Dawkins vs. Bayes

Recently we've taken a look into several atheist arguments that seem valid on the surface, but are actually circular reasoning.  In particular, we've discussed Richard Dawkins, Bart Ehrman, and David Hume.  Dr. Dawkins assumes God cannot exist in his attempt to prove God does not exist.  Dr. Ehrman assumes miracles are probability zero (i.e., impossible) in his attempt to show that miracles are not accessible historically.  Hume assumes one must see a miracle in order to prove a miracle.  But besides circular reasoning, what do each of these have in common?  Each of their arguments are easily defeated by simply applying rigorous probability theory.