Fans of the NFL anticipate the release of the schedule every year. Typically in May, the release gives the fans time to size up their opponents, monitor the current roster, and prepare for new players from the soon-to-be draft. The NFL attracts masses in the stadiums. Over 18 million fans were in attendance last season according to Statista. But how exactly does the NFL manage to make a schedule for a league with over a quadrillion possible combinations? The answer lies in an algorithm. An algorithm that works with the processing power and advanced machine learning of Amazon (AWS).
Before the NFL began utilizing computers, there was a pin board. On the board was every team from the league and it took the NFL’s team nearly an entire year to create the schedule for the following season. There are rules that the schedule must follow which eliminate some of the possibilities. This includes: teams playing their division, and all teams play each other once every 4 years, and must consider BYE weeks. These rules helped form the very foundation of the algorithm.
With a billion scheduling options per team, the NFL was looking for a way to save money and time creating the schedule. They turned to Amazon to identify ways to create schedules in minutes and not days. Amazon accomplished this by creating an algorithm that runs with machine learning. With thousands of different instances, they help create the final 80,000 schedules that will be reviewed by six NFL executives.
Some of these instances include:
With the help of Amazon, the NFL has been able to save over $2 million of work and hours. The final schedule is still handpicked from the 80,000 possibilities. The NFL takes into consideration the trends of teams. With the new flexible schedule, a team that may not have had a chance to play in prime time can earn their way in with a good win-loss record. All of these are done to give the fans the best experience possible. That may be done through key matchups, international games, or flex-scheduling.
This is all done to give the fans, the main product consumers, the best experience possible.The collaborative efforts between two-titans of the industry proves how strategic projects and partnerships can pay off for both parties. Solutions may not seem clear in the onset but with participation, planning, and execution, they become evident. The NFL wanted to maintain some of their previous scheduling and planning powers and used machine learning to alleviate the heavy lifting of manual scheduling. In doing so, they still have full control over the schedule and can even switch games with flex-scheduling but save 2 million dollars a year.