Almost every college score in circulation makes the same quiet choice: it ranks all schools on one list. That sounds neutral until you notice what it does to a community college, which gets measured against research universities it shares almost nothing with and sinks accordingly. The UCD Score starts from a different premise. It sorts all 3,839 colleges into three peer groups and scores each school only against its own kind, so a two-year college competes with two-year colleges and an open-admission public competes with its equals. The payoff shows up in one number. Across 3,837 scored colleges, the median score lands within half a point of 65 in every one of the three peer groups, which is exactly what a fair score should do.
Does Every Type of College Get a Fair Shake
Yes, and the medians prove it. When each peer group is scored against itself, the center of each group lands in the same place: a median of 64.8 for selective four-year schools, 65.2 for open-admission four-year schools, and 65.4 for two-year colleges. A spread of under one point across three very different kinds of institution is not an accident. It is the signature of a score that refuses to rank a welding program against a medical school.
How the Scores Spread Across the Three Groups
Each peer group is built from a distinct slice of the data, then scored on a curve that places its median near 65. The three groups are not equal in size, but they are equal in where their typical school lands. The table below shows how the 3,837 scored colleges divide, with the average and median score for each group.
| Peer group | What it holds | Colleges | Average | Median |
|---|---|---|---|---|
| A | Four-year, selective | 1,636 | 64.9 | 64.8 |
| B | Four-year, open / online | 794 | 63.8 | 65.2 |
| C | Two-year | 1,407 | 64.9 | 65.4 |
| All | Every scored college | 3,837 | 64.7 | 65.1 |
The median lands at 65 no matter the type of school
Median UCD Score by peer group, across 3,837 scored colleges
The four bars are nearly the same length, which is the point. A score that let the selective four-year group sit far above the two-year group would just be selectivity wearing a new name. Here the groups are level at the center, and the differences between schools live inside each group, where the comparison is fair.
What the Label Bands Actually Look Like
The five labels are stacked, not evenly split. The score is curved so the median sits near 65, which pushes most colleges into the two middle bands and leaves the top and bottom thin. Good and Strong together hold 80.2 percent of all scored colleges. Excellent and Average, the two extremes, hold barely 3 percent between them.
That shape is what makes the label useful. Because Excellent covers only 79 schools, about 2 percent, the band means something when a college earns it. And because the bands are identical for every peer group, an Excellent two-year college and an Excellent selective university have each cleared the same bar of 85, each against its own peers. The top two-year college scores 89.7 and the top selective university scores 96.5, and both wear the Excellent label honestly, because neither was forced to beat the other to get it.
How We Measured This
The UCD Score is computed from federal data, College Scorecard and IPEDS, on four sub-scores: outcomes, which combines graduation and retention rates; value, the ratio of 10-year earnings to net price; affordability, low net price paired with a high share of Pell students; and selectivity, low acceptance rate and high test scores, which applies to four-year schools only. Each college is percentile-ranked against its peer group on each sub-score, the sub-scores are blended on group-specific weights, and the result is curved so the median lands near 65. Two of the 3,839 colleges lack enough data to score and are dropped, leaving 3,837. The full method, the weights, and the source vintages are on the methodology and data sources pages.
What the Numbers Do Not Say
A median that lines up across peer groups means the score is fair across types of school. It does not mean every type of school is interchangeable. A two-year college scoring 80 and a selective university scoring 80 have each beaten their own peers by the same margin, but they still offer different credentials, timelines, and paths, and the score does not collapse that difference. The peer-group design also means a number is only meaningful next to its group: a 70 in the two-year group is not the same achievement as a 70 in the selective four-year group, because the competition differs. And the four sub-scores reflect who enrolls as much as what a school teaches, the same caveat that shadows every outcomes figure built on earnings and graduation data.
What This Means for Students
Read a college's score next to its peer group, never on its own. A 72 means a school sits above most of its true competitors, and that is the comparison that should shape a list, not a raw rank that mixes community colleges with research universities. Because the median is 65 everywhere, anything in the Strong band is genuinely above the middle for its type, which is why an open-admission school can clear that bar the same way a selective one does, a pattern spelled out in why the score does not just reward selectivity. Put two or three schools you are weighing side by side in the Compare tool and read the sub-scores, not only the headline number, since two schools at 72 can get there in very different ways.
What This Means for Parents
Treat the label as a peer-relative grade, not a national rank. A community college labeled Strong has outperformed most two-year schools in the country, which is real information even though it will never top a list that starts with the Ivy League. The fair-scoring design is what lets a school like Victor Valley College earn an Excellent that a single national ranking would hide entirely, and the same logic surfaces the schools that punch above their weight on outcomes their price tag would not predict. Before reacting to any school's score, use the Match Quiz to see which type of school fits the student first, then read each candidate's score against its own group, where the number was built to be fair.
