Your description of genes being 'set' is innacurate. You have to appreciate that in any gene pool there is a huge number of 'bad' genes knocking around. This is why so many (relatively) albinos, leucistics etc pop up. These are genes that would more than likely prove lethal to a particular animal, just as a mutation giving the snake no eyes would.
Anyway, I won't go too far into it, but suffice it to say inbreeding will lead to the eventual expression of these 'bad' genes. In some cases (when people are hoping for unusual colour / pattern mutations) this can be a good thing, but in others it will lead to reduced reproductive success etc.. The question is, how much inbreeding can you do before such mutations begin to surface. If you have access to a copy of the Reproductive Husbandry of Pythons and Boas (Ross & Marzec) this is described very well.
In some cases, such as island (or highly localised) populations, many of the deleterous genes have already been eliminated from the population, simply because the population size is so small. Then, you may get the 'setting' of genes Jeff describes, although there's still going to be some genetic diversity present.