Bay Area School-District Homes: What Actually Separates a GreatSchools 10 From a 9?

The Direct Answer

GreatSchools' 1-10 is a within-state relative rank, not a proficiency rate. A 10 is the capped top-decile band, which makes 10 versus 9 one tick on a scale rather than one step in school quality. Four Peninsula and South Bay high schools all rated 10 ranged from 68.80% to 92.70% on 2025 CAASPP grade-11 math — a 23.9-point spread.

Who this article is for

If any of the following describes you, this article was written for you:

• You are touring the Peninsula or the South Bay — particularly at the $5M+ tier in Atherton, Palo Alto, or Los Altos Hills — and you have been glancing at the 1-10 school score on a Zillow or Redfin listing page, then ruling a neighborhood in or out on the strength of it;
• You have noticed that GreatSchools, the California School Dashboard, Niche, and U.S. News reach different conclusions about the same school, and you cannot tell which one deserves the weight;
• A school you have tracked for a year just moved from 10 to 9 (or the other way), and you want to know whether the school changed or the formula did;
• You are planning a cross-state or cross-city move around a long education path, and you want to upgrade from reading a score to reading the data before you decide which district the budget lands in.

One note on scope. What follows takes apart how the rating itself is computed. It stays inside school and assessment data: it does not rank schools or communities against one another, and it draws no comparison beyond academic measures.

Three core decision dimensions

To read a 1-10 correctly, you only have to ask three questions. What is it measured against, how is it weighted together, and what does it compress out. Answer those three and the difference between a 10 and a 9 stops being mysterious.

Dimension one: what the score is measured against — a within-state rank, not an absolute proficiency rate

GreatSchools states it plainly in its own methodology. The rating is a 1-10 measure of how well a school supports student success relative to other schools in the same state: 1-4 is below average, 5-6 is average, 7-10 is above average. The conversion for each component is relative in the same way. A school's statewide performance is first expressed as a percentile, then mapped into a decile band to produce the 1-10 — the top 10% statewide earns a 10, the next 10% earns a 9, and so on down.

That single design choice settles two things. First, a 10 does not mean proficiency near 100%; it means the school falls in the state's highest band. Second, a 10 is a ceiling. Every school sitting in that top decile receives the same component 10 no matter where inside the band it actually lands, and the summary rating — a weighted average of those components, rounded to a whole number — flattens the differences further. The line between 10 and 9 is a tick mark on a percentile scale, not a step change in school quality.

Dimension two: how the score is weighted — two passes, and renormalization when a component is missing

The GreatSchools summary rating is not a single measure. It is a weighted average of at most three themed ratings: the Student Progress Rating (state growth data; the Academic Progress Rating substitutes when growth data is unavailable), the Test Score Rating (performance on state standardized assessments), and the College Readiness Rating (high schools only — built from SAT/ACT results, the four-year graduation rate, and participation in advanced coursework such as AP and IB, with California additionally folding in the share of graduates meeting UC/CSU entrance requirements).

The weighting runs in two passes. The first is the base weight: whichever of Student/Academic Progress and Test Score is higher receives 0.45, and the remaining components receive 0.27 each. In other words, within a single school, the stronger of those two measures is handed the larger starting weight. The second pass is the information weight, set by how much data exists (how many grade, subject, and measure combinations) and how well that data discriminates among schools in that state, with a cap ensuring no component contributes more than the Student Progress Rating. The two passes are added, renormalized across whatever components are available, then summed and rounded to a whole number.

The most consequential sentence in the methodology is a single line: if any themed rating is missing, the remaining weights are rebalanced to sum to one. That is the main mechanism behind a school's rating moving between years. The state reports one more category of data, or one fewer; a component appears where there was none, or disappears where there was one; and the entire weight structure is rewritten. The school itself can be unchanged while the number moves. GreatSchools says as much directly — components update as new data arrives, and the summary rating can change with them.

Dimension three: what the score compresses out — the absolute proficiency rate sits underneath

Decile mapping plus a capped top band means the real differences among the highest-performing schools get flattened. That is the honest answer to "what separates a 10 from a 9": rather than comparing the two digits, pull the two schools' absolute CAASPP met-or-exceeded percentages and set them side by side. Those are the raw signal, before anything was made relative.

Growth measures deserve the same early look. GreatSchools uses the Student Progress Rating to capture year-over-year improvement, while the California School Dashboard colors a school by combining current-year performance (Status) with the change from last year (Change). The absolute proficiency rate tells you where this cohort stands right now. The growth measure tells you how far the school moved them. Read together, they are a complete picture; read alone, neither is.

The numbers, laid out

Four 10-rated high schools and one 8 for contrast: same digit, different proficiency

The core numbers first: Lynbrook High, Henry M. Gunn High, Los Altos High, and Palo Alto High all carried a GreatSchools summary rating of 10 as of September 2026, and all four carried a 10 on both published components — Test Score Rating and College Readiness Rating. Yet on 2025 CAASPP grade-11 data, their math met-or-exceeded rates run from 68.80% to 92.70%, a spread of 23.9 percentage points, while English language arts runs from 75.78% to 91.99%, a spread of 16.2 points. One identical digit, four materially different levels of performance underneath it.

The counterintuitive point worth keeping: Gunn and Paly sit inside the same district, Palo Alto Unified. Same summary rating of 10, same 10 on both components — and an 8.6-point gap in ELA met-or-exceeded, a 6.6-point gap in math. A composite rating can bury a difference that exists inside a single district. The 8 in the table makes the same argument from the other direction: Menlo-Atherton's Test Score Rating is 9 and its College Readiness Rating is 6, and the weighted blend of those two produces an 8. A lower summary rating can mean one component pulled the average down, not that the whole school sits a tier lower — and conversely, a 10 can simply mean two components are both pinned against the ceiling. The clearest evidence of that ceiling effect is sitting in the same table: four high schools share a Test Score Rating of 10 while their CAASPP math met-or-exceeded rates remain 23.9 points apart. The component is capped; the absolute gap is not. The only way to know what is happening behind a summary rating is to open the components, then return to the absolute percentages.

One denominator problem is easy to overlook. A proficiency rate is a ratio, and ratios are only comparable when their denominators line up. When you compare absolute percentages across schools, hold the grade, the subject, and the school year constant, and record the Students Tested count from the same row of the CDE public data. Two schools with visibly different tested populations deserve extra caution before their rates are read against each other.

Four platforms, four denominators

The core difference first: when one Peninsula high school produces four different conclusions on four platforms, it is not because somebody computed wrong. It is because the four are not asking the same question. GreatSchools asks where the school ranks within its state. The California School Dashboard asks how far the school sits from standard and whether it improved on last year. U.S. News asks where the school ranks nationally on six weighted measures. Niche blends government data with user reviews.

What to remember: the California School Dashboard has improvement built into its color by construction — current-year performance combined with the change from last year, resolved on a 25-cell grid — so a school whose absolute level is not the highest but which keeps improving can land a better color than intuition expects. U.S. News, meanwhile, puts 30% of its weight on college readiness, so a school with high AP or IB participation will place noticeably better there than a pure state-assessment proficiency read would suggest. The four scores are not in conflict. They are four answers to four different questions. If you have already learned to hold third-party valuations at arm's length — we took that layer apart in Can You Price a Bay Area Home Off Zillow or Redfin? — apply the same skepticism to third-party ratings: ask what is being measured before you read the number.

Data source: GreatSchools Ratings Methodology Report and the GreatSchools school profile pages for the five schools cited; California Department of Education (CDE) CAASPP public data, 2025 administration (2024-25 school year, grade 11, all students, Smarter Balanced ELA and mathematics met-or-exceeded percentages and Students Tested counts); CDE California School Dashboard academic indicator and 5x5 color grid documentation; U.S. News 2026-2027 Best High Schools methodology; Niche Best Public High Schools methodology
Updated: 2026-09 (GreatSchools summary and themed ratings retrieved 2026-09-02; CAASPP figures are the 2025 administration — as of the retrieval date the 2026 administration had not been released)
Scope: Peninsula and South Bay public high schools; private schools excluded, as GreatSchools rates public schools only

What MK Group sees on the ground

Families who genuinely decide around education are never comparing one score against another. MK Group worked with a dual-income couple who both did AI research at a large Seattle employer, relocating across states with an eight-year-old daughter. Their employer ran teams of equivalent scale in both places, so the job did not change and neither did the pay. What moved them was the long education path — the depth of the school ecosystem, and the entire University of California system their daughter might eventually reach. They landed in Palo Alto because four hard requirements held at once: both parents still working, a public-school floor they could rely on, walking or cycling distance to Stanford, and no need for a large lot. What the team did through that process was not hand them a ratings table. It was explaining attendance-area boundaries among Palo Alto Unified's elementary schools, how those feed into the middle schools, and how Palo Alto High and Gunn High differ — so they could see the real distinctions inside a single city. (For how those two high school assignments are actually drawn, see How Do I Know Whether a Palo Alto Address Feeds Gunn or Paly?)

The harder lesson runs the other way: however high the rating, confirm first which school the house actually feeds. MK Group once walked a family through a home in Atherton Oaks, in the center of town. Atherton as a whole spans three elementary districts — Las Lomitas, Menlo Park City, and Redwood City — with assignment decided street by street, and the 94027 zip code proving nothing on its own. The house the client first fell for sat on the wrong side of the school boundary, and comparable homes on the two sides of that line carried a transaction gap of roughly $1.5 million. After the diligence, the client redirected in time and locked onto the other side of the boundary. On that center-of-town boundary purchase, the first move by Marie Wang (DRE# 02110980) and Kevin Mo (DRE# 02127623) was to verify the attendance area by exact street address; the rating discussion came after that step, not before. (The full verification method is laid out in How Do You Verify Which District a House Actually Belongs To?)

One more path worth flagging: if private school is on the table for your family, this trail of evidence ends abruptly. GreatSchools' methodology states it rates public schools only, because the ratings depend on publicly reported state education data. Comparing private against public requires a different frame — see Private vs Public School in Silicon Valley.

Common mistakes

Mistake one: "a 10 school must be a full tier better than a 9 school"

It does not hold. Each component is mapped from a statewide percentile into a decile band: a 10 corresponds to the highest band, a 9 to the one below it, and between them sits a single tick that may represent one or two percentile points. Run the logic the other way and the gaps inside the top band can be large: among the four 10-rated high schools checked here, 2025 CAASPP grade-11 math met-or-exceeded runs from 68.80% to 92.70%, a 23.9-point spread. To compare two schools properly, place their absolute percentages side by side rather than comparing two integers that have already been made relative and then capped.

Mistake two: "the rating dropped from 10 to 9, so the school got worse this year"

Not necessarily. The summary rating is a two-pass weighted average of at most three themed ratings: the base weight gives 0.45 to the stronger of the two academic components and 0.27 to the rest, an information weight reflecting data volume and discriminating power is layered on, and the whole thing is renormalized across whatever components are available. So when the state reports one more or one fewer category of data, or a component appears or disappears, the entire weight structure is rewritten. The school can be unchanged while the number moves. When you see a rating change, open the components first — find out which one moved, and whether a component was newly added or newly absent — before drawing a conclusion.

Mistake three: "several platforms rank it highly, so the conclusion is confirmed"

Not quite, because they are answering different questions. GreatSchools asks about within-state rank. The California School Dashboard asks about distance from standard plus the change from last year. U.S. News builds a national ranking with 30% of the weight on college readiness. Niche blends government data with user reviews. Even on a single GreatSchools page, the 1-10 school rating and the 1-5 star user reviews are two parallel displays on entirely different scales — do not read them as corroborating each other. When several high scores line up, first confirm they are answers to the same question.

Mistake four: "the school is rated 10, so this house feeds that school"

This is the most expensive mistake in the set. A school rating describes a school; it says nothing about a house. On the Peninsula, two sides of the same city — even two sides of the same neighborhood street — can fall into different districts, and the zip code does not settle assignment: Atherton alone spans three elementary districts, decided street by street. No amount of precision in a rating substitutes for verifying the attendance area by exact street address. The order is always the same: confirm assignment first, discuss ratings second.

Next steps

1. Pull the target school's absolute CAASPP met-or-exceeded percentages: search by school name on the California Department of Education's CAASPP public data platform, record the grade, subject, school year, and Students Tested count, and hold all four constant whenever you compare across schools.
2. Check the California School Dashboard color as a second pass: it combines current-year performance with the change from last year, which supplies the direction of travel that an absolute percentage cannot show.
3. Open the GreatSchools components instead of stopping at the summary rating: note the Test Score Rating, the College Readiness Rating, and the Student Progress Rating (where present) separately, and you will see immediately which one is carrying or dragging the composite.
4. Before you write an offer, verify the attendance area by exact street address: confirm it with the district registration office or the city's official GIS tool, and do not let a zip code, a neighborhood name, or the school tag on a listing page stand in for that step.
5. Put ratings last in the decision chain: settle the education path (public, private, or a hybrid) and the list of hard requirements first, then read absolute data and growth trends, and only then use ratings as a fast filter. Run that order backwards and the boundary is where the most expensive mistake gets made.