The spread of competing narratives—such as misinformation and fact-checking efforts—on social media platforms presents a complex sociotechnical challenge with far-reaching implications across political science, communication studies, network science, and computational modeling. This study introduces a stance-based epidemiological framework, \(SEI_A I_D Z\) , designed to capture how users adopt, contest, or remain skeptical of circulating narratives, reflecting the socially contagious nature of narratives. We apply this model to TikTok data from Taiwan’s 2024 presidential election, a high-stakes context in which electoral misinformation and coordinated counter-narratives evolved in parallel. By explicitly incorporating narrative stance into the diffusion process, the model offers a more nuanced and interpretable account of real-world narrative dynamics. Compared to baseline models \(SEIZ\) , \(SEI_A I_D Z\) demonstrates significantly improved accuracy and reveals the critical influence of transmission rate ( \(\beta \) ) and stance-transition rate ( \(\psi \) ) on the trajectory of narrative spread. These parameters directly shape the basic reproduction number and provide actionable levers for intervention: reducing \(\beta \) through content throttling and increasing \(\psi \) via timely fact-checking are shown to effectively suppress the amplification of harmful content. This work offers a multidisciplinary modeling approach for analyzing and managing the spread of competing narratives in complex digital ecosystems.