How the document reads your list

What you give it.

Four things, and nothing else. There is no account, so nothing is remembered between visits and no profile is built.

• Your list — between 1 and 15 US universities, in the order you enter them. It is the list you intend to apply to, not a wishlist we score you against.
• One test score — an SAT total or an ACT composite. Nothing is converted between the two. Your SAT is read only against the SAT range a school published and your ACT only against its ACT range, so a school that published a range for the other test is reported as unmeasurable rather than estimated across.
• Your family income band — optional. It selects which of the school's published net-price rows is read. Skipping it costs you the cost figures and nothing else.
• The round — which of the rounds below the document is for. Early Decision II additionally asks what happened in round one, because a school that already denied you is not a school that can still say yes.

It does not ask for, read or rate your GPA, your essays, your activities, your awards, your recommendations or your transcript. Nothing in this product evaluates writing or achievement.

What it reads about each school on your list.

Per school, five things — each one a figure that school published, carrying the source and the cycle it was filed for:

• The early-round split. How many applicants the school admitted in its binding Early Decision round, against how many it admitted from the rest of the same applicant pool. Both counts come from the school's own filing; the division is done once, in one function, and the result is shown as a multiple.
• The overall admit rate. The share of all applicants the school admitted in the most recent cycle held for it.
• Where your score sits inside the 25th-to-75th percentile range of the students that school admitted — at or above the 75th, inside the middle 50%, or below the 25th. That is a position in a published range, not a score we compute for you.
• What the school cost families in your income band.
• Its application deadline for the round, as the school itself entered it with the Common App.

Where the figures come from.

• IPEDS (Integrated Postsecondary Education Data System) — federal data from the U.S. Department of Education.
• College Scorecard — federal data from the U.S. Department of Education.
• Common Data Set — the standardized survey universities publish about themselves, which is where the early-round counts come from. This is the colleges' own reporting, so we never label it federal.
• Early-round press releases published by the schools, for early-round figures a Common Data Set does not carry.
• The Common App first-year requirements grid, for deadlines — each member school's own dates as it entered them.

We cite the source for every school-level number. IPEDS and College Scorecard are attributed as federal; the Common Data Set is attributed to the colleges directly. Coverage is not uniform — schools publish on different schedules, and a figure that is current for one school may be a year older for another. Every acceptance rate, SAT range and cost figure on a school page shows the academic year it describes; where we could not establish the year, no year is shown. Where a school published nothing, the document prints the absence rather than filling it in.

The chance that every school says no.

One figure per list, and the simplest arithmetic that could produce it: multiply each school's published rejection rate. A school that published no admit rate is counted out and reported as counted out, never assumed to be average.

It is computed twice — over the list as you entered it, and over the list plus the schools the document offers to add — so the pair shows what the change would do.

This multiplies each school’s published admit rate as if the decisions were independent. They are not — the same transcript is read by every office — so treat it as a look at the shape of the list, not a forecast about you.

The rounds, and the call in each.

Three rounds, each on sale in its own window on the Eastern Time clock, each scored on its own date. Every date below is a rule, not a calendar we retype each August:

• Early Decision answer — on sale Sep 1 – Nov 15, scored Dec 15.
• Regular Decision list check — on sale Nov 2 – Jan 15, scored Apr 1.
• Early Decision II call — on sale Dec 12 – Jan 15, scored Feb 15.

Which school gets the binding Early Decision application.

A school on your list is a candidate only when all of this holds, from its own filing:

• it runs a binding Early Decision round;
• it filed the early-round counts, so the split can be measured at all;
• your score is not below the 25th percentile it admitted — a school that admitted nobody at your score is not where a binding application goes;
• its measured early-round advantage is at least 1x. A school whose published figures show its early applicants admitted at a lower rate than everyone else is not a candidate, and the document says so with the number rather than hiding the school.

Candidates are then ordered: first by whether your score is at or above the school's 75th percentile rather than inside the middle 50%, then by the size of the published advantage, then by what the school cost families in your income band. Ties break on a fixed key, because a ranking that can return two answers for one input cannot be graded later.

This compares two rates the school published for two groups of applicants. The Early Decision pool is not a random slice of the applicants — recruited athletes, legacy applicants and QuestBridge matches are counted inside it — so this is a difference between the school’s two pools, not a multiplier on your own chances.

The Regular and Early Decision estimates.

The paid Early Decision document prints two estimates for every school on your list, and says how each was made:

• Regular starts from the school's admit rate outside Early Decision — everyone who did not apply early: its admits less its Early Decision admits, over its applicants less its Early Decision applicants, from the same filing as its early-round figures — and is multiplied by one of three factors, chosen by where your SAT or ACT sits against the range that school admitted. The table is below. A school that filed no Early Decision counts starts from its published admit rate instead, the same figure the chance that every school says no is computed from, above.
• Early Decision starts from the Early Decision admit rate the school filed — its Early Decision admits divided by its Early Decision applicants — and is multiplied by the same factor. It is not multiplied by the school's early-round advantage.
• Neither estimate is printed above 90%, and neither is printed above the rate the school reported: a reader inside or above the admitted range gets that rate as it is.
• A school that runs Early Decision and publishes no early-round counts gets a Regular estimate and the words “ED offered, counts not published”. No early figure is estimated for it.

Because both estimates take one factor, the Early Decision estimate divided by the Regular one is the school's filed Early Decision rate divided by its admit rate for everyone else — exactly the early-round advantage printed on the school's card. Nothing multiplies by that advantage; the two estimates stand in that ratio because of where each one starts.

The document then orders the schools that filed an early-round breakdown by the gap between the two estimates, in percentage points. That order can differ from the call above, which ranks by your position in each school's admitted range and then by the advantage; where it does, the document names the rule that put them apart, directly under the call.

The three factors, and what they rest on.

What is known. A published 25th–75th percentile range is a statement about the students it describes: a quarter of them scored below the 25th percentile, half inside the range and a quarter at or above the 75th. That column is arithmetic, not a measurement.

What is assumed. How a school's applicants split across the same three positions. No school publishes that, and we hold no source for it. We assume applicant pools are wider at the bottom than the classes admitted from them, and set the split to the figures in the third column. This is a judgement of ours and the estimate is only as good as it is.

The factor is the first share divided by the second, and never above ×1.00. If a position holds a quarter of the admitted and 35% of the applicants, applicants in it were admitted at ×0.71 the school's overall rate. Inside the range and at or above the 75th, the same division gives more than ×1.00; we print ×1.00 there instead, because the reported rate counts every admit, including those admitted for reasons a test score can't show. So the three estimates are deliberately conservative: weighted by where applicants sit, they average to less than the rate the school published, not back to it.

The same table is applied to the early pool. That is a second assumption: that a school's Early Decision applicants split across its admitted range the way all its applicants do. Nothing published says so. The rate a school filed for its early pool also includes recruited athletes and legacy applicants, and the estimate does not take them out.

A school that publishes no admitted range for the test you took gets no factor: its published rate is printed as it is, and the row says so. Nothing is converted between the SAT and the ACT.

These are estimates. Regular starts from the school’s admit rate outside Early Decision — everyone who did not apply early, from the same filing — or from its published admit rate where the school filed no Early Decision counts. Early Decision starts from the Early Decision admit rate the school filed. Both take one factor for where your test score sits against the range that school admitted: ×0.71 below its 25th percentile; inside the range or at or above its 75th, no adjustment — we never print an estimate above the rate the school reported, because that rate counts every admit, including those admitted for reasons a test score can't show. The ×0.71 rests on our assumption about who applies (35% of applicants below the range, 45% inside it, 20% above) — no school publishes that split, and we apply the same split to the early pool. The rate a school filed for its early pool includes recruited athletes and legacy applicants; the estimate does not take them out. No estimate is printed above 90%. Because the factor is the same on both, the two estimates stand in the ratio the school’s own filing shows — its Early Decision admit rate over the admit rate of everyone else, the multiple printed on its card. Your grades, essays and activities are not inputs to either figure — they move your own result, and no published number says by how much.

The two schools it offers to add.

This is the one place a school you did not name enters the document, and the selection rule is not a figure source: U.S. News Best Colleges decides which schools are even looked at. Its national-universities and liberal-arts-colleges ladders are separate and are never mixed — a college ranked 30th among liberal arts colleges is not the 30th national university, and reading one column as though it were one ladder is how a top-25 national list was once offered three liberal arts colleges. So: the list you entered decides which ladder is searched, by majority, and every band printed names the ladder and the edition it was measured against.

On that ladder, a candidate must satisfy all of:

• ranked below your own lowest-ranked school, within the next 40 places and never past rank 80;
• your score at or above its published 75th percentile less 20 points, on your own test. A school that published no range for your test is not rejected for it — the check is skipped and the omission is printed, so the reason line drops that clause instead of inventing one;
• an overall admit rate between 15% and 50%. Below that it is a long shot wearing the word “safety”; above it, adding it barely moves the figure above;
• it runs a binding Early Decision round. A school whose own policy says it does not is excluded: there is no early application to spend there.

Those are ordered by the published early-round advantage first (from 1.3x up), then by merit awards, then by rank. At most two are offered, and a slot is left empty rather than filled with a school that failed a check.

Which schools come off a Regular Decision list.

A school is beyond your band only when both hold: your own score is below the 25th percentile that school admitted on the same test, and it admitted fewer than 25% of its applicants in the latest cycle held for it. A school that admits half its pool is not out of reach because a score is low.

At most two of those are named as ones to take off, never leaving fewer than three schools on the list, and only when taking them off does not RAISE the chance that every school says no. Taking any school with a published rate off a list raises that figure; the additions usually pay for it and sometimes do not, so the document names the largest number for which it does not rise — and names none when no number satisfies that. The rest keep the finding on their own row.

Early Decision II.

The same call, over a smaller set: the schools on your list that hold an Early Decision II deadline for this season which has not passed, minus the school that answered in round one. A school that deferred you stays on the list and out of that set — it will read the application again in its Regular Decision pool, and you cannot apply to it twice. A school that denied you leaves both, and the document says so in a line rather than dropping it silently. Where a school files one combined figure for both of its Early Decision rounds, that is what is shown, labelled as combined. Nothing is split, scaled or estimated into an Early Decision II-only rate.

What it deliberately does not do.

• No estimate from anything but your test score. The Regular and Early Decision estimates read one thing of yours: where your SAT or ACT sits against the range each school admitted. Nothing in the document turns your grades, essays or activities into a number. The chance that every school says no belongs to the list and is computed from published rates alone, with its caveat printed beside it every time.
• No multiplying by a school's early-round advantage. That advantage is the ratio of two pools the school reported. No estimate is multiplied by it: the Regular estimate starts from the rate of the pool outside Early Decision, the Early Decision estimate from the rate the school filed for its early pool, and both take the same factor.
• No reading of your writing or your record. Nothing here reads an essay, rates an activity or an award, or reads a recommendation.
• No person. There is no consultant, no counsellor and no advising call attached to any purchase.
• No figure nobody published. An absence is printed as an absence.

And what published data cannot tell anyone.

• Test-optional distortion. When a school does not require test scores, only the students who chose to submit appear in its published band. Those students typically have stronger scores, so the range often reads higher than the full admitted class.
• Nothing about the rest of your application. Essays, recommendations, demonstrated interest, institutional priorities, and recruited-athlete or legacy status move real decisions and appear in no public dataset.
• A year of lag. Published figures describe a cycle that has already closed. The class you are applying to has not been admitted yet.

Recorded prediction, and the grading.

Every placement in a document is written to an append-only record when the list is scored. That record refuses edits and deletions, so the call cannot be quietly revised once decisions are out, and each document reports how many of its own schools are in it rather than assuming all of them are.

Each round is scored on its own date — the table above — and the first public report is the grading of this season's Early Decision calls in December, including the ones we got wrong. What gets graded is the narrow claim we actually make, never a chance of admission. What is sealed, and when it is scored.