Methodology
How the index is built
KINKSpec places every human sexual and non-sexual interest on one honest, comparable scale. This is how each entry is defined, classified and scored: in plain terms, with the research it rests on.
- 486
- Entries catalogued
- 11
- Categories
- 264
- Subcategories
- 0–100
- Popularity index
What we catalogue
A broad, encyclopedic view of human interests: named by the everyday word people actually use, with the clinical term kept alongside it.
Fetish
A strong, reliable erotic interest in a specific body part, object, material, act, role or scenario.
Paraphilia
The clinical umbrella term. A paraphilia is not a disorder in itself , it only becomes one when it causes real distress or harm.
Non-sexual fixation
Everyday, non-erotic fixations too: collecting, sensory interests, material obsessions, for the fuller picture.
Every entry sits in exactly one of 11 categories, from Body Parts & Partialism to Clinical Paraphilias, and is written as a neutral, clinical article: never explicit, never instructional.
Three catalogs, three leaderboards
Interests, identities and relationship structures are documented the same way, but never ranked against each other.
Interests
Kinks, fetishes and paraphilias, what someone is drawn to. The main catalog, at /browse.
Identities
Sexual and romantic orientations, who someone is. A separate vocabulary, at /orientations.
Structures
Monogamy, polyamory, open relationships, how a relationship is arranged, at /relationships.
All three use the same five-signal 0–100 index described below, and all three are ranked strictly within themselves. A structure is never ranked against a kink, because “more common than” only means something between comparable things. Confusing the three is the most common error in this subject area: an orientation is not a preference you pick, and polyamory is a relationship arrangement rather than an orientation.
The 0–100 popularity index
A single headline number per entry: how prominent an interest is across the world, not a measure of worth.
Real-world prevalence is wildly uneven. A handful of interests are held by tens of percent of people; the long tail by well under one in a hundred. Numbers that span so many orders of magnitude pile up near zero on a normal scale, so we measure prevalence on a logarithmic scale. On that scale, prevalence forms a clean bell curve, and the index spreads evenly across the full 0–100 range.
The self-assessment reads this same index back: it places the interests a reader selects on the curve described here and returns their weighted average position. Nothing in it is a separate instrument.
The base-10 logarithm of an interest's prevalence is normally distributed, a bell curve.
index 0–10
33 entries · 7%
index 10–20
40 entries · 8%
index 20–30
79 entries · 16%
index 30–40
93 entries · 19%
index 40–50
98 entries · 20%
index 50–60
87 entries · 18%
index 60–70
35 entries · 7%
index 70–80
15 entries · 3%
index 80–90
4 entries · 1%
index 90–100
2 entries · 0%
A clean, slightly right-leaning spread, mean 39.2, ranging 2–92. The log scale doing its job.
From prevalence to points
Prevalence is stretched across the index between two anchors, one person in ten thousand at the floor, three in five at the ceiling, so the rarest and most common interests both have room to breathe.
If an interest is held by…
- 0.1%26
- 1%45
- 5%71
- 14%83
- 50%98
Five signals behind every score
Prevalence anchors the index, but four more signals shape it, so a score reflects real-world presence from every angle.
Prevalence
How much of the adult population holds the interest
40%Search interestEST
Online search and adult-content demand
20%Community sizeEST
Size of dedicated communities and forums
18%Cultural visibilityEST
Mainstream and pop-culture awareness
12%Research attentionEST
Volume of academic and clinical study
10%
Where the numbers come from
Prevalence is computed; it is derived arithmetically from the entry's cited population estimate, and where a study genuinely measures that figure, the entry cites it. The four signals marked est : search interest, community size, cultural visibility and research attention, are indicative editorial estimates: scored 0–100 against a fixed anchor rubric (reference interests pin the scale, the catalog's giants sit at 85–95, mid-tier interests around 50–70, obscure paraphilias at 5–25: and everything else is placed relative to them), informed by public signals such as search-trend data, community sizes and publication volume, but not traceable to a per-entry source. Together they carry 60% of the composite weight, which is why the entry's confidence rating and this label travel with every score. Treat the index as a documented editorial judgement, not a measurement.
The five signals are combined with a geometric mean, not a simple average. That choice matters: an interest can't reach the top on one spike alone; it has to score across the board. Something prevalent but invisible in research lands in the middle, where it belongs.
Prevalence tiers
A plain-language band for each entry, from near-universal to vanishingly rare, with how many entries fall in each.
- Ultra-common≥ 20%26 · 5%0
- Very common8 – 20%47 · 10%1
- Common2 – 8%152 · 31%2
- Uncommon0.5 – 2%130 · 27%3
- Rare0.05 – 0.5%96 · 20%4
- Very rare< 0.05%35 · 7%5
Grounded in research
Figures aren't guesses. Every estimate is anchored to published prevalence studies and clinical literature, then kept consistent across the whole catalog.
~50%
of adults report interest in at least one paraphilia.
Feet
are the single most common body-part focus by a wide margin.
4–7%
of people have never had a single BDSM-themed fantasy.
2
fantasies are statistically rare, and both involve a party who cannot consent.
How sure we are
Prevalence research is thin and uneven, so every figure carries a confidence level, and we say so plainly.
- High
- Multiple converging surveys or community datasets agree.
- Medium
- One solid dataset, or several strong indirect proxies.
- Low
- Anecdotal or proxy-only, read it as an order-of-magnitude estimate.
Ethics & care
A reference about human sexuality has to be handled with care. Here's the line we hold.
Clinical and non-explicit
Every entry is written in neutral, encyclopedic language. No pornographic detail, no how-to: descriptions, not depictions.
Completeness, never endorsement
A small number of clinically-recognized paraphilias are inherently harmful or illegal because they require a non-consenting party. These 22 entries appear for completeness only, behind a clear content notice, with harm and legal context, never as endorsement or instruction.
Symbolic illustration
Each entry is illustrated with a single graphite still-life, an object or metaphor that stands in for the idea. Never a person, never an act.
Reference, not advice
KINKSpec is educational. It documents what interests exist and how common they are: it does not give medical, psychological or legal advice.
Limitations
What this index is, and isn't. We'd rather state the caveats than imply false precision.
- Prevalence figures are directional estimates synthesized from heterogeneous sources: clinical surveys, community sizes, search data, not census counts.
- Online and search-based signals over-represent Western, English-speaking people with internet access.
- The catalog aims for breadth over exhaustiveness. New entries are added and the whole index re-scored as better evidence arrives.