Start with the body
Exact band and cup availability is a hard constraint, not a cosmetic filter.
A recommendation should be
possible to disagree with.
So we show the source, the rule and the limitation. Found / Fit does not ask an AI to invent a persuasive answer; it applies a visible scoring system to a deliberately small dataset.
Open all 40 sources ↗Match the movement
Running, HIIT and daily wear demand different compromises. Support claims are interpreted in context.
Weight the failure
Bounce, chafing, breathing, heat and shoulder load are scored separately because one “best bra” cannot optimize all of them.
Expose uncertainty
“Strong” means repeated, specific evidence—not certainty that it will fit your individual body.
The scoring model
Every point has a job.
Scores are deterministic for the same answers and dataset. They are comparative, not clinical.
The v2.3 model below runs in shadow mode first. Until the audit is accepted, public ordering remains on v2.2 while both scores and ranks are stored for comparison.
Size and body fit20 pointsExact brand mapping; sister sizes only when explicitly approved
Activity and support10 pointsCompared with the selected impact requirement
Prioritized pain outcomes30 points50/30/20 weighting; approved product-specific discussions only
Feature priorities15 pointsMust-have, important and nice-to-have preferences
Returns and access8 pointsDestination, returnability and policy quality
Price fit7 pointsScored only after the budget hard gate
Evidence quality10 pointsReview state, specificity, quantity and freshness
Counter-evidenceup to −10Non-pain context-matched negative claims are deducted after scoring
Before scoring, products are excluded when the approved size is unavailable, support is insufficient, wire preference conflicts, the price exceeds budget, the product is discontinued, or no in-stock returnable offer serves the destination. Evidence confidence is also shown separately from the fit score. Product pain outcomes remain neutral until at least three independent, human-approved discussions exist; six discussions have medium influence and twelve allow full influence.
The source standardWhat gets in.
What gets in.
What stays out.
Included
- One unique original post ID per evidence row
- A specific body, activity or fit failure
- Purchase, replacement, trial or recommendation intent
- Direct Reddit page or same-post Reddit RSS
- Negative results from specialist products
Excluded
- Search snippets counted as source evidence
- Duplicate domains or tracking URLs for one post
- Generic complaints without enough context
- Brand promotion and unsourced summaries
- Assumed willingness to pay when no price exists
What this site cannot honestly tell you.
Fit certainty. Torso length, root width, tissue distribution and personal tolerance vary beyond band and cup.
Market size. Forty discussions are qualitative evidence, not a representative demand survey.
Permanent prices. Prices, policies, stock and color-specific size ranges can change after review.
Medical advice. Pain, wounds or breathing problems deserve qualified care when they persist.