The inbreeding itself isn't what causes the deformities, I think its the fact the odds of the mutation showing up if its recessive increase. The mutation can be 'good' as in a new color or pattern. But it could be bad as well, so adding new bloodlines is a good way to help keep the strain robust. I don't know if this made sense or not but I hope that it helps.
Mutations are very rare though - they happen so infrequently that it's not worth worrying about, what is worth worrying about are deleterious genes that your initial stock may carry already. By nature of the beast, genetic problems tend to be genetic (there are dominant genetic problems, but it is very easy to eliminate a dominant problem in one generation, and also animals displaying the type of genetic problems you would want to eliminate don't tend to be able to breed)
There are two main schools of thought - the first one is outcross, don't inbreed, to narrow the chances of recessive defects rearing their heads. The second way of doing things is to inbreed to make sure your animals don't carry any deleterious mutants - if they do, you can stop the line or fashion a test mating program to clear your lines of the "bad gene". The problem with just outcrossing is that in a finite world, outcrossing every generation forever is just not possible, and also, if you were breeding a line with a recessive defect in it, then by outcrossing you're producing loads of unidentified potential hets, which could end up being a problem way down the line.
Inbreeding doesn't cause genetic illness, and if you inbreed and get mutants and monsters then your original breeding stock had all those genes to cause the problems you're seeing now.
The only problem that is really caused by inbreeding is termed inbreeding depression. Once a population reaches a certain point of relatedness (I think it's due to the levels of homozygosity in the mating) then fertility decreases, clutch sizes will decrease, and the overall size of the animals can also decrease. It's impossible to say that X number of generations of inbreeding will cause inbreeding depression, and especially difficult when matings are not always brother x sister, but looser inbreedings such as aunt x nephew can skew the overall inbreeding percentage.
Unfortunately this is difficult to track in snakes as, over here at least, pedigree documents showing heritage seem to be non-existant. Just to show a comparison, in my rats I keep full family tree documents going back as many generations as possible (normally between 15-25 generations), and I have noticed that once you work out inbreeding calculations for a decent number of generations, a mating between brother and sister can actually be a lesser inbreed than a mating between two animals where there are no common names in the first three generations of the pedigree. So when you buy "unrelated pairs" of snakes - how unrelated are they really? You may find in some cases a brother-sister pair might be more genetically diverse