Honors Program
Honors Program: Embargoed Theses
First Advisor
Rob Nickolaus
Date of this Version
Spring 5-6-2025
Document Type
Thesis
Citation
Ruth, J., Smiley, J, Ridha, Z., Flatley, R., Moloney, C. 2025. AI-Powered Sport Storytelling from Event Data. Undergraduate Honors Thesis. University of Nebraska-Lincoln
Abstract
Hudl helps teams and players reach their full potential, and the Hudl Fan Platform is meant to do the same for fans. While fans can find streams, tickets, and rosters on the site, there’s often no context or story to get them excited about the game. As a result, fans struggle to stay engaged, family members miss key moments, and players don’t get the recognition they deserve. With the rise of AI and access to rich game data, our team saw an opportunity to fix this. We built a system that uses large language models (LLMs) to automatically generate pregame articles. These summaries help tell the story of each matchup, highlight key stats, and guide fans to streams and tickets, bringing more energy and attention to every game.
We chose to focus on a more defined scope to ensure we could deliver high-quality work with greater depth and precision. This year, we focused on pregame articles for Boys Varsity Basketball. Overall, the team experienced success in generating pre-game articles while also discovering roadblocks that would prevent the widespread rollout of pre-game content. At this point, we have an understanding of what is needed and how to implement widespread AI content generation across the Hudl Fan platform.
Comments
Copyright Jace Ruth et al 2025