Regarding a standard for evidence:
Evidence should be:
1) Directional and specific. It should point to what you are suggesting it points to, and you should be able to establish just why it points that direction.
2) It should be testable. Once it is established that it points a specific direction, it should be shown to be reasonable and sound under the scientific method (which doesn't just apply to science, but to all things where a question of "proof" exists).
3) It should soundly exclude rival theories. In other words, what Bob Enyart presents as evidence of the existence of God should not be able to be used by Zakath as evidence of the non-existence of God, nor should any alternative explanation be able to reasonably explain the evidence.
4) It should be empirical and observable, not merely mythical. When relying on accounts of things long past, they should be reliable 1st Person eyewitness accounts, not 3rd Person third-hand hearsay ("he said that she said that he said") like what we tend to see in the Bible.
5) It should be corroboratable by disinterested third parties, especially if it concerns accounts of the "miraculous" sort. I mean, if the only witnesses to alleged Christian miracles are Christians, then it really cuts the reliability way down if there is no one else to confirm that such-and-such really happened.