This talk is based on reflections and narratives arising from DoomEd, a research and development project that produced a science-based Half Life 2 mod and unleashed it on the worldwide modding community. Using evidence gathered from the 50,000+ downloads and numerous player comments and observations, we examine what we have learned from the process that can inform both gaming and education domains and explore the boundaries between perceptions of learning and of playing. We share our insights into what gamers want from games, looking at their identity as gamers and as learners, what they like and dislike and why they play. We examine how games might develop learning and skills for the 21st century, by motivating players through risk, challenge, failure, and fear; and question if situated learning and gaming experiences like DoomEd can ever be transferable to an educational context or whether we are barking up the wrong tree.

We look at the coming together of learning and digital games not as silver bullets that can cut through all educational ills, nor as motivation-rich curriculum supplements, but as both a distinct sociocultural activity (and a potentially commercially viable gaming genre in its own right) that enables learners to play, to be amused, and to experiment and think; and as an approach to education in which educators can learn how to be better in what they do within their professional role and effect transfer.

We illustrate how the development of learning within games contrasts with formal learning and argue for the transfer of game-based learning strategies to formal learning through process and project-based approaches as well as separate activities outside formal or mainstream education. We suggest that both learner and teacher education programs need to recognize and harness risk, challenge (to overcome Welfarist notions), failure, and fear as key factors connected to learner motivation, innovation, and success.