Less mysteriously than most applicants assume, and more systematically. At selective colleges a reader usually works piece by piece through the academic record, the writing, the activities, the honors and awards, and the recommendations, and forms a view about each before forming one about the whole. At many schools more than one person reads a file before any decision is discussed. Dartmouth's dean of admissions describes its committee as a jury of sorts: the officer who read a file presents the candidate, and the group debates and votes. Columbia says no candidate is admitted without discussion by multiple admissions officers. Vanderbilt says it trains its readers to look for reasons to admit rather than reasons to deny.
Every college that participates in the Common Data Set reports how much weight it gives 18 different factors in first-year admission decisions: very important, important, considered, or not considered at all. Most colleges also publish a good deal about what they are looking for within those factors.
The factors schools weigh most heavily vary enormously, and so does what they say they want within them. There's no one-size-fits-all application, which is why guidance specific to the schools you're applying to is worth more than generic advice.
Some examples are public and easy to verify. The University of Florida does not consider letters of recommendation at all. The University of Washington states that its reviewers don't see test scores even when submitted. Notre Dame says its most competitive applicants are well-rounded, which cuts directly against the specialize-deeply advice most students receive. MIT lists the ability to prioritize balance between work and play among the things it looks for. Penn's admissions office names five traits inspired by Benjamin Franklin that it looks for, including being “imperfect but self-improving” — a quality most applications are built to conceal.
None of this is secret. But it's typically spread across multiple sources and surfaces (webpages, podcasts, blog posts, governmental data, etc.) and mostly not updated on a fixed schedule — so almost no family assembles it.
The honest answer is that it depends on the school, and the landscape has been shifting enough every year since 2020 that last year's advice is frequently wrong. Four situations now co-exist side by side.
Several colleges have also announced dated returns to requiring scores in future cycles. If you're a sophomore or junior, the policy that will apply to you may not be the one in effect today — check the effective year, not just the status.
One important gotcha: a college's Common Data Set describes the cycle that just ended. For any school that's changed its testing policy, the published weighting is a cycle out of date. The policy page is current; the grid is historical.
Most feedback a student gets on an application is evaluative — this essay is strong, this activity description is thin. That's useful, but it stops a step short of the thing that actually helps: knowing what to spend the next three weeks on when six things could be improved and there's only time for two.
Prioritization is school-specific by nature. A thin activities list matters more at a school that rates extracurriculars very important than at one that rates them considered. A missing score is decisive at a school that requires them and irrelevant at one that never looks at them. A recommendation you can't control is not worth worrying about at a school that doesn't read letters. The same file has a different to-do list at two different colleges, and the ordering is where most of the value sits.
Yes, and more than applicants tend to believe. Several schools single out the senior-year program as something they pay particular attention to — Middlebury states that no admission decision is made without midyear grades, and Duke requires first-quarter grades from Early Decision applicants. A lightened senior schedule is one of the most visible things in a file and one of the least often explained.
It varies more than almost any other factor. Some colleges rate recommendations very important. At least two well-known universities do not consider them at all. One large system tells applicants explicitly not to submit them. Before spending effort briefing a recommender, it's worth knowing which of those describes the schools on your list.
That a college considers more than grades and scores — essays, activities, talent, character, and the context a record sits in. It doesn't mean every factor counts equally, but it also doesn't mean the process is unknowable. Many colleges that describe their review as holistic also publish a report stating how much weight each factor carries.
No. Nothing honestly can. An odds estimate has to be built from what can be measured and quantified, and colleges say plainly that this isn't how they make decisions. The University of Virginia says it doesn't use a formula, equation, or algorithm during its review. A Brown admissions officer told the student paper there is no single quality in every successful admit.
What is important is that of the colleges we cover — most of the selective private schools and many of the top publics — more than half rate essays “very important”, and about half say the same of character and personal qualities. By comparison, only about a quarter give test scores that weight. What can be known — which is what this tool has been built to do — is how your materials read against what a school says it weighs, and that's specific and actionable in a way a probability is not.
At the same time, there are considerations in admissions decision-making that are unknowable to outsiders: things like institutional priorities, and the quality of the overall applicant pool in a given year. This tool, like any other, can't account for these factors — which is yet another reason “chance me” tools are about as reliable as Magic 8-Balls.
An essay review looks at one document. This reads your whole file — personal statement, supplemental essays or UC personal insight responses, activities list, honors and awards, transcript, class rank and test scores where they count, plus what you expect your recommenders to say — against a chosen school's priorities, factor by factor. It's explicit about what came from the school and what our reading of it is, and it says what the read cannot see as well as what it can. There are full sample reads for Swarthmore, MIT and UC Santa Barbara if you'd rather see the output than read about it.
They're a good indication of one thing: whether your grades and scores fell in the range a school admitted in a previous cycle. They're a poor guide to what it wants more broadly. A profile reflects the outcome of thousands of decisions, not the reasons for any of them, and it blends in every priority a school had that year, including ones it never publishes. Middle-50% ranges mean a quarter of admitted students scored below the bottom of the band. At test-optional schools, score ranges come only from students who chose to submit, which tends to push them up. And a profile can't tell you what separated two applicants with similar numbers, which at selective schools is where much of the decision happens. Duke's dean of admissions said in a 2024 interview that the school very rarely distinguishes between applicants on academic grounds. What separates applicants is better found in what each school says it weighs, in its own words, which is what this tool reads your file against.
See how your file reads, school by school About 3 to 4 minutes to run. A full sample read is linked at the top.