This paper discusses the effects of multiple viewpoint videos for metacognition of experiences. We present a system for recording multiple users' collaborative experiences by wearable and environmental sensors, and another system for viewing multiple viewpoint videos automatically identified and extracted to associate to individual users. We designed an experiment to compare the metacognition of one's own experience between those based on memory and those supported by video viewing. The experimental results show that metacognitive descriptions related to one's own mind, such as feelings and preferences, are possible regardless whether a person is viewing videos, but such episodic descriptions as the content of someone's utterance and what s/he felt associated with it are strongly promoted by video viewing. We conducted another experiment where the same participants did identical metacognitive description tasks about half a year after the previous experiment. Through the experiments, we found the first-person view video is mostly used for confirming the episodic facts immediately after the experience, whereas after half a year, even one's own experience is often felt like the experiences of others therefore the videos capturing themselves from the conversation partners and environment become important for thinking back to the situations where they were placed.