IBM collects 1.2bn data points over a typical US Open match, including biometric data on 21 joints.
IBM launches its new serve quality stat in weeks, from a June proof of concept to an August release.
IBM's likelihood to win updates after every point using 20 years of historical data and pundit commentary.
How IBM Turns 1.2 Billion Data Points Into Fan Experiences
IBM's sports practice turns raw event data into fan-facing experiences. At the US Open it collects 1.2bn data points per match, then uses AI to explain them, because fans value accuracy and context over speed. The same playbook runs across Wimbledon and the Masters, where phones are banned and the app is the only way to follow play.
Takeaways
Fandom shifted to interactive, personalized experiences. Ian cites a survey finding that 72% of sports fans use an app and 40% want information in one place. Monica says apps moved from a secondary screen to the primary screen, with betting feeding the equation.
Fans value accuracy over speed in data. About 44% ranked accuracy as more important than speed, versus a little over 20% who preferred speed. Monica says fans value accuracy over speed, so IBM prioritizes correct, contextually relevant insights.
IBM processes 1.2bn data points per match. At the US Open IBM collects 1.2bn data points over a typical tennis match, from traditional stats to biometric data on 21 joints. Monica stresses data is not the end result and Gen AI surfaces contextually relevant insights.
Serve quality stat went from concept to release. A June proof of concept became an August release. IBM used Hawk-Eye cameras for biometric data, built a model with IBM Bob, and used Gen AI to summarize the scoring criteria. Subject matter experts were essential.
Likelihood to win updates after every point. IBM made the feature live two years ago using 20 years of historical data and live stats. Pundit commentary also feeds in. Key points explain momentum shifts. Monica says performance factors like mental toughness are worth capturing.
Masters app is the only fan experience. Phones are banned at Augusta National, so the Masters app is a completely virtual fan experience. Monica calls it the best example of making complex data simple, and says it takes an army to pull off.
AI cuts ideation from weeks to hours. Monica says tools like Claude and IBM Bob let teams prototype quickly, and ideation with Wimbledon and the Masters used to take weeks or months. IBM has used AI for personalization for almost a decade while balancing privacy.
US Open in-venue operations
IBM usually brings 15 to 20 people on-site to the US Open, backed by a large remote team, and brought a bigger team this year because of a large scope.
Usage spikes to over 14m users, so IBM ensures infrastructure can handle the traffic and that the USTA's data stays secure.
The in-venue mode of the US Open app shows which courts have open seating. It helps fans buy tickets and check line and concession queues.
Tap a name to go there · tap a line to see why they are linked
On this episode 2
Show 1
The USTA owns the US Open data as IBM's technology partner relationship.