That would literally be a religious argument, so I won't name examples, but many religious beliefs can either be tested empirically, have implications that can be tested empirically. If the religion exists in a social or organizational context where there are obvious vested worldly interests for the leaders of an organization to further certain beliefs (money, power, etc.) and those beliefs are designed to further those worldly interests, that would be a signal of improbability as well. Furthermore, for two sets of religious beliefs that share the same base texts and doctrines, the theological and exegetical doctrines can be judged on their own logic and decide between the two.

To put it bluntly (and give a contrived, fictitious example), a religion that believes in a flat earth that rests on the back of a unicorn that was created by a committee of gods 34 months ago (complete with false evidence and memories of what happened before) based on an idiosyncratic interpretation of the New Testament that demands that you give 75% of your money to The Glorious Spiritual Leader and then have sex with him, would be particularly improbable compared to almost any other religion.