Be on the lookout, when arguing with skeptics and atheists, for a rhetorical technique I call the Shermer Shuffle. I call it that because I noticed it most clearly watching a debate with infamous atheist Michael Shermer and because it is not so much an argument as a fancy jig that distracts the theist interlocutor while entertaining the crowd.
It goes like this:
Demand evidence for God's existence.
When the Christian provides evidence, insist that the only real evidence is that which is repeatable under experimental conditions and accessible to empirical methods developed for the natural sciences.
Declare that God does not exist because the Christian has not provided any evidence.
(Optional) Point out that even if the Christian could provide evidence, then it would be empirical, and repeatable, and scientific, and therefore God would not be a supernatural being at all, but a natural phenomenon.
This last step is the flourish that gives the Shuffle Shermer's name. It is implied in most such shuffles, but usually the atheist is not honest enough, or not self aware enough, to admit that this is the only conclusion that is allowed by his absurd argument: his demand for scientific evidence entails that God is either unprovable in principle or else is not God.
You'll see the same pattern when a skeptic argues against the resurrection. They demand evidence. When historical evidence is given, they up the ante and demand scientific evidence, at which point they declare the case closed. "Dead men don't rise from the grave. It's scientifically impossible," they say. Well if there were scientific evidence of dead men rising from the grave to the extent that the skeptic would feel comfortable declaring such an event "possible", then it would be happening all the time, making it a regular, natural occurrence, thus destroying the miraculous nature of the claim.
A recent episode of the UK radio show "Unbelievable" had another example of this pattern. The show was about healing miracles reported by Ken Fish. I'm not sure I believe the man, or that I credit the miracles he presented. They are a bit dubious, so in practice I sided more with the skeptic interlocutor. In fact, most miracle claims I am skeptical of, and there are a select few that I would be willing to credit with being truly, unambiguously miraculous. The key, though, is that since I have philosophical arguments that convince me that God exists and probably cares about us, and historical arguments that demonstrate to my satisfaction that the resurrection happened, I can be skeptical, but open to the evidence.
That is why in principle I disagreed with the skeptical guest on “Unbelievable” completely, because ultimately her case came down to the Shermer Shuffle. The miracles for which there was medical documentation were plausibly in the category of "happens often enough that it could be natural", like a rare cancer not known to go into remission going into remission. The one miracle mentioned that she might have agreed was miraculous was the report of a Chinese woman's finger growing back in front of Ken Fish's eyes. But, she said, if she were to believe in that sort of thing she would want to see evidence of fingers growing back all the time in response to prayer. And there's the trick: if they did grow back all the time in response to prayer, then that would be just like the case of cancer remissions, and she could chalk it up to a simple post hoc, ergo propter hoc fallacy.
Such arguments are analogous to a man ignorant of triangles demanding that you show him a triangle with four sides, or else he won't believe they exist since all shapes have four or more sides. If you show him a triangle, then he will insist that it really has four sides, it just doesn't look like it, since all shapes have four sides. And, if you met his challenge and produced a four-sided triangle, he wouldn't be troubled, since a shape with four sides is called a square. So, triangles don't exist.
The central mistake is that the skeptic has limited reality to a narrow band of categories, namely "scientific" categories, and declared everything outside those categories "impossible" or "nonsensical" or "unreal" or "without evidence". But the problem is that the thing being argued is by definition not contained in his categories.
The skeptic forgets: "There are more things in heaven and earth, Horatio, / Than are dreamt of in your philosophy."