The answer probably isn’t as easy as any of us would like. You don’t.

My father taught college math until I was in junior high, and he used to talk to my brother and me about math whenever we went out to dinner. (I don’t know why the subject didn’t come up when we were having dinner at home, but virtually every time we went out for dinner, he’d pull his mechanical pencil out of his pocket and work us through math problems on paper napkins until our food arrived. And sometimes after.) When I was in fifth grade or so, I came home talking about interpolation and extrapolation. He drew me a graph with numbers on it and whatnot and asked me tell him what the Y-value of some number on the graph was. The answer was easy; it was right there on the graph. Then he asked me to give him the Y-value of a number not on the graph. That wasn’t so easy.

“Extrapolation is dangerous,” he said to me. “Just because this looks like a straight line between these two numbers doesn’t mean that it continues to behave like that beyond those points.”

This is a roundabout way of saying that you don’t really know how the pattern would behave beyond the range of sizes in the pattern. There are reasonable guesses, though. Patterns, unlike some math functions, tend to behave logically. In the example I gave in the last post, the trend would say either to decrease the rate of change by two or increase by two. That difference of four stitches could be tremendous, or it might not be, depending on your gauge. So yeah, some of it comes down to guesswork. If you want a tight fit, try going down. If not, go up. The nice thing about the guesswork here is that you’re really unlikely to mess something up irrevocably.

In other news, I finished my third of the baby sunhat in Holiday Knits. It’s such a sweet little pattern, and kind of unconventional for a knitted baby hat. Will have to get pictures of the second and third hats (the first has already been given away) tomorrow, along with an attempt at a ballet slipper for my former-ballerina-cousin’s baby girl (not sure if it’s working) and the baby blanket I’m blocking for my mom. I might post the pattern for it too, as it’s something we accidentally made up together.

Anyway, this is something I’ve explained more than once over at LJ’s knitting community – how to retool a pattern to work with a yarn that doesn’t match the given gauge.

I’ll start by saything this isn’t always the greatest idea in the world. While you certainly could take a sturdy-looking aran sweater with chunky cables and work it with a fingering-weight yarn… well, why would you? In my opinion, this works best if you have a yarn that’s not substantially different from the prescribed yarn, especially if you’re making an object that has a lot of detail in it. If the detail is in the shaping, as in the hourglass sweater from Last-Minute Knitted Gifts, you can get away with a drastic change in gauge more easily than if it’s an intricate twisted-stitch garment.

I am one of those bad knitters, by the way. I rarely check my gauge, either out of laziness or impatience. But for this process is critical. Pick the yarn you want to use and work up a swatch, probably using the needle size recommended on the ball band. This swatch needs to be fairly sizable, because you want to be able to judge the behavior of it as a piece of fabric – how does it look? Is it comfortable to work with at this size, or are you having trouble keeping a consistent gauge? What’s the drape like? What kind of stretch do you get? If you’re doing something with a stitch pattern other than a simple stockinette, make sure you swatch in that too, whether or not the pattern gives the gauge in that stitch pattern as well. Sometimes a needle that works well for stockinette or garter or seed stitch is exponentially harder when you’re working with a more complicated pattern. You really don’t want to kill yourself over a sweater, so make sure you pick something reasonable.

For the purpose of this post, I will be using Elizabeth Morrison’s Lucky at Knitty as my example. It’s a classic men’s sweater, but at a finer gauge, it could easily be a beautiful, feminine garment. So how do you do it?

The gauge of the sweater in the pattern is 16 stitches/26 rows over 4 inches on US 7 needles. That’s on the thick side of worsted, so what if you wanted to make it with, say, a sport weight?

Let’s make it easy on ourselves and presume we want to make the small size. The instructions say to cast on 83 stitches for the back. Divide that by the number of stitches per inch in the original gauge:

83/4 = 20.75

That’s the width of the hem of the front panel of this sweater, and since you work it straight up, there’s no shaping to worry about on this one. That’s the width all the way up to the armholes. So let’s figure that you want to work with a yarn that’s 6 stitches to the inch. Multiply your dimension by your gauge:

20.75*6 = 124.5

Sometimes when you’re doing this, you get lucky and land right on another number in the cast-on directions. Most of the time, however, you won’t. Depending on how many sizes are available and how much difference there is between the original gauge and your gauge, you may have to start extrapolating sizes. In this case, the largest size on the example pattern only casts on 115 stitches. However, as there’s a difference of 8 stitches between each size, you can easily size up to 123 stitches. At that point, you’re close enough to the number our arithmetic gave us that you can probably use that without a problem. This isn’t an exact science, and generally a stitch or two in one direction or the other isn’t a problem.

If you were lucky enough to land on a number given in the pattern, go forth and knit. There are still a few more things to consider, but you won’t need this next part. If you’ve had to extrapolate your size, you may want to take a moment to go through the pattern and write out your extrapolated instructions throughout. Might just be easier on you down the road.

Now, sometimes patterns don’t size up evenly the way this one does, but usually there’s some rhyme and reason to it. I’ve seen some patterns where the difference between the smallest and next-smallest size was the same as the difference between the next-largest and largest size. If that’s the case, I would look at what the pattern of increase is to extrapolate beyond the given sizes. (For those of you who know something about calculus, this would be like looking at the derivative or rate of change.) For example, if you were told to cast on 20 (30, 38, 44, 48, 54, 62, 72) stitches and you need one size bigger, look how the sizes vary. The differences run this way: 10, 8, 6, 4, 6, 8, 10. They’re increasing or decreasing by two stitches with each size. So if you need one size beyond the pattern, add 12 stitches – cast on 84.

This, of course, is the kind of math you need to do throughout the garment. And here’s where those who were lucky enough to land on a number in the pattern have to pay attention again. You have to do the same thing with your row gauge, or you’ll end up with something very wonky.

On a pattern like our example, this isn’t a huge deal in the body of the sweater. It says to work in pattern until the piece measures 15″. When you have a garment that tells you to work until the piece measures a certain length, frequently each size has a different length. In that case, follow the size that fits you, not the size you’re working. Otherwise you’ll end up with a bare midriff or a really baggy sweater. This may or may not be the look you’re going for.

Of course, some patterns give you a number of rows to work. This is especially true of sleeve patterns, and we’ll be looking at Lucky’s sleeves next. Figure out your row gauge. This pattern says to work 16 rows straight before you begin your shaping, so how much should you do? The pattern’s original gauge gives you 6.5 rows per inch. Row gauge can vary wildly (I’m a combined knitter, and my row gauge never matches patterns), so let’s just say you’re getting about 8 rows per inch. How long is 16 rows in the original gauge?

16/6.5 ≈ 2.46

So that, in our hypothetical gauge, would be:

2.46*8 ≈ 20

You would work 20 rows before starting the shaping, and would repeat this process throughout the sleeve.

ETA: Robin Dodge reminded me of a shortcut involving proportions. Say the pattern gauge is 5 stitches per inch and theirs is 4 – divide theirs by yours and multiply that by the number of stitches throughout the pattern. Most of the time, it works pretty well. But if it’s your first time going through this process, I would recommend doing it the long way, just to make sure you really understand the process and how everything interacts. Just my opinion, though.

One more thing: sometimes this takes some experimentation. For example, if you follow the directions for the raglan shaping exactly on this pattern, you could end up with something wildly misshapen. (Knitty’s Mariah, for example, had an enormous yoke on the larger sizes because the designer didn’t originally take into account the fact that there isn’t nearly as much variance in neck size as in bust size.) You have a lot more stitches on the needle than the pattern thinks you do. You may need to rethink the decreases in order to get a proper fit up top. This may include decreasing more stitches at a time or decreasing on every other right-side row, depending on how your stitch and row gauge interact.

As a side note, there’s an interesting byproduct of this process. You can resize a garment entirely with gauge. Holiday Knits has an adorable little baby sun hat in it. It’s about the size for a sixth-month-old baby, and is made with a worsted yarn. I wanted to make it for a younger baby, so I used a sock yarn on a smaller needle. I didn’t have to do any math at all once I had my initial gauge and knew the circumference that would give me. (And made it out of a self-striping sock yarn, so it’s super-cute.)

Comments and questions more than welcome. I’ve composed this over a few days, so this may not be the most coherent explanation ever.

Actually, I should start further back than that. Sure, I’m thinking about a baby cardigan, but it’s really more in the context of eventually thinking about a cardigan for me. I’ve discovered, however, that baby garments are a good place to experiment. I seem to have no shortage of friends having babies (which is a completely different discussion), so it’s a good time for me to start poking around with new techniques and design ideas.

Then the other night I started reading Eunny Jang’s steeking tutorial, and even though the thought of cutting my knitting up is vaguely terrifying, I kind of want to. I’ve recently fallen in love with Fair Isle anyway, and this seems like something I should try at least once.

So a simple yoked cardigan, the front opening steeked open. I prefer making raglan sweaters because I haaaaaate seaming, but in this case it’ll depend on whether or not I can figure out how to handle the increases within the pattern. And I’m thinking about using this simple snowflake as my motif. As for colors, the parents of the recipient aren’t finding out if they’re having a boy or a girl ahead of time, so I’m thinking something neutral and low-contrast, probably in Dale’s Baby Ull. I might even go as subtle as white and off-white, even though that’s kind of crazy considering it’s a baby gift. At least it’s superwash?

Time to dig out pattern books, I suppose. Have to start figuring things out somewhere.

And it’s the last bit (the fearless part, not the stupid part) that I really think is most important to one’s knitting. A fear of screwing up irreparably can stop you from doing a lot of beautiful projects out there, but here’s the thing: in my experience, it’s almost impossibly to screw up irreparably.

So I’ll start off with pictures of a project I finished about a year ago, but I’m still very proud of it. This lace tablecloth was a wedding gift to my childhood best friend. It was my first foray into lace. (Also, a crazy foray into lace.)

Click image for large version.

The finished dimensions of this piece were 90″ by 68″. The recipient was given her grandmother’s dining room table, a very large dark wood piece. The doubled strands of mercerized cotton in off-white made for a striking piece up against a dark surface like that, but also made for a headache or two (hundred) in the process.

The lace pattern was adapted from the lace shawl pattern in Sarah Dallas Knitting, worked on size 6 needles. Blocking this was a bit of a nightmare – there’s a dry cleaner in my area to whom I would entrust anything I’ve ever made, but his equipment was nowhere near large enough to take a piece this big. There was much crawling around on the floor the day I decided it had to be done right then or the gift would be horribly, horribly late.

Things I learned? Knitting with crochet cotton is generally a bad idea. No, really, it is. Especially doubling up strandes that are smaller than fingering-weight yarn. I had fun with the project when I wasn’t wanting to rip my hair out, but in retrospect, a less complicated lace pattern probably would have been a more appropriate first attempt.

Will I do it again? …Probably.

You know what they say. Nothing ventured…