As the NFL Kicks Off, AI Scores a Touchdown

When fans tune into NFL games this fall, they may not realize how much of what they’re watching is being influenced by artificial intelligence (AI). That hot rookie receiver from a small school in Iowa who’s outplaying rookies from “football factory” schools, that linebacker who was in exactly the right place to stuff a fourth-and-goal run, and those “completion probability” stats that broadcasters show every time a receiver makes an acrobatic catch … all are influenced by AI.

And moving forward, when you see two players walk away from a violent collision uninjured, that too may be the result of AI. The NFL recently announced that it has joined forces with Amazon Web Services (AWS) to launch a new challenge for AI developers: a $100,000 prize to the data scientists who come up with winning models for automatically identifying players on the field who are involved in helmet collisions using game footage. (Think of Madden NFL – where you can keep track of all the players on the field with nametags and arrows.)

Eventually, this collaboration will develop the Digital Athlete, which the NFL says will be “a virtual representation of an NFL player that can be used to better predict and hopefully prevent player injury.” The goal is to identify defensive players who may need coaching to change their tackling technique, or offensive players who need to change how they enter collisions, with the ultimate goal of reducing injuries to both.

There are many other areas where AI will play a role in the NFL this year:

  • Draft Classes: Teams have learned that emotion based-decisions can result in bad draft picks (“I loved watching that guy in the national championship game – we need to draft him!”), so more and more teams are taking emotion out of the process and instead using AI for data-driven decisions. It is the concept of Moneyball, enabled with software that can process millions of data points, versus Billy Beane scanning papers manually. Because the technology uses computer vision to analyze patterns and performance, and compare those against player performance through the years, this “Moneyball” approach stands to deliver more hits than misses for draft picks.
  • Play Calling: AI and computer vision can identify “tip offs” based on where players are positioned, so the defense can be alerted to a likely play. Likewise, the defensive alignment can be used to inform the quarterback on the likelihood of a blitz, etc. 
  • Stadium Security: And, we’d be remiss if we didn’t talk about how AI can improve the fan experience at the game. Systems like Patriot One’s AI enabled threat detection platform can eliminate the delays caused by walk-through metal detectors at gates, and alert security guards to developing incidents throughout the stadium, so they can intervene at the earliest possible point and ensure the maximum patron experience. The technology most recently was used at the Giants-Browns pre-season game

Fewer injuries, and better players, play calling and stadium experience. Looks like a championship year for AI in the NFL!