In this video, William Lane Craig critiques Bart Ehrman on his applications of the "criterion of authenticity" in historical Jesus studies. Craig repeatedly emphasizes the point that the criterion can only be properly used to make a postive argument that a claim is more likely to be historical. He says the criterion cannot be used to negatively argue that an event is less probably historical as Ehrman does.

Is this characterization of the criterion's usage accurate? Can the criterion only be used positively? It seems as though Craig is attemting to rig the rhetorical game so that any data can only be used as evidence in favor of his claims and to dismiss anyone arguing otherwise.

If Craig's view is correct, what may be considered criterion of "unauthenticity" that can be used to negatively argue that a claim is less likely to be true?