The universe is a rather unlikely place. There are dozens (to our knowledge) of finely tuned measurably-constant values in the universe, each of which must be accurate to within a tiny tolerance (about ${10}^{-10}$%) for human life to be possible. Added to this, our current understanding of quantum mechanics does not actually allow us to predict exactly what takes place in first moments of the universe; even if the nature of the universe were just right at the instant of the big bang, we could not actually guarantee that the same Earth would pop out each time we watched; it is in fact rather unlikely to do so.

A common counter-argument often runs somewhat along these lines:

1. The universe we are in is precisely a universe capable of supporting (our) life.
2. Though these universes are very improbable, we know that we are in fact in one such universe.
3. Thus, because they have not happened, we can discard all the non-life supporting universes.
4. Thus when considering the probability of the universe, we should really be thinking of the probability of our universe, given that the universe supports life.

Another common idea is that scientific laws need only make predictions about the future, and that it is valid to use a law to justify a past situation, even if it occurs only with very low probability (if all explanations are equally likely, this of course does not hold).

These seem to be a common argument among certain atheists, yet it is obviously flawed. The flaw is in fact reminiscent of the ‘Monty Hall problem’, in which the key mistake is confuse the distinguishing point about the situation, instead using another property which is not independent from the previous one. In both cases, the flawed argument muddies the waters by concocting a second reason to consider the goat’s door, or our universe, as distinguished, then by considering the probability of the second distinction is mistaken on the overall probability.

In our universe, we have to remember that the hypothetical observer is detached from the universe conceptually (by the nature of the problem, in which we consider multiple possibilities). The distinguished element of the universe is the fact that it, and it only, is the actual outcome of everything; that it happens to have life in it is a secondary issue. That such pattern as we see, much of it ultimately coming from the microwave background fluctuations, exists, cannot be explained by the flawed argument above. One could perfectly well pronounce the universe noise (as the atheist must); yet this does not solve the question of a source for the pattern; the likelihood of the universe, based on current knowledge, the scientific laws notwithstanding, existing as they do to make predictions, still fails to answer this question.

Of course this does not indicate proof, but empiricism and rigour being as yet insufficient to explain our origins fully, a justification for further methods of investigation is produced; further, a sound probabilistic argument is given by the example of a creator God, who, of whom, unlike the big bang’s after-effects, we have no evidence to argue for the creation, can aptly be a source of order.