Sport these days, like it or not, is big business. Given the large numbers now involved, managers and owners are increasingly turning to data analytics in order to mitigate the risks they take when bringing in players. In 2007, the annual MIT Sloan Sports Analytics Conference hosted 175 attendees on the MIT campus. In 2014, the conference filled the Boston Convention & Exhibition Center.
Traditionally, teams have relied on the gut instinct and intuition of scouts and management to judge a player's merits. Philadelphia 76ers coach Doug Collins perhaps most famously exemplified the resistance to a change in this methodology, answering one reporter who asked if he was an ‘analytics guy’ by saying: ‘I’d blow my brains out. There's 20-page printouts after every game - I would kill myself.’ ‘My analytics are here…’, he continued, pointing to his gut.
However, many teams are now recognizing the benefits that come from using advanced data analysis. Sports have always had a wealth of statistics to draw on, although it is only since the turn of the millennium that teams have begun using advanced analytics to divine which are most relevant to the outcome of a match. Take up has been particularly high in baseball, spurred by Billy Beane’s success using sabermetrics with Oakland Athletics, famously depicted in Michael Lewis’s 2003 book, ‘Moneyball: The Art of Winning an Unfair Game’ and the subsequent film starring Brad Pitt. Almost all MLB teams now employ at least one statistical analyst. Traditional scouting is still recognised as an important part of the process, however it is now accepted that it would be foolish to abandon a wealth of player-performance data in order to maintain the romantic vision of the solitary genius judging talent, the wanderer above a sea of fog.
American football is clearly a different beast to baseball, and the use of data analytics is arguably a far more difficult proposition for a number of reasons. Last week’s NFL draft, however, involved nine teams who have embraced analytics, including last season’s Super Bowl champions, the New England Patriots. Seven more teams use it to some degree.
Data analytics is used in American football in a number of ways, such as developing strategy and evaluating opponents. One of the most challenging ways for it to applied in the sport, particularly in comparison with baseball, is in its evaluation of individual players. Football is a far more complicated game than baseball, one less reliant on individuals and more on the coordination of all 11 players’ efforts. This also makes it exceptionally difficult in the context of the draft in that NFL teams must judge whether a player’s success playing at college will translate into the pro leagues, where the step up in terms of quality is arguably greater than most other sports.
The response to this challenge has been greatly aided by a surge in new technologies, especially wearable technology. GPS units and RFID tags are now being used by teams to measure speed and track the distances players run in practice. The NFL joined forces with Zebra Technologies last season to put two RFID tags in every player’s shoulder pads. This provided real-time player statistics to use for quantitive analysis. Such data can then be compared with other players coming into the league in order to judge whether they represent a good pick. There are a number of examples of clubs paying over the odds for players as a result of their failure to utilize these resources, one being Georgia outside linebacker Jarvis Jones. Jones was shown to be slow in all measures of analysis, but was still selected as Pittsburgh Steelers’ 17th overall pick in the 2013 NFL draft. Since then, he has played 21 games and made three career sacks.
Teams are also using data analytics to get the inside track in the draft process itself, using data to predict when a player may be picked. Teams take into account the mock drafts to determine when a prospect might be drafted, and use this information to work out how to get the best value when trading early picks for more later in relation to this.
When it comes to buying a human being to play in a high contact physical sport like American football, it’s always a gamble. Injury issues, personal crisis, and a sudden drop in form are all factors that can be impossible to predict. And while it may be true that data cannot show heart, and there is still definitely room for intuition and gut instinct when finding players, with wearable technology and data analytics constantly improving teams must take every advantage they can get in order to predict MVPs of the future and get value for money in the ultra competitive draft arena.