Estimates only β see methodology below for how these are calculated.
How this is calculated
Every self-rating on this page (Activities, Awards, Essay, Letters of Recommendation, Hardships) uses the same 1β7 scale: Extremely Weak, Weak, Below Average, Average, Above Average, Great, Excellent. Leaving a category unrated (or, for Activities/Awards, leaving the list empty) never counts against you β it's treated as neutral, not weak.
Each school gets an Academic Index (GPA percentile within that school's own reported GPA distribution β weighted by year if you gave a breakdown of either weighted or unweighted GPA β class-rank percentile against its reported rank bands, SAT/ACT composite percentile against its 25thβ75th range, course rigor from your AP/IB load and weighted-GPA strength, minus a small penalty for failed courses) and a Holistic Index (your activities β structured entries blended 50/50 with your self-rating β awards β level blended 50/50 with your self-rating β essay, letters-of-recommendation confidence, and hardships). If a school doesn't report a GPA distribution or test-score range at all, we fall back to a rough generic estimate rather than dropping that factor entirely β noted per-school below when it happens.
The two are blended using weights derived from that school's own Common Data Set "Admission Factors" table β a school that marks test scores "Very Important" and essays "Considered" leans more on your Academic Index; a school that rates both "Very Important" blends them evenly.
That blended score is compared against a base acceptance rate using a logistic model to produce a probability, then mapped to one of seven tiers. The base rate uses that school's own reported Early Action/Early Decision or Regular Decision rate when available (about a third of tracked schools report this split); otherwise it falls back to the school's overall rate adjusted by a disclosed, generic industry-wide early-round boost. If you flagged a competitive major, that rate is further reduced by a disclosed generic multiplier β we don't have real per-major CDS data, so this isn't school-specific. We also apply a small upward calibration shift throughout: a Common Data Set distribution describes the admitted class, not the full applicant pool (which includes many unrealistic "reach" applications that drag the raw acceptance rate down) β so a profile that matches the admitted-class median is treated as better odds than the bare acceptance rate would suggest.
Small additive adjustments apply for: in-state residency at public schools (scaled by how much that school says it considers residency β different for every target school, so a CA school and an NY school are judged independently), legacy status (scaled by that school's "alumni relation" importance), first-generation status (scaled by that school's "first generation" importance), community-service/work activities (scaled by that school's "volunteer"/"work experience" importance), and international applicants at need-aware schools.
- This is a statistical estimate, not a prediction β admissions decisions involve factors this form can't capture.
- Race/ethnicity is not used in scoring.
- Activity role weights, the research bonus, and the competitive-major adjustment are general heuristics, not school-specific CDS data.
- Schools with incomplete Common Data Set submissions fall back to a generic 55/45 academic/holistic weighting, noted per-school below.