How We Calculate Your Match
Match scores are probabilistic estimates, not predictions. They reflect how a student’s academic profile compares to what is typical at each school — using a Bayesian model that weights eight measurable signals against national benchmarks.
The 8 Signals
All weights sum to 100%. Signals with more data are weighted more heavily; signals you haven’t provided yet default to national averages.
| Signal | Weight | What it measures | Data source |
|---|---|---|---|
| GPA | 28% | Compares your GPA against the typical GPA of admitted students at each school, estimated from the school's acceptance rate and national GPA distribution data. | College Scorecard (US Dept of Education) |
| Test Scores | 23% | Compares your SAT or ACT score to the school's reported 25th–75th percentile range. Scores above the 75th percentile add a meaningful positive signal. | College Scorecard / Common Data Set |
| Course Rigor | 18% | Measures the number of AP, Honors, and IB courses you've taken relative to what is expected for the school's selectivity level. | Student profile (self-reported) |
| Activities | 8% | Evaluates extracurricular involvement, sustained participation (3+ years), and leadership roles. Demonstrates depth and commitment outside the classroom. | Student profile (self-reported) |
| School Selectivity | 8% | Uses the school's overall acceptance rate as a baseline prior. More selective schools start with a lower base score that other signals can raise or lower. | College Scorecard (US Dept of Education) |
| Graduate Outcomes | 7% | Compares median earnings 10 years after enrollment to the national median, normalized by the fraction of students who received federal financial aid to correct for selection bias. | College Scorecard / Bureau of Labor Statistics |
| Major Fit | 4% | When you specify an intended major, compares the school's program-level earnings for that CIP code against the national average for the same major. | College Scorecard (CIP-level program data) |
| GPA Distribution Fit | 4% | When Common Data Set GPA distribution data is available, estimates where your GPA falls within the actual distribution of enrolled students. | Common Data Set (school-published) |
Data Sources
College Scorecard (US Dept of Education)
The primary source for acceptance rates, test score ranges, enrollment, earnings, debt, and cost data for all Title IV institutions. Updated annually, typically with a 1–2 year lag from the academic year.
IPEDS (Integrated Postsecondary Education Data System)
NCES survey data covering graduation rates, faculty ratios, program offerings, and campus safety (Clery Act). Updated annually on a similar schedule to College Scorecard.
Bureau of Labor Statistics (BLS)
Occupational Employment and Wage Statistics used for projected salary and growth data by occupation. Updated annually.
US Census / American Community Survey
ACS 5-year estimates used for regional cost-of-living context and demographics. Typically released in December with a 1-year lag.
Common Data Set
School-published CDS forms provide GPA distributions, standardized test policies, and admission factor weights for participating institutions. Availability varies by school.
Most data is updated annually, typically with a 1–2 year lag from the academic year. Each data point on school profile pages is labeled with its source year.
Profile Completeness and Accuracy
The more data you provide, the more signals the model can use — and the more accurate your match becomes. Missing data falls back to national averages, which works as a prior but reduces precision.
High Completeness
- •GPA (weighted or unweighted)
- •SAT or ACT score
- •AP/Honors/IB courses
- •Activities logged
- •Intended major
All five signals present. Match scores are most reliable.
Medium Completeness
- •GPA provided
- •Some activities or courses
Partial data. Test score or rigor additions improve accuracy.
Low Completeness
- •No GPA
- •No test scores
Match is based on national averages. Add your profile for real results.
Disclaimer: Match scores are probabilistic estimates based on public data. They do not predict admission outcomes. College admissions involve holistic review beyond quantifiable metrics — essays, letters of recommendation, demonstrated interest, and other factors that are not captured in this model.