Real-effort tasks are widely used in economic research to enhance external validity while maintaining experimental control. However, existing research employing physiological measures suggests that individuals exhibit distinct affective responses when engaging in such tasks, challenging fundamental assumptions about their neutrality. This study investigates whether different real-effort tasks elicit divergent emotional responses and how these affect performance. We conducted a laboratory experiment with 18 participants, measuring heart rate (HR) while they completed two distinct real-effort tasks: a monotonous math task and an engaging ball-catching task. Using time series clustering and feature-based clustering, we identified two distinct HR pattern clusters, indicating systematic differences in physiological responses. Our findings suggest that task characteristics influence emotional engagement, which partially impacts performance, depending on task characteristics. This study contributes to experimental economics by demonstrating systematic affective variations in real-effort tasks and emphasizing the need to reconsider their assumed comparability.