The Problem With Knowing There Is No God

Steve Hays posted a link on Triablogue to a book reviewon the Denver Seminary site for James Spiegel’s The Making of an Atheist.

After reading the review, I had an observation and realized that there was an epistemological problem with atheistic naturalism:

I've always said that I don't believe in atheists. I don't think they exist, despite anyone's profession to be one.
In a world where too many Christians are too afraid to point out what the Bible clearly says about disbelief, I appreciate Spiegel's Biblical observation that unbelief is rooted in a desire to self-justify personal sin.
Aside from this, I've just checked around because I've never thought of this before, but it occurred to me that atheism has no positive arguments. A quick web search of "positive arguments for atheism" led to some sites that claim to have positive arguments, but they were anything but positive. All the arguments I see have to with disproving God. That would seem obvious given the label "a-theist" except that if it were possible for this universe to exist without God, there must be an epistemology that someone could verbalize for it that doesn't include a reference to something that doesn't exist.
The best argument I found so far is that naturalism predicts the universe... except that it doesn't. The argument doesn't prove a lack of God. It merely assumes that the universe functions in a predictable way and further assumes that it follows that if the universe is predictable that God could not exist (which is also begging the question of a negative argument and therefore employs two healthy fallacies).
But this only demonstrates that if there is no God we wouldn't be able to know it since our ability to know anything is bound up in what is naturally predictable. If there is a God, then the only way we could know it is if he reveals himself to us in a way that is naturally unpredictable (miraculous) - which he has.