I'd say look into sources (websites/books whatever you prefer) coming from the legal plant breeding world, not cannabis specific.
all the cannabis specific stuff that a lot of people read is very basic (like only talking about the mendellian recessive/dominant stuff), sometimes even plain wrong use of terminology (like I once encountered some text from one of RC Clarke's books, despite his good reputation in the scene, in that text 'qualitative' and 'quantitative' traits were explained in the wrong way).
to make sense of the general info you need to have a bit of an overview image in your head about plants in general, and what would change specifically for weed.
for example one difference between plants that matters a lot on the breeding strategies used is 'allogamous' vs. 'autogamous'. autogamous plants are plants that naturally self-pollinate most of the time (like beans and tomatoes), so you can easily self/inbreed them, even grow them right next to eachother with minimal crossing between the lines. but the hybridising step might be more finicky (like with beans you have to peel open the flower before it opens).
allogamous on the other hand are outcrossers like cannabis. since they usually outcross they are more likely to suffer from inbreeding depression, and they often have bariers to prevent self-pollination. like seperate male/female flowers which mature at different times, or even completely seperate plants for the sexes like weed.
I'd start off by making sure you understand the mendel-stuff.
then you could for example look into qualitative vs quantitative traits. (qualitative is a yes/no kind of situation, quantitative is a spectrum kind of situation).
then with that and the specifics of the cannabis reproduction cycle (seperate sexes but selfing easy, most traits only scored in female plants, etc) you can look into breeding strategies that are used in other plants and see which you could use and with what goal. (for example, backcrossing is a nice tool, but the way some people in the weed scene use it won't work well. i.e. crossing back to an unstable parent, and expecting to get more stability out of it.)
you can also look into specific traits how they work. like weed is not the only plant product where aroma matters for example.
just some random breeding terms that you could google:
-single seed descent
-mass/bulk selection
-G x E interaction (genotype x environment)
-heterosis (=hybrid vigour)
-introgression (what you do with backcrossing), recurrent parent
- H^2, or heritability
- (specific/general) combining ability