Transparency

How We Score Colleges

The UCD Score turns official federal data into a single number you can trust. This page shows exactly how it is built, what it measures, and what it cannot tell you.

Last updated June 2026

What We Stand For

Choosing a college is one of the most expensive decisions a family makes. The numbers behind it should be straight. These four commitments shape everything on this site.

Federal Data Only

Every statistic comes from public U.S. government datasets. Nothing is invented or estimated.

Scores Are Not for Sale

No college can buy a higher UCD Score, change its data, or move up a ranking earned on the merits. The numbers are reported exactly as the government publishes them.

Fair Peer Comparison

Colleges are ranked against genuine peers, not forced onto one national list that ignores their differences.

Open Methodology

This page exists so anyone can see how a score is produced. Nothing about the formula is hidden.

What the UCD Score Is

The UCD Score is a number from 0 to 100 that ranks a college against the colleges most like it. It is a percentile, not a fixed grade. A score of 80 means a college performs better than roughly four out of five of its peers on the measures that matter most: whether students finish, what they earn, and what it costs to get there.

Because the score is relative to a peer group, it is most useful for comparing similar colleges. Read it alongside the raw numbers on each college profile rather than as a standalone verdict.

The UCD Score is our own measure. We created it, and we calculate it from public federal data using the formula on this page. It is not an official government rating, and no government agency produces or endorses it.

Peer Groups

It would not be fair to rank a two-year community college against a research university. They serve different students with different goals. So every college is placed in one of three peer groups and scored only against the others in its group.

Peer groupWho is in it
A. Four-year selectiveFour-year colleges that report an acceptance rate below 100%.
B. Four-year open or onlineFour-year colleges that admit nearly all applicants or do not report an acceptance rate.
C. Two-yearCommunity and technical colleges offering two-year programs.

The Four Sub-Scores

Each college is measured on up to four things. Every sub-score is itself a percentile within the peer group, on a 0 to 100 scale.

Outcomes

Do students finish, and do they stay? Combines the six-year graduation rate with the first-year retention rate.

Value

What do graduates earn relative to what the college costs? Compares median earnings ten years after entry against average net price.

Affordability

Can lower-income students afford to attend? Combines net price (lower is better) with the share of students receiving Pell Grants.

Selectivity

How competitive is admission? Combines the acceptance rate with typical SAT and ACT scores. Applied to four-year selective colleges only.

How the Score Is Calculated

The four sub-scores are combined into one overall score using weights set for each peer group. Selectivity applies only to four-year selective colleges, so the other groups spread that weight across the remaining measures.

Peer group Outcomes Value Affordability Selectivity
A. Four-year selective35%30%20%15%
B. Four-year open or online45%35%20%0%
C. Two-year45%30%25%0%

Keeping the Scale Honest

A raw percentile would mark half of all colleges as "below 50," which misrepresents how the underlying data really performs. To keep the scale readable, the result is adjusted so a typical college lands near 65, strong performers reach the 80s and 90s, and the weakest still score around 30 rather than 0.

Small Colleges

Colleges with fewer than 500 students are capped at a score of 80. With very few students, a single graduating class can swing the percentile math sharply, so the cap prevents tiny institutions from topping a list on thin data.

Missing Data

When a college has not reported one of the inputs, that measure is set aside and the remaining weights are rescaled. A college is never scored as zero for data it simply never filed.

Which Colleges Receive a Score

Every college that reports enough data to its federal sources receives a UCD Score. A small number of institutions report almost no usable data. Rather than guess at a score for them, we show their profile without one. A missing score means missing data, not a judgement about the college.

Score Labels

Every UCD Score also carries a plain-language label, so the number is easy to read at a glance. Each label covers a band of the 0 to 100 scale and, like the score itself, reflects how a college compares with its peers.

UCD ScoreLabel
85 to 100Excellent
70 to 84Strong
55 to 69Good
40 to 54Fair
Below 40Average

Admissions Difficulty Labels

Separate from the UCD Score, each college carries a label describing how hard it is to get in, based on its acceptance rate. Colleges that admit nearly everyone, including most community colleges, are shown as Open admission.

Acceptance rateLabel
Under 10%Highly Selective
10% to 25%Very Selective
25% to 50%Selective
50% to 75%Moderate
75% to 90%Easy
90% or moreOpen

What the Score Cannot Tell You

A single number is a useful starting point, not the whole picture. Here is what the UCD Score deliberately does not capture.

  • It does not measure campus culture, social life, or whether a college will feel right for you.
  • It does not rate individual programs. A college that is strong overall may be average in your specific major, and the reverse is also true.
  • Federal data runs roughly one to two years behind. A college that has recently changed will not show that change yet.
  • Earnings figures cover students who received federal financial aid, which is most but not all of a graduating class.
  • A high score is a strong signal, not a promise. Use it to build a shortlist, then read the full profile and visit if you can.

About Our Editorial Content

The underlying figures on this site, such as earnings, net price, and graduation rates, come directly from the federal datasets behind them. Composite measures like the UCD Score are calculated by us from that same data, using the formula on this page.

The written summaries beside those numbers, such as the paragraphs describing a college or a major, are written by us or composed directly from the underlying federal data.

Numbers are never generated or estimated by AI. Reported figures are shown exactly as the federal source publishes them, and calculated figures follow the published formula. The wording explains the data; it never replaces it.

How Often We Update

Federal education and labor datasets are refreshed on a regular cycle, generally once a year, though the exact timing varies by source. After each release we rebuild the site from the new files, so every page, score, and label moves to the latest published figures together. Until the next release, the site reflects the most recent data available.

Methodology Questions

What is a good UCD Score?
A score of 70 or higher (Strong or Excellent) places a college in the upper range of its peer group. Because every score is a ranking against genuine peers, compare a college to others in the same peer group rather than across groups. A 75 at a community college and a 75 at a selective university both mean "well above the typical college of that kind."
Can a college pay for a better UCD Score?
No. The UCD Score is calculated entirely from public federal data by a fixed formula, with no advertising, sponsorship, or manual adjustment for any college. No one can pay to change a score, the underlying data, or where a college ranks on the merits.
Why do two colleges with similar statistics have different scores?
The UCD Score is a percentile ranking within a peer group, not a fixed grade. A college is measured against the other colleges most like it. Two schools with similar raw numbers can land in different peer groups, or sit at different points relative to their own peers, which produces different scores.
Why is the score curved instead of shown as a raw percentile?
A raw percentile would label half of all colleges as "below 50," which misreads how the underlying data actually performs. The curve keeps the 0 to 100 scale honest and readable: a typical college lands near 65, strong performers reach the 80s and 90s, and the weakest still score around 30 rather than 0.
How often is the data updated?
Federal education and labor datasets are released once a year. After each release we rebuild the site from the new files so every page, score, and label moves to the latest published figures together. Until the next release, the site reflects the most recent data available.

Still have a question about how we score colleges? Contact us and we will walk you through it.

Next See where the data comes from Every dataset behind the UCD Score, named, dated, and linked to its official source. Data Sources →