Wagering on Free Will
If you don’t think the evidence can decide the question on free will, you might run a wager style argument, as some studies have suggested that belief in free will encourages moral behavior (Vohs KD, et al. Psychol Sci. 2008).
Ah, but you object that wager-style arguments cannot motivate belief because you think doxastic voluntarism is false. Well, give it a shot and try to believe on the basis of this wager. And if you succeed, you will have more than pragmatic reasons for holding your belief. As William James puts it, “[t]here are… cases where a fact cannot come at all unless a preliminary faith exists in its coming” (The Will to Believe, 1896).
Belief in free will may be just the sort of belief that verifies itself, if one is able to believe in it while admitting the evidence isn’t sufficient on its own to compel belief. If one chooses to believe because one thinks it is not a possibility closed off by science, and assesses the merits of the belief from pragmatic concerns, then one has the sort of first-person experience of freedom that libertarians tout even in the face of Libet tests.
In other words, see if you can bring yourself to believe in free will by wagering on it, and thus experiencing direct evidence of doxastic voluntarism, i.e. direct control over your own beliefs.