Agent forecasting systems have been explored to investigate agent patterns and improve decision-making in various domains, e.g., pedestrian predictions and sports analytics. Badminton represents a fascinating example of a multifaceted turn-based sport, requiring both sophisticated tactic developments and alternate-dependent decision-making. Recent deep learning approaches for player tactic forecasting in badminton show promising performance partially attributed to effective reasoning about rally-player interactions. However, a critical hurdle lies in the unclear functionality of which features are learned for simulating players’ behaviors by black-box models, where existing explainers are not equipped with turn-based and multi-output attributions. To support these functionalities, we propose a turn-based feature attribution approach, ShuttleSHAP, for analyzing badminton forecasting models based on variants of Shapley values. ShuttleSHAP is a model-agnostic explainer aiming to quantify contribution by not only temporal but also player aspects in terms of fine-grained cues. Incorporating ShuttleSHAP into the state-of-the-art turn-based model on the benchmark dataset reveals that it is, in fact, insignificant to reason about the initial few strokes, while conventional sequential models have greater impacts.