No, the probability is terrible no matter what format one is using. Actuality doesn't infer a good probability, it merely means that /pink flamingos/, are possible. Even if I tried to better the probability with assumptions of repetition, in /0/, or a different definition of /flamingo/, ie a broader unit or subject, it's still horrendous probability. When one might get shall we say, 'almost believable' numbers, is when I really am skewing every variable in my direction. We wouldn't assume randomness anyway, the problem being lack of comparable ''assumed random'' comparison models, we don't really have them. That's why it is /0/, and then each equation depending on the persons bias, or whatever, affects how they deal with that. I'm perfectly ok with leaving /0/ as it is, in other words I wouldn't compound the accumulated data in order to bolster my argument, it's not necessary. Where one gets ''better /still awful/, probability is when one really generalizes the units, but then we aren't talking about something that relates to any real meaning for how one would label things, etc.