Military training may benefit from using game-based strategies to aid in instruction. This study evaluated GameBased Training (GBT) by investigating a control condition and two feedback conditions: leaderboard and excessive positive feedback. Leaderboards produce competition using a ranking system. Excessive positive feedback praises correct behaviors. The control condition lacked a feedback strategy. Specifically, the GBT task involved detecting kinesic cues in a virtual simulation. Detecting kinesic cues, or non-verbal gestures that convey one's true emotions, help Soldiers interpret their environment. The kinesic cues trained in this study were aggressiveness and nervousness. The research investigated correlations between involvement and performance per each strategy; as well as correlations between game usage and involvement. A Kruskal-Wallis test was also conducted to assess differences in involvement between strategies. Involvement was operationalized as flow, immersion, and engagement surveys based on Likert scales. Performance comprised detection accuracy and response time in the kinesic cue detection task. Significant correlations were found between some select flow subscales and performance; and video game usage and the flow subscale of concentration on task. The KruskalWallis test revealed the flow subscale of unambiguous feedback was highest in the excessive positive feedback condition. The research findings offer design considerations relevant to select flow states and their relation to performance. Unexpected flow outcomes suggest extrinsic motivation may have played a part in increasing performance within the excessive positive feedback condition. Additional considerations for GBT may be explored for educational applications. This research lays the foundation for defining which flow subscales map to specific feedback elements. As a limitation, the immersion and engagement questionnaires may require modification to increase sensitivity.