The top of the scale

How to get from 1400 to 1600 on the SAT

A 1400 already means most questions are going right. The last 200 points are a different problem from the first 400 — they are not hiding in material the student has never seen, and no quantity of additional practice papers will find them on its own.

Why 1400 is where progress stops

Almost every SAT course teaches the same handful of shortcuts: eliminate two options, plug in numbers, work backwards from the answers, keep to a pace. They work. They are the reason a student moves from 1100 to 1400 in a few months, and no one should give them up.

The problem is that the hardest questions on the test are written by people who know exactly which shortcuts are being taught. The standard design is a question that presents as a familiar type, invites the automatic move, and then breaks the pattern one step in. A student who has practised the shortcut until it is automatic has also practised walking into the trap until that is automatic.

This is why the plateau is so consistent, and why it feels like effort has stopped converting. It has not. It is being spent reinforcing the thing that is now the limit.

Where the last 200 points actually are

It helps to translate the score back into questions, because the scale disguises how few are left. These are rough figures and the caveats below the table matter:

Section score	Roughly how many questions missed	What is usually causing them
650	10–13	Genuine content gaps — topics not yet secure
700	6–9	A mix of thin content and avoidable error
750	3–5	Mostly avoidable error, plus the hardest few items
780–800	0–2	The hardest items only, with no room for a slip

Approximate only. The conversion differs between Math and Reading & Writing, moves from form to form, and — because the test is adaptive and four questions per section are unscored pretest items — depends on which questions were missed, not only how many. Read the middle column rather than the left one: a student at 1400 is missing something like twelve to eighteen questions across the whole test, and by this stage most of them are not questions they could not do.

The diagnosis that has to come first

Below 1400, reviewing by topic works: the wrong answers cluster, a topic is weak, teach the topic. Above 1400 that method fails, because the wrong answers stop clustering by topic and start clustering by cause. Two missed questions on quadratics can have nothing in common.

So the review has to change shape. Every missed question gets classified into one of five causes, and the classification — not the topic — is what gets counted:

Content. The student genuinely could not do it, and could not do it with unlimited time. Above 1400 this is usually the smallest category.
Misread. The question was answered correctly, but it was not the question asked. Solving for x when it wanted 2x. Choosing the sentence that is true rather than the one the passage supports.
Slip. A sign error, a transcription error, a right answer marked in the wrong place. Trivial individually, and collectively the single largest barrier to 1600.
Timing. Rushed, or never reached. Distinguish the two: they have opposite fixes.
Judgment. The shortcut was applied where it was not safe, or the student committed to a plausible option before checking the discriminating detail. This is the interesting category and the one that most repays teaching.

Keep the log for three or four full mocks and the picture is usually unambiguous — one or two causes account for most of the loss. Almost every student is surprised by which. The instinct is to assume content, and the evidence rarely agrees.

Math: what the hardest questions are actually testing

Setting up, not solving. At this level the algebra is not the difficulty; turning three lines of English into the right equation is. Practise translation as a separate skill from computation.
Knowing when Desmos helps and when it costs. Built into Bluebook and excellent for systems, intersections and roots. Also a reliable way to lose ninety seconds typing a function that could have been reasoned out in fifteen.
The student-produced responses. Roughly a quarter of the Math questions have no options at all, so elimination is unavailable and a slip cannot be caught by noticing that your answer is not offered. Accuracy discipline matters most here.
The pattern-break questions. Nonlinear systems, function transformation and anything asking for a condition on a parameter rather than a value. These are where the last two or three marks in Math sit.
Checking as a habit, not a phase. A student aiming at 800 does not check at the end; they verify each answer against the question as they go, because there is no time budget for a second full pass.

Reading & Writing: the two-plausible-answers problem

The recurring experience above 700 in this section is two options that both look defensible. Students then choose the one that reads better, which is exactly what the question is designed to punish. In practice there is nearly always a discriminator, and it is nearly always something textual — a qualifier, a scope word, a tense, a claim that is true in the world but not supported by this passage. Finding it is a trainable habit; picking the nicer sentence is not a method. The Reading & Writing page goes into how that is taught.

Two structural facts about this section are worth exploiting. First, College Board states that questions testing similar skills are grouped together and arranged from easiest to hardest, so difficulty rises predictably within each group rather than at random — which tells a student where in the module the risk is. Second, the grammar questions, the Standard English Conventions group, are the most finite thing on the whole test. The rules are countable, they repeat, and a student who is losing marks there is losing them to a fixable list rather than to a talent gap. That is usually the fastest available gain for anyone in the 1400s.

Timing at the top of the scale

Bank time early, spend it late. In Reading & Writing, where questions are grouped and ordered easiest to hardest, the early questions in each group are the cheapest place to save seconds and the last are where seconds are worth most. Math is not ordered by difficulty, so there the judgment is per question.
Take the first module seriously. It decides whether the second module is the harder version, and only the harder version carries the top of the scale. See how the adaptive format works.
Practise the recovery, not just the paper. Everyone meets one question that will not come out. The skill is leaving it inside twenty seconds without the next four questions suffering for it.
Rehearse in Bluebook. On-screen reading, the built-in Desmos and the flagging tool all behave differently from paper, and none of that should be new on the day.

A realistic timeline

Move	What it usually takes	The work that does it
1400 → 1500	Two to four months	Cause-classified review; closing the Conventions list; slip discipline
1500 → 1550	Two to four months	The hardest item types, taught individually; near-perfect accuracy under time
1550 → 1600	Open-ended	Sustained flawlessness, and a degree of luck in the form you sit

Timelines assume regular weekly work alongside school, not full-time preparation. Plan backwards from a sitting that leaves room for a second attempt — see the test date table.

An honest word about 1600 itself

Very few students need one. A score above 1500 clears the published band at the great majority of universities that report one — at the most selective handful it sits at the bottom of the range rather than above it — and beyond that the application turns on essays, references, subject depth and everything else that is not a test score. The step from 1550 to 1600 is the most expensive 50 points available and among the least valuable.

That is worth saying plainly, because the students who ask this question are usually the ones with the most to lose from spending another three months on it. If the last 50 points are fun, chase them. If they are coming out of essay time, they are a bad trade.

Common questions

Why am I stuck at 1400 on the SAT?

Because the strategies that got you to 1400 cannot reach past it. Elimination, plugging in numbers and pacing rules carry a student efficiently through easy and mid-range questions, and the hardest questions are written specifically to look like the textbook case and then break the pattern. A student who has automated the shortcuts has also automated the mistake. Progress past 1400 comes from diagnosing which specific judgment calls are going wrong, not from sitting more papers.

How many questions can you miss and still get 1600?

Usually none, and on some test forms one. As a rough guide a 780 means one or two questions missed, a 750 three to five and a 700 six to nine, but treat those as indicative only: the conversion differs between the two sections, changes from form to form, and is affected by the adaptive routing and by the unscored pretest questions. That is the real difficulty of a 1600: it is not a knowledge threshold but an accuracy one, and it leaves no margin for a careless slip anywhere in either section.

How long does it take to go from 1400 to 1600?

Longer than the same effort took to get from 1200 to 1400, and the two halves of the climb are not equal. Moving 1400 to 1500 is usually a matter of months of consistent, diagnosed practice. Moving 1500 to 1600 can take as long again, because at that level the remaining errors are rare, individually different and hard to reproduce on demand. Plan a timeline in months rather than weeks, and register for a sitting that leaves room for a second attempt.

Is a 1600 worth aiming for?

Often not. The admissions value of the step from 1550 to 1600 is close to zero at almost every university, because both scores clear any published band and the application is then decided on everything else. Chasing the final 50 points can cost time that essays, references or a subject olympiad would repay far better. Aim for 1600 if the process itself is enjoyable and the rest of the application is already strong; otherwise bank a score above 1500 and move on.

Do more practice tests raise a score above 1400?

Not on their own. Sitting a paper measures the score; it does not change it. What changes the score is what happens after the paper, when every wrong answer is classified by cause rather than by topic and the most frequent cause is worked on directly. A student who sits ten mocks and reviews none will finish where they started. Two mocks reviewed properly beat ten sat and filed.

Stuck in the 1400s?

Send Lewis the score report and a couple of questions that went wrong. The free 20-minute demonstration lesson is spent on those questions, so you get a straight read on which of the five causes below is actually costing the marks.

Request a free demo lesson Ask Lewis on WhatsApp

Prefer email? Write to lewis@altairprep.com.