I'm incredibly fascinated by the ghost comparison. Is the probability that ghosts are a real physical phenomenon higher or lower than the probability that aliens exist or have visited us? That's an extremely interesting question, and I'm sure someone could do a statistical meta-analysis comparing the incidence of, say, UFO sightings with the incidence of paranormal experiences (if such an analysis doesn't already exist). Both questions seem like the things that should be generally empirically falsifiable (and indeed, specific instances certainly are), but humanity's curiosity about both has proven remarkably durable despite centuries of curiosity and myriad efforts to settle (negatively) both questions once and for all.