Skip to Main Content

STATS’ Data and Automated Language Generation Lead to Unmatched Fantasy Content

By: Andy Cooper

Long gone are the days of fantasy football scores lagging behind NFL games. Team owners have expected scoring to update in real-time for a number of years, but individual player news and analysis have often been many hours – if not days – behind because of the exclusive reliance on humans.

At STATS, the process is changing in a way that’s differentiating the company’s player-specific fantasy content from the market. Along with parent firm Vista Equity Partners, STATS acquired Automated Insights in 2015 and is now leveraging the marriage of STATS’ comprehensive data with Ai’s natural language generation technology to create automated, player-specific content with unmatched efficiency.

STATS is already the first to use these methods for baseball, and its first season doing so gives an idea of how it will similarly impact fantasy owners’ Sundays and midweek fantasy matchup prep this fall. MLB just passed the midway point of the season, and STATS’ player news entries stood at 26,602 through July 4. A year ago at that point, it was 5,536 with a season total of 12,580 produced entirely by humans.

“Traditionally, data has been our differentiator,” said Brian Orefice, who oversees STATS’ editorial content as the company’s head of production for the Americas. “Now, we’re layering on top of that an automated solution that is 100 percent scalable and unmatched in the industry. What has traditionally been a labor-intensive product is now relatively turnkey with improvements to latency and scale that have given us a value proposition above the strict kind of data strength that we’ve had previously. It does the amount of work that, frankly, nobody is in a position to do. It has exponentially increased our scale and minimized our timing.”

No, that doesn’t mean robots are writing your game previews, injury updates or player advice. The complexity of sport makes expert analysis necessary. But with fantasy platforms continually becoming more detailed and evolving from the traditional QB-2RB-2WR-Flex-TE-D-K format, the associated player news and analysis needed to expand. Providers must create more content to meet needs of clients with deep and daily platforms, and it comes with a growing expectancy to do it much faster.

It’s possible to meet those demands at STATS because an automated system allows for prompt reporting on everyone on the field rather than selective, time-consuming output created exclusively by humans parsing box scores and individual player game logs. Rather than eliminating the human element, automation frees up writers to shift focus from straightforward performance reporting to providing clients and fantasy owners with useful advice and analysis as quickly as possible. In other words, leave the humdrum to the machines. Let the experts dig in.

“It allows the writers to focus on even more detailed stuff,” Orefice said. “We’re constantly pushing out more and more insightful and contextual information.”

2017 will be the first NFL season STATS’ clients see this spike in player news, and the content created will make for analysis that no longer overlooks the performance – or lack of performance – by your WR3 or TE2. Trends and insights will be identified with a new level of productivity which allows team owners to size up a player performance while it’s fresh in mind rather than trying to do so between sips of coffee from the office chair on Monday morning.

And this is a process that will only become more intuitive and immediate in coming seasons as data science allows for further automation.

“This is just the tip of the iceberg,” Orefice said. “Eventually what’s going to happen is there will be an insights engine that sits on top of that logic. I’m excited for what we’ve done. I’m even more excited to get this insights layer on top of it to really add additional context to what we’re doing.”