Yarn calculator with every assumption on the page

It comes from a shared table of worsted-weight averages, and it is a starting point rather than an answer — calibrate before you buy, because readings taken off real finished projects routinely land well above the prefilled figures and the difference is measured in whole skeins. Nobody publishes length per stitch: it is not on a ball band, it is not in the Craft Yarn Council standard, and it cannot be, because it depends on stitch pattern and tension as much as on the yarn — a cabled panel eats far more than the same stitch count in plain stockinette, and a loose hand on a large needle eats more than a tight one. The calibration box takes about a minute: give it the stitch count of anything you have already finished and the yardage it actually consumed, and it hands back a figure measured on your own work.

Because that number was written for a specific yarn in a specific put-up, and one of those three things has usually changed by the time you buy. Substituting a yarn changes it, since two skeins of identical weight can hold very different lengths. Changing size changes it, and not proportionally — a garment two sizes up is more than two sizes' worth of yarn because area grows with both dimensions at once. Working the piece longer, adding a border, or knitting at a gauge other than the printed one all change it. The stitch count is the thing that does not lie, which is why this page starts there.

Yes, and by more than most people allow for — this is the largest single source of error on the page. A cable pulls the fabric in, so the same width takes more stitches and each of those stitches also carries more yarn through the crossing; estimates in the region of a third more for a heavily cabled fabric are common and are not an upper bound. Stranded colorwork is worse in a different way: two yarns run along every row, so the length consumed is roughly double even though the stitch count is unchanged, and floats behind the work add more. The way to handle all of it is to calibrate against a piece worked in the same technique rather than to scale a plain-stockinette figure by guesswork.

Because the interval around the estimate crosses a skein boundary, and printing one number there would be a guess dressed as arithmetic. If your yardage lands between 1,150 and 1,400 and your skeins hold 200 yards, the honest answer is six or seven, not six. Two ways out: narrow the interval by measuring your own length per stitch instead of accepting the prefilled one, or buy the larger figure. Buying up is not waste in the way it feels — an extra skein bought in the same dye lot is insurance you cannot buy later, and the leftover figure printed beside the result tells you how much of it you would expect to have left.

No — a stitch consumes the same length whichever direction the fabric is going, so a hat of 10,000 stitches and a flat piece of 10,000 stitches at the same gauge want the same yarn. What changes is everything around the stitches. Working flat means seams, and a seam eats yarn and needs tails left long enough to sew with, which is what the waste percentage is for. Working in the round means no seams but often means steeks, a join at every round, and in colorwork a float carried across areas you would have left out flat. If you are moving a pattern between the two, adjust the waste allowance rather than the per-stitch figure.

Run the arithmetic backwards: multiply your yardage by 36 to get inches, divide by your length per stitch, and that is the stitch budget. At 400 yards and a calibrated 0.85 inches a stitch that is about 16,900 stitches, which at 5 stitches and 7 rows to the inch is a piece around 22 inches by 22 — or, held to 8 inches wide, a scarf of 422 rows finishing about 60 inches long. Take a tenth off the budget before you plan against it, for the same seams and ends the waste field covers here, and remember that the last skein is the one you run out of in the middle of a row.

For a single piece with two ends to weave in, ten percent is generous. For anything sewn together it is often not: a cardigan in five pieces has ten seams' worth of tails, a fringe consumes yarn that never appears in any stitch count, and a garment with a separate button band, collar and pockets can lose closer to a fifth. Two cases deserve zero rather than ten: a scarf worked in one piece and cut nowhere, and any project where you have already added the border stitches to the count instead of leaving them out. The field is there to be changed, and the interval beside the result shows whether the change matters.