The percentages are percent likelihood, sorry. It means you have a 25% of any individual snake being a Pastel Pied (so if you have 8 eggs you have a .75^8 chance of no Pastel Pieds and a .25^8 chance of getting all Pastel Pieds (the math behind getting a certain number is trickier and I havent worked with it in a long time, so yeah.... Probability (nPr and nCr) is not my strong suit))).

Anyway, The offspring could be (using single genes in the format):

Het Pied Het Pied Pastel
Het Pied Het Pied Normal
Het Pied Normal Pastel
Het Pied Normal Normal

(This is how I do this all in my mind)

So that comes down to Pastel Pied, Pied, Pastel Het Pied, Het Pied. Comes out as 100% Het Pied 50% Het Pied 50% Pastel, thus .5x.5 (in this case this is the multiplication for all of them, this doesnt always happen) = .25

So you have the following probabilities of resulting snakes:

25% Pied Pastel
25% Pied
25% Pastel Het Pied
25% Het Pied

Again that last set is probability of the genotypic results. Phenotypic results are 25% Pastel Pied, 25% Pied, 25% Pastel, 25% Normal (what they look like, not the genes that are represented).

So, again, the bolded text are the % likelihoods of any snake having some genotype.
Sorry, I do things in a confusing way... This is why I usually dont show my work in math...