The short answer is that even though spider may be a completely dominant mutation, heterozygous spider X heterozygous spider does not produce 100% spiders - each egg has a 25% chance of producing a normal.

The following is the long answer on why and how I find it easier to understand inheritance by ignoring mutation types and looking at genotypes and then use the mutation type later to figure out the phenotypes.

There seems to be a lot of confusion over the term "dominant".
 

For one thing, people use "dominant" to describe both a proven completely dominant mutation and any type of non-recessive mutation before we know yet if it's co-dominant.
To avoid this confusion I prefer to use "some type of dominant" for mutations such as spider where we know it's not recessive but we really aren't completely sure yet if it's co-dominant or what I like to call "completely dominant". I don't think "completely dominant" is really a text book term but then in the scientific community they probably would wait until they knew the inheritance for sure before they described it and we need a way to talk about it for the years that it's sold before being completely figured out.

The next point of confusion is between the mutation types (recessive, co-dominant, and completely dominant) and the inheritance. The mutation type tells you what the mutant genotypes (heterozygous mutant, homozygous mutant) look like relative to each other and normal. It doesn't change the basic rules of how the genes are inherited which apply to all non sex linked mutations and are pretty simple until you start throwing in mutant alleles (different mutations of the same gene).

With the recessive mutation type, the heterozygous mutant (one mutant copy of the gene and one normal copy of the gene) should look normal. Heterozygous refers to the genotype (describing it's genes) of the animal and it's appearance is it's phenotype.

With co-dominant (or perhaps better referred to as incompletely dominant) the heterozygous genotype does not look normal and the homozygous mutant genotype looks different yet. Usually the phenotype of the homozygous mutant genotype is more extreme than the heterozygous genotypes (i.e. super pastel vs. pastel).

With the completely dominant mutation type we would expect the homozygous mutant to look just like the heterozygous mutant but different than normal. If it turns out that spider is completely dominant then homozygous spiders will look just like heterozygous spiders. Even just one copy of the mutant gene is completely dominant over one normal copy of the gene so the heterozygous spider is just as spider as the homozygous one with two copies.

As far as inheritance of the genotypes, you can forget for a minute what type of mutation you are talking about. If we are talking about single non-sex linked mutant genes without multiple mutant alleles then the same simple rules apply regardless of the mutation type.

Each egg has the following odds:

1. Heterozygous X Normal = 50% chance heterozygous, 50% chance normal

2. Heterozygous X Heterozygous = 25% chance homozygous, 50% chance heterozygous, 25% chance normal

3. Heterozygous X Homozygous = 50% chance homozygous, 50% chance heterozygous

4. Homozygous X Homozygous = 100% chance homozygous

5. Homozygous X Normal = 100% chance heterozygous

If you can remember these 5 rules then all you have to do is figure out what you are breeding.

The vast majority if not all spiders so far have been heterozygous for the spider gene. From #1 above you can see that heterozygous spider X normal should produce babies that each have a 50% chance of being heterozygous spider and a 50% chance of being normal for the spider gene. From #2 you see that heterozygous spider X heterozygous spider produces babies that still have a 25% chance of being normal. However, here is where it gets a little interesting, will 1/3 of the spiders produced be homozygous super spiders and the other 2/3 heterozygous spiders like the parents? If the mutation is completely dominant the homozygous spiders will look just like the heterozygous spiders and you will not know for sure that they have been produced until you grow them up and produce a large number of only spiders breeding one of the homozygous ones to normals as in #5. This takes time and is why we can't be sure yet if spider is completely dominant.

Maybe spider will turn out to be co-dominant after all and NERD just hasn't been lucky enough to produce a homozygous spider from heterozygous spider X heterozygous spider breedings so far. Maybe two copies of the gene are lethal and there will never be a homozygous spider (this will take even longer to reasonably prove).