- The 54 Reading and Writing questions come from 12 templates; learn the move for each and you stop reading passages cold.
- College Board orders every module by question type, so you always know what’s coming next and can run the matching move.
- The 8 punctuation rules cover every Boundaries question; a student at 650 to 700 who learns them typically stops missing three to four questions a test.
- For every template, read the question before the passage. The question tells you what to look for; the passage is where you find it.
- Predict before you look at the choices on Words in Context and Transitions, the two templates that punish reading by sound.
Your Reading and Writing score has sat between 650 and 700 for three tests, you read well, and you can’t tell why the same section keeps producing the same number. Every one of the 54 Reading and Writing questions on the digital SAT is one of 12 templates, and every punctuation question is one of 8 rules, so the way out of 680 is to stop treating it as a reading test and start running the move for the template in front of you.
Students at your level have the reading. What they’re missing is recognition. Here are the 12 templates and the 8 rules, each with an example we wrote ourselves (none of these are College Board questions) and the move that answers it.
How the section is built
College Board publishes the structure at satsuite.collegeboard.org: 54 questions in 64 minutes, split into two modules of 27 questions in 32 minutes each, with the second module chosen by your performance on the first. Each question has its own short passage of 25 to 150 words. There is one question per passage, which means there is no passage worth reading before you know what you’re being asked about.
The questions come in a fixed order by domain: Craft and Structure, then Information and Ideas, then Standard English Conventions, then Expression of Ideas. Inside each domain they run roughly easy to hard. College Board names 11 skills across those four domains; we teach 12 because one of them, Form, Structure, and Sense, needs two different moves.
That order is a gift. You always know roughly which template is coming, so you can run its move without hunting for the question type.
Craft and Structure: three templates
1. Words in Context
The stem asks which choice “most logically completes the text” at a blank, or what a word “most nearly means.” Our example: “The committee’s report was notably ____ : it ran to four pages where the previous year’s had needed forty.” A student reading by tone might pick “thorough.” The sentence’s logic (four pages instead of forty) demands something like “concise.” Logic beats tone here. The move is to cover the choices, predict a word from the sentence’s own logic, then eliminate everything that doesn’t match the prediction.
2. Text Structure and Purpose
The stem asks for the “main purpose of the text,” the “function of the underlined sentence,” or the “overall structure.” Our example: a passage describes a gardener’s failed attempt to grow citrus in a northern climate, then explains the greenhouse she built. The underlined sentence is the one about the failure. Its function is to establish the problem the rest of the passage solves. The move is to name what each sentence does (sets up, contrasts, gives an example, concludes) in two words before reading the choices.
3. Cross-Text Connections
Two short texts on one topic, and the stem asks how the author of text 2 would “most likely respond to” a claim in text 1. Our example: text 1 argues that a city’s tree-planting program improved air quality; text 2 notes that air quality improved citywide in the same years, including in districts with no new trees. Text 2’s author would say the improvement can’t be credited to the trees. Start with the claim. The move is to locate text 1’s specific claim first, then find the sentence in text 2 that speaks to that claim and nothing else.
Information and Ideas: four templates
4. Central Ideas and Details
“Which choice best states the main idea” or “according to the text, what is true about X.” Our example: a passage says a bridge was celebrated at its opening, criticized within a decade for its narrow lanes, and widened in the 1950s. The main idea is the change in how the bridge was judged, not the widening. The widening is a detail. The move is to answer in your own words before looking at the choices, then pick the choice closest to yours.
5. Inferences
“Which choice most logically completes the text” at the end of a passage that builds an argument. Our example: a passage explains that a species of moth is drawn to a specific flower’s scent, and that the flower has recently spread into a new valley. The logical completion is that the moth is likely to follow. The trap is a completion that’s true in the world but not supported by the passage. The move is to find the completion that the preceding sentences force, not one they merely allow.
6. Command of Evidence, textual
A claim or hypothesis is stated, and the stem asks which finding “would most directly support” it. Our example: a researcher proposes that children who read aloud to a pet improve their reading fluency. The supporting finding is one that compares reading-aloud-to-pets children against otherwise similar children who didn’t, and shows a fluency difference. A finding that children enjoy reading to pets is pleasant and irrelevant. The move is to state what evidence the claim would need, then look for it.
7. Command of Evidence, quantitative
A table or graph with a passage making a claim, and the stem asks which choice “most effectively uses data from the table to complete the statement.” Our example: a table shows two towns’ recycling rates in 2015 and 2023, and the passage claims the smaller town improved faster. The correct choice names the two numbers that show the smaller town’s larger change. A wrong choice will quote a true number from the table that doesn’t make the comparison. The move is to check two things: is the number actually in the table, and does it support this claim specifically.
Standard English Conventions: three templates
8. Boundaries
Punctuation between and within sentences: periods, commas, semicolons, colons, dashes. These are the 8 rules, and they get their own section below. The move is to identify what sits on each side of the punctuation (a complete sentence, a fragment, a list, a nonessential phrase) before judging the mark.
9. Form: agreement
Subject-verb and pronoun agreement, usually with a distracting phrase between the subject and the verb. Our example: “The collection of letters, photographs and maps ____ now housed in the city archive.” The subject is “collection,” singular, so the verb is “is,” even though “maps” sits right beside the blank. “Maps” is the distraction. The move is to cross out every prepositional phrase and find the bare subject.
10. Structure and Sense: tense, modifiers, pronouns
Verb tense consistency, dangling modifiers and pronoun reference. Our example: “Having finished the survey, the results surprised the team.” The results didn’t finish the survey; the team did. The sentence needs “the team was surprised by the results.” The move is to ask who or what is doing the action in the opening phrase and make sure it’s the subject that follows.
Rick’s note: I make every student say the rule out loud before they answer a Conventions question. “Two complete sentences, so semicolon or period.” It feels slow for a week. Then it becomes the fastest thing they do on the test, because there’s nothing to reread.
Expression of Ideas: two templates
11. Transitions
A blank at the start of a sentence, and the stem asks for the “most logical transition.” Our example: “The orchard’s older trees produced less fruit each year. ____ , the farmer decided to replace a third of them.” The second sentence is a consequence of the first, so “accordingly” or “as a result,” not “however” or “for example.” The move is to label the relationship between the two sentences (addition, contrast, cause, example, sequence, concession) before looking at the choices.
12. Rhetorical Synthesis
A bulleted list of notes and a stated goal, and the stem asks which choice “most effectively uses relevant information from the notes” to achieve the goal. Our example: notes about a photographer’s birthplace, her first exhibition in 1962, her move to Texas, and a technique she pioneered; the goal is to “emphasize the photographer’s innovation.” The correct choice is about the technique. Three wrong choices are true sentences about her birthplace or her exhibition. The move is to underline the goal’s key word and pick the only choice that does that specific thing.
The 8 punctuation rules
Boundaries questions are the most learnable points on the section, because the test uses a small, fixed set of rules and the wrong answers break them in predictable ways. A student at 650 to 700 who has been answering by ear is usually missing three or four of these per test. All examples are ours.
Rule 1. Two complete sentences need a real divider
A period, a semicolon, or a comma plus a coordinating conjunction (for, and, nor, but, or, yet, so). A comma alone is a splice, and it is the single most common wrong answer on the section. “The storm passed by noon, the roads stayed closed until evening” needs a semicolon or a period at the comma.
Rule 2. A dependent clause in front takes a comma; behind, usually not
“Because the storm passed by noon, the roads reopened.” Comma. “The roads reopened because the storm passed by noon.” No comma. The test will offer both orders, and the comma rule flips with them.
Rule 3. Nonessential information is bracketed by a matching pair
Two commas, two dashes, or two parentheses, never one of each. “The lighthouse, built in 1871, still operates.” If you can lift the phrase out and the sentence survives, it’s nonessential and needs the pair. If the sentence changes meaning without it, no marks at all: “The lighthouse that was built in 1871 still operates” (as opposed to the one built later).
Rule 4. A colon follows a complete sentence
What comes after a colon can be a list, an explanation or a single word, but what comes before it must stand alone. “The report named two causes: the drought and the price of feed” is correct. “The two causes were: the drought and the price of feed” is wrong, because “The two causes were” cannot stand alone.
Rule 5. Semicolons separate list items that contain commas
“The team visited Austin, Texas; Santa Fe, New Mexico; and Tucson, Arizona.” The semicolons do the job commas can’t, because the items already contain commas. Outside this case, a semicolon on the SAT is almost always doing Rule 1.
Rule 6. No single comma between a subject and its verb, or a verb and its object
“The researchers who studied the reef, found new species” is wrong. A pair of commas around a nonessential phrase is fine (Rule 3). A lone comma splitting the sentence’s spine is not, and the test offers it often because it sounds like a pause.
Rule 7. Apostrophes mark possession and contraction, never plurals
Singular possessive adds ‘s; plural possessive adds s’ after the s. “Its” is possessive; “it’s” is “it is.” “The archive’s records” (one archive), “the archives’ records” (several). “The 1990s” takes no apostrophe. The test loves “its/it’s” and “their/they’re/there” in the same question.
Rule 8. A dash can replace a colon or a comma pair, but not half of one
A single dash after a complete sentence works like a colon. A pair of dashes works like a pair of commas around a nonessential phrase. One dash opening a nonessential phrase and a comma closing it is wrong. Match the marks, and the question becomes Rule 3 or Rule 4.
Using the list at 650 to 700
Take your most recent Bluebook test review and label every Reading and Writing miss with its template number, 1 through 12, and for Boundaries misses, the rule number. Then count. Students at your level typically find that two or three templates account for more than half the misses, and that the rest are scattered singletons.
Work the two or three. For each, do ten questions of that template in a row, untimed, saying the move out loud before you answer. Then ten timed at 71 seconds each. The point is to make the recognition automatic, so that on test day you see the stem and the move starts before you’ve finished reading.
Then check the first module. Reading and Writing questions in the 650 to 700 range are often lost in module 1 to rushing, and module 1 decides whether you see the harder second module where the 700s are. Finish module 1 with four minutes to spare and check your flagged questions. Our guide to the hard module explains why that matters so much.
What to do this week
- Label every Reading and Writing miss on your last test by template number, and every Boundaries miss by rule number.
- Pick the two templates with the most misses. Do ten untimed questions of each, saying the move aloud, then ten timed.
- Learn Rules 1, 3 and 4 cold. Those three carry the section. They cover most Boundaries questions a student at 680 is missing.
- On your next full-length test, read every question stem before its passage, and finish module 1 with four minutes to check.
- Bring the labeled review to a free first hour and Rick will tell you which templates are yours and what a realistic Reading and Writing target is.
Questions people ask
How many question types are on the SAT Reading and Writing section?
College Board names 11 skill areas across four domains. We teach 12 templates because Form, Structure, and Sense splits cleanly into agreement questions and sentence-structure questions, which need different moves. Every question on the section fits one of the 12.
What order do the questions come in?
College Board orders each module by domain: Craft and Structure first, then Information and Ideas, then Standard English Conventions, then Expression of Ideas, with questions roughly easy to hard inside each group. The same order in both modules, every test.
Which templates cost students at 650 to 700 the most?
Typically Words in Context, Transitions, Boundaries and the quantitative Command of Evidence items. These four share a cause: the student reads by feel rather than running a move. The error log from a recent test tells you which of the four are yours.
Do I need to memorize vocabulary?
Less than you’d think. Words in Context tests whether you can work out a word’s meaning from the sentence’s logic, so prediction beats flashcards. A short list of common words with a second, older meaning (temper, want, discriminate, novel) is worth knowing; a thousand-word list is not.
How is this different from what my class taught?
A class usually teaches the domains. This is one level down: the specific question stems, what each one is actually testing, and the two-step move that answers it. The section rewards recognizing the template before you finish reading the question.