The National Merit Selection Index: Why Reading and Writing Counts Twice
The National Merit Selection Index: Why Reading and Writing Counts Twice
Two juniors in California walk out of the October PSAT/NMSQT with the same 1440. One is named a National Merit Semifinalist. The other misses the state cutoff by eight points and is Commended.
Nothing about that is a mistake, and it is not close. National Merit does not rank students by PSAT total score. It ranks them by the Selection Index, and the Selection Index counts your Reading and Writing score twice.
If you are chasing Semifinalist status, this single piece of arithmetic should reorganize your entire study plan. Most students chasing it have never done the calculation.
The formula
> Selection Index = (2 × Reading and Writing score + Math score) ÷ 10
Each PSAT/NMSQT section is scored 160–760, which puts the Selection Index on a 48–228 scale. A 760 in both sections gives you the maximum 228.
The consequence falls out immediately. Section scores move in 10-point steps, and:
| Move | Change in PSAT total | Change in Selection Index |
|---|---|---|
| +10 Reading and Writing | +10 | +2 |
| +10 Math | +10 | +1 |
| +40 Reading and Writing | +40 | +8 |
| +40 Math | +40 | +4 |
A point of Math is half a point of Reading and Writing in this competition. Every hour you spend is worth double on one side of the test.
What that does to identical totals
Here is the same PSAT total built four different ways, with the class of 2027 result attached to each.
| Reading and Writing | Math | PSAT total | Selection Index | Semifinalist in how many states, class of 2027 |
|---|---|---|---|---|
| 760 | 680 | 1440 | 220 | 45 of 50 — everywhere except NJ, MD, MA, VA, WA |
| 720 | 720 | 1440 | 216 | 34 of 50 |
| 700 | 740 | 1440 | 214 | 27 of 50 |
| 680 | 760 | 1440 | 212 | 22 of 50 — Commended elsewhere |
Eight index points of spread, from one total score, and half the country's cutoffs sitting inside that spread. The math-heavy student in the last row is not worse at the PSAT than the student in the first. They are worse at the thing the index pays for.
This is why "will a 1450 qualify?" is an unanswerable question, and why every forum thread arguing about it goes in circles. The answerable version is: will a 750 Reading and Writing and a 700 Math qualify in my state?
The class of 2027 numbers, and why they moved
Semifinalists for the class of 2027 were announced on September 9, 2026, from the October 2025 PSAT/NMSQT. They were the lowest cutoffs in the program's history — all 50 states came in below the prior year, most by at least three points.
| Benchmark | Class of 2027 | Class of 2026 | Class of 2028 (projected range) |
|---|---|---|---|
| National Commended level | 208 | 210 | 207–212 |
| Highest state (New Jersey) | 223 | 225 | 222–225 |
| California | 220 | 224 | 219–224 |
| Texas | 219 | 222 | 218–222 |
| Florida | 216 | 219 | 215–219 |
| Lowest states (WV, MT, WY, ND) | 208 | 210 | 207–212 |
Roughly 16,230 students were named Semifinalists this year out of about 1.3 million entrants, down about 5% from the prior year.
The cause of the collapse is the part worth internalizing. It was not students getting worse at math. High Math scores were close to flat; the number of students scoring 700–760 on Reading and Writing fell by about 27% from the previous administration, after having spiked by 18% the year before. Because Reading and Writing is double-weighted, that swing dragged all 50 cutoffs down with it.
Two things follow for anyone sitting the October 2026 PSAT/NMSQT:
- The scale is unstable at the top, and it is unstable on the Reading and Writing side specifically. Treat your state's cutoff as a range, not a line. Plan to clear the top of the projected range, not the bottom.
- Reading and Writing is where the competition is actually decided — both because of the doubling and because that is where the score distribution keeps moving.
Where the index points actually are
The Reading and Writing section is 54 questions across two modules, of which roughly 50 are scored — the rest are unscored pretest items. Near the top of the scale, where 720 and 730 and 740 are separated by very little, a single question can be the difference between one index point and three.
Those 54 questions are not one skill. They split into four scored domains, and they are not equally learnable:
| Reading and Writing domain | Share of section | Approx. questions | How it behaves |
|---|---|---|---|
| Craft and Structure | 28% | 15–16 | Vocabulary in context, text structure, cross-text connections — trainable, but slowly |
| Information and Ideas | 26% | 13–14 | Command of evidence, inference, quantitative evidence — the hardest to move quickly |
| Standard English Conventions | 26% | 13–14 | A finite, closed rule set. The fastest points on the entire test |
| Expression of Ideas | 20% | 10–11 | Transitions and rhetorical synthesis — pattern-based and highly drillable |
Standard English Conventions and Expression of Ideas together are about 46% of the section — roughly 23 to 25 questions. They are governed by rules and recurring patterns rather than by reading comprehension, which means they are the part of the PSAT that responds fastest to targeted work. And because of the doubling, every one of those questions is worth twice what a Math question is worth to your index.
Compare that to the Math side, where the 44 questions split into Algebra (13–15), Advanced Math (13–15), Problem Solving and Data Analysis (5–7), and Geometry and Trigonometry (5–7). A student who is already at 740 Math is fighting for maybe two questions in Advanced Math or Geometry and Trigonometry — and each of those is worth one index point.
The strategic conclusion for most index-chasers is uncomfortable and correct: if your Math is above 730 and your Reading and Writing is below 730, almost nothing you do in Math is worth doing.
What to do before the October PSAT/NMSQT
The 2026 PSAT/NMSQT in-school testing window runs October 1–30, 2026, with some schools also offering a Saturday administration in October. You do not register with College Board — your school does. Confirm your exact date with your counselor now, because it varies school by school and the makeup options are limited.
Scores are released in waves based on when your school submits answer documents, generally late October through mid-November 2026. Semifinalist lists for the class of 2028 do not reach high schools until the end of August 2027, with the press embargo lifting in mid-September. That is an eleven-month wait; plan on it.
With the time you have left:
- Convert your practice scores into an index, not a total. Every time you take a practice test, compute
(2 × RW + M) ÷ 10and compare it against the projected range for your state. Your total is a distraction. - Get a domain-level breakdown of your Reading and Writing misses. Not "I need to read more" — a count of misses in Information and Ideas, Craft and Structure, Expression of Ideas, and Standard English Conventions. Most students at 700+ RW are solid in two domains and a band down in the other two, and those two weak domains hold nearly all of the index points still available to them. A diagnostic that reports at the domain level takes about an hour and tells you which of the four to spend October on.
- Start with Standard English Conventions. A closed rule set, about a quarter of the section, double-weighted. Nothing else on the test has that return profile.
- Do not chase percentiles. College Board's percentile ranks are computed from the prior three years of results and are set before you sit the exam, so a 99th-percentile score tells you nothing reliable about this year's cutoff. The historical cutoff record is a far better guide.
If your Reading and Writing score is the thing standing between you and your state's number — and for most students reading this, it is — the same four-domain analysis is the whole basis of how to raise a Reading and Writing score at the top of the scale.
The honest caveat
Cutoffs are not a standard you meet. They are the point at which a fixed number of Semifinalist slots, allocated to your state by its share of high school graduates, happens to run out. The class of 2027 proved how far that line can move for reasons that have nothing to do with you: a scaling shift on one section dropped every state in the country.
So set your target above your state's projected range rather than at it, put the work where the index actually pays — and remember that the same Reading and Writing domains that decide National Merit are the ones that decide your SAT score in the spring. Nothing you do here is wasted if the index does not come in.