Originally posted by Ushgarak
Why do you say we have no idea about how probable the conditions for life are? I think we have some idea.
I've heard people break down the Drake equation before, looking at all of the variables individually and trying to find values to it. However, in all cases, these values are built by inferring facts about life in the universe from the single instance of life originating on Earth.
I suppose some people might be able to come up with numbers from that type of analysis, but in terms of real statistics, it has no power.
The findings of extrasolar planets will help this immensly, because if we can figure out how likely "Earth-type" planets are (our size, distance from star, atmpspheric make-up, colission that caused a moon [tidal effects], and a host of potentially billions of other variables), we can make better assumptions, but how would we ever know the probability of life originating on "earth-type" planets? We would need to discover dozens before we could make those inferences, and even then, they are not very statistically powerful. There would almost certainly be a sampling bias too, ie, the earth-type planets we are able to discover probably don't represent a random distribution of all earth-type planets
Originally posted by Ushgarak
But we do have some idea; we have lots of information about the circumstances that support life just by looking at our own planet. That gives us something to work from.
the problem is the inference from earth to other planets. We don't know how representative of other "earth-type" planets earth is.
Say there are 160000000000000000000000000 earth like planets in the universe, we don't know that the specific chain of events that produced life here aren't 1/160000000000000000000000000, an instance in which we would only expect to find ourselves alive in the universe
though, this brings up the complication of, "how do we know we need "earth-type" planets?", which makes the ability for us to draw inferences from earth life even weaker