While we do not know the probability of abiogenesis, we can make an educated guess as to the range of probable likelihoods. ie, the probability of the probabil

While we do not know the probability of abiogenesis, we can make an educated guess as to the range of probable likelihoods. ie, the probability of the probability. I read a paper on this a while ago, which I cannot find now, but it places the likelihood at some meaningful percentage per million years, which in aggregate across several orders of magnitude of planets, makes life elsewhere in the universe very likely. It uses the statistical notion that Earth should not be modeled as a single data point, but as a series that you integrate over, and concludes that since life emerged so soon after Earth became habitable, that the emergence of life is likely likely (the probability of the probability) - within the lifespan of a similar star. To give you a more clear example... Imagine a handing a subject a bag containing a die with an unknown number of sides. You ask a subject to reach into the bag, pull out a die and keep rolling it until he gets a 1. ...and then report to you only the number of times it took to roll a one. In statistics, can you infer a probability on the number of sides of that die, based upon how long it took the subject to reach a 1 - even if the experiment is only performed once? The answer is YES. If the subject only had to roll the die twice to get a 1, you can estimate that the die probably had very few sides. Similarly, if the subject that it took a million tries to roll a 1, you can very comfortably say that the die was not 2 sided (a coin). In the same vein, since life appeared on Earth in the few hundred(s) of millions of years after it became habitable, then the likelihood is probably not close enough to zero to rule out life in the cosmos. Treating Earth as one data point is an unwise oversimplification.