Gamification is a promising approach to reducing misinformation susceptibility. Previous research has found that “inoculation” games such as Bad News and Harmony Square help build cognitive resistance against misinformation. However, recent research has offered two important nuances: a potentially inadvertent impact of such games on people’s evaluation of non-misinformation (“real news”) and exponential decay over time if no memory-strengthening exercise is provided. We address these issues in two preregistered lab experiments (N1 = 191, N2 = 321) and four quasi-experimental in-game surveys implemented in Harmony Square (N3 = 559) and Bad News (N4 = 2,558, N5 = 419, N6 = 882). In Experiments 1 and 2, we test if providing different types of feedback after playing Bad News enhances discriminative ability of misinformation and real news 1 week postgameplay and find that doing so resulted in homogeneously better accuracy at identifying both misinformation and non-misinformation compared with a control condition, which played Bad News without feedback. In Experiments 3–6, we implemented two different types of feedback exercises in the Harmony Square and Bad News games and find that this significantly boosts discernment compared with playing the game without a feedback exercise, primarily by improving accuracy at detecting real news. We confirm these results using signal detection theory. We conclude that feedback exercises boost the effectiveness of gamified misinformation interventions, likely due to an improved learning environment.