For most of the world, football brings to mind powerhouse players like Lionel Messi of FC Barcelona, or Diego Maradona of Argentina. But for North Americans, football is more likely to bring up a personality like Tom Brady of the New England Patriots, or coach Dan Marino of the Miami Dolphins franchise.
While both sports produce absolutely fanatical devotees from all over the world, they are decidedly different—especially in the realm of analysis. Because the games are so different, so are the methodologies and application of data collected from the matches.
Unlike soccer, many fans of football understand their fair share of data analytics just by interacting with their team on a regular basis, monitoring star players, and attempting to predict what will happen during the annual NFL draft.
Pundits and diehards stay up to date on stats like receiving yards, rushing yards, passing yards, touchdowns, tackles, sacks, interceptions, kicking, punting, and yards allowed. That was ten different categories. Go into any local dive bar in a city with an NFL team, and listen to one of the US’s 160 million football fans start citing stats from their team’s run in the mid-70s all the way to the current season.
On the other side of deafening cheers filling a stadium, classic food that Americans associate with the game of football, and being surrounded by thousands of other passionate fans, is hard mathematics. Of course, betting on a game is a common part of American fanfare—in 2019, Nevada alone dealt with $5.3 billion in sports betting. But many Americans may not realize just how much math and data analysis is required to make informed NFL future picks and predictions that so many interact with. From biometrics to trajectory, analytics changes football from a full-contact sport to a scene from Good Will Hunting.
The Recent Rise of Sports Analytics
In short, the gathering and summary of sports data allows for teams to reflect on their play and make concentrated efforts that are specific to a lacking sector. Though it may be surprising, the term ‘sports analytics’ is actually a fairly recent term that has taken off in recent years, popularized by the US film Moneyball in 2011, though the concept is based on the book which was published in 2003.
Since the rise of sports analytics, gone are the days of a ‘gut feeling’ that may drive a coach or manager to make pivotal decisions on the field. Today, all 32 NFL franchises have a separate sports analytics department. Hard data is collected and handed over to a group of mathematicians who will crunch these numbers to assess patterns and deficits that the team will utilize in a variety of matters.
For instance, a team may collect data on a potential recruit. By collecting this data, a mathematician can assess these numbers and compare them to leading players to see how a potential recruit stacks up. Decision-making is, quite simply, driven by data and the subsequent data profiles that are created from hard numbers. These profiles determine whether or not a recruit will be offered a spot on the team, as well as potential drafts and free agents.
However, these profiles aren’t only being used by NFL teams to help build all-star offenses and defenses. Sports analysis is a separate industry in which larger segments of data are analyzed—not only player stats. By analyzing data on entire teams throughout a season, and applying these numbers to multiple season runs, experts will create odds and choose likely outcomes for NFL events, like the Super Bowl and the annual NFL Draft.
Zoom out farther and this expert analysis does more than predict futures—it also directly informs the entire sports betting industry, as aforementioned. In offshore sportsbooks alone, NFL sports betting is a $100 billion-dollar industry. It is the most wagered on sport in the US, with the NBA, MLB, and NHL bringing in only $75 billion dollars collectively.
So, it makes sense that NFL franchises would employ their own data analytics department. And, in terms of keeping players at optimum fitness, data can also help improve injury prevention. By investigating the finest details of their team’s performance, like how long a player sprints a 40-yard dash or how often a player intercepts a pass, teams can make direct changes to improve play.
Modern Programs for Data Analysis
Though it isn’t yet used by the NFL, the NBA has begun to institute ‘Player Tracking’ programs. These programs, such as SportVU, involve placing cameras at strategic points around a basketball arena and court. Over the course of a game or workout, these cameras can take up to 25 photos per second and create data about how fast a player runs, how many times the player touches the ball, how many rebound opportunities the player has, and how many times they capitalize on these—on and on the list goes.
With such technology available over such an unfathomably broad scope of action, the question shifts to: how do analysts know what data to crunch and how to apply it? These finer questions will likely be hammered out in the years to come, and only after the misapplication of such information.
Another recent advancement is the placement of minute sensors inside of sports equipment, such as a golf ball, which can determine how fast and how far a ball is going in real time. The same can be said for the swing of a baseball bat, or the trajectory of a tennis ball. This type of immediate data can be applied for analyzing sports or informing live betting at the most breakneck pace possible.
It can also be used to create massive multi-year sets of data that compare current activity and stats to data available from decades ago. At this time, unlike programs such as SportVU, data sensors are expensive to obtain and maintain—especially considering the treacherous reality of surviving inside a golf ball or baseball bat, which are battered every day.
However, all the data and even most expert analysis in the world cannot account for an athlete’s human tendency. At the end of the day, strong data points can’t prevent an accident or an off day in even the most conditioned athlete. Not yet, at least.
