Following on from the last post, this from George Hunsinger is also worth a read on the historical reliability of the Gospels. He says that rather than having to prove the historical reliability of the Gospels as such, it is only necessary to show that the claims made in the Gospels have not been disproved.
Hunsinger agrees that more can be shown than this, but it is not necessary to do so. Historical probability (which is what we are always dealing with in historical research) is a certain kind of knowledge, but the New Testament accounts of Jesus invite a knowing that is not limited to historical probability. So, for example, he makes the observation that the key claim of the resurrection of Jesus is beyond historical proof, and that Christianity is asking for a response beyond what can be "proven" by historical probability.