I recommend Victor Stenger's blog, which is on the Huffington Post. His latest article is a synopsis of his argument that the god hypothesis is disfavored by the available evidence. He adopts a simple, pragmatic, evidenced based approach. We can do the same from a higher level perspective and substitute testing the closely related, but more general, supernatural hypothesis instead of the god hypothesis. If our universe appears to exist and operate for the purpose of harboring humans, if our universe appears to exist and operate under the design and control of an intelligent agent to further that intelligent agent's goals, then that is evidence for supernaturalism. If our universe appears to exhibit an indifference to humanity and exhibits a chaotic, random operation then it is naturalistic.
We have determined that our universe operates according to laws. The laws that govern our universe's operation exhibit an indifference to humanity. Humanity lives at risk of annihilation from a large asteroid, a volcanic storm, a gamma-ray burst, contagious disease, supernovae, or even a solar flare eruption from our sun. The patterns and constraints that define the laws governing our universe exhibit no intended purpose, no goal. Our universe's overall organization and operation is haphazard, unstable, and random. Galaxies are built from a small minority of all of the particles, and after they are born they all eventually disintegrate. The immortality of the Greek gods is a fiction given that our universe dissapitates at an accelerating rate. Furthermore, the natural laws governing our universe are not violated.
The aforementioned considerations are the most relevant empirical evidence available to us for deciding this question. The best fit conclusion that follows is that our universe is exclusively naturalistic. This can easily sound sad or tragic to us but that is entirely irrelevant to determining what is true. It may be that the popularity of supernatural belief is motivated at least in part by wishful thinking, but that is an ineffective, parochial, self-centered way of reaching conclusions about how our universe works.
One of the points Victor Stenger emphasizes in his article is that we arrive at conclusions about how our universe operates by constructing models. Ultimately, the only thing that matters here is whether the models we construct are successful. Does the model describe what we observe accurately and consistently? Does it make predictions which are subsequently verified? The reason that success is our ultimate criteria is that we have no way of measuring the truthiness of our assertions by any other criteria. If our model is successful then we have achieved everything that we reasonably can want to achieve. Thus, it is a mistake to worry that we are missing some additional ultimate truth as a result. Since we do not have access to that ultimate truth we have not in any sense failed by not acquiring it. So when some people argue for believing in an unevidenced god on faith they are making an intrinsically silly argument.