We can make a judgement of the probabilities by using past examples. The god-did-it explanation has been used for a very large number of things. We can examine the subset of those for which we have arrived at a reasonably sure answer and see how they came out.
I'd judge there have been at least 100 different phenomena for which god-did-it has been used as the answer. (lightening, the movement of planets, diseases, forms of life on earth, ...) Now let's see what the talley is comparing god-did-it to other natural explanations:
mmmm check one for natural, another for natural, and another, god yet? - nope, another for natural.
Tally seems to be about 100 to zero. Based on the I'd say the odds of god being the explanation for the next phenomenon we pick is less than 1%; perhaps much less. This isn't exactly a sound way to calculate odds but no less stupid than the OP's version.
™ Version 4.2