But that would only work under assumption that in any group of people at least one of them has to be a ghost, or at the very least the chance that there is a ghost in a group of people is greater than 0, right? Is it something about the chance of someone being a ghost being truly unknown, and thus all possible values of probability being taken as equally rational, and with infinite number of possible values for probability of someone being a ghost for infinite number of them observing that no one in a group of people you’re in is a ghost… No, that wouldn’t work either, because it would require an assumption that this specific group of people might have a ghost among them. Assuming anyone can be a ghost with unknown probability still only works when the group you’re observing is entire population, does it not? Limiting it to specific group of people relies on it being representative of entire population, and random groups are not. Especially if you were to be a ghost, that would already make a group you’re in rather unique. Or not, depending on what’s the unknown value of probability of someone being a ghost.

I mean, what???