There are various ways of trying to justify induction by appeal to certain a priori principles of probability. But all such appeals fail, either by being irrelevant to the issue or by begging the question. A seemingly more promising approach would be to take induction to be a basic form of sound reasoning, which, like deduction, does not stand in need of justification. I call this the simple view. But this view too can be shown to fail. For instance, consider a case in which we know that a regularity that has held in our experience so far has done so by mere coincidence (i.e. for no reason). Because we know that the regularity is just a coincidence, we know that we cannot rationally use it as a basis for predicting outcomes in the future. But this is not something that can be adequately accounted for if we accept the simple view.
Keywords: a priori, coincidence, induction, justification, predicting, probability, reasoning, regularity, simple view,