← Sorting Hat

Methodology

How this works

Sorting Hat is a psychographic neighborhood recommender. It asks 16 questions about how you actually live, scores you on 18 lifestyle dimensions, and ranks 115neighborhoods across the NYC metro by fit. The goal is to surface places where you’d genuinely thrive, with honest tradeoffs called out — not broker copy.

The dimensions

Every neighborhood is scored on these 18 dimensions. Each score lives in [-1, 1]. 8 are symmetric (both poles are real lived preferences — landing on the wrong one hurts in either direction). 10 are asymmetric needs(only the high end is an active preference; landing low just means “not a driver” and the neighborhood’s abundance can’t hurt you).

The questions

Across the 16 questions, three formats appear:

The vector of dimension scores is derived from your answers, not accumulated. Editing an earlier answer recomputes everything from scratch — no drift across back-navigation.

The math

For each neighborhood, the engine measures how far its profile sits from yours, dimension by dimension. Symmetric dimensions contribute the squared difference in either direction. Asymmetric-need dimensions contribute the squared shortfallonly — if the neighborhood over-delivers what you’d already deemed unimportant, no penalty. Distances are mapped to a 0–100% match score.

A small cultural-tag boost (+8% per match, capped) rewards alignment between cultural communities you selected and the neighborhood’s tagged communities. Non-negotiable must-haves operate as hard filters — neighborhoods that fail any one are excluded from the ranked list entirely, with the failing constraint surfaced so you can see why.

Where the data comes from

Archetypes

Each result page opens with a named archetype: a 8-way clustering of common lifestyle profiles in this metro. The match is the archetype whose profile sits closest to your vector. The archetype label is descriptive, not prescriptive — your ranked neighborhoods come from your actual vector, not the archetype’s.

Honest limitations

Want to see how individual answers move neighborhoods up or down in the ranking? Open the sandbox — same engine as the real quiz, with a live ranking panel. Or browse a static per-question paths report showing which neighborhoods each answer pulls toward. For the inverse view — what fraction of quiz combinations land each neighborhood in the user’s top results — path reachability.