Chris Anthony is obsessed with fantasy football. So much so, he set out to find a way to beat his friends and families at the game. Ten years ago, he built a custom app with Claris FileMaker to leverage real-world data to bring artificial intelligence (AI) to the fantasy team selection process. Thus, FantasyOmatic was born — an app that powers a popular web service, providing football enthusiasts with weekly ratings and rest-of-season projections.
https://www.claris.com/customers/stories/building-a-winning-fantasy-football-team-on-data-not-luck
When setting off to build an intelligent fantasy football app, Chris knew just where to start, having used FileMaker for low-code development to support some personal hobbies. With FileMaker, the application, FantasyOmatic, used machine learning (ML) to assess a variety of data points, such as player career performance, game location, and defensive matchups, to calculate player rankings and predictions. The approach has evolved over the years, incorporating new sports data and allowing the predictive performance service to get better with new FileMaker releases.
“Claris keeps releasing new FileMaker technology, allowing us to continuously modernize our platform,” Chris continues. “For example, we originally used an API to connect to a system to run our ML model and create the ratings for us. Now, with FileMaker’s native API, we integrate directly and use its core ML functionality to pull data right out of the application, running predictions on our website and allowing subscribers to access for their own fantasy teams.”
Earning industry respect.
To put FantasyOmatic to the test, back in 2012 Chris entered the app’s projections into a fantasy football industry accuracy contest. At the end of the 2012 NFL season, the projections ranked as the 12th most accurate of the year, besting experts at known fantasy sites, such as ESPN.com, NFL.com, and CBS.com. Today, FantasyOmatic is routinely recognized for its prediction accuracy, fantasy tools, and weekly rankings — powered 100 percent by machine learning approach.
“The first few years our predictive game model was in place, we went on to accurately predict three of four consecutive Super Bowl winners,” says Isaac, a developer with FantasyOmatic. “Our system outperformed many of the big data predictive companies and their models, earning us respect right out of the gate.”
Fantasy tools for all.
FantasyOmatic offers a mix of free and subscription-only content, providing users with valuable and easy-to-use insights related to team’s strength of schedule, defensive ratings, individual team matchups, and individual player comparisons.
“We use a calculation field to write a weekly preview article highlighting players to look out for each week,” adds Luke, a developer with FantasyOmatic. “Readers probably don’t realize that it’s the application’s AI functionality authoring the content. It’s that user-friendly.”
Competing against the big dogs.
Organizations spend big bucks to compete in the fantasy football arena, throwing millions of dollars behind applications and prediction tools. FileMaker and the power of core ML allows FantasyOmatic to play right alongside companies with larger staff and more extensive budgets.
“We are a small team but making a big impact on the industry,” Chris adds. “With FileMaker, we’re able to do a better job predicting fantasy football outcomes — and do it more efficiently, cost effectively, and with more agility.”
