The bad advice this corrects

A recurring pattern across dating tools and advice content is the numeric score — a compatibility percentage, an “interest level: 7/10,” a single figure meant to compress an entire, ambiguous human situation into something that looks precise and feels reassuring. It looks like clarity. It’s manufactured certainty wearing a decimal point.

The DEZAIR standard

This library will not attach a number to how much someone is interested in you, how compatible two people are, or how likely a relationship is to work. Not because the underlying question doesn’t matter, but because a single number implies a level of precision that doesn’t exist in the evidence, and presenting it anyway teaches readers to trust false certainty over their own honest uncertainty.

Why this is the actual standard

A percentage carries an implicit claim: that the underlying factors have been measured accurately enough to produce a specific figure, and that the figure means something stable and comparable — a 73% is meaningfully different from a 68%. Nothing about reading human interest supports that level of precision. The actual evidence available — a handful of behavioral signals, weighed against context and baseline — supports a rough, qualitative read: strong, weak, mixed, insufficient. Compressing that into “73%” doesn’t add information. It removes the honest fuzziness of the real evidence and replaces it with a false sense of resolution.

There’s a second problem underneath the first: a number invites a kind of decision-making this library specifically argues against — treating the read as a finished verdict rather than a provisional, evidence-weighted belief. Once something is expressed as “73% interested,” it stops feeling like a hypothesis to keep testing and starts feeling like a fact to act on. That shift is exactly the opposite of the “read, don’t guess” discipline this whole library is built around.

What this looks like in practice

Consistent with the standard:

  • Describing evidence in qualitative terms — strong, weak, mixed, still forming — tied to what’s actually been observed
  • Naming specifically which markers are and aren’t present, rather than compressing them into one output
  • Treating any read, however confident, as open to revision with new evidence

Inconsistent with the standard:

  • A single number presented as a summary of someone’s interest or compatibility
  • Language that implies false precision — “she’s 80% into you” — even informally, even as a rhetorical shorthand
  • Treating a qualitative read as though it had the same fixed, comparable precision a number would imply

Edge cases worth naming

This isn’t a rejection of structure or of trying to be specific — the five-marker reciprocity model, the evidence-strength labels used throughout this library, are both genuine attempts at precision. The difference is that they describe what kind of evidence exists and how much of it, rather than compressing everything into one deceptively tidy figure. Specificity about the evidence itself is the goal; false precision about the conclusion is what’s being avoided.

The standard

If the honest answer is “strong evidence across several independent signals” or “genuinely mixed, more observation needed,” that’s the answer this library gives — never a number dressed up to look like it knows something it doesn’t.