← Back to College Results

How the numbers work

What each figure on the results page counts, and why none of it is a prediction. Describes the site as of October 2026.

In short: every number comes from voluntary, self-reported submissions. Nobody checks them against school records, the students who share are not a random sample, and many groups are tiny. Treat the results as stories from classmates, not odds.

Where the data comes from

"Students" means submission groups

Each browser that submits gets its own random ID, and all the colleges submitted from it are grouped together. The site counts each group as one student. That's usually right, but:

Per application vs per student

Some figures count applications (one row per student per college); others count students (submission groups). They answer different questions and can't be compared directly.

Figures on the results page and what they count
FigureCountsHow it's calculated
StudentsStudentsSubmission groups in the selected school category.
Got in somewhereStudentsGroups with at least one "Accepted" in the selected category.
4-yr accept rateApplicationsAccepted ÷ (accepted + rejected + waitlisted), at colleges that aren't open admission. Pending decisions are left out. Community colleges are left out because they admit everyone.
Avg UW GPAStudentsAverage unweighted GPA, one value per group.
College details: RateApplicationsAccepted ÷ all decisions reported for that college, including waitlisted and pending ones.
CompareApplicationsSame as college details, plus averages over accepted applications only.
Where people ended upStudentsStudents reporting "enrolling" at each type of school. If nobody in the view has marked enrolling, acceptances are used instead, and the heading says so.
GPA range by destinationApplicationsMiddle half (25th–75th percentile) and median unweighted GPA of the rows behind "Where people ended up". Only shown with two or more values.
Acceptance rate by weighted GPAApplicationsAccepted ÷ all decisions in each GPA band at non-open-admission colleges. Pending decisions count in the denominator here.
Early vs RegularApplicationsAccepted ÷ (accepted + rejected + waitlisted) at non-open-admission colleges, by EA/ED vs RD.
Most applied schoolsApplicationsColleges with two or more reported decisions, spelling variants grouped.
Popular majorsStudentsIntended major as typed, one per group.

Waitlists, pending decisions and missing values

GPA

The form asks for an unweighted GPA (1.0–4.0) and a weighted GPA (1.0–5.0). It doesn't say which grades, years or weighting rules to use, so students may have calculated them differently. Treat small GPA differences as noise.

School categories

Students type college names freely, so categories are worked out from the name: UC and CSU by name pattern; "Public" from a list of public universities plus name patterns; "Private" for four-year colleges that aren't on the public list; "Community" by a list of local community colleges or the school type the student picked; T20, Ivies and liberal arts colleges by exact name lists. Because these are name matches, a college spelled differently from the list (for example "Stanford University" instead of "Stanford") can land in a different category or none.

Overall admission rates ("unverified estimate")

College details, Compare and "Hardest schools" show an approximate overall admission rate for many colleges. These come from a table built into the site. The table records no source or year for its figures, and a note in the site's code says several UC and CSU figures were found to be out of date and corrected in August 2026 while the rest were not re-checked. That's why they're labeled unverified estimates. For a college's real figure, use the college's own Common Data Set or the U.S. Department of Education's College Navigator.

Why these aren't your chances

Which controls change what

Known inconsistencies

These are documented rather than "fixed" because fixing them means changing how existing submissions are grouped or counted, which is under review: