What Is an Asymmetric Compound Raglan?

I love a good raglan sweater. Whether to wear, to knit, or to design, it is my favorite sleeve construction in knitting patterns. Guilty Pleasure was the first sweater pattern I designed using a compound raglan construction. My upcoming pattern Wychwood ups the ante: it's an asymmetric compound raglan. What does that even mean? A raglan's a raglan, right?

In this blog post I discuss the prevalence of the top-down raglan construction in knitting patterns and why that lead to the loss of finesse in raglan shaping. You'll also learn what an asymmetric compound raglan is, three major differences between an asymmetric compound raglan and a conventional raglan, and how that affects the fit of raglan sweaters. At the end of the post I'll also give you a few pointers on finding compound raglan knitting patterns.

---

The Prevalence of Top-Down Raglan Knitting Patterns

A raglan sleeve is a type of sleeve construction in which the sleeve seam runs diagonally from neck to the underarm. This sleeve construction was named after FitzRoy Somerset, the first Baron Raglan whose arm was amputated after the Battle of Waterloo in 1815. Legend has it that he (or most likely his servant or tailor) invented a sleeve type that was either easier to pull on with the use of one hand or in which a missing limb wasn't as obvious. Maybe both.

Whatever the origin, top-down raglan sweaters have become wildly popular among knitters in the past few years. Topping the most-popular patterns on Ravelry is tin can knits' Flax with — adding up all the different yarn-weight variations — an impressive 47K projects. Another most-knit beginner-friendly freebie is Florence Miller's Step by Step Sweater which comes with an accompanying video tutorial with a running time of nearly two hours.

Raglan knitting patterns released in 2000–2024.

This wasn't the case up until a decade or so ago. Pulling some statistics from the Ravelry pattern database I plotted the number of all raglan knitting patterns (dark blue) released yearly from 2000 to 2024, divided into patterns knit flat & seamed (green) and those made seamlessly from the top down (blue).

As you can see, patterns knit flat, made in pieces, then seamed made up of more than a half of all raglan patterns released up until the year 2006. Top-down raglans started their steady rise around the same time, matching the release numbers in 2011, and then surpassing flat patterns in 2012. Incidentally, Flax was released the following year.

Three Limitations of the Conventional Raglan Construction

As popular as these patterns are, conventional raglan construction makes three over-simplifications that make them easier to knit even for beginners… but not necessarily well fitting on any body.

Designer Jen Parroccini makes a great distinction between patterns that are designed to knit and those that are designed to fit. Designed-to-knit patterns prioritize knitter enjoyment and ease of pattern writing at the expense of fit. They use simple techniques and minimal shaping, and result in something wearable but not necessarily fitting.

Designed-to-fit patterns, on the other hand, ensure you get a well-fitting and tailored garment out of the pattern but the knitting process — and the pattern — may be more complex. These patterns may include the dreaded at-the-same-time instructions or customized shaping for which the knitter may need measure their body, do some math, or follow a formula to achieve a great fit.

I'm not saying Flax is the cause — it's a symptom — but somewhere around the year 2012 top-down, seamless, in-the-round raglan knitting patterns started making three unfounded assumptions:

1. Body and sleeves grow at a constant rate. (They don't).

2. Body and sleeves grow at the same rates. (They don't).

3. Back and front body are the same. (They aren't).

To improve the fit of raglan sweaters we need to get rid of those assumptions. Here's where asymmetric compound raglans come in!

Constant-rate vs. Compound Raglan

Anyone who's ever knit a raglan sweater is undoubtedly familiar with the standard increase rate: 8 stitches on every other round (or row, if you're knitting a cardigan). This creates a roughly 45-degree angle running from the neckline to the underarm. Although Barbara Walker writes in her book Knitting from the Top that

the standard rate of increasing […] will serve quite well for almost all garments

this is not the case in today's size-inclusive environment. In fact, the standard increase rate serves well only for a few sizes somewhere towards the middle of the size range.

What happens in smaller sizes is that the desired stitch count for bust circumference is reached before the desired armhole length. If you'd separate body and sleeves at the right circumference, the upper yoke of the garment would be too short and the underarms burrow into your armpits. A simple solution — which you've no doubt encountered in knitting patterns — is to work a few rounds even until the right upper yoke length is reached, cutting the angle of raglan increases short.

The opposite occurs on the larger end of the size spectrum. When the desired armhole length is reached, there are not yet enough stitches to accommodate the bust. If you'd separate body and sleeves at the right length, the upper yoke of the garment would be too tight across the bust. If we're sticking with the standard increase rate, there are three solutions to this, all equally hacky:

1. cast on more stitches at the neckline so that there are enough rows to fit all increases

2. keep increasing until the desired stitch count is reached even if it means the yoke length will go over and result in a swoncho that restricts your arm movement

3. cast on more stitches at the underarms in the hopes that it'll cover the difference

And as you know, massive necklines and drooping underarms are surefire indications of large-size grading gone wrong. I'm sure you've encountered examples of these.

A compound raglan is characterized by a changing angle of raglan increases.

The more elegant solution is to relax the assumption that increases always have to happen on every other round. Here's where compound raglan shaping comes in. Instead of one constant increase rate, you need two: a faster rate and a slower rate. (You can actually have more than two.)

Compound raglan shaping is one in which the angle of increases changes. In knitting patterns the faster and slower rates usually translate to increases placed on every other round (EOR) and every fourth round (E4R). This works well on the smaller end of the size range where you'd otherwise have to do those work-even rows after having reached the desired stitch count.

On the larger end, however, these rates might not be rapid enough. Your faster and slower increase rates might be every round (ER) and every other, or every other and every third, respectively. There is no reason for the increase rate to be an even number… other than a convention in knitting patterns to avoid doing shaping on wrong-side rows or at odd-number intervals. Ultimately it's the body measurements and stitch-to-row gauge ratio that dictate the best increase rate pattern, not some unspoken rule in pattern writing.

In conventional, constant-rate raglans the increase rate is the same all the way from the neck to the underarm. In compound raglans the increase rate changes either once or twice.

There are different ways of dispersing the faster- and slower-rate increases over the yoke length. The simplest way is to go linearly from faster to slower. This is the approach taken in Ann Budd's The Knitter's Handy Book of Top-Down Sweaters: placing all EOR increases first, followed by all E4R increases. (Some of the larger sizes even have three increase rates!)

The more complex but ultimately better-fitting approach is to create an S-curve pattern with the increases. Because the human body doesn't grow at a constant rate from the neck to the underarm, an S-curve shape better follows the contours of the human anatomy. In this approach increases are made more rapidly at first to accommodate the shoulders. After this initial stage the rate slows down before picking up again near the underarm.

Compound raglan shaping on square versus sloped shoulders.

A mathematically easy way is to divide the faster-rate increases in half and have, for example, half of the EOR increases first, followed by all of the E4R increases, then the remaining half of the EOR increases. But this half-and-half split is just a jumping-off point and can be tweaked further as necessary.

Both Barbara Walker (Knitting from the Top Down) and Maggie Righetti (Sweater Design in Plain English) propose for using the faster increase rate up to the tip of the shoulder bone, then switching to the slower rate. If you have very square shoulders, you may need to do more than half of the faster-rate increases for the garment to cover your shoulders. On the other hand, if you have very sloped shoulders, you may need less than half and can switch to the slower rate earlier.

Same-rate vs. Different-rate Raglan

Now let's look at the other side of the coin in the conventional raglan increase rate: why do we increase 8 stitches on every increase round?

There a four increase points on a raglan sweater. On a conventional raglan all increase rates and raglan lines are the same.

This comes from the over-simplification to tie body and sleeve increase rates together. A raglan sweater has four shaping points where the body meets the sleeves: front and left sleeve, front and right sleeve, back and left sleeve, and back and right sleeve. These are usually marked with stitch markers, and when you increase one stitch on each side of each marker, you get 4 × 2 = 8 stitches increased on every increase round.

But human arms and busts don't grow at the same rates and in the same proportion. Again, the standard rate works only if you're somewhere in the middle of the size range but creates ill-fitting sweaters on the extreme ends.

In smaller sizes marrying the body and sleeve increase rates results in sweaters that have sleeves that are too narrow for the wearer: the desired bust circumference stitch count is reached before the upper sleeve is of the right size. And on the large end the opposite happens: sleeves are too wide for the wearer. This is because the desired sleeve circumference is reached well before the desired stitch count over the bust.

The solution is obvious: relax the assumption that body and sleeve increase rates must be the same.

On smaller sizes this means making more increases in the sleeves than in the body. In knitting patterns this might be, for instance, increasing on the sleeve side of the stitch markers on every other round while, at the same time, making some of the increases on the body only on every fourth round. Conversely, you need the opposite on larger sizes: more increases in the body versus the sleeves.

Sounds complicated? It doesn't have to be. There are ways to write separate body and sleeve increases in an understandable, knitter-friendly way without having to resort to "at the same time" instructions, spreadsheets, or checklists. Here, for example, is a snippet of my upcoming Wychwood pattern. This section applies to sizes 3X–6X in which more-frequent increases are needed in the body and less-frequent increases in the sleeves.

Snippet of compound raglan-shaping instructions from Wychwood.

It does add to pattern complexity and length, though, but it's yet another compromise between designed-to-knit and designed-to-fit patterns.

Symmetric vs. Asymmetric Raglan

Last, let's examine the 'asymmetric' part in asymmetric compound raglans.

In an asymmetric raglan, raglan seams are shorter in the front and longer in the back.

The third over-simplification made in today's raglan patterns is that the back and front body are the same. This can result in ill-fitting necklines in top-down raglans but it also means back and front raglan seams that are of equal lengths. Amy Herzog uses the term 'standard raglan' to refer to these patterns. I'd call the shaping solution used in most patterns today 'symmetric raglan': it assumes that human bodies are symmetric front and back.

But as she writes in the Ultimate Sweater Book

On real people, back armhole depths and front armhole depths are different.

This should be reflected in knitting patterns as well: the back raglan seams should be longer than those on the front. An asymmetric raglan allows for this asymmetry in the human body. In knitting patterns this means doing a few extra increase rows on the back shoulders only. The difference is quite small and relatively stable across the size range: usually just a couple of more rows or 1–2 extra increases on the back is needed.

Wychwood is an asymmetric compound raglan.

The term 'asymmetric raglan' might be new but the concept isn't. In The Principles of Knitting, June Hemmons Hiatt doesn't even make the distinction between constant-rate and compound raglans or symmetric and asymmetric raglans (emphasis mine):

The triangle that defines the front and back armhole will have the same width, but a different length, which means each slope requires a separate pattern. The same is true of the triangle that that defines the sleeve cap, and in addition, these measurements are entirely different from those of the front and back armholes.

In other words, by her definition a raglan sleeve construction is comprised of four entirely different angles in raglan lines, and the increase-rate calculations for each should be made separately and independently of each other. There isn't even a mention in the book that these four lines — front, back, left, and right — could or should be forced equal. Why did we start to assume so?

On a compound raglan all increase rates are different and the raglan lines are of different lengths. On an asymmetric raglan the top of the sleeve cap is asymmetric: higher in the back than in the front.

It is relatively easy to achieve this asymmetric shaping when knitting flat in pieces: you'll use different rates of on the left and right edges of the sleeve, for example. But it isn't difficult in top-down raglans, either, whether working entirely in the round or starting the upper yoke flat.

If you're knitting in the round and doing short-row shaping to lower the front neck (as you should), start the short rows on the sleeves (as opposed to the fronts) and work a couple of increases only at the back shoulders before extending the short rows to cover all fours raglan increase points.

If you're knitting back and forth instead, such as when knitting a cardigan or doing the "staggered start" approach, cast on stitches only for the back and sleeves, and work the first one or two increases only at the back shoulders before gradually starting to shape the front neck.

Add to that variable increase rates and increase rates calculated separately for the body and sleeves, and you've got a true asymmetric compound raglan.

Compound Raglans Today

Take another look at the chart on raglan knitting patterns released in 2000–2024. That tiny blip of purple at the bottom right of the chart? That's the number of compound raglan patterns released annually. It's been averaging around 10 or so patterns released each year since 2014. The number has been steadily on the rise over the past two years but it's still a drop in the ocean of all raglan patterns.

The number of compound raglan patterns released annually is minuscule compared to all raglan patterns.

There are probably multiple reasons for an increase in compound raglan patterns but the biggest driver is easy to identify: the call for more size inclusivity in knitting patterns. As was noted in the discussion above, the standard raglan shaping rate of 8 stitches on every other round only works for a limited number of sizes somewhere towards the middle of the size range.

Up until now designers have been able to get away with constant-rate increases, same rates on body and sleeves, and symmetric front and back bodies. As the demand for size-inclusive patterns increases, this is no longer enough. On the extreme ends of the size range — both small or large — you need more finesse in raglan shaping. Rather than have all sizes follow the same formula, raglan shaping instructions can to be broken down into grading blocks that all behave a little differently: slower increase rates and more increases in the sleeves in the smaller sizes, faster increase rates and more increases in the body in the larger sizes. The exact numbers will, of course, depend on the stitch-to-row gauge ratios of the pattern.

Example of grading blocks for compound raglan shaping.

As that little blip on the chart indicates, more and more knitting designers are beginning to incorporate fine-tuned raglan shaping in their patterns. Aimee Sher of Aimee Sher Makes and the designer duo Selena Shepard and Abbye Meiklejohn from Wool & Pine are on the forefront in compound raglan patterns but there are many others, of course. While there isn't (yet?) a tag for this type of sleeve construction in the Ravelry database, you can tease them out by using the pattern notes filter in the advanced search and typing in "compound raglan".

References

These books were essential in writing this blog post. Many of them are decades old — Knitting from the Top was originally published in 1972! — but the principles of compound raglan shaping don't change with age.

• Ann Budd, The Knitter’s Handy Book of Top-Down Sweaters (Interweave 2012)

• Amy Herzog, Amy Herzog’s Ultimate Sweater Book: The Essential Guide for Adventurous Knitters (Abrams 2018)

• June Hemmons Hiatt, The Principles of Knitting: Methods and Techniques of Hand Knitting (Simon and Schuster 2012)

• Maggie Righetti, Sweater Design in Plain English (St. Martin's Press 1990)

• Barbara G. Walker, Knitting from the Top (Schoolhouse Press 2022)

---

Pin this post!