BleacherBot: AI Agent as a Sports Co-Viewing Partner
* denotes equal contribution.
Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems
Conversational AISportsCo-ViewingLLM
What makes an AI a good companion for watching sports?
Watching a game together can be about sharing excitement as much as following the score. But viewers do not all want the same energy from a companion. BleacherBot explores how an AI co-viewer can respond to those differences.
Design around different levels of excitement
A formative study with ten participants highlighted the importance of arousal, or how activated and excited a viewer wants the experience to feel. We developed a sports-viewing agent with different interaction styles, then evaluated it in a main study with 27 participants.
Matching the viewer mattered
Participants reported greater engagement and enjoyment when the agent’s interaction style matched their preferred arousal level. This suggests that a highly energetic companion is not a universal default: the fit between the viewer and the agent is important.
A companion should suit the moment
The study offers design guidance for co-viewing agents that account for individual preferences and viewing contexts. We position these agents as complements to human social interactions. The findings describe the studied viewing experience and do not establish that an AI companion can replace watching with other people.
Read the original abstract
Co-viewing, traditionally defined as watching content together in the same physical space, enhances emotional connections through shared experiences. With the rise of remote viewing during the COVID-19 pandemic, existing solutions, such as second-screen platforms and rule-based AI companions, struggle to facilitate meaningful social interactions. This study explores the potential of Large Language Models, which offer human-like interactions and personalization. Our formative study with ten participants revealed the importance of managing arousal levels, highlighting the need to balance between high- and low-arousal levels across different viewing contexts. Based on these insights, we developed ‘BleacherBot’, a sports co-viewing agent with distinct interaction styles that vary in arousal levels. Our main study with 27 participants demonstrated that matching users’ preferred arousal levels with the agent’s interaction style significantly enhanced their engagement and overall enjoyment. We propose design guidelines for AI co-viewing agents that consider their role as complements to human social interactions.