ESPN’s Fantasy Football Turns to IBM Watson For Answers
For all the braggadocio, silly names, outlandish bets and long-lasting rivalries, fantasy football is, at its core, an exercise in data. Which players will produce the best output? Does past performance indicate future results? Is this trade equitable?
There is no shortage of data. From the pure stats to the analysis, it can be a bit dizzying to sort through it all to make a rational decision.
To help make all that data actionable, ESPN’s Fantasy Football product has turned to Watson. Watson, IBM’s cognitive computing platform that can answer natural language questions and is perhaps best known for dominating Jeopardy!, has been working with ESPN’s fantasy platform for four years, crunching the numbers and helping players make decisions about who to start and who to sit. Last year, for example, Player Insights with IBM Watson analyzed 228 million articles to deliver 25 billion insights for fantasy users.
And after three seasons of predicting which players are likely to boom or bust, this year IBM and ESPN are utilizing Watson’s skills to help with trades.
Trade Assistant with IBM Watson is meant to help take the angst out of trading by suggesting trades that are equitable. “Most people have a negative experience with trades,” says Chris Jason, the senior director of product management for ESPN Fantasy & Interactive. “It’s, ‘I keep getting trades from this crazy person in my league.’ And for beginner players, that’s a huge hurdle to climb. So the idea we’re most excited about is the match-making piece—using Watson to find mutually beneficial trades.”
That’s the sort of differentiating feature that helps make ESPN’s Fantasy Football product a knockout.
“The product is as healthy as it’s been,” says Jason. “We put up record numbers, we’re growing the number of players and the time they spent in the product. And Watson is contributing to that.”
In addition to suggesting fair trades, Watson helps players make decisions by assessing the value of a player, the cost of losing that player and the overall equity of a trade.
“The premise we set out with trades is that we were focused on fairness and value,” says John Kent, Program Manager for IBM Sports & Entertainment, Partnerships. “We didn’t want those lopsided trades. We wanted some realistic trades that are good for both teams and hopefully improve the likelihood of it being accepted. And we’re taking into account multiple perspectives. We put market valuations on the players, there’s the cost of ownership, the cost of losing the player—there’s a lot that goes into it. And the Player Insights set the foundation for trades.”
Integrating Trade Assistant is another in a long line of IBM-led collaborations that have helped augment ESPN's platforms and content. The two companies have worked for years on the tennis side, so it made sense to deepen the partnership. For both companies, Watson has helped to make ESPN’s Fantasy Football offering a standout in the market, all while raising Watson’s profile.
“IBM is primarily a business-to-business company,” says Kent. “Consumers see the Watson ads, but they don’t know the capabilities. So there’s nothing like putting Watson into action for consumers.”
Adds Jason: “The typical fantasy player is a very valuable user. They tend to be more tech savvy. They tend to use mobile more. They tend to have a higher median household income. If you want to get the brand in front of folks, this is a great way to do that. That person playing fantasy on Sunday might be making important business decisions on Monday.”
And since Fantasy Football is based on making decisions according to available data, integrating an AI assistant that specializes in that is a natural extension of Watson’s capabilities. “We’re solving a data problem,” says Elizabeth O’Brien, IBM’s program director, WW Sports & Entertainment Partnership Marketing. “How do we manage data and find the needle in the haystack? When we started out with Player Insights, we were using natural language processing to understand where those nuggets really were. Trade is just another level of complexity beyond roster setting.”
One roster decision can be the difference between fantasy glory and last place (which, depending on the league, can come with a score of punishments, ranging from tattoos to taking the SATs and pretty much anything else you can think of.) To be able to leverage the information provided by Watson to avoid embarrassment is every fantasy player’s best friend.
