Story by
ESPN NFL ‘Data-Cast’ to Rely on AI Tech From Ex-Cowboys All-Pro
- Story by
- ESPN NFL ‘Data-Cast’ to Rely on AI Tech From Ex-Cowboys All-Pro
(Illustration/Grace Hughes, Photos by Getty Images)
ESPN's next NFL altcast, debuting Monday, was built with the future in mind.
The analytics-heavy MNF Playbook with Next Gen Stats feed will include run-vs-pass and blitz predictions to help fans view pre-snap action the way a quarterback or linebacker would. Former QB Dan Orlovsky and LB Luke Kuechly , along with Field Yates, will provide commentary during Monday's 49ers-Colts matchup, while the underlying tech is tied to another former star linebacker.
Advertisement
More from Sportico.com
Playbook will air on ESPN2 for the next two weeks, before moving to the ESPN app when the Manningcast returns on Jan. 3 (Joe Buck and Troy Aikman's main broadcast will stay on ESPN). Playbook 's fate is uncertain after that, but ESPN executives jumped at the chance to test how machine-learning models could provide additional insight for viewers.
"It took me two minutes to be like, 'OK, if you can do this in real time, this could work,'" ESPN VP of production technology and innovation Phil Orlins said of the first time he reviewed the tools now at his disposal.
The underlying tech is provided by startup Adrenaline, which has built a FieldVision AI product that promises to offer a "gridiron crystal ball." Its play prediction model is trained on more than a million plays, with decades of NFL and college data informing predictions.
Advertisement
Cowboys 2016 first-team All-Pro linebacker Sean Lee joined the company within a year of his 2021 retirement to provide an on-field technician's expertise. As the company's head of sports data, he has taken his intuitive play anticipation skills and guided engineers on how to convert that process into code, putting particular emphasis on tracking opponent personnel and alignments. "I'm really teaching football; they were also helping us learn about models and things that I had no clue about, being a meathead football player," Lee said.
"We're trying to replicate Sean's football brain in an artificial intelligence," Adrenaline CEO Casey Huke said.
Huke has spent close to a decade studying how analytical tools could help teams, betting companies and fans at home. Adrenaline is also working on drive and game-based prediction models. Beyond media, the tech could have applications in the world of sports betting, where fans are increasingly drawn to prediction markets that challenge them to identify future outcomes.
Adrenaline was introduced to ESPN in August via a connection at the NFL. ESPN has leaned into analytical altcasts in recent years, from Statcast presentations of Sunday Night Baseball to the NBA InsightCast during last year's Western Conference Finals.
Advertisement
Amazon has pioneered its own NFL nerdvana with its Prime Vision with Next Gen Stats feed, though ESPN's new offering will differ in its presentation. ESPN has opted for camera views more similar to the typical broadcast rather than an overhead look; it has also created its own booth for the show instead of largely using the main audio feed.
"The audience's baseline football knowledge has gotten higher, and we're trying to serve that audience who love football, who know football," Lee said. "We say we want to bring the meeting room to the living room."
Best of Sportico.com
Sign up for Sportico's Newsletter . For the latest news, follow us on Facebook , Twitter , and Instagram .
