So, here's what I ended up doing: My prototype is a 38" bust worked at 4.5 sts per inch. I started with 68 sts in the neck and worked until I had the correct number for my body (give or take a few to make the pattern work out). Then I divided that number in half to determine how many sts I had between the shoulders in the neck and how many I ended up with in the body. I did this because I needed to make sure I ended up with the same number in the front after working the V-neck. This involved me looking at the number of sts on either side of the neck, and manually adding the number of increase sts I'd have after every increase row, and after I started adding sts along the neck edge. I did this for hours until I finally figured out the algebraic formula to make it go faster, lol. At the end of this process, I had to adjust the cast on numbers, and adjusted the number of increase rows before starting the neck increases...in every size, lol.
Once I figured out the cast on numbers, I had to go address the sleeves. I had a different number of sleeve sts to pick up, so I had to recalculate the rate of decreasing for each size. I was pleased to see that my estimates for the traditional sizes weren't off be very much, nor were my estimates of the number of sts that would need to be picked up. And after the math was done, the numbers for my size still stayed the same (yay!). From there, all I had to do was put in the traditional lengths for the sleeves and the body, and the pattern was complete!
I realized a few things during this process:
- sizes do not increase or decrease by a set number of stitches. Trying to increase the next size by exactly one set of numbers only ended up with sleeves that were way too small in the upper arm, or way to large.
- when sizes do go up by a set number of sts, they won't all be in the same place. for a few of the sizes, the cast on numbers are the same but the amount of sts in a particular section may vary. One size may have more sts across the back, while the other has more in the shoulders.
- it's much easier to draw a sketch of the design (which I did), then decide on the sizes and plot out all of the numbers based on the percentages (which I did not do ), than it is to knit the prototype first and then figure out all the numbers for the other sizes. I know this because I had an idea for another sweater, I sketched it and looked up the information regarding neck percentages and sleeve percentages, and plotted each size based on the gauge of the design. From there the pattern almost wrote itself!
- I'm probably going to need to knit this again, because my cast-on numbers changed slightly after I did the revisions, and I need to be sure that the final sample matches my pattern stitch for stitch. Looks like I'm going to get to make this in red after all ;)
Now, after taking the advise of some very knowledgeable and experienced designers, I've sent the pattern out for the first round of editing before I start test knitting. This will catch all of the major issues, and potential errors before it goes out to the testers. From that point I'll get to see if any of the sections are unclear, based on the feedback that I get, and I can make any further adjustments before sending it back to the editor. I've also contacted a woman about creating the schematics for this design. However, I realized that I have the same program as she does, so eventually I'll have to learn to create my own...but for my first design I want to have someone more experienced do that for me.