Even in a probabilistic sense, I could argue that an atheist's strongly held belief would be matched by the religious observer's strongly held belief, and in that way would have the two would have the same probability to lie.

This poll, though probably just for light amusement and discussion, brings up many questions about the philosophy of science. Science is a human endeavor and as such is very much limited to how the world is perceived by humans. There will always be bias, not all of it intentional. Data has no bias, though, so reproducibility is very important before putting too much stock in an observation.