manager. Oakland was a small-market team and did not have the financial resources to compete with big-market teams like the New York Yankees or Los Angeles Dodgers. Team ownership had also instructed Beane to keep the A’s payroll low. Unable to spend money on the best free agents, Beane began looking for a more cost-effective way to field a competitive team.

He turned to the ideas found in sabermetrics, particularly the idea that getting on base was more important than how a player got on base. He started to identify players who had a high on-base percentage and low strikeout rate. These players were patient at the plate and willing to draw a walk rather than swing at a difficult pitch trying to get a hit. Beane constructed the A’s around sabermetric ideals and turned Oakland into a perennial playoff team. Oakland won more than one hundred games in 2002 and 2003.

In 2003 Beane’s sabermetrics strategy was the subject of the book Moneyball: The Art of Winning an Unfair Game by author Michael Lewis. The book introduced sabermetrics to a wider audience and prompted several baseball teams to adopt the analytic strategy. One of those teams was the Boston Red Sox who hired Bill James as a senior advisor on baseball operations in 2003. James would stay with the Red Sox until 2019, during which time the team would win four World Series titles.

Beane’s strategy would receive even more attention in 2011 when Moneyball was made into a film of the same name. The film proved very popular and further boosted the profile of sabermetrics. By this time, data analysis was being used by teams in other sports, prompting the adoption of the term sports analytics to describe the process. By the late 2010s, teams in all the major professional sports had hired their own analytics staff and used data analytics in some form to guide team-building and on-field strategy.

The use of analytics was greatly enhanced by the growth of technology in the twenty-first century. Advancements, such as artificial intelligence technology, have allowed for faster and more efficient data collection, giving teams more information to process.

Applications

Although sports analytics are widespread in professional sports in the United States, they are used in different ways depending on the type of sport. Because baseball depends on individual matchups more than other sports, analytics play a larger role in crafting an in-game strategy than in a team sport like football. Analytics are still important for in-game decisions in football, but in that sport, they are used more often for player evaluation.

Boosted by technology, baseball analytics have evolved to include a myriad of statistics powered by mathematical data. One of the most widely used statistics is a metric called wins above replacement (WAR). This considers a player’s offensive and defensive capability and his position to determine how many more games that player will help his team win than the average Major Leaguer at the same position. The mathematical formula for WAR is complicated but takes into account the number of runs that a player scores through hitting and baserunning and the runs he prevents by his defense. For example, over the course of his career, Barry Bonds, who played from 1986 to 2007, has a career WAR of 162.8, meaning his presence on the field was responsible for that many more wins than an average player.

Another hitting statistic is called on-base plus slugging plus (OPS+), which is a combination of a player’s on-base percentage and slugging percentage adjusted for the ballpark he plays in. On-base percentage is simply the number of times a player gets on base—either by hit, walk, or error—divided by his at bats. Slugging percentage is the total number of bases a player gets per at bat. For example, a home run is worth four bases, while a single is worth one. This gives more weight to a home run than a single. OPS+ illustrates a player’s ability to produce runs and removes outside variables, such as the dimensions of a home ballpark.

Similar metrics are used to evaluate pitchers. ERA+ takes the standard ERA formula—number of earned runs, divided by innings pitched, multiplied by nine—and adjusts it for ballpark and opponent faced. The home run to fly ball rate (HR/FB) examines how many home runs a pitcher gives up per fly ball that he allows. This can identify a pitcher who leaves his pitches in spots where batters can hit them more easily. It can also identify a pitcher who just made a mistake and gave up a home run because of bad luck.

Following the success of Billy Beane and his Moneyball system, several National Basketball Association (NBA) teams also began to look at analytics to improve their on-court performance. In 2007, the Houston Rockets hired Daryl Morey as general manager. Morey’s career background was in sports strategy and statistics and not basketball. Like Beane, Morey revolutionized the NBA by focusing on statistics and analytics. Among his innovations was the true shooting percentage, a measure of a player’s efficiency in shooting from two-point range, three-point range, and the free-throw line. This metric measures more than shots taken and shots made. It weighs three-point shots as more valuable and includes free-throws made, which also result in points.

In the past, the NBA was dominated by tall players who put up shots close to the basket. With defenses clamping down, players often took two-point shots away from the basket. Analytics showed that taking a two-pointer away from the basket was less efficient than taking a three-pointer from even farther away. The added risk associated with the more difficult shot was outweighed by the additional points scored. This philosophy changed the style of the NBA, with teams looking for quicker, more accurate three-point shooters over tall, slower players. In the 1980s, after the three-point shot was adopted in 1979, players took an average of 2.4 three-point attempts per game. In the 2021–2022 season, only one team averaged less than 30 three-point attempts per game.

Many National Football League (NFL) teams also rely on analytics, although their use is more apparent in player evaluation than on-field decisions. Different teams have different styles of play and prioritize a certain caliber player over others. For example, teams with an up-tempo passing offense may value speed and quickness, while teams built to tun the ball may prioritize power and strength. Many teams use analytics to determine the players who best fit their offensive and defensive schemes.

One common metric used by the NFL is expected points added (EPA), which is a tool to determine how many points a team will score on a possession based on stats like down, yards gained, field position, etc. For example, if a team has a first-and 10 at their own 25-yard line, they may have an EPA of 1.06, meaning they are expected to score that many points on the possession. If the team loses yards on the first play, that value will decrease. If they gain 10 or more yards for a first down, the value will increase.

An area where analytics plays a significant role in on-field decision making is when going for a first down or fourth down or deciding to go for two points rather than one after a touchdown. Prior to the late 2010s and 2020s, coaches were often reluctant to go for it on fourth down for fear of turning the ball over to the other team. However, analytics showed that depending on the field position and yards needed for a first down, it was more advantageous for teams to attempt to gain the yards than to kick the ball away. Analytics also demonstrated the optimal times to go for two-points after a touchdown. As a result, coaches who rely more on analytics tend to go for it on fourth down more often than their counterparts in the past. They also typically use an analytics-based chart to see when they should go for two points as opposed to one.

About the Author

Richard Sheposh graduated from Penn State University in 1989 with a Bachelor of Arts degree in communications and journalism. He spent twenty-three years working in the newspaper industry as a writer and an editor before entering the educational publishing business.

Bibliography

Abbas, N. M. (2019, August 13). NBA Data Analytics: Changing the Game. Towards Data Science. towardsdatascience.com/nba-data-analytics-changing-the-game-a9ad59d1f116.

Advanced Stats. (2022). MLB.com. www.mlb.com/glossary/advanced-stats.

Fox, L. (2021, August 12). How The NFL Uses Analytics, According to the Lead Analyst of a Super Bowl Champion. Forbes. www.forbes.com/sites/liamfox/2021/08/12/how-the-nfl-uses-analytics-according-to-the-lead-analyst-of-a-super-bowl-champion/?sh=69a53555424e.

A Guide to Sabermetric Research. (n.d.) Society for American Baseball Research. sabr.org/sabermetrics.

Lindbergh, B., & Sawchik, T. (2019). The MVP Machine: How Baseball’s New Nonconformists Are Using Data to Build Better Players. Basic Books.

Shin, D. (2020, October 7). Evolution of Analytical Data in Sports. Medium. dabidsheen.medium.com/evolution-of-analytical-data-in-sports-8a096ace464c.

Steinberg, L. (2015, August 18). Changing the Game: The Rise of Sports Analytics. Forbes. www.forbes.com/sites/leighsteinberg/2015/08/18/changing-the-game-the-rise-of-sports-analytics/?sh=343a8cd14c1f.