The Origin of Analytics: Charles Reep
The origins of soccer analytics, albeit in a rudimentary form, are typically traced to Charles Reep, an Englishman, war veteran and dyed-in-the-wool soccer fanatic. Legend has it that during one game in 1950, Reep became upset at a team’s pathetic scoring attempts and began recording observations and trends he saw in a notebook. In that game, he noted the large ratio between attacks and goals, concluding that even a modest boost in scoring efficiency would result in an additional goal per game. Reep developed a system to notate spatial information for each play, quickly scribbling as the match went on, allegedly spending 80 hours after each game analyzing and interpreting his shorthand.
Reep became soccer’s first known analytics employee when he was hired part-time by the Brentford team, which avoided relegation after winning 13 out of 14 matches and doubling their goals scored. He would bring his message to different teams, instilling a culture of attacking that was allegedly borne out by data. While he stopped advising teams in the late 1950s after one of his teams took a nosedive, Reep continued his observations and calculations. His legacy was primarily rooted in pushing clubs to reduce passing and emphasizing the benefit of using fewer long passes to propel the ball downfield—a notion based on his observation that most goals are scored on plays with fewer than four passes.
This conclusion, however, was logically flawed. He had always preferred watching the long-ball game and had a preconceived notion that it must be the best strategy, so his shoddy analysis would naturally confirm his initial assumption. His calculations relied on counting the number of passes and splitting such sequences into different categories, documenting the number of goals scored for each category. But most plays have three or fewer passes in the first place, so it naturally follows that more goals will be scored on the most common type of play. When the proportionality of each sequence type is taken into consideration, the better strategy is actually the opposite of what Reep suggested. In reality, analysts have found, more passing is tied to more scoring.
Analytics would grow more complex and commonplace with the advent of companies putting technology and improved computing to use in analyzing play on the pitch. Gone was the observer furiously scribbling player positions and goal shots on the sidelines, and in his place was a more technologically sound system. An interesting development in soccer analytics was the creation of rival companies, each using different tracking mechanisms or measuring different actions to inform teams’ choices. But to the public—and even to prospective teams that debated which company’s number-crunching to employ—the differences between these firms was not obvious.
One such company called Opta was founded in 1996 and quickly partnered with the English Premier League to build infrastructure to provide data insight from matches. The business began collecting real-time data in 1999 and making that information available for individual players in 2000. Five years later, the company would streamline its tracking technology to provide live position tracking information for subscribers.
One year before Opta’s beginning, the rival company Prozone also began collecting its own data and measurements to make insights. Interestingly, while both companies contract with the English Premier League and share information, the actions and metrics tracked remained highly secretive. Unlike baseball, where the latest sabermetrics are subject to continuous theorizing and debate, soccer’s newest metrics companies were shrouded in secrecy. Opta and Prozone (purchased by STATS LLC, now Stats Perform, in 2015) tracked separate player actions—though fans and even prospective users were unsure what differentiated one from the other.
Prozone was the brainchild of engineer Ram Mylvaganam, who was introduced to the intersection between technology and soccer when the consulting firm he worked for got a contact with Derby County football club. The initial “Prozone”—or professional zone—was a group of massage chairs that emitted electrical pulses where players would sit as coaches went over film from the last game. Mylvaganam talked with the team’s coaching staff and remarked on the inefficiencies of the video systems Derby County was using. He may not have been familiar with the intricacies of soccer, but he did know about technology and data. Aiming to apply new technology to the sport, he purchased part of Video Sports, which had developed pixel-tracking cameras, and installed eight of them around the club’s stadium. There were literal blips in implementing the system at first, as some players would be completely missing from tracking records, but it was an unprecedented step in harnessing the power of cameras to track player movement.
The Movement Spreads
When the Derby County coach, Steve McClaren, was hired by Manchester United in 1999, the change of settings allowed Mylvaganam to bring his ideas and improved tracking metrics to a brighter stage. Manchester United brought in Prozone, which had been working at no charge, and offered a cash payment if the club won a trophy. They won three, and Prozone’s business began expanding. In 2000, six teams from the Premier League could be counted among the company’s customers, benefitting from the opinions of Prozone consultants dispatched to address team needs. Bolton, a lower division team, requested the new firm’s services and would be a proving ground for whether the system could boost a second-tier team; after adopting the technology and notching a championship, Bolton was promoted to the Premier League. They even excelled in the Premiership and earned a UEFA Cup berth in 2005 for the first time in club history. Analytics firms and teams began to realize the importance of ball possession and passing in the last third of pitch, which is strongly correlated with team performance.
Soccer was more resistant to the analytics revolution than other sports. European teams largely shunned the onset of numbers and graphs, preferring the deeply imbued emotional aspect of the game and intangible grit of players on the pitch. But in the first decade of the 2000s, teams began establishing nascent analytics departments tasked with analyzing data from Opta or Prozone, and analytics had become common fare in the top leagues by the end of the decade.
The increased consumption pushed analytics giants to recruit more analysts and trackers to supply a greater volume of teams. In 2014, Opta employed around 350 part-time workers tasked with documenting passes, headers and goals. The recording was done by human eye, with employees marking on a screen where a pass was received or deflected–analysts would confer on controversial decisions and come to a consensus or summon a supervisor. Sophisticated hand-eye coordination is a must, as workers are required to rapidly indicate player positioning with a mouse click using their right hand, while typing in player name and details using their left. The advanced motor skills weed out many soccer fans eager to get paid to watch and analyze matches.
However, Prozone’s optical tracking system means that the company doesn’t need analysts to consistently key in player positions like a modern-day Charles Reep. Instead, such tracking is automatically taken care of by cameras, and only requires human intervention to clean up data muddled by clustered players or other technical inconveniences. Opta, on the other hand, touts its advanced data analytics packages. The company also has inked deals with television providers to feed live statistics and tidbits to commentators throughout a broadcast; the announcers and analysts message one another with questions and relevant responses.
Outside firms have also sprung up to provide tailored consulting services to teams. 21st Club, so named because of the 20 clubs in the EPL, provides data insights primarily to teams seeking talent on the transfer market. As big-name players on prominent clubs cost a fortune as transfers, some teams have shifted their attention to first-tier players in second-tier leagues. For example, Red Star Belgrade—one of the elite teams in Serbia’s league—used 21st Club’s services to find Lorenzo Ebecilio, an obscure player from Cyprus. In his first season with the team, Ebecilio notched a Champions League qualifier goal and later recorded an assist. Thus, improved analytics and metrics can not only help teams devise strategy and improve the players they have, but also seek out undervalued players on less prominent teams.
Analytics in American Soccer
The onset of analytics in the American soccer realm has been somewhat slower. While the United States was the birthplace of sabermetrics and more advanced baseball analytics, that enthusiasm did not carry over to many Major League Soccer clubs. Perhaps this is due to less money and smaller fanbases than in the EPL and other European soccer leagues—there may be less reward for employing a well-trained analytics staff or company to modestly increase a team’s competitive edge.
In 2015, some clubs explained that they didn’t find soccer analytics useful. An Orlando City Soccer Club spokesperson said, “nothing can replace actually meeting a player, understanding his character on and off the field, and seeing if the chemistry works well with the organization.” FC Dallas and the Chicago Fire said they had toyed with analytics but did not find them to be particularly helpful. However, other teams were more numbers-focused. The San Jose Earthquakes count themselves among data-conscious clubs, perhaps unsurprisingly, given that Billy Beane of Moneyball and Oakland Athletics fame was a partial owner. Sporting KC and the New England Revolution, two MLS teams that found success in the mid-2010s, were also harnessing data to make insights at the time. Sporting KC tends to use analytics with fitness and performance angles, evaluating whether players’ stamina and statistical profile would make them a good fit in Kansas City. The New England Revolution also uses such a system to analyze potential recruits and signings, but unlike Sporting KC, takes a more theoretical approach in evaluating whether certain strategies work and using analytics to scout opposing teams.
Overall, in the mid-2010s, the MLS was far behind the EPL and other top-flight European leagues in the soccer analytics revolution. In the last several years, however, American soccer has made strides in contributing to a more vibrant analytics scene. Read more about MLS’ newest partnership with Stats Perform, in addition to the latest soccer technologies and metrics, here.