College Applications · Inside the Read
Selective colleges do not read your application and form an impression. They score it. Court filings from the Harvard admissions case exposed the actual machinery: six numbered ratings per applicant, on a scale where 1 is best and 6 is worst. Most files get a first read in under ten minutes, by an officer who owns your geographic territory and has roughly 800 other files to get through.
The clearest public window into a selective admissions office came from litigation. Documents filed in the Students for Fair Admissions case against Harvard described a rating system that had never been laid out publicly before, and the structure is worth understanding even if you are not applying anywhere close to Harvard, because it shows what a trained reader is actually doing with a file.
A first reader assigns a number in each of six categories. The scale runs from 1, the strongest, to 6, the weakest, and readers can append a plus or minus to separate candidates inside a band.
| Category | What a "1" means |
|---|---|
| Academic | Near-perfect grades and standardized test scores |
| Extracurricular | Truly unusual achievement, not simply long participation |
| Athletic | Recruitable talent at the varsity level |
| School support | Counselor and teacher letters that are outstanding, not merely positive |
| Personal | Maturity, integrity, leadership, kindness, courage |
| Overall | The reader's summary judgment, informed by the other five |
Two figures from those filings are worth committing to memory. Applicants with an overall rating of 1 were always admitted, and applicants at 3-minus or below were almost always rejected. And although fewer than 20% of applicants received a personal rating of 1 or 2, that group accounted for roughly 80% of the admitted class.
Read that second statistic again, because it undercuts the most common assumption about elite admissions. The personal rating, not the academic one, was doing much of the sorting. Grades and scores get you into the conversation. Something less measurable decides how it ends.
Ratings detail as reported from court filings by The Harvard Crimson. Harvard's system is Harvard's, not a universal standard, but the shape of it recurs across selective offices.
Applications are not distributed at random. Files are sorted into queues by geographic territory, so one officer reads everything from your region and reads it against the other schools in that region. An officer may carry around 800 files in a season.
Territory assignment matters more than most families realize. Your student is not being compared against every applicant in the country. The comparison that counts is against other students from the same area, and often from the same school.
Since the University of Pennsylvania pioneered the approach in the 2013-14 cycle, many selective colleges have replaced the solo first read with committee-based evaluation. Two officers sit together, each with a screen. One works the academic material, transcript and testing and course rigor. The other works everything else, essays, activities, recommendations. They talk as they read and reach a recommendation together, typically in four to ten minutes.
Four minutes sounds brutal until you consider what it rewards. A reader working that fast needs the record to be legible immediately: clean transcript, obvious course progression, a grading scale they do not have to decode. Anything that forces them to stop and work something out is time taken from the parts of your application you actually want them reading.
Committee-based evaluation background via Inside Higher Ed.
Beyond the numbered ratings, offices attach shorthand as a file moves. Four abbreviations come up often enough to be worth knowing: IC (intellectual curiosity), LTE (likelihood to enroll), DI (demonstrated interest), and HSI (high school index). Exact codes vary between institutions and most are never published, but the categories they capture are consistent.
Whether the record shows a student who pursues subjects past the requirement. Course choices in the final two years say more here than any adjective in an essay.
Whether you have shown real signals of wanting to attend, through visits, sessions, or engagement. Colleges that track it use it to protect their yield. See how to show demonstrated interest.
A prediction of whether you would actually come if admitted. Nobody wants to spend an admit on a student who was always going elsewhere, which is part of why early rounds behave the way they do.
Context on the school itself: what courses were available, how previous applicants from there performed, what its grading is worth. This is the piece most homeschool families never think about, because there is no institutional history for a reader to draw on.
Recruited athlete, legacy, first generation, and similar flags that route a file down a different path with different comparisons.
None of these is a secret score you can game. Each is a reader trying to answer one question quickly: given where this student came from and what was available to them, how much did they make of it?
Two of these categories land differently for a homeschool applicant, and both are fixable.
School support is a scored category, and your counselor is a parent. A reader who sees a glowing letter from the applicant's own mother cannot weight it the way they would weight a letter from an independent counselor. That is not prejudice, it is the same discount anyone would apply. The fix is at least one substantial letter from someone outside the family: a co-op instructor, dual-enrollment professor, employer, or coach. See recommendation letters for homeschoolers.
The school index has nothing to draw on. No course catalog, no history of past applicants, no institutional grading record. Your transcript has to supply that context itself, which is why the grading-scale key and clear course titles carry more weight for a homeschooler than for anyone else, and why class rank is usually reported as none.
Most of this system is not addressable. You cannot make a reader spend twelve minutes instead of five, and you cannot manufacture a personal rating. Four things are genuinely within reach.
For what colleges themselves report weighting, our analysis of 601 colleges' Common Data Sets ranks all nineteen admission factors, and the Common Data Set guide shows how to look up any single college's own weighting.
At selective colleges, readers score applicants in defined categories rather than forming a general impression. Harvard's system, exposed in court filings, rates six: academic, extracurricular, athletic, school support, personal, and overall, on a scale of 1 (best) to 6 (worst). Fewer than 20% of applicants earned a personal rating of 1 or 2, but that group made up about 80% of admitted students.
Files are sorted by geographic territory, so one officer reads all applicants from your region and compares them against each other. Many selective colleges now use committee-based evaluation, where two officers read side by side, one covering academics and one covering everything else, and reach a recommendation in four to ten minutes. An officer may handle roughly 800 files in a season.
Under committee-based evaluation, a first read and recommendation typically takes four to ten minutes. That is why a legible transcript matters: anything a reader has to stop and decode consumes time that would otherwise go to the rest of the application.
IC stands for intellectual curiosity, LTE for likelihood to enroll, DI for demonstrated interest, and HSI for high school index. Each is internal shorthand an admissions office may attach to a file: whether the student pursues subjects past the requirement, whether they would actually attend if admitted, whether they have signaled real interest in the college, and what context the applicant's school provides. Exact codes vary between institutions and are rarely published.
School support is a separately scored category covering the strength of counselor and teacher recommendations. A rating of 1 requires letters that are outstanding rather than merely positive. Homeschool applicants should expect a parent-written counselor letter to carry less weight, which is why at least one letter from an independent recommender is worth arranging.
Many do. Demonstrated interest and likelihood to enroll are used to predict whether an admitted student would actually attend, which protects the college's yield. Not every college tracks it, and a college's Common Data Set will state whether it considers interest at all.
The formal six-category rating system belongs to highly selective offices. The underlying behavior is broader: files sorted by territory, read quickly, and judged against the applicant's own school context. A clear transcript and outside corroboration help at any level of selectivity.
Automatic weighted and unweighted GPA, a grading-scale key readers do not have to decode, and courses grouped by year. The math and the on-screen build are free to start, with no credit card.
Start Free