This paper examines webcam-enabled head tracking for games in a home setting. A new head interaction technique was developed based upon prior laboratory-based research, with a focus on making it robust to the variable conditions of a home setting. Our technique was integrated into a test-bed game and 550 hours of gameplay data was collected from 2500 users, many of whom also provided formal feedback. The head tracking performed creditably and players reported that the experience was more immersive. Head tracking failed to enhance competitive playing performance, perhaps owing to familiarization effects. Nevertheless, the data revealed evidence of learning amongst users, suggesting that performance would improve with continued use. Key lessons that emerged in the home setting in contrast to the earlier laboratory study were a demonstrated need for clear guidance and feedback during system set-up, and greater caution regarding its deployment, having discovered a small population of users who became nauseous.