The basic thing to remember, and if you get it you can figure everything else out, is that genes come in pairs - one from mom and one from dad.
With the exception of some gender chromosome genes each animal has two of everything. Each offspring gets one and only one of the two copies of each gene from dad and likewise from mom. That pair, one from dad and one from mom then make up the set the offspring has available to pick from when passing on a single copy to the next generation.
Another important concept is that there are many many different gene locations. So there are many different "normal" gene versions, at least one for each location. In general, each of the mutations we have seen so far in ball pythons is at a different location with its own different normal version. A notable exception is the white snake complex (lesser, mojave, phantom, Vin Russo, latte, possibly others) where there appears to be a number of different mutations of the same gene.
When the pair of genes for a certain location match that is refereed to as the homozygous genotype. When the pair is different (i.e. one mutant and one normal) then it's the heterozygous genotype. A homozygous normal for albino has two normal versions of the albino gene, one from each parent. A homozygous albino got the albino mutation from each parent. And a homozygous pastel (aka super pastel) got the pastel mutation from each parent. We can always figure out which version of a gene an animal that is homozygous will pass on because it doesn't have a 2nd type to pick from. A super pastel will always give the pastel mutant version of the pastel gene because it doesn't have the normal version of the pastel gene to give. Likewise, you can't produce a super pastel without the pastel gene being present on both sides because each parent can only give one copy and both must give pastel to create super/homozygous pastel.
Now spider has proven to be at a different gene location from pastel. So each parent gives one copy of the gene at the pastel location and also one copy of the gene at the separate spider location. A just plain pastel is homozygous normal for the spider gene so will always give the normal for spider copy along with one or the other of its pastel copies (either normal for pastel or the pastel mutant copy). A bumblebee can give either known versions of the pastel gene (normal for pastel or mutant pastel) and either known versions of the spider gene (normal for spider or spider mutant).
A Punnett's square is a way of looking at all the possible different combinations each parent could give and putting them together in all possible combinations. Each square then represents a possible outcome for one egg and the number of like squares compared to the total number of squares then represents the likelihood of that outcome. But as long as you can remember the two copies each parent has and if the mutations you are interested in are different genes or not you should be able to figure anything out using the principle that each parent contributes one of their two copies of each gene to the offspring.