Trust in AI for fantasy football grows as full automation looms

Trust is growing in AI technology used by fantasy football players to make weekly lineup decisions, but expert opinions differ on a fully automated future.

More fantasy football players are using artificial intelligence software for lineup decisions when competing against friends, coworkers, family members and strangers in their virtual leagues. File Photo by Tannen Maury/UPI
More fantasy football players are using artificial intelligence software for lineup decisions when competing against friends, coworkers, family members and strangers in their virtual leagues. File Photo by Tannen Maury/UPI

MIAMI, Oct. 15 (UPI) -- Trust is growing in AI technology used by fantasy football players to make weekly lineup decisions, but expert opinions differ on whether enthusiasts will see a fully automated future.

"With what Moneyball did to baseball, that is now happening to fantasy sports because we have a democratization of analysis tools and capabilities that everyone can access," said Eoin McMillan, founder of Sourcetable, a powerful AI spreadsheet.

In Moneyball, an unconventional baseball manager and a young economics expert use data-driven analytics, in contrast to traditional scouting to build a competitive team on a shoestring budget.

Some 29 million people in the United States play fantasy football, with 14 million playing this year through ESPN. Millions more battle friends, family members, co-workers and strangers in fake football fracases through Yahoo, CBS, Sleeper and other platforms.

But now, have more help than ever -- in the form of artificial intelligence. That assistance is automatically provided by mainstream fantasy football platforms, but some fans also discretely download apps, consult Chat GPT and utilize other software to gain an edge.

Recent surveys show increased trust in AI-generated sports content. File Photo by Tannen Maury/UPI
Recent surveys show increased trust in AI-generated sports content. File Photo by Tannen Maury/UPI

ESPN and IBM are in their ninth year of a partnership behind AI-powered watsonx, which will provide more than 36 billion insights for fans this NFL season. Watsonx data is presented in the form of personalized player overviews, waiver and trade grades, boom and bust probabilities, a trade analyzer and more.

"We are not doing it to replace anything," said Elizabeth O'Brien, program director for IBM sports and entertainment partnerships. "We are doing it to basically find the proverbial needle in the haystack. ... What we found with our research is that fantasy fans always want another source of information, another source of input, another source of reassurance they are making the right decisions."

Fantasy football fans are turning to AI technology used on Wall Street at an increasing rate to make weekly lineup decisions. File Photo by John Angelillo/UPI
Fantasy football fans are turning to AI technology used on Wall Street at an increasing rate to make weekly lineup decisions. File Photo by John Angelillo/UPI

A 2024 study conducted by Morning Consult found that 58% of respondents aged 18 to 29 believed AI would have a positive impact on sports, while 40% of those over 55 agreed with that assessment.

An August survey of more than 20,000 sports fans from a dozen countries revealed that 63% expressed trust in AI-generated sports content.

ESPN and IBM partnered to deliver AI-powered watsonx to provide data for fantasy football players, like the Week 7 boom-or-bust likelihood for Cincinnati Bengals wide receiver Ja'Marr Chase. File Photo by John Sommers II/UPI
ESPN and IBM partnered to deliver AI-powered watsonx to provide data for fantasy football players, like the Week 7 boom-or-bust likelihood for Cincinnati Bengals wide receiver Ja'Marr Chase. File Photo by John Sommers II/UPI

While ESPN's implementation of watsonx has grown to full implementation for its users, other platforms present AI assistance to varying degrees. Yahoo's AI immersion, which is in its early stages, leans toward draft preparation and custom draft and matchup recaps, which are delivered to millions of users.

Yahoo experienced a 200% year-over-year increase for subscribers to Yahoo Fantasy Plus, which includes cheat sheets with AI-powered player insights that continuously update.

Atlanta Falcons running back Bijan Robinson has
Atlanta Falcons running back Bijan Robinson has "boomed" three times and has yet to "bust" this season, according to IBM watsonx, which is used on ESPN's fantasy football platform. File Photo by AJ Sisco/UPI

Sleeper allows users to conduct mock drafts against AI. And CBS provides AI-generated information to its users through SportsLine's machine learning model.

"Ultimately, we're in the first quarter of our AI journey," said Andrew Machado, head of product, design and technology at Yahoo Sports.

"We're excited by how we're using the technology today to create better experiences for our fans, but we're just scratching the surface. We look forward to continuing to build with AI, guided by our north star of helping fans have more fun and be smarter fantasy managers."

IBM and ESPN initiated their fantasy football union through a pilot program during the 2017 NFL season. O'Brien said ESPN had a "business problem" due to users not completing enough trades -- an indicator used to gauge platform engagement-- and that led to the AI advancements.

Since then, ESPN and IBM increased data sources and the accuracy of their analysis, driving consumption and leading users to inhale ESPN content streams.

"To maintain their position as the top fantasy platform, they need to keep innovating ways to have better features for their fantasy managers," O'Brien said.

The advancements have made leagues accessible to more fans, especially those without time to scour the Internet for accurate, up-to-date information on thousands of football players.

They've also prompted a question: Is full automation on the horizon? O'Brien said the topic has not been part of discussions between IBM and ESPN.

"There is no interest in replacing the human element here at all," O'Brien said. "The people on the network are valuable. The fantasy managers are valuable. Their engagement is valuable. What we are trying to do is give them tools to manage what is essentially a data problem."

While the idea of a click-it-and-forget-it AI toggle on mainstream fantasy football platforms has yet to reach fruition, experts in the field have varying opinions on its likelihood.

"I think it's an inevitable certainty," Sourcetable's Eoin McMillan said. "The question is what remains interesting?"

Many fans who didn't rely on watsonx or AI-related tools from mainstream platforms turned to Sourcetable and other untraditional avenues for AI assistance. McMillan's Sourcetable has become increasingly popular in fantasy football, with users plugging the Wall Street-used data science technology into their league platforms to make more informed decisions.

"There was an element of 'If I'm not doing this and my friends are, I'm going to be at a disadvantage,'" McMillan said. "The game itself has changed."

A quick Google search, or voice activated request to Chat GPT through Alexa for some, reveals an endless menu of free AI-related applications for fantasy football. One of those is rotobot AI, branded as a fantasy football agent.

The app uses AI to scour the internet for information and brings it to fans fingertips.

"It feels like sports fans today need to make more and more decisions than ever before, and there is more and more sports content out there to help them -- but also confuse them," said Aishik Lala, co-founder and CEO of RotoBot AI. "It just becomes very cluttered. You don't know who to believe and who to trust.

"Our angle is to understand sports fans and cater right content for them to make the best decision possible."

RotoBot AI launched in 2024 and had 10,000 downloads last fantasy football season. This year that total reached 50,000, including 25,000 in September.

Lala said users seek RotoBot AI, which hopes to launch a voice mode in the coming months, for more specific data than provided by mainstream fantasy football platforms.

"Our initial user discovery process and one of the big reasons we built this was that we realized a lot of people saw the AI watson stuff on ESPN, but they didn't really believe it. They didn't really know which ones were specific for their lineup. They didn't necessarily agree with values over time," Lala said.

"We just saw a good opportunity to really drill down on this and make sure it makes sense for people and train the performance based on what user feedback we generate."

While McMillan believes full automation is inevitable -- and even happening already in some cases -- he says it will most likely lead to leagues needing more defined rules and terms related to AI usage.

Yahoo's Andrew Machado said fans should maintain "ultimate control" over their fantasy teams, but AI can aid that effort.

"Sometimes, AI-powered tools can help us better deliver this expert analysis and data at scale to our millions of users," Machado said. "But at the end of the day, it should be up to fans to make their own lineup decisions, which helps preserve the fun of fantasy sports."

Lala doesn't see a future of full automation because of the feeling of gratification fantasy football gives fans when they make the correct decisions. But he also recognized the possibilities of fully automated competition.

"Ultimately, the interest comes from the people," he said. "The fun of fantasy is very important. You want to feel empowered to make the right decisions.

"If I were just to leave that up to AI, it would simply be financially driven," Lala said. "I would probably use it to enter tons and tons of leagues and maybe there would be full-on AI leagues where people would just be competing with their algorithms. For users to adopt this, ultimately the fun and interest has to come from them.

"If it were to enter full automation, it would taint a lot of the fun. Once the fun is gone, there's nothing."

Category: Football