Yeah, it can be a bit confusing at first. The predictor programs all use percentages, but if you do very many punnett squares to figure your own probabilities, you quickly find that percentages are sometimes tough to "see" in your mind's eye.
However, if you take that 1 and divide it by the 64, you get the same 1.5625% possibility for obtaining a triple homozygous animal from a pairing of triple heterozygous parents. So, it's not that difficult, once you get the hang of it.
As for the number of animals you would want to keep, I like to think in big numbers. So, if I had triple het animals, and I wanted to produce triple homo offspring, I know that I need to produce (statistically) at least 128 babies to produce two of them, and that is hoping that they would be a 1.1 pair!
128 babies comes out to about 13 clutches having the same genetic possibilities. Breeding all of the females to several males with the same genetics, means that I would want to hold back at least 15.15 of those triple het animals, if I wanted a real chance at getting what I desired!
Whew! That's a lot of normals to sift through! It's a lot easier to just start with double homozygous animals, or at least animals with compatable hets in the parent generation.
Have fun with your plans, however they work out for you!
