The rules behind the result
This page lists the exact FFIT formulas the calculator runs, in the order it runs them, with the source papers and the reasons the answer can change. No scoring, no hidden weights.
femalebodyshapecalculator.com turns four measurements — bust, waist, high hip and hip — into a published FFIT body-shape result and shows the exact rule that produced it.
What FFIT is
FFIT — the Female Figure Identification Technique — was developed at North Carolina State University from 3D body scans for clothing fit. It compares four circumferences: bust, waist, high hip and hip. The first version, published by Simmons, Istook and Devarajan in 2004, described nine shapes. A 2007 paper by Lee, Istook, Nam and Park turned seven of those shapes into the arithmetic rules that most online calculators still use.
The method does not use height, weight, age, photographs, shoulder width or facial features. That is a limit as well as a simplification: two bodies with the same four circumferences get the same label even if they look different in a mirror.
The four measurements
| Measurement | Where the tape goes | Used for |
|---|---|---|
| Bust | Around the fullest part of the bust, tape level with the floor, not compressing the tissue. | All rules |
| Waist | Around the narrowest part of the torso, usually just above the navel, with a relaxed exhale. | All rules |
| High hip | Around the top of the hip bones, about 7–9 cm below the waist, before the body widens to the fullest hip. | Bottom Hourglass and Spoon |
| Hip | Around the fullest part of the hips and seat, feet together, tape level. | All rules |
The 2007 formulas
These are the seven-shape rules used by most calculators. They are checked in the order shown, and the first match wins.
| Order | Shape | Rule (inches) |
|---|---|---|
| 1 | Hourglass | (bust − hip) ≤ 1 AND (hip − bust) < 3.6 AND (bust − waist ≥ 9 OR hip − waist ≥ 10) |
| 2 | Bottom Hourglass | 3.6 ≤ (hip − bust) < 10 AND (hip − waist) ≥ 9 AND high hip ÷ waist < 1.193Needs high hip. |
| 3 | Top Hourglass | 1 < (bust − hip) < 10 AND (bust − waist) ≥ 9 |
| 4 | Spoon | (hip − bust) > 2 AND (hip − waist) ≥ 7 AND high hip ÷ waist ≥ 1.193Needs high hip. |
| 5 | Triangle | (hip − bust) ≥ 3.6 AND (hip − waist) < 9The 2007 rule assumes the waist is smaller than the bust and hips. |
| 6 | Inverted Triangle | (bust − hip) ≥ 3.6 AND (bust − waist) < 9 |
| 7 | Rectangle | (hip − bust) < 3.6 AND (bust − hip) < 3.6 AND (bust − waist) < 9 AND (hip − waist) < 10 |
What changed in 2020
Sokolowski and Bettencourt found that the 2007 formulas assumed the waist was always smaller than the bust and the hips. On plus-size scans with a larger abdomen, that assumption produced labels that did not match what the scans actually looked like — and some scans matched no rule at all. The 2020 revision adds two shapes (Diamond and Oval) and non-negative lower bounds to Triangle, Inverted Triangle and Rectangle.
| Order | Shape | Rule (inches) |
|---|---|---|
| 1 | Hourglass | (bust − hip) ≤ 1 AND (hip − bust) < 3.6 AND (bust − waist ≥ 9 OR hip − waist ≥ 10) |
| 2 | Bottom Hourglass | 3.6 ≤ (hip − bust) < 10 AND (hip − waist) ≥ 9 AND high hip ÷ waist < 1.193Needs high hip. |
| 3 | Top Hourglass | 1 < (bust − hip) < 10 AND (bust − waist) ≥ 9 |
| 4 | Spoon | (hip − bust) > 2 AND (hip − waist) ≥ 7 AND high hip ÷ waist ≥ 1.193Needs high hip. |
| 5 | Triangle | [(hip − bust) ≥ 3.6 AND 0 ≤ (hip − waist) < 9] OR [(bust − waist) < 0 AND (hip − waist) ≥ 0]The second branch captures a smaller bust than waist while the hips still lead the waist. |
| 6 | Inverted Triangle | (bust − hip) ≥ 3.6 AND (bust − waist) < 9 AND (hip − waist) ≥ 0 |
| 7 | Rectangle | |bust − hip| < 3.6 AND 0 ≤ (bust − waist) < 9 AND 0 ≤ (hip − waist) < 10 |
| 8 | Diamond | (hip − waist) < 0 AND (bust − waist) < 0Waist larger than both bust and hips. |
| 9 | Oval | (hip − waist) < 0 AND (bust − waist) ≥ 0Waist larger than hips, not larger than bust. |
The calculator uses the 2020 set by default. It also computes the 2007 set and displays it when the two disagree, so you can see which version your numbers depend on.
Why the order matters
Some bodies meet the conditions for more than one shape. FFIT checks the rules in a fixed order and stops at the first match. Spoon, for example, overlaps with Bottom Hourglass and Triangle on purpose: the high-hip ratio decides which one wins. If you reorder the rules, the same measurements can produce a different label. The order above is the order the code uses.
When high hip is missing
Bottom Hourglass and Spoon both need the high hip. Without it, the calculator does not guess. If your hips lead your bust and your waist is at least 7 inches smaller than your hips, it returns the Pear family and asks for the missing measurement. That is why a result can say “Pear family” instead of “Spoon” or “Triangle”.
How the result card works
- Primary result. The first matching rule under the 2020 formulas.
- Rule text. The conditions that fired, with your numbers substituted.
- Closest boundary. The nearest threshold that is currently satisfied — the edge a small change would cross.
- Sensitivity. The calculator re-runs the rules with each measurement moved up and down by 1 cm, then names any shape that appears.
- 2007 comparison. If the older formulas would give a different label, both are shown.
Limits you should know
A 2021 study by Parker, Hayes, Brownbridge and Gill measured 1,679 women using different tape-placement definitions. Moving a measurement by about 1 cm moved up to 40% of the sample into a different FFIT category. The paper also notes that FFIT omits shoulder measurements, even though shoulder width matters for garment construction.
Two practical consequences: treat a result near a boundary as provisional, and treat the label as a description of four circumferences rather than a fact about the body. The measurement guide is written to reduce placement error, but it cannot remove it.
Sources
- Simmons, Istook & Devarajan (2004). Simmons, K., Istook, C. L., & Devarajan, P. (2004). Female Figure Identification Technique (FFIT) for Apparel. Journal of Textile and Apparel, Technology and Management, 4(1). https://www.ergo-eg.com/uploads/books/devarajan_full_106_04%20(3).pdf The original nine-shape FFIT system, developed at North Carolina State University from 3D body scans.
- Lee, Istook, Nam & Park (2007). Lee, J. Y., Istook, C. L., Nam, Y. J., & Park, S. M. (2007). Comparison of body shape between USA and Korean women. International Journal of Clothing Science and Technology, 19(5), 374–391. The seven-shape formula set used by most online calculators. It uses bust, waist, high hip and hip.
- Sokolowski & Bettencourt (2020). Sokolowski, S. L., & Bettencourt, C. (2020). Modification of the Female Figure Identification Technique (FFIT) Formulas to Include Plus Size Bodies. Proceedings of 3DBODY.TECH 2020. DOI 10.15221/20.22. https://doi.org/10.15221/20.22 Adds Diamond and Oval, and non-negative lower bounds for Triangle, Inverted Triangle and Rectangle. This site uses these formulas by default.
- Parker, Hayes, Brownbridge & Gill (2021). Parker, C. J., Hayes, S. G., Brownbridge, K., & Gill, S. (2021). Assessing the female figure identification technique’s reliability as a body shape classification system. Ergonomics, 64(8), 1035–1051. DOI 10.1080/00140139.2021.1902572. https://doi.org/10.1080/00140139.2021.1902572 1,679 3D scans; up to 40% of women changed shape category when measurement placement changed by about 1 cm.
- Istook / SizeUSA distribution (2005). North Carolina State University (2005). Fashion designers ignoring shapes of women’s bodies. Newswise / NC State news release; Newcomb, B., & Istook, C. (2006). Using SizeUSA to improve apparel fit. https://www.newswise.com/articles/fashion-designers-ignoring-shapes-of-womens-bodies 6,318 US women from the 2003 SizeUSA scan survey: 46.1% Rectangle, 20.9% Spoon, 13.8% Inverted Triangle, 8.4% Hourglass, 10.7% other.
Method questions
Which formula version does the calculator use?
The default is the 2020 revision by Sokolowski and Bettencourt, which adds Diamond and Oval and gives Triangle, Inverted Triangle and Rectangle non-negative lower bounds. The 2007 seven-shape formulas are also computed, and the result card shows them when they disagree.
Why are the thresholds in inches?
The published formulas were written in inches. Converting centimetres to inches before applying them keeps the thresholds exact; applying 9 or 3.6 directly to centimetres would change the rules.
Why does the order of the rules matter?
Some bodies meet the conditions for more than one shape. FFIT checks the shapes in a fixed order and stops at the first match. The order is part of the method, not an implementation detail, so it is listed on this page in the same order the code uses.
What does the sensitivity check do?
It re-runs the classification with each measurement moved up and down by 1 cm, then reports any shape that appears. It is a practical version of the Parker et al. finding: if a small tape-placement change flips the result, the page says so instead of presenting one label as certain.