Methodology

We score every candidate school across three explicit dimensions — academic fit, financial fit, and location fit — then anchor an admission probability against the school's published acceptance rate, then write 80–120 words of per-school reasoning that cites the actual numbers. No invented stats; if a school is missing data, the card shows “n/a”.

Data sources

• IPEDS HD (NCES) — name, state, city, public/private, urbanization, level. Public domain.
• College Scorecard (US Department of Education) — admission rates, SAT/ACT mid-50%, tuition, net price, cost of attendance breakdowns. Public.
• Common Data Set— collected incrementally per school; gaps surface in the list as “n/a”.

What we do NOT do

• We don't produce a global ranking — we produce a list scoped to your profile.
• We don't scrape paid sources or violate terms-of-service to enrich the index.
• We don't store your profile in a database (yet — coming in the paid release).
• We don't invent admission stats. If a school is missing data, the card shows “n/a”.

Known limits

Our admission probability is a heuristic, not a learned model — it will improve once we collect anonymized outcome data. Early-decision and early-action recommendations may lag this year's policy changes — always confirm with the school's admissions office.