Ordered Network Analysis (ONA) helps us to understand the dynamics of specific events over time. Recent work has leveraged these benefits of ONA to study affect dynamics, the shifts in emotions over time, theorizing differences that might be driven by both motivational factors and even learning experiences. This study uses ONA to investigate these relationships within the context of the affective transitions made by students of four behavioral archetypes—Scanners, Worksheet Users, Conversers, and Roamers—in Crystal Island, a game designed to support inquiry-based learning in microbiology. Results indicate that while both Scanners and Worksheet Users exhibited more positively valenced affect dynamics, only Worksheet Users showed measurable learning gains—likely due to their greater engagement in systematic exploration and reflection, particularly while experiencing and regulating confusion and frustration. In contrast, Conversers and Roamers experienced persistent boredom and prolonged confusion, accompanied by low motivation and poor performance. These findings suggest that successful learning in game-based environments depends not only on positive affect but also on the presence of productive struggle and effective emotional regulation, demonstrating the potential of an analytical approach centered around ONA to elucidate the relationships between behavior and affect during learning.