How it works
The cast-on figure comes from your gauge alone: stitches = target width x (your stitches / swatch size), rounded to a whole number, with rows worked the same way from your row gauge. Swatch size is 10 when you choose Metric (cm, per 10 cm) and 4 under Imperial (inches, per 4"). The pattern's own gauge is used separately, to produce the stitch and row ratios and the pattern-as-written width comparison in the table.
Your gauge stitches is the number that matters most, and it should come from a washed and blocked swatch rather than the ball band. Defaults are a 22-stitch, 30-row pattern gauge against your 20 stitches and 28 rows, with a 50 by 60 target. The needle hint compares the two stitch counts: more than 1.5 stitches apart suggests a full size change, more than 0.5 a half size.
Stitches and rows are scaled independently, which is honest but not always sufficient, because row gauge frequently refuses to match even when stitch gauge does, and shaping worked over a fixed number of rows needs rethinking rather than rescaling. The result is not rounded to a stitch-pattern repeat either, so ribbing, lace or cables may need a stitch or two added by hand. Ease, blocking growth and yarn relaxation are not modelled at all.
Frequently asked questions
My gauge is 20 stitches where the pattern says 22, what do I cast on? +
For a 50 cm width at 20 stitches per 10 cm, 100 stitches. The tool multiplies your target width by your own stitches per unit instead of scaling the pattern's number, so it works from finished measurements rather than a ratio.
Should I go up or down a needle size? +
Fewer stitches per 10 cm than the pattern means your stitches are too large, so go down. More stitches means you are knitting too tightly, so go up. The tool states which, and whether the gap warrants a full size or a half.
Does the converter work in inches? +
Yes. Selecting Imperial switches the swatch basis to 4 inches, the standard American gauge square, and relabels the target width and length fields. Enter every measurement in the same system you selected, since nothing is converted between the two.
How wide would the pattern come out at my gauge? +
The bottom table row answers exactly that, dividing the pattern stitch count by the pattern's gauge and then by yours. With the defaults, 120 stitches measures 54.5 cm as written but 60 cm at your looser gauge, a 5.5 cm difference.
Why does my row count still look wrong after fixing stitch gauge? +
Row gauge is notoriously stubborn, and changing needle size to correct stitch gauge often leaves rows out. The tool scales rows from your own row gauge and target length, so use that for plain sections and follow the pattern's shaping instructions elsewhere.
Gauge is the measurement the pattern depends on
Patterns are written in stitches and rows, not in centimetres, so the finished size depends entirely on your stitch size matching the designer's. The defaults here — around 22 stitches and 30 rows to 10 cm — are a common worsted-weight gauge, and a difference of one stitch per 10 cm compounds into several centimetres across a garment body.
Two knitters with identical yarn and needles routinely produce different gauge, because tension is personal. That is why the swatch exists and why skipping it is the most common reason a garment comes out the wrong size.
Swatching properly, and row gauge
A swatch has to be large enough to measure away from the edges — at least 15 cm square — worked in the pattern stitch the garment uses, and blocked exactly as the finished piece will be. Measuring an unblocked swatch is measuring the wrong fabric, and some fibres change dramatically on blocking.
Adjust gauge with needle size rather than by trying to knit tighter or looser, which does not hold over a whole garment.
Row gauge is usually secondary because length is often worked to a measurement rather than a row count — but for anything shaped over a set number of rows, such as a raglan yoke or a sleeve cap, it matters as much as stitch gauge.
