Fair enough… Let me dust off the keyboard LOL
Bane hit some of it, especially the parts about feeding and size being somewhat dependent on the mom.
I will focus in a bit on those and also discuss a different side that is often overlooked
With regard to the mom playing a part in size, this is basically a physiological feedback system. A smaller female will, naturally, only be able to contain smaller eggs in her body. If the embryos were to grow to a size that jeopardized the female then she would die and so they would die as well. As such, there are embryogenesis controls that regulate embryo size and also, consequently, effect eventual adult size. However, those controls are not as strict on the size regulation once the animal hatches, which is why the feeding can result in a larger than anticipated animal.
The other factor that is frequently overlooked is that the way the retic breeders calculate these percentages are, to some respects, totally irrelevant.
Obviously a Mainland x SuperDwarf would be labeled as 50% SuperDwarf. But a 50% SuperDwarf x 50% SuperDwarf would also be labeled as “50% SuperDwarf”. Likewise a “50% SuperDwarf” from a 50% SuperDwarf x 50% SuperDwarf bred to another “50% SuperDwarf” from a 50% SuperDwarf x 50% SuperDwarf would also be labeled as “50% SuperDwarf”
The realist however, it that those latter pairings are more likely to produce larger animals
Let me explain
There are multiple genes responsible for determining size. For simplicity sake here, I am going to reassign our values from Mainland and SuperDwarf to “Red” and “Yellow”. And also for simplicity sake, we are just going to pretend that there are only two genes involved R1 or Y1 and R2 or Y2
When you make the initial breeding, your animal genotype is very simple:
R1Y1 = Orange
R2Y2 = Orange
But if you breed two of the 50% animals together, your genotypes are more varied:
R1R1 = Red
R2Y2 = Orange
R1R1 = Red
R2R2 = Red
R1Y1 = Orange
R2Y2 = Orange
Y1Y1 = Yellow
R2Y2 = Orange
R1Y1 = Orange
Y2Y2 = Yellow
Y1Y1 = Yellow
Y2Y2 = Yellow
Now grab any two of those at random and breed again. If you grab the “wrong” two, you have now biased the next generation toward one or the other outcomes. Repeat it again and your deviation becomes even more extreme. And given the breeder is selecting their holdbacks first, the likelihood is that, whether accidentally or intentionally, they have pulled the favourable genotype (e.g., the smaller animals) first so the buyers end up picking the less desirable of the options (e.g., the larger animals)
Make sense?