Honest answers
Accurate enough to shape your list. Not accurate enough to judge your application. They are good at sorting schools into reach, target, and likely. They are unreliable about your personal odds, and least reliable at the selective schools where students check them most. Here is the honest breakdown, from a company that builds admissions AI and refuses to sell it as a predictor.
01 · Credit first
This page is not a hit piece. The banding job is real, and some tools do it well for free.
Sorting a long list into bands
Given your GPA and scores against a school’s published ranges, a calculator can usually tell a likely school from a target from a reach. That sorting is genuinely useful when you are staring at forty schools and need a shape for the list.
Catching obvious mismatches
If a school’s enrolled students almost all sit above your numbers, a calculator will say so, and you should listen. The reverse read is useful too. Schools where your record sits comfortably above the middle are real options, not consolation prizes.
Being free and fast
Some good chancing calculators cost nothing. CollegeVine’s is free and reasonable at the banding job. We compared ourselves to them honestly, including what they do well.
Trajecta vs CollegeVine →02 · The limits
Four structural problems. None of them are fixed by a better model, because none of them are modeling problems.
01
You apply once. If the calculator said 40% and you got in, was it right? There is no rerun of the year to find out. Accuracy for one applicant is close to unfalsifiable, which is worth remembering whenever a tool advertises how accurate it is. Averages across thousands of applicants say little about the one file you care about, which is yours.
02
Calculators run on numbers and checkboxes because that is what fits in a form. Essays, recommendations, context, and whether the whole file tells one story or three are absent. Colleges say those things matter, in the admission factors they file publicly every year, and for selective schools they are often the deciding layer.
03
At a school admitting most applicants, a calculator is fine, and also barely needed. At a school admitting one in ten, the honest answer for nearly every qualified applicant is “unlikely, and it depends on things no form captures.” A precise-looking percentage at a reach school is confidence theater. The tools are weakest exactly where students lean on them hardest.
04
Many calculators tune themselves on outcomes users report. Students who get in report more than students who do not, admits are more fun to share, and profiles arrive unverified. None of that makes the tools useless. It does mean the training signal underneath the percentage is softer than the decimal point implies.
The deeper question, whether any AI can predict admissions at all, gets its own page: can AI predict college admissions?
03 · A different shape of tool
Our conclusion after building admission math ourselves: the honest move is not a better percentage. It is a different output.
Trajecta’s Committee Review reads your finished application the way a committee would, instead of scoring it the way a form would. A specialist essay reader files a report first. Four officers then read the entire file through different lenses, academics, narrative, impact, and context, and react to each other in writing. When they disagree, and they genuinely do, the split is recorded, and a chair resolves it into a decision with next steps.
Every read is anchored to the school’s own Common Data Set numbers, the same public data behind our free college pages. The output is not a percentage. It is the argument your file starts, which is the thing you can actually fix before you submit. And because we know how this class of tool gets oversold, we published a complete run on a real application so you can judge it before creating an account.
04 · FAQ
At sorting schools into reach, target, and likely bands, reasonably good, and that is a real service. At telling you your personal odds, unreliable, and least reliable at selective schools where most applicants are qualified and most get denied. No calculator sees your essays, your recommendations, this year’s applicant pool, or the college’s internal enrollment needs.
There is no trustworthy public ranking, because individual-level accuracy is nearly impossible to verify from outside. Free options like CollegeVine’s chancing calculator do the banding job fine. Our advice is to pick by what the tool is honest about, not by the accuracy it claims. A tool that gives you a band and says why is being straight with you. A tool that gives you one confident percentage at a reach school is not.
Treat it as one signal about your numbers, nothing more. When a school admits a small fraction of applicants, strong numbers make you competitive, not safe. Committees deny files with excellent numbers every cycle for reasons that live in the parts of the application calculators cannot read.
Two things, and we are explicit about what neither one is. The college list scores fit and admission probability from each school’s published data, shown as bands. Committee Review then simulates the committee itself on your finished application: an essay specialist files a report, four officers read the whole file and disagree in writing, and a chair resolves the split into a decision with next steps, anchored to the school’s Common Data Set. It is calibrated against real, publicly shared admissions outcomes, and it is a rehearsal of the conversation, not a prediction of the outcome.
It answers a different question, so accuracy is not the comparison. A calculator estimates where your numbers sit. The simulation shows what a committee argument about your whole file could sound like, including what the readers push back on. One is a thermometer, the other is a rehearsal. We built the rehearsal because it is the one you can act on before you submit.
Sort your list with real data, then rehearse the committee conversation on the application you actually wrote.