You check each school one at a time. Stanford, 3.6%. Low, but people get in. Columbia, 4.0%. Low, but people get in. Penn, 5.4%. Cornell, 8.8%. Every number is small, and every number feels survivable, because you are reading them one at a time and each one leaves a door open.
Then you hit submit on fifteen of them and wait.
Here is the question almost nobody asks before that moment: what are the odds that every single one says no?
The number
We looked at the college lists real students actually built on this site — 71 students, 657 school selections. The distribution is not subtle:
| Selectivity | Share of picks |
|---|---|
| Under 15% admit rate | 68% |
| 15-40% | 17% |
| 40-70% | 11% |
| Over 70% | 5% |
Two out of three picks go to schools that reject more than 85% of applicants. The single most-selected school was Stanford.
So take the fifteen most-picked schools on this site and multiply the rejection odds together:
Stanford (3.6%) · Columbia (4.0%) · Penn (5.4%) · Harvard (3.7%) · Princeton (4.6%) · Brown (5.4%) · Cornell (8.8%) · Yale (3.9%) · Berkeley (11.0%) · UCLA (9.0%) · Duke (5.7%) · MIT (4.6%) · Chicago (4.5%) · Michigan (15.6%) · Dartmouth (5.4%)
Probability all fifteen reject you: 37%.
Not 37% that it goes badly. 37% that in late March, after fifteen applications, fifteen essays, and a stack of fees, there is nowhere to go.
Trim it to the top twelve — drop Berkeley, UCLA, Michigan — and it climbs to 49%. A coin flip.
Why this is invisible
Nobody hides this number. It just never gets computed, for a specific reason: acceptance rates are published per school, and lists are experienced per school.
You research Stanford on Tuesday and Cornell on Thursday. Each session ends with the same feeling — hard, but possible. There is no page anywhere that says "given these fifteen together, here is your floor." College sites have no reason to run it. Ranking sites are organized one school per page. Even most counselors talk school by school, because that is how the conversation naturally goes.
The arithmetic only appears when someone multiplies. And multiplication is exactly what a list is.
The honest caveat, because it cuts against us
That 37% assumes each decision is independent — that Yale flipping its coin tells you nothing about how Princeton flips its own.
That assumption is wrong, and it is wrong in the direction that makes things worse, not better.
The same transcript, the same test scores, and the same essays go to all fifteen schools. Highly selective colleges weigh broadly similar things. If your application sits just under the bar at one of them, it probably sits just under the bar at several. Rejections cluster. So do admits.
Which means 37% is a floor, not an estimate. The real number for a list this top-heavy is higher, and we cannot tell you how much higher without data no college publishes.
We would rather say that plainly than hand you a precise-looking number we cannot defend.
What actually moves it
Here is the part worth sitting with. The lever is not effort. It is not another essay draft, not a fourth revision of your activities list, and not one more reach added at the bottom of the page.
Adding one school that admits around half its applicants takes 37% to 18%.
Adding a second at around 70% takes it to 5.6%.
Two additions. The odds of a completely empty spring fall by a factor of six. No test retake does that. No essay does that.
And notice what this is not saying. It is not "lower your sights." All fifteen reaches stay on the list. Nothing gets removed. The shape of the list changes, and the shape is what was doing the damage.
The uncomfortable corollary: if your list is fifteen schools and twelve are under 10%, several of those twelve are not doing separate work. The eleventh sub-10% school adds very little the tenth did not already add. It costs another fee and another supplement, and it moves your floor by about a percentage point. Those are the slots worth trading.
The question to bring to your list this week
Not "is this school realistic?" — you have already asked that fifteen times and gotten fifteen reasonable-sounding answers.
Ask instead: "if I multiply these together, what is my floor, and which two changes move it most?"
That is a different question, and it has an actual answer.
---
Where these numbers come from
Acceptance rates are institution-reported figures published through the U.S. Department of Education (IPEDS / College Scorecard), most recent available year. Each school page on this site links its own source.
The list-composition figures (71 students, 657 selections) are aggregated from lists built on PrepToDone and reported only in aggregate.
The probability calculation is straightforward multiplication of per-school rejection rates under an independence assumption, stated and qualified above. It describes the shape of a list. It is not a prediction about you.