Stephen Penn, Deepeka Gurunathan, Vaishnavi Govind Bhamare
Small game companies must test new products carefully, particularly when releasing expansions to existing games. As the product line grows, testing demand outpaces what small teams can absorb through hiring, creating a scaling problem. AI augmentation is a potential solution, but its effects on designer and tester behavior in small game companies are not well understood. This paper investigates how AI augmentation in card game testing reshapes designer and tester behavior, using an exploratory case study (Yin, 2018) of an existing company. Python code was developed to simulate matches between AI-controlled decks, and the resulting data were presented through two dashboards to one of the company’s owners. Qualitative analysis of the meeting transcript surfaced three propositions describing what designers and testers need from AI-augmented testing, alongside a three-step Validation-Interpretation-Internalization (V-I-I) cognitive cycle through which the dashboard user comes to trust, understand, and incorporate the information. This V-I-I cycle is itself the behavioral change AI introduction produces, and the resulting framework offers small specialty firms a planning tool for phased AI adoption. The findings contribute to the limited academic literature on AI in product testing for small firms.