Methodology

How our data and estimates work

Parents and students use this app to make consequential decisions. We owe you transparency about where the numbers come from and what they can - and cannot - tell you.

Data sources

• College Scorecard (U.S. Department of Education): acceptance rates, median earnings 10 years post-enrollment, debt, completion rates. Refreshed annually. Source: collegescorecard.ed.gov.
• Common App Member Requirements: per-school deadlines, application fees, test policy, recommendation requirements, supplemental writing requirements. Sourced from the public PDF Common App publishes for the current cycle. The PDF doesn't include CEEB IDs, supplemental-essay counts, or interview policy - those fields stay blank in our data.
• Bureau of Labor Statistics OES (May 2024 release): occupation-level median wages, 25th and 90th percentile, and employment growth projections (2023-2033). Used in Careers by Major. Source: bls.gov/oes.
• O*NET OnLine: major-to-occupation crosswalks via Classification of Instructional Programs (CIP) and Standard Occupational Classification (SOC) codes. Used for the major-to-careers mapping.
• Static enrichment data(~200 schools): hand-curated values for graduation rate, campus size, student-faculty ratio, financial aid percentages, demonstrated-interest weight, ED advantage multipliers, supplemental essay counts. Sourced from each school's public website, IPEDS, and the school's Common Data Set when available.

The chance estimate

Our chance estimate (visible in /compare) is deliberately simple and explainable, not a black box. The math:

1. Start with the school's overall RD acceptance rate.
2. Adjust for stats fit.We treat ~100 SAT points and ~0.3 GPA as one band on either side of the school's reported averages. Above the median multiplies the baseline by up to 1.7x; below multiplies down to 0.4x.
3. Adjust for application timing.When applying ED, we honor the school's reported earlyDecisionAdvantage multiplier (typically 1.5x to 3x).
4. Adjust for demonstrated interest.Schools that mark interest as "Important" weight tracked touchpoints meaningfully: 3+ touchpoints adds 15%, zero touchpoints subtracts 25%. "Considered" schools see smaller adjustments. "Not Considered" schools ignore touchpoints.
5. Clamp to 1-95%.The number never reads as "guaranteed" or "impossible" without strong evidence.

What the chance estimate does NOT account for

• Essay quality
• Recommendation strength or specificity
• Hooks (legacy, athletic recruitment, faculty connections, first-generation status, geographic diversity)
• Major-specific admit rates (engineering and CS programs at universities are typically more selective than the university overall)
• Application portfolio strategy (shotgunning highly selective schools weakens the average odds, not the per-school odds)
• Application year volatility (acceptance rates have moved 30-50% across cycles for some schools)
• Institutional priorities for any given cycle (budget, demographic targets, geographic balance)

Treat the number as one signal, not THE answer. If our estimate disagrees with your school counselor's read, weight the counselor higher.

Cost and earnings figures

Tuition figures are sticker price, not what most students actually pay. The Cost vs Earnings chart in /compareattempts a fairer comparison by subtracting average financial aid from sticker, then multiplying by 4 years - but this still doesn't reflect your specific family'said award. For that, use each school's Net Price Calculator (linked from the school detail page).

Earnings figures (median, 10 years out) come from College Scorecard, which uses federal tax data. They reflect a single number for the entire school across all majors - earnings by specific major are visible in /outcomes.

How fresh is the data?

Federal data (Scorecard, BLS) refreshes annually, around spring. Common App requirements refresh each summer for the upcoming cycle. The 200 hand-curated school records are reviewed manually each summer, but individual fields (e.g., demonstrated-interest weight) may go a few years between updates. Look for the "Updated" or "cycle" badges next to data fields - they show how fresh each value is.

Bugs, gaps, or wrong data?

We document everything we can verify; we don't guess at fields we don't have sourced. If you spot something wrong or stale, let us know and we'll correct it.