Changelog

Scoring v2.1 improves how we handle the ~33,000 schools that lack direct state test data — mostly private schools and specialized public programs.

• Tiered proxy academic scores: Schools without test data receive an estimated academic score based on directly-measured schools in their area. City-level proxies (from schools in the same city) carry 80% of normal academic weight. State-level proxies (statewide average) carry 65%. Every school page clearly indicates when a proxy score is used.
• Holdout validation: We validated these discount levels by holding out 10% of directly-measured schools, imputing their scores via proxy, and comparing to actuals. City proxies predicted within 1.5 points on average (10-point scale), state proxies within 1.7 points.
• Private school minimum components: Raised from 2 to 3. Private schools with only 2 thin components (representing ~45% of weight space) are now shown as “Not Yet Rated” instead of receiving misleading scores.
• Equity display change: Equity metrics (diversity index, demographics, FRL%) are now shown as informational context only — no longer displayed as a numeric score. Our poverty-adjusted academic score already accounts for socioeconomic context.

Full details at /methodology.

Rankings on MySchoolScout are now bylined to Ryan Kallander, the person actually running the ranking pipeline, with Person-type JSON-LD structured data on every page. We also added a methodology sidebar that links directly to the full scoring breakdown, so any parent (or any crawler) can trace a score back to the exact formula and data sources that produced it.

This is the E-E-A-T — experience, expertise, authority, trust — layer that distinguishes primary-source school data from the many sites that just resell someone else’s numbers.

School, city, and district pages now display a visible “last reviewed” date so parents can see exactly how fresh the underlying data is. We also emit Schema.org Review structured data for each school rating, which tells Google and LLM crawlers that our scores are editorialized ratings tied to a specific methodology — not scraped numbers.

Why it matters: families comparing schools deserve to know whether the rating they’re reading reflects this year’s data or last year’s. The freshness badge makes that explicit on every page.

Publishers, real-estate sites, and community blogs can now drop a single iframe into their pages and show the top-rated K-12 schools for any U.S. ZIP code, powered by our rankings. The widget pulls live from our rankings database, respects our ranking weights, and links back to the full school profile on MySchoolScout.

For parents reading our pages, this means the data we publish is the same data showing up wherever the widget is embedded — no stale mirrors, no inconsistent scores.

Twenty of our highest-traffic city guides — Houston, Dallas, Austin, Phoenix, and more — were rewritten with deeper local context: neighborhood breakdowns, district-level commentary, and direct links to the top-ranked schools and tracts in each metro. Each guide now passes a TF-IDF content-quality gate before publishing, so every article earns its place in the index rather than being boilerplate.

Parents searching “best schools in [city]” should hit a page that actually answers the question instead of a thin listicle.

The map was originally desktop-first. This update brings it to parity on phones: a new mobile toolbar puts the layer controls and legend inside a bottom sheet, the map index page shows thumbnail previews of each metro, and the homepage now features a promo card linking into the map so first-time visitors can discover it.

Phones are where most parents actually research schools, so shipping these was non-negotiable.

We self-hosted our display and body fonts (DM Sans and Fraunces) to remove a render-blocking external CSS fetch, and fixed a cumulative layout shift on map containers that was hurting our Core Web Vitals scores. Pages load faster, feel snappier, and stop reflowing as the map initializes.

Google uses Web Vitals as a ranking signal, so this pays off in both user experience and search visibility.

High school profiles now display AP course availability and IB program status, pulled from the U.S. Department of Education’s Civil Rights Data Collection (CRDC) for the 2020-21 school year. This is the most comprehensive federal source for advanced-course access, and it lets us show parents not just whether a school offers AP, but how many students are actually enrolled in those courses.

For families weighing college-prep options, this fills in a gap that NCES alone doesn’t cover.

The Neighborhood Decision Engine is a new interactive map that overlays school quality against housing data — home values, rent, homeownership rate — at the census-tract level for every major U.S. metro. Click a tract to see the specific schools serving it, jump to the full school profiles, and compare adjacent neighborhoods side by side.

This was the founding vision of MySchoolScout: a map that lets a parent answer “where should I actually live?” instead of “how good is this one school?” It’s the most ambitious thing we’ve shipped so far, and we’ll keep iterating on it.

Private school coverage is now complete: we imported 22,345 private schools from the NCES Private School Survey (PSS) and supplemented missing schools from state-level sources (e.g., the California Department of Education’s private school directory). Every private school now has a dedicated profile page with enrollment, grade range, religious affiliation where available, and a private-school-specific ranking model that doesn’t rely on public-school test data.

Private schools are a real option for many families, and they deserve the same transparency we give public schools.

Scoring v2 is a significant upgrade to the ranking algorithm. The biggest changes:

• Level-specific weights: elementary, middle, and high schools are now scored with formulas tuned to what actually matters at each level (e.g., college-readiness for high schools, growth for elementary).
• Poverty adjustment: we account for the correlation between student poverty and raw test scores so schools serving higher-poverty populations aren’t unfairly penalized for factors outside their control.
• College readiness: a new component incorporating AP/IB access, graduation rates, and SAT/ACT participation for high schools.

The full methodology is published at /methodology — including the exact weights, what each component measures, and the data vintages used.