This study analysed the conflict expression and polarisation in digital communication through the YouTube comment section on TV One’s coverage of riots between the Persib Bandung and Persija Jakarta supporters. Through a qualitative content analysis approach with NVivo, the study systematically examined 1,200 user comments collected from the video titled ‘Mob Attack on Supporters During Persija vs Persib Match, Two People Arrested’, which was released on 20 February 2025. The analysis combined three techniques: frequency analysis, sentiment analysis, and Social Network Analysis (SNA). Frequency analysis revealed the dominance of conflict-related terms such as supporters, violence, provocation, fanaticism, and anarchy, reflecting how language is used to reinforce group identity and deepen polarisation between Bobotoh and Jakmania. Sentiment analysis showed that 60% of comments carried a negative tone, 25% were neutral, and 15% were positive, indicating strong public emotions of anger, disappointment, and demands for legal action. SNA further identified key accounts functioning as hubs of discussion, amplifying narratives of conflict and shaping collective perceptions through interaction patterns and subgroup debates. These findings confirm that digital platforms such as YouTube do not merely reflect public opinion but actively serve as arenas for discursive conflict, where language, identity, and polarisation are reproduced and intensified. This study underscores the urgency of stronger content moderation on digital platforms, balanced framing by mass media, and preventive monitoring by relevant authorities. It provides theoretical insights for discourse analysis in digital communication and practical recommendations for managing conflict in both online and offline contexts.