Running simulation...
A Simulation in Seven Acts

One Thousand Decisions

Every spring, millions of high-school seniors learn where they'll spend the next four years. We simulated 1,000 of them — from 20 elite feeder schools to 55 colleges — to see what the system looks like from above.

1,000 Students
55 Colleges
6 Admission Rounds
Applications
Scroll to begin
Chapter I

Meet the Class

Before a single application is filed, the class already has a shape. Our 1,000 students hail from 20 feeder schools spanning seven states — elite boarding schools in New England, competitive magnets in New York, prep schools in California and Florida.

No two applicants are alike. Each carries a behavioral archetype — the physics olympiad champion, the literary magazine editor, the recruited lacrosse midfielder — and a structural position shaped by family wealth, parental education, and access to opportunity.

Average SAT across all 1,000 students. The national average is 1060. This is a feeder-school population.

The applications are in. First up: the binding early round.

Chapter II

Early Decision: The First Commitment

In November, hundreds of students make a binding promise: accept me early, and I will attend. No shopping around, no comparing offers. It's the biggest gamble in the process — trade all your optionality for a statistical edge.

The advantage is real. Across our 55 colleges, ED acceptance rates are consistently higher than Regular Decision — sometimes by a factor of 3-5x. But the commitment is binding: accepted ED students withdraw all other applications.

Students committed after Early Decision

But most students are still in the game. The non-binding round is next.

Chapter III

The Waiting Game

Early Action is the non-binding cousin — learn your fate in December, but keep your options open until May. At Harvard, Yale, Princeton, and Stanford, Restrictive Early Action lets the strongest applicants signal serious interest without locking in.

But for many, the early round ends not with an answer but a deferral — pushed to the RD pile to be reconsidered alongside tens of thousands more.

Students deferred from early rounds — still waiting

The deferred join the regular pool. March will settle most fates.

Chapter IV

Regular Decision: The Main Event

March arrives, and with it, the deluge. The vast majority of applications — and the vast majority of rejections — land in Regular Decision. Every seat not locked down by ED is now contested by the full national pool.

At HYPSM schools, RD acceptance rates drop below 3%. Even at Selective schools, the rates compress as the applicant pool swells far beyond available seats.

Total admission decisions rendered across all rounds

Envelopes are open. Now it's the students' turn to choose.

Chapter V

Decision Day

For the first time all year, the power shifts. Students — not colleges — hold the cards. Those with multiple acceptances must choose one and reject the rest. Yield — the share who actually show up — is now the metric that keeps deans awake at night.

Cross-admit battles are fierce: when a student holds offers from HYPSM and Ivy+, the higher-tier school almost always wins. Financial aid, archetype fit, and personal preference create the exceptions.

Students with zero acceptances after all rounds

Not everyone is settled. The waitlist still has cards to deal.

Chapter VI

The Waitlist

Welcome to admissions purgatory. The waitlist exists so colleges can hedge their bets — if yield falls short, they pull names off the list. But most students languishing here will never get that call.

Summer melt adds a final twist: students who committed fail to show up — visa issues, financial gaps, family changes. More waitlist spots open. The cycle continues until the class is finalized.

Students lost to summer melt after committing
Chapter VII

By the Numbers

How well does the simulation match reality? We compare our simulated acceptance rates against published Common Data Set figures for all 55 colleges.

Points near the diagonal line indicate calibration accuracy. The simulation uses a logistic admission model with Cholesky-correlated GPA/SAT, archetype-based EC quality, hook multipliers in logit space, and phantom applicants to account for the 98.7% of the national pool we don't model.

of students ended up unplaced — shut out of every modeled college
Student Composition
SAT vs GPA colored by archetype, with distribution breakdowns
Early Decision Results by Tier
Applications, acceptances, and fill rates
Early Action Outcomes
Accepted, deferred, and rejected by college tier
Acceptance Rates by Round
Early vs Regular Decision rates across tiers
Where Students Chose to Enroll
Enrollment by tier, showing yield patterns
Where 1,000 Students Ended Up
Each square is one student, colored by enrollment tier
Simulated vs Real Acceptance Rates
Each dot is a college; diagonal = perfect calibration