How the score works
No single ranking agrees with the others — each optimizes for a different mix of reputation surveys, citation counts and research output. Analects Proverbs doesn't pick a winner between them; it normalizes and combines them, using a default weighting based on each ranking's real-world recognition. You can adjust those weights yourself on the university rankings and high school rankings pages.
University ranking sources
QS World University Rankings
22% default weightWeighs academic & employer reputation surveys heavily alongside faculty/student ratio and internationalization.
Publisher: Quacquarelli Symonds · 2024 edition
Times Higher Education World University Rankings
22% default weightBalances teaching, research environment, citations, industry income and international outlook.
Publisher: Times Higher Education · 2024 edition
US News Best Colleges (National Universities)
13% default weightThe ranking most consulted by US high schoolers and parents: undergraduate experience, admit rate, class size, graduation rate. US institutions only.
Publisher: U.S. News & World Report · 2024 edition
US News Best Global Universities
13% default weightEmphasizes global & regional research reputation, citations and international collaboration.
Publisher: U.S. News & World Report · 2023–24 edition
Academic Ranking of World Universities (Shanghai Ranking)
10% default weightResearch-output focused: Nobel laureates/Fields medalists, highly cited researchers, and papers in Nature & Science.
Publisher: ShanghaiRanking Consultancy · 2023 edition
Center for World University Rankings
9% default weightPurely outcomes-based — no reputation surveys. Ranks on education quality, alumni employment, research and faculty citations.
Publisher: CWUR · 2024 edition
Niche Best Colleges
7% default weightStudent-voice driven: campus life, academics, value and diversity, built from reviews by actual students and alumni. US institutions only.
Publisher: Niche · 2024 edition
CWTS Leiden Ranking
4% default weightPurely bibliometric — no surveys, no reputation input. Ranks strictly by citation impact and scientific collaboration.
Publisher: Centre for Science and Technology Studies, Leiden University · 2024 edition
High school ranking sources
A separate, smaller composite for our best high schools list — same normalize-and-weight formula below, independent from the university score.
US News Best High Schools
40% default weightRanks US public high schools on college readiness, math/reading proficiency, and graduation rates. Public schools only.
Publisher: U.S. News & World Report · 2024 edition
Niche Best Schools (K-12)
25% default weightStudent and parent reviews plus academic data, covering both public and private high schools separately.
Publisher: Niche · 2024 edition
Washington Post Challenge Index
20% default weightRunning since 1998: ranks by AP/IB/Cambridge exam participation relative to graduating class size, regardless of scores. Rewards access, not just top scorers.
Publisher: The Washington Post · 2024 edition
Newsweek America's Best High Schools
15% default weightCombines graduation rates, college readiness metrics and academic performance data across US public high schools.
Publisher: Newsweek · 2024 edition
The formula
- 1. Normalize. Each source's rank is converted to a 0–100 score relative to that source's total published pool, where rank 1 scores 100 and the bottom of the pool scores 0. Beyond roughly the top few hundred, publishers themselves report bands (e.g. "601–650") rather than exact positions — we score those using the band's midpoint, not an approximation of our own.
- 2. Weight and combine. A school's aggregate score is the weighted average of its normalized scores from every source it appears in, using the weights above by default. Sources it isn't ranked in are simply excluded, not counted as zero — and its weight is redistributed proportionally among the sources that do rank it.
- 3. Re-rank. All schools are sorted by aggregate score to produce the Analects Proverbs rank. Two schools with the same displayed score (rounded to one decimal) share the same rank — the next distinct score picks the rank back up where a strict count would have left off, the same convention published rankings use for ties.
Both datasets are curated snapshots of each publisher's most recent rankings, refreshed manually — around 467 universities and 159 high schools, with exact positions where we're confident and published bands beyond that. They're intended as illustrative aggregates, not official rankings. The default weights reflect general public recognition of each ranking, not an endorsement of its rigor — which is exactly why they're adjustable. High schools also carry a "Subject strengths" panel on their profile pages that's explicitly an editorial estimate derived from each school's specialty, not a named subject ranking.
Admissions data
University profile pages also show acceptance rate, test-score ranges (or the local equivalent — UK A-levels, Australian ATAR, and similar, for countries where SAT/ACT isn't the admissions currency), average GPA, applications received per year, and a link to the school's own admissions page. This is separate from the ranking sources above — it isn't aggregated or weighted, just the most recently published figures for each field. Coverage is intentionally partial: it's only shown for universities we have genuine confidence in a real, current figure for, never a guess dressed up as data. Where we don't have it, the page says so plainly rather than showing an invented number.
Campus & outcomes data
The campus photo and overview paragraph on each profile page are fetched live from Wikipedia, the same way logos are fetched live from Google's favicon service — never manually chosen or hosted by us, so it's always a real image and real text, and simply doesn't appear for a school without a usable Wikipedia photo. Class size, graduation rate, tuition and early-career salary follow the same partial-but-never-fabricated policy as admissions data above, and are US-specific concepts, so international universities only show student-faculty ratio and enrollment.
Standardized testing policy
Some university profiles carry a testing-policy label — SAT/ACT required, test-optional, test-flexible, or test-free (scores not considered at all) — for the fall 2026 entry cycle, and the same labels appear on the score-check page. These policies change year to year, and a stale "test-optional" tag on a school that has since reinstated testing would be actively misleading, so the label only appears where the policy is well documented and stable. Most universities carry no label at all; that means we aren't confident, not that a policy doesn't exist. Always confirm on the university's own admissions page before deciding whether to submit a score.
Earnings after graduation
Where salary figures exist, a profile shows median early-career and mid-career earnings, the growth between them, and a "cost payback" figure — how many years of early-career median earnings equal four years of net price. The growth and payback numbers are arithmetic on figures already shown, not separately sourced statistics.
Four limits matter more than the numbers themselves. These are medians, so half of graduates earn less, and the spread inside one university is far wider than the gap between most universities. They are institution-wide across all majors, so much of the difference between two schools reflects what their students study rather than how well they teach — an engineering-heavy institute out-earns a liberal arts college largely on major mix. Geography and selectivity are baked in, since graduates often work near where they studied and selective intakes would earn well anywhere. And they are US-reported figures, so universities outside the US show nothing here rather than a number that doesn't translate. We deliberately publish no per-subject earnings: field-of-study earnings exist in the US federal College Scorecard for federal-aid recipients, aren't in this dataset, and splitting an institution-wide median across majors would be invention.
Rankings by subject
The subject pages draw on real, published QS World University Rankings by Subject positions, shown as bands. QS builds those tables from academic reputation and employer reputation surveys plus research citations per paper and h-index, weighted differently by discipline — citation metrics count for more in medicine and the sciences, the reputation surveys dominate in law and business. Two consequences are worth holding onto: a subject table rewards research and reputation, which is not the same thing as undergraduate teaching quality, and it ranks institutions within a broad field, not individual degree programmes. Coverage is narrow by design — a university missing from a subject page is outside the tier we're confident about, not "unranked". This is entirely separate from the "Subject strengths" radar on profile pages, which is our own editorial estimate and is labeled as such wherever it appears; the two are never mixed.
International study costs
The study-abroad page, and the cost card on non-US university profiles, describe what an international student pays in each destination country. These are deliberately country-level figures, not per-university ones. National tuition policy — whether public universities charge non-nationals at all, and who is exempt — is stable and verifiable. Per-school, per-programme international fee schedules change every year and vary by faculty; publishing hundreds of them would produce exactly the kind of quietly-stale data this site exists to avoid.
Figures are bands, given as approximate US dollar conversions so that countries can be compared against one another at all — the university itself will bill in local currency, and the amount it bills is the one that counts. Bands describe a typical bachelor's degree at a public university; private institutions, medicine and dentistry routinely cost far more. Living-cost ranges are national typicals, and capital cities run well above them. Visa fees, health insurance, flights and proof-of-funds deposits are excluded. Because tuition policy and post-study work rights are set by governments and do change — Norway began charging non-EEA students in 2023 after decades of free tuition — treat all of it as a starting point for comparison and confirm against the university and the destination country's official immigration source before committing.
Graduate & professional programmes
The MBA, law and medicine pages rank programmes, and are a separate data layer from everything above. This distinction is the whole point of those pages: business, law and medical schools admit separately from the university, are funded separately, and are ranked by different publications on different criteria, so a university's overall position on this site tells you very little about its professional schools. They are also distinct from the QS-by-subject bands — a strong QS Medicine position is an institutional research-and- reputation signal, not an "MD programme rank".
Positions are bands reflecting where a programme consistently sits across the recognised rankings for its field — US News Best Business Schools and the Financial Times Global MBA Ranking for business, US News Best Law Schools and Best Medical Schools (Research) for the US tier, QS and THE for the international tier — rather than one publication's exact number in one year. Coverage is the well-documented top tier only; a programme's absence means it sits outside the tier we're confident about, never that it is unranked or weak. Each page leads with the structural facts that decide whether a programme is open to you at all — work-experience expectations for an MBA, the jurisdiction limits on legal qualification, and the graduate-versus-undergraduate entry split and international-admission caps in medicine — because those matter before any ranking does.
Value score
Where a university has both net price and mid-career salary data, its profile shows a "value score": mid-career salary divided by four years of net price. A score of 2.0 means one year of mid-career earnings alone would cover twice the full 4-year cost of attending. This isn't a separate published statistic — it's simple arithmetic on the two real figures already shown elsewhere on the page, meant as a rough, transparent shorthand for return on investment, not a precision instrument. It ignores financing costs, opportunity cost, and everything about a degree that isn't salary.
Ranking history
University profile pages show a real, growing history of this dataset's own rank and score for that school — one snapshot recorded every time the data is refreshed, starting from whenever that university's history section first appears. We don't have verified year-by-year historical ranking data to backfill, and won't invent a trend line, so a school's history starts empty and builds up genuinely over time rather than showing a fabricated past.