US Open Uses AI to Explain the Science Behind the Serve
IBM’s new Serve Quality metric turns player-tracking data into near-real-time insight, joining AI match analysis and conversational tools across the tournament’s digital platforms.
The US Open is giving fans a closer look at one of tennis’s most complex movements. Working with IBM, the USTA has introduced Serve Quality, an AI-powered metric that analyzes serving mechanics and translates them into a score fans can follow during matches.
Available across all 254 singles matches at the 2026 tournament, the feature extends the role of AI in the US Open’s digital experience, bringing movement analysis into the app and website alongside personalized updates and interactive match coverage.
Reading the Motion
Serve Quality uses spatial tracking to examine the positions of the ball, racquet, and a player’s joints. The system evaluates eight phases of the serving motion, including movements such as knee flexion and shoulder rotation, to assess the serve’s efficiency and effectiveness. IBM Bob supported development of the AI solution.
IBM says the tracking captures 21 points across the body and racquet 50 times per second, generating approximately 1.2 billion data points over the tournament. Confluent manages the live data stream that supports the near-real-time score.
For fans, the potential is a more accessible view of technique. Serve speed provides an immediately recognizable number; movement analysis adds context about how a player produces that delivery. The product challenge is translating a complex physical action into information that helps viewers follow the tennis without interrupting it.
Following the Turning Points
The expanded digital experience also includes Key Moments, which explains changes within the existing Likelihood to Win feature, and an enhanced Match Chat, powered by watsonx Orchestrate. The assistant combines live and historical information, with some answers now incorporating photos and video. Its AI agents and models are tailored to tennis terminology and the USTA’s editorial style.
The features appear together in the tournament’s match pages. On the SlamTracker page for Jessica Pegula’s first-round match against Elena-Gabriela Ruse, fans can access Match Chat, follow the probability display, and review descriptions of pivotal passages of play. A suggested question asks how serve quality affected the match, connecting the new metric with the conversational experience.
That arrangement gives spectators several ways into the same contest: a quick statistical check, an explanation of a turning point, or a question about what they have just watched.
Building on an Established Workflow
The new features build on earlier work to embed AI into the USTA’s digital and editorial operations.
“The digital experience of the US Open is of enormous importance to our global fans and, therefore, to us.” – Kirsten Corio, the USTA’s chief commercial officer.
There is already evidence of expanded editorial output. IBM says USTA editors used generative AI in 2024 to increase first-round Match Reports from 20 to 64—3.2 times the previous output. That provides a historical productivity benchmark, although it does not measure the performance of this year’s Serve Quality feature.
For sports properties evaluating similar investments, the US Open offers a concrete product-development example: connect proprietary competition data to a question fans want answered, then make the answer accessible within the live experience.
The next test is audience response. Repeat usage, engagement with explanations, and whether fans find the new metric understandable will reveal more about its value than the volume of tracking data alone.
