This research project proposes to investigate the comparative effectiveness of digital game-based learning (DGBL) and embodied learning in mathematics education. It will address a gap in the literature by directly comparing a digital learning game, Decimal Point, to its embodied counterpart developed using the WearableLearning platform. The research aims to assess differences in learning outcomes and student affect in the embodied and DGBL conditions. Furthermore, it will examine how gender and prior knowledge influence learning outcomes in both the DGBL and embodied learning contexts, following up on a previously observed gender effect in Decimal Point. The project will also explore the potential of large language models (LLMs) to translate and design embodied learning activities from a digital learning game. The expected contributions include insights into the specific benefits and limitations of each approach, guidance for designing more equitable educational technologies, and a novel methodology for leveraging AI in educational design. In short, this research bridges learning science, game design, and artificial intelligence to advance the understanding of active learning in mathematics.