Proving History

A sensible approach

to evaluating the probability of assertions about history. Carrier's approach using Bayes's Theorem to cut through specious arguments is a valid scientific method, an important analytical tool, and can be helpful pedagogical technique.
While I would rate the main thesis and content of the book as 5 stars, overall readability is perhaps more like 3 stars. Hence my rating of 4 stars.
The first difficulty Carrier faces is that of trying to explain mathematical concepts of probability and statistics that are thoroughly garbled most people's minds to a nontechnical audience. I have 3 degrees in mathematics, so I mind this and don't view this as adversely affecting readability directly, but it does mean that he does face a difficult problem as an author.
There were three main problems with readability that I noted.
(1) Typos and other small errors: these were mostly minor and just a little annoying. For example, missing parentheses in formulae, numerical errors in cross-references in explanations, and the introduction of mathematical techniques in notes (specifically Laplace's Rule of Succession) that are not explained in the book itself.
(2) Philosophical prolixity: an unfortunate tendancy to lapse into philosophical jargon when it is not really needed. As a reader with considerable mathematical training, I find this simply to be annoying noise. I can only conjecture that the effect on a nontechnical reader is not likely to be good. I think it marks a lapse in the author's awareness of the needs of his target audience.
(3) Diffuse reasoning in analysis: this is a problem in presentation style. In Carrier's analysis of historicity criteria the presentation of the analytical reasoning tends to be too diffuse. In presenting his analysis he also presents and refutes all common counterarguments as he goes along. This sometimes makes it comically difficult to follow the overarching thread of analytical reasoning over the course of 80 pages say, when a dozen or more examples of counterarguments are brought up and refuted in the course of the explanation. It would be much easier to follow the reasoning if it were first succinctly outlined an argued in 10 to 20 pages. Then examine the examples after the argument is completed and in the reader's mind to exhibit their fallacies.