I was not speaking of the probability or disprovability of religions itself, but of the set of guidelines that exist for interpreting particular religious texts accurately.
For example, there are guidelines to help ensure one is actually interpreting the original, intended meaning of a passage. For Christianity, these include: don't conflate description and prescription, a passage "can't mean what it never meant" (that is, the only applicable interpretations are those the original recipients could have gotten out of a passage - so, for example, no attack helicopters in Revelation, since John's audience wouldn't have been able to get that interpretation), don't assume that a passage is the Lord speaking directly to you/your group (present-day Christians can eat pork, for example, even though the ancient Israelites were forbidden to, because that restriction isn't renewed for those under the New Testament - namely, Christians), pay attention to a passage's genre (the descriptions of Hell/Sheol in Psalms are intended to be poetic, for example, while Hebrew's argument that Christ descended to Hell is not), and so on.
Most of these guidelines revolve around the principle that the point of interpreting is to find out what the author actually meant to convey, which is a single, objective thing that one can attempt to find out through concrete rules such as those listed above - thus, one may not have all the answers (someone else may have a good, falsifiable argument against what you thought a passage meant), and, consequently, you can be proven wrong. Hence the logic behind my original statement, that one can be proven wrong or right when debating aspects of a religion:
Anything where you have all the answers, and cannot be proven wrong.
...So not religion.
Also: oh hey, it's the Marathon guy.