I have noticed that some sweaters, often top-down raglans, end up producing ill-fitting underarm areas. Often there is a lumpy poof of extra fabric on the body, arm, or both. I often thought this was because I’m not very busty or because it was an issue with ease. However, even if I’m making the correct size for my torso, I still get these lumps on certain sweaters. I see them on other peoples’ sweaters too. So why are they there?

After looking over measurements and formulas for many different styles of sweaters, I think I may have come up with the reason. I’ll begin with two examples which don’t produce lumps and then look at one that does.

Example 1: Elizabeth Zimmermann's EPS system uses a mathematical formula to plan for customizable fit. K is the term used to express the body stitches needed, calculated by multiplying your gauge by your body circumference. The body is 100% of K, arm circumferences 35%, underarm stitches 8% of K (there is some flexibility allowed in the percents, but I’m going to use a simple version here). On a bottom-up sweater, the yoke is created when 8% of K stitches are removed from each sleeve (35-8=27%) and twice on the body for the corresponding underarms (100-8-8=84%). The sleeves are then knit onto the body with the underarm stitches on holders to be seamed later. Mathematically, the EZ yoke is 84+27+27=138% of K or 1.38K. This produces a sweater without the dreaded lumps.

Example 2: Stefanie Japel uses a top-down method that increases to the full arm and full body circumferences. Mathematically this ended up being somewhere around 166-169% of K on her sweater (100% for the body, and 33-34.5% for each sleeve). This method also produces a nice sweater, without the lumps.

So, why can two methods get great results and be so different? Why are other methods making the lumps? I think it is because of a term I like to refer to as total circumferences (T). If I were to add up EZ's and Japel's total arms and body circumferences, they both come in around 166%-170% of K or 1.66K-1.7K.

- EZ’s: add the yoke percents 27+27+84 to the four sets of underarm percentages 8+8+8+8 for a total of 170% of K or 1.7 K. (If using the smaller percentages of EZ’s EPS method, you can get a T number closer to 1.66K.)
- Japel’s: yokes are typically 166-169% of K or 1.66-1.69K

Although they are constructed differently and are the extreme opposites for underarm stitches cast on, they both arrive at a similar T values. I have looked at sweaters that had no underarm stitches (0% of K), a few (1-2% of K), several (3-4% of K), and many (5-8% of K) and all seemed to fit into this theory.

The Math: Now look at a top-down raglan sweater:

1. Calculate the body stitch count for the sweater=sweater circumference x gauge=K.

2. Look at the number of stitches for the yoke before dividing for the body and sleeves. Add the underarm stitches you are to add four times to this number. Why add the underarm stitches four times? Because you are casting those stitches on to make the underarms on the body of the sweater (two times) and later they will be stitches picked up for each sleeve (two times). This is T (total circumferences stitch count).

3. Divide T by K. If the number is 1.7 or below, the sweater will fit nicely at the underarms. If it is over 1.7, it will be lumpy. The closer you get to 1.8, the lumps get bigger.

Example 3: the math from a lumpy sweater top-down raglan

1. A sweater with a gauge of 6 stitches/1 inch and a circumference of 37” would have 222 stitches=K.

2. Increase yoke to 370 stitches. Add 6 underarm stitches four times=370+6+6+6+6=394 stitches=T.

3. 394/222=1.77K (177% of K)

Result: this sweater did not fit as well as I would like in the underarm area.

Data Collection: Here’s where I’d like a bit of Citizen Mathematician Assistance. Look at your top-down raglans (or sweaters in general) and look for one with the lumps and look for one that fits well in the underarm area. Do the math described in the steps above and see if my theory works for your body shaping. I’m curious if this theory works for others or just my build. Perhaps for your shape you can wear something with a T of 1.75K better than a 1.68 with no resulting underarm issues. Or, perhaps you can go up to 1.72K will no ill effects. Feed back and comments welcome!

Conclusion: If this theory works for you, then it may be of use in the future. Before beginning a project, run through the calculations and see what kind of a T measurement you get. Does it work for your body type? If not, then make some alterations. Reduce or add underarm stitches or alter the number of stitches you need to increase to for the yoke. Hopefully the knowledge will result in better fitting sweaters.