At one point in the interview/debate, Dawkins made the point he has made elsewhere (as have others in recent years) regarding the problem that the apparent fine-tuning of the laws of physics to enable the existence of life has posed for naturalists:
TIME: Both your books suggest that if the universal constants, the six or more characteristics of our universe, had varied at all, it would have made life impossible. Dr. Collins, can you provide an example?
COLLINS: The gravitational constant, if it were off by one part in a hundred million million, then the expansion of the universe after the Big Bang would not have occurred in the fashion that was necessary for life to occur. When you look at that evidence, it is very difficult to adopt the view that this was just chance. But if you are willing to consider the possibility of a designer, this becomes a rather plausible explanation for what is otherwise an exceedingly improbable event--namely, our existence.
DAWKINS: People who believe in God conclude there must have been a divine knob twiddler who twiddled the knobs of these half-dozen constants to get them exactly right. The problem is that this says, because something is vastly improbable, we need a God to explain it. But that God himself would be even more improbable.
This remark only goes to show how limited and wedded to his own philosophical naturalism Dawkins truly is - it is a straw man, but judging from what I have read of Dawkins, he is incapable of seeing this due to his commitment to naturalism. Why is it a straw man argument? First, the concept of "vast improbability" needs to be contextualized. When we say this fine-tuning is "vastly improbable," we mean it is improbable given a closed system of natural causes. In other words, probability (in loose and simplified terms) refers to the likelihood that such-and-such would happen given the laws of nature. We cannot lose sight of this context in this discussion.
The second and third points follow from the first: no thinking Christian theist worth his salt is going to claim that God is subject to natural causes for an explanation of his existence. Hence, labeling God as an "improbability" is simply a category error - God is not the kind of thing that is subject to the laws of nature, even in principle (note: I am not arguing for the existence of God here; I am saying that God as conceived and argued by Christians, which is what Dawkins is trying to refute, cannot in principle be refuted by appealing to improbabilities. Though the probability of a naturalistic explanation for the laws of physics being fine-tuned to allow for life can be argued against validly, in that it fulfills the category requirements of probabilities - which is the starting point for Intelligent Design). So Dawkins is arguing against a straw man in that he is positing a God no Christian theists (barring possibly Mormons) believe in, though it seems to me that there is a bit of slight-of-hand going on here on Dawkins' part that makes this hard to see.
So, to sum up: 1) Probabilities are things that apply only to natural phenomena; 2) God does not exist as a natural phenomenon; therefore 3) probabilities do not apply to the existence of God, and any arguments against the existence of God based on his supposed improbable nature commit a category error and are fallacious. Moreover, if all natural explanations for the fine-tuning problem fall prey to vast improbabilities or unverifiable/unfalsifiable assertions (see my next installment for more on this), and invoking an intelligent designer as an explanation does not in fact appeal to an even greater improbability, I think it is reasonable to conclude that the theist offers the better explanation for this problem.