Home / How it works
The Unified Model

We don’t just give you a score.
We show you the math.

Admit Path recreates the admissions process, then uses evidence rather than anecdotes to coach you through it. Here is exactly how it works, because we believe you deserve to see how every conclusion is reached.

Layer 1

Theory-based model

What schools say they want. We study each program’s stated admissions practices and encode what matters and how much.

  • Published criteria and Common Data Sets
  • Academic research and admissions office publications
  • Interviews with current and former admissions professionals
  • Fit frameworks such as Holland Codes (RIASEC)
Layer 2

Practice-based model

What schools actually do. We analyze real outcomes to find which factors truly correlate with admission, and by how much.

  • Anonymized outcomes shared by applicants, with consent
  • Verified documents for higher-confidence data
  • Public datasets such as IPEDS
  • Integrity checks: no duplicates, reliability thresholds
Synthesis

The Unified Model

Where theory and practice meet. A factor joins the model only when the evidence supports it. Where they diverge, you get a strategic insight.

  • Statistical testing against holdout data
  • Updated as committees and policies change
  • Human scores feed back into the model
  • Every output explained with a “Why?”
Finding the gap

When what schools say and what they do diverge, that’s your opportunity.

Suppose a program’s published materials suggest that a certain factor, say demonstrated interest, matters only a little. If the practice-based evidence shows it is strongly associated with admission at that school, we flag it and advise you accordingly.

It works the other way too. A factor that schools mention often but that shows little effect in real outcomes gets less weight in your plan. You spend your limited time where it counts.

How a factor earns its place

  1. Theory proposesResearch suggests a factor matters, for example legacy status at a particular college.
  2. Practice testsWe check the outcome data: all else being equal, does that factor actually change the likelihood of admission?
  3. The model decidesIf the evidence confirms it, it becomes part of the Unified Model. If not, it doesn’t.
  4. We keep it currentWhen an admissions director changes or a school announces new policies, we update.
The process

Discovery, analysis, committee, report.

We collect information the way an admissions committee does, and we review it the way they do: with humans acting as an independent committee, supported by statistics.

1

Discovery

Three interactive questionnaires: a basic profile common to all applicants, a supplementary one for your program type (like LSAT or MCAT details), and an application questionnaire modeled on the official form.

2

Analysis

Your profile is compared with the practice-based profiles of admitted students and weighted using the Unified Model.

3

Committee review

For Application Review, independent trained readers score essays, recommendations and mock interviews using committee-style rubrics. Their scores feed the model.

4

Report

You get clear results, the reasoning behind them and a prioritized plan, plus a recommendation on whether and when to apply.

What you receive

Results you can act on

Admit Preparation results

A high-level view early in the process, when there’s time to act.

Admission probability, on a seven-level scale

Lower extreme
Well below average
Below average
Average
Above average
Well above average
Upper extreme

Culture match

Good matchReasonable matchPoor match

Recommendation

Apply nowApply laterReconsider

Application Review results

A specific, comparative metric once your application is drafted.

Probability as a multiple of the average applicant

5.2× more likely than the average applicant to be admitted
(illustrative example)

A multiple compares you with the real applicant pool. It avoids false precision while still telling you something meaningful.

Recommendation

Apply nowApply with changesApply laterReconsider

Committee feedback

Reader scores on criteria like intellectual vitality, personal character and narrative cohesion, with specific fixes before you submit.

Glass box, not black box

Every insight comes with a “Why?”

Algorithms you can’t question aren’t worth trusting. Whenever the Unified Model produces a result, we show the factors behind it, so you can see what is driving it and act.

  • Explained results: each score shows the evidence that shaped it.
  • Human-reviewed badge: you always know when independent readers have verified a recommendation.
  • Honest limits: we never present a single percentage as certainty.
Culture match: Good match
?Why? Your interest profile is strongly Investigative (Holland Code), and this university has a high share of research-active undergraduates.
?What would change it? Your fit would be even stronger with evidence of independent inquiry, such as a summer research project or a science fair entry.
Reviewed by independent committee
Where we focus

Statistical modeling where the data is deep

Each of these programs admits hundreds of students every year, which gives us enough evidence to be confident. We don’t model outcomes where the data is too thin to be reliable. For high school admissions and lifelong learning, our guidance combines published criteria with expert review, and we always say which approach applies.

College

Selective undergraduate admissions, including Early Decision and Early Action strategy.

MBA

Work experience quality, leadership trajectory, GMAT/GRE and round strategy.

Law (JD)

LSAT and GPA positioning, plus scholarship versus ranking trade-offs.

Medicine (MD)

AMCAS, clinical and research experience, secondaries and interviews.

How we use data

Consent first. Never scraped, never sold.

Our practice-based model is built through a give-to-get approach: applicants who use Admit Path can choose to contribute their anonymized results, including acceptances, rejections and waitlists, so that the next applicant gets better guidance. Contributors unlock deeper insights in return.

Because data from the people we serve is what makes the model work, protecting it is non-negotiable.

Your consent, your choice

You decide whether to contribute outcome data. Nothing is added to the model without permission.

Anonymized before modeling

Personally identifying information is removed before any data enters the aggregate model.

You stay in control

Grant or revoke advisor access at any time, and ask us to delete your data whenever you like.

Never sold to advertisers

We use data only to improve the model and generate insights. We do not reveal any client relationship without your consent.

See where you stand.

Start with a free readiness assessment. No obligation, and no tax returns needed for sliding-scale pricing.

Start free assessment