What a school proficiency percentile actually measures
Every state writes its own test and picks its own pass mark. That single fact decides what a school score can honestly be compared with — and it is the reason this site never ranks a district against one in another state.
There is no national school exam in the United States. Each state writes its own tests, sets its own pass mark, and reports the share of students who clear it. A district where 80% of pupils are "proficient" in one state and a district with the same 80% in another have not been measured against the same thing — the pass marks are not the same height, and they move when a state revises its standards.
This has a hard consequence for any site that scores places, including this one. A single national school score, built by putting those percentages side by side, is confidently wrong. It would rank a state that sets a gentle pass mark above a state that sets a demanding one, and no amount of arithmetic afterwards fixes it.
So we rank inside the state, and only inside it
The schools figure here is a percentile among districts in the same state. A district at the 90th percentile did better than 90% of districts in its own state, measured on the test its own state set. That comparison is sound, because everything in it sat the same exam.
It also means the number travels badly in one specific way: the 90th percentile in one state and the 90th in another are both "near the top locally" and are not a claim that the two are equally good. When you compare suburbs in different metros on this site, treat the school score as "how it stands among its neighbours", which is usually the question you are asking anyway.
A percentile answers "compared with what?" honestly. A national grade pretends the question does not exist.
Districts, not schools
The data is published by district, and a district is not a school. Two things follow.
A large district contains schools that differ from each other by more than districts differ from each other. A strong district average is not a promise about the particular school at the end of a particular street.
And a town is not always one district. Plenty of suburbs are split between two or three, sometimes with one district covering most of the town by area and another covering a corner of it. Where that happens, this site names the districts and the share of the town each covers, and weights the score accordingly — but the address you buy decides which one you get, and the boundary does not follow the town line.
What the other two numbers mean
Graduation rate is the four-year adjusted cohort rate: of the students who started ninth grade together, the share who finished with a diploma in four years. It is a genuine outcome measure and a slow one — the cohort it describes started school four years before the figure was published.
Student–teacher ratio is enrolment divided by teaching staff. It is not class size. A district with a ratio of 14 does not have classes of 14; it has specialists, part-timers and small-group teachers in the same divisor. The ratio is still worth reading as a rough measure of staffing, which is what it is.
What none of it measures
Test scores track the intake of a school at least as strongly as they track its teaching. A district serving families with high incomes and high levels of education will post high proficiency whatever happens in its classrooms, and a district serving poorer families will post lower proficiency while adding more to its pupils' learning year on year. The public data cannot separate the two, and anyone telling you their score does is selling something.
The practical use is narrow and real: the percentile sorts a metro's districts, and sorting is what a shortlist needs. After that, look up the specific schools your address would attend, check whether the district has rezoned recently, and visit. That last step tells you more than every number on this page put together.