It would be wrong and there isnt a reason to put faith in a guess we think will be wrong. How much we would wager shows how much trust we put in an idea. Something that is not existent isnt anything to put trust in. Lack of evidence doesn't give much to chew on.
As Willemena correctly pointed out, justification is what distinguishes (accidental) true belief, from knowledge. If I believe, for no apparent reason but just as a silly guess, that it is raining today in London, and it is actually raining today in London, did I
know that it was raining today in London? No, it was a lucky guess.
On the other hand, if I believe that it is raining in London because I saw it on the weather channel, and it actually is raining in London, not only is my belief
true, but it is
justified- it is based on sufficient reason to hold that belief.
Thus, in this latter case, my belief would count as knowledge.
***
Look, I'm not making any radical claims about having discovered some new ground-breaking evidence that proves God doesn't exist- rather, I'm making a point about what does or does not constitute knowledge. It is often held that knowledge, to count as knowledge, must be proved beyond all possible doubt. While in mathematics this is possible, this is simply impossible in other domains. Thus, if absolute certainty is the criteria, then it turns out we know nothing outside of the tautologies of mathematics and logic.
But we
do know stuff- how to change a tire, what my brothers birthday is, what time (approximately) the sun will rise tomorrow, and so on. And what distinguishes
these cases (i.e. of us
knowing stuff), is that we have defeasibly true and justified beliefs.
And I'm saying atheism, as belief that theism is false, satisfies this criteria. Thus, atheism can, under the right circumstances, be said to constitute knowledge, whereas belief in the existence of God/gods- i.e. theism- can not (because it is false).