Most atheists and theists believe that the logical problem of evil has been answered. Since the free will defense is a species of the greater good defense, the greater good defense is a defeater for the logical POE. Now comes along the evidential POE. The evidential POE then claims, among other things, that regardless of the greater good theodicy, which might resolve logical worries, there are evils for which we cannot fathom what the greater good could be. These evils don't seem to affect free will (in the case of those who employ the FWD to answer the problem of evil full stop), they don't seem to justify other good reasons that have been proffered, viz., soul building, loving discipline, whatever. So, there just isn't a God-justifying reason for these (whatever they are, there's different examples) evils. The argument seems to be that if we can't see a God-justifying reason, there probably isn't one. One response to this has been the response of so-called Skeptical Theism. ST argues that this is a bad inference to make because one feature of the Christian worldview is that God and many of his reasons are beyond our ken. We would be justified in saying there is no elephant in the room because after surveying the room we could reason from "noseeum" to "thereisnun." Obviously, this is because the area we have to search is sufficiently small for us to justifiably conclude that there is no elephant in the room. On the other hand, to survey the room and say that there are no elephants in the world because there are none in the room, is to make a bad inference.
So, that is a very short introduction to the discussion. I think it is sufficient for my purposes.
Now, most atheists also believe in determinism. For all our actions there are causes that we could trace ultimately back to the "big bang" (or whatever).
If atheists and determinists Daniel Dennett and Owen Flanagan are right, then the reasoning in the "noseeum" arguments proffered by many atheists should lead them to believe in libertarian free will too. As Dennett says,
"Whatever else we are, we are information-processing systems, and all information processing systems rely on amplifiers of a sort. Relatively small causes are made to yield relatively large effects.... Vast amounts of information arrive on the coattails of negligible amounts of energy, and then, thanks to amplifications powers of systems of switches, the information begins to do some work,...leading eventually to an action whose pedigree of efficient...causation is so hopelessly inscrutable as to be invisible. We see the dramatic effects leaving; we don't see the causes entering; we are tempted by the hypothesis that there are no causes." (Dennett, Elbow Room, 1984, 76-77, cited in Goetz & Taliaferro, Naturalism, 2008, 17).Dennatt is arguing that due to a failure to be aware of the causes of our choices (noseeum), we cannot reasonably conclude that there are no causes (thereisnun). His justification is that the causes are beyond our ken. And so our failure of awareness is to be expected given the broader metaphysics of metaphysical naturalism.
Owen Flanagan agrees,
"[T]he myth of a completely self-initiating ego, an unmoved but self-moving will, [is] simply a fiction motivated by our ignorance of the causes of human behavior." (Flanagan, The Problem of the Soul, 2002, 112, ibid).This isn't to argue for libertarian free will, obviously. It's simply to point out that some atheists may be inconsistent in (a) not holding to libertarian free will and (b) holding that there are no God justifying reasons because you can't see the reasons.