@Bimmylee
Those analogies are misleading, for an evident reason: they postulate right in their description the complexity sources known as human beings that are perfectly capable of cheating or making up a story on the spot and known to do so.
That's why in both cases anyone will find it more likely to believe that the dealer manipulated the cards or that the guy is really a thief: it's the simplest option given the known hypothesis including the dealer or the man standing in my house.
In the case of creationism, god is part of the thesis. In other words you're putting the complexity of there being a god that is somehow able and willing to create the universe or drive its evolution on the scales of plausibility against strictly natural laws. Since this god is more complex than a human being and harder to explain, this nets to being more complex than the godless option that originated humans.
Or in simpler words, if I met god in the street and (s)he demonstrated to be able to create universes on a whim, like I know a man can cheat at cards or lie about stealing, then I could say that the creation of our universe would be a very likely conjecture. But that was a big if, hiding all the implausibility of the situation.
PS I was going to say something about the anthropic principle and the misusue of statistic, but Rath already did. Darn.
