In discussions on the Modal Ontological Argument (MOA) it is often pointed out that there is only one real premise; everything else is an entailed conclusion. That premise is, more or less:
1. It is possible that God exists.
Now from this premise it follows that God does exist, by the rules of modal logic. Some people have suggested that the MOA does not truly succeed, since the opposite premise could be made:
2. It is possible that God does not exist.
From the rules of modal logic and what we mean by God, this argument will have the conclusion that God does not exist. Since, for all we know (so the argument goes), either (1) or (2) is true, then the MOA must not be a good argument, since the proper response on these premises will be some kind of agnosticism (this would be true, so the argument goes, even if we have independently good reasons for thinking theism to be true, though I’m not sure how that works). But it’s worth noting that (2) is not the opposing premise to (1). So what is? Actually, it is this:
3. It is not possible that God exists.
This, I think, frames the discussion in the appropriate way. For now we want to weigh the opposing premises (not necessarily premises that entail each other’s falsehood) to see which of them bear the most plausibility, or from which we gather the most modal intuition. We may have an equal amount of intuition, initially, for (1) and (2) (I don’t, but that isn’t the point!). However, most people I know do not share an equal amount of intuition or prior plausibility for (1) and (3). Most people think that God’s existence is at least possible, at least stronger than they do think that God’s existence is impossible.
So what happens? If we deem the premise more plausibly true than false, or more plausible than its negation, then we find that the entailed conclusion of the MOA is that God actually exists! If that is the case, we actually have acquired a reason to think that (2) is false, or to prefer (1) to (2), thus breaking the stalemate we might have had. Why do I say that? Well, any reason for thinking God exists is a reason for thinking that the postulate that God’s existence is impossible is mistaken; but any reason for thinking that “God’s existence is impossible” is mistaken is a reason to think that (2) is false. This is because of the entailment commitments of (1) and (2). You could just as easily say by acquiring a reason to say God exists you’ve acquired a reason to say (2) is false, and because of entailments, any reason to think (2) is false is a reason to think (3) is false.
The takeaway is that someone will have to come up with a reason to embrace (3) in order to run the anti-MOA that results from (2). Someone will have to come up with a good reason to think God’s existence is impossible, or else we will be within our epistemic rights in believing that God’s existence is possible and, hence, God exists.
 Thus, we actually see the opposite of (2) is:
4. It is not possible that God does not exist.
Since (2) and (4) are entailments of (3) and (1), respectively (after all, if it is not possible that God exists, then, possibly, God does not exist; and if it is possible that God exists, then it is not possible that God does not exist, since God’s necessary existence turns out to be entailed by the MOA), then it only makes sense that it is (1) and (3) that oppose each other.