I agree with Czcibor on the constraints at hand (habitability instead of inhabitation et cetera), but I have a problem with the "guessing" part.
We do have one example of life (duh), so we know it isn't pure guessing.
Moreover, unless the process of emergence of life was _extremely_ finetuned (unnatural), it will have ended in more than one instance. Even inflation, what resulted in our local universe, wasn't finetuned enough to likely result in only one universe (of which our observable universe is a small part of). I don't know of any such finetuned process, so I would say that it is extremely unlikely.
Reversely, we know from observation that life emerged rapidly. E.g. there was a habitable Cool Early Earth @ 4.4 Ga bp (billion years before present), and with a similar impact mass flux as the late bombardment life could arise at least @ 4.35 Ga bp. Otherwise unconstrained phylogenies place the first dateable clade split at average 4+ Ga bp. (Check Timetree for archaea vs bacteria).
That means the process was frequent and/or easy. In either case, that is the benchmark to use whenever an environment like Hadean Earth was present. That seems to be the case in terrestrials in the radiative habitable zone on 0.5 - 1.4 Earth radius bodies. (Though I would think most < 1 Earth radius bodies would dry out too much for a surface biosphere (Mars, Venus).)
If it is not guessing, is it quantitative? Well, not in ordinary statistics. But stochastic processes has funny properties of observability and being controllable as all other processes. And for control under loose constraints of a modeled process (say, Poisson) just 1 sample is enough. So I would claim, arguably, that the one observation we have is enough constraint to be quantitative.