Also in the clips in the video is fooled all _three_ of the participants tested. Couldn't find anything in the paper about sample sizes... but hopefully it was more than three...