What you believe about Silver's model, however, is still ultimately a matter of common-sense judgment, and unless you think that you have a model so good that it should be used in a shut-up-and-calculate way, your ultimate best prediction of the election results won't come with any numerical probabilities, merely a vague feeling of how confident you are.

Want to make a bet on that?