How we model
Every figure we publish is labeled by how we know it.
Construction figures come from production records. Wear counts come from real wearers. Estimates come from our physics model. When our testing produces measurements, they replace the estimates on this page.
The three labels
Fact
Construction and origin, backed by production and import records.
Reported
Real wearers, quoted as given, on record with dates.
Modelled
Estimates from our physics model. Never called tests. Replaced by measurements.
What stands behind each claim
The ledger we keep on ourselves.
| Claim | How we know | Label |
|---|---|---|
| 0.07 mm film | Manufacturer engineering specification | Fact |
| About 1 g per piece | Derived from shipment records | Fact |
| Thousands of micro-perforations | Manufacturer engineering specification | Fact |
| Crafted in Korea | Import invoices on file | Fact |
| Latex-free | Material specification | Fact |
| 15+ wears | Internal wear testing to 15; one professional wearer reported 20-30 from a single pair (on record, April 2026) | Reported |
| Holds through heat, perspiration, dancing | Construction (perforated ultra-thin film) plus modelled scenario estimates below | Modelled |
| Comes off clean | Modelled residue estimate below | Modelled |
| Not waterproof | Design limitation, disclosed | Fact |
The model
What it simulates, and what it cannot.
Our adhesion model simulates how a cover behaves on skin: peel strength of the adhesive, edge behaviour as a function of film thickness, the effect of perspiration, heat and skin oils, and degradation across wash cycles. Each scenario runs 800 simulated wear profiles spanning different skin types, activity levels and care habits. Inputs are anchored to our suppliers' engineering specifications.
The estimates it produces are exactly that: estimates. We publish them because a labeled estimate is more honest than an unlabeled promise, and because publishing them commits us to replacing them with measurements.
Where the model disagrees with our own experience, we keep the lower number in check. The model's median wear count currently sits below our tested 15+ and below the 20-30 one professional wearer reported. We have deliberately not tuned it to match. Only the wash-cycle measurements will move it.
Constructions compared
The physics belongs to the construction, not the logo.
| Construction | Typical thickness | Survives a summer event | Survives an office day | Functional wears | Residue after heat | Under silk |
|---|---|---|---|---|---|---|
| Ultra-thin perforated film (ours) | 0.07 mm | 8 in 10 | 7 in 10 | 10 typical; 15+ tested; 20-30 reported with care | None expected | Invisible: matte, no edge |
| Molded silicone dome | 1.5-2 mm | 7-8 in 10 | 7 in 10 | ~20 | Minimal | Dome can print through fitted fabric |
| Fabric-backed gel insert | ~3 mm | 9 in 10 | 9 in 10 | ~10 | Frequent | Visible structure under thin fabric |
| Paper-mesh disposable | ~0.13 mm | 6 in 10 | 5 in 10 | 2-3 | Occasional | Low profile |
| Marketplace generic molded | ~1.2 mm | 4 in 10 | 3-4 in 10 | ~10 | Frequent | Blunt edge shows |
What replaces the estimates
Measurements, in this order.
- 1Wear-cycle counts across three care routines, on real pairs, until failure.
- 2Peel-strength measurement of the adhesive, fresh and after repeated wears.
- 3Day-wear diaries: full days, logged conditions, logged outcomes.
- 4Heat sessions: summer-condition wear with perspiration, timed.
- 5Professional wardrobe feedback from live stage use, as it arrives.
As each completes, its results publish here and the label upgrades from modelled to measured.
Questions
Why publish estimates at all?
Why compare constructions instead of brands?
What would change these numbers?
July 2026 · This page changes as evidence lands.