This dissertation presents a study that explores a specific implementation of concreteness fading used in a serious game that teaches programming. Concreteness fading consists of first presenting concepts with concrete representations before swapping them gradually with their concrete, normal counterparts. The goal is to figure out how concreteness fading should be applied to a programming game to have it increase learning. Expert interviews are performed to discuss different aspects of how the technique is utilized in the game Reduct. Participants also play through the game before discussing it. Results show that some found the individual representations of mechanics within the game to be the biggest flaw while others pointed to how it handled the fading aspect. These generally come down to a lack of clarity, and should be considered when developing future games of this style.