The fourteen days before the digital exam are the stretch where most students lose points they had already earned. Not in the obvious way, by forgetting a formula, but in the quiet way: they keep cramming new material, their accuracy on the things they already knew starts to slip, and they walk in tired, jittery, and convinced that one more topic will be the one that saves them. The SAT math final review is not about adding knowledge in the last fortnight. It is about protecting the knowledge you already have and arranging it so that it fires cleanly under timed pressure on the morning that counts.

Here is the claim this guide will defend, and it is a specific one: in the last two weeks, the highest-return activity is not learning, it is consolidation and tapering. A student who spends these days diagnosing exactly where points leak, drilling those leaks shut, rehearsing the timed rhythm of a module, and then deliberately easing off so they arrive rested will outscore an identical student who studies harder and longer right up to the night before. Effort in this window is not free. Past a point it costs you, because a fatigued brain makes the careless slips that the quantitative section punishes hardest, and a panicked one second-guesses correct first instincts. The plan below treats the taper as preparation, not as a reward for finishing, because peaking on a single date is a thing you engineer, not a thing you hope for.

What this guide gives you that a generic “study hard and relax” tip cannot is a dated, day-by-day calendar with one job per day, each job tied to a specific diagnostic output and to the companion article that teaches the underlying skill. You will not be told to “review your weak areas.” You will be told which day to find them, how to rank them, which to fix and which to abandon, when to stop touching new content entirely, and what the final seventy-two hours should look like down to the morning routine. By the end you should be able to print the countdown, tape it above your desk, and execute it without deciding anything new under stress, which is exactly the state you want your decision-making reserved for on the day itself.

Where the final two weeks actually sit in your preparation

The two-week window is a distinct phase with its own logic, and treating it like an extension of the months that came before is the first mistake. During the long middle of preparation, the goal is acquisition: you are learning linear systems, you are learning how exponential models behave, you are learning to read a Desmos regression. Acquisition is slow, it tolerates mistakes, and it rewards volume. The final fortnight inverts every one of those properties. The goal becomes retrieval under constraint, the tolerance for new mistakes drops to near zero, and volume becomes actively harmful past a modest ceiling. If you carry the acquisition mindset into these days, you will keep opening new topics, each of which you half-learn, and half-learned content is worse than no content because it produces false confidence that collapses on a hard Module 2 item.

The first thing to understand about this window is that your score is, for practical purposes, already mostly determined. The skills that will carry you through the quantitative section were built over weeks and months, and you cannot meaningfully rebuild them in fourteen days. What you can do, and what this phase is for, is recover the points that sit just outside your grasp: the topics you almost know, the careless slips you keep repeating, the timing decisions you make badly under pressure, and the calculator techniques you have not yet made automatic. These are the recoverable points, and there are usually more of them than students expect. A typical test-taker leaves a meaningful number of points on the table not because the underlying idea is beyond them but because the idea is fragile, the execution is sloppy, or the clock ran out before they reached an item they could have solved. The final two weeks are an exercise in converting fragile knowledge into reliable knowledge and in tightening execution so that what you know becomes what you score.

The second thing to understand is the role of the diagnostic. You cannot triage what you have not measured, and so the entire plan hinges on an honest, timed practice assessment taken at the very start of the window, with its errors sorted into a structure you can act on. This is where the tier system earns its place. Borrowing the logic from the broader analysis of how the quantitative section distributes its content, you sort every miss into one of three tiers. Tier 1 holds the high-frequency, lower-difficulty content that appears constantly and that you should never miss: linear equations, slope and intercept, percentages, basic data reading. Tier 2 holds the moderate-frequency, moderate-difficulty material that separates a solid middle score from a strong one: systems, quadratics, functions, ratios, the common geometry. Tier 3 holds the rare, hard content that shows up sparingly and mostly in the harder second module. The tier of a missed item, far more than the topic label, tells you what to do about it, because a Tier 1 miss is a five-alarm fire worth fixing first while a Tier 3 miss may be worth deliberately ignoring in a two-week window where time is the binding constraint.

The third thing to understand is that the digital format changes what final review should rehearse. The exam is adaptive at the module level, which the deeper treatment of how Module 1 routing gates your score ceiling lays out in full, and the short version is that your Module 1 performance determines whether you are routed into an easier or a harder Module 2, and the harder route is the only one that opens the top of the scale. This single fact reshapes the priority of your final days. It means Module 1 accuracy is not just important, it is the gate, and rehearsing clean, unhurried, error-free execution on the easier and medium items matters more in the last two weeks than chasing the occasional hardest problem. A student who drills exotic Tier 3 content while still dropping a Tier 1 percentage question is optimizing exactly backward. The taper, the rhythm sets, the careless-error audit, all of it bends toward the same end: lock down the points that gate the route, then attempt the ceiling.

How the digital math section behaves in your last fortnight

To plan the final two weeks well, you need a precise picture of the machine you are walking into, because the plan rehearses behaviors specific to that machine. The quantitative portion of the digital SAT runs as two modules, each timed at thirty-five minutes, taken on the Bluebook application on a screen rather than on paper. The calculator is permitted throughout, with the Desmos graphing calculator embedded directly in the testing application, and the reference sheet of geometry formulas is available on screen at all times. The average time budget works out to roughly ninety-five seconds per item, though that average hides the real distribution, because the early items in a module tend to move fast and the later ones devour the clock. Understanding this rhythm is part of what the final-week timed simulations are meant to ingrain, so that on the day your pacing is a habit rather than a calculation.

The adaptive structure is the feature that should shape your priorities most. The first module presents a mix of difficulties, and your accuracy on it routes you to a second module that is either more difficult or less difficult. The harder second module carries the higher scoring ceiling, so reaching it is the precondition for a top result. The practical consequence for your final two weeks is that you should treat Module 1 as a place where accuracy beats speed, every time. There is no medal for finishing the first module with eight minutes to spare if you rushed and dropped two questions you could have nailed. The full reasoning behind accuracy-first pacing lives in the dedicated treatment of the three-pass system for a thirty-five-minute module, and the final-week simulations exist precisely to make that three-pass instinct automatic: clear the certain points first, return for the medium ones, then spend whatever remains on the hardest.

A second behavioral fact worth internalizing before the taper is that the calculator does not save you from conceptual error, it only saves you from arithmetic error, and only if you have rehearsed it. Desmos will graph a system and hand you the intersection, it will find the zeros of a quadratic, it will fit a line to a table, but it does none of this unless your fingers already know the moves. The full catalogue of those moves belongs to the dedicated Desmos strategy for the digital exam, and one day of your final two weeks is reserved for nothing but rehearsing them until they are reflexive, because a technique you have to think about under time pressure is a technique you will not use when it counts.

The third behavioral fact is the one students most underestimate: the reference sheet is generous but incomplete. It supplies the area and volume formulas and the special-right-triangle ratios, but it does not supply the slope formula, the midpoint formula, the vertex form of a parabola, the quadratic formula, the exponent rules, or the common Pythagorean triples, and reaching for a formula you assumed would be provided and finding it absent is a confidence killer in the first minutes of a module. The complete accounting of what the sheet gives you and what you must carry in memory is the job of the formula and concept reference sheet, and one day of the countdown is devoted to a final pass through it so that nothing on the day is a surprise.

Finally, the digital format rewards a particular kind of error discipline, because the careless slips that survive into the final weeks are usually not random. They cluster into recognizable patterns: misreading what the item asks for, solving for the wrong variable, dropping a negative sign, answering in the wrong units, or selecting an answer that solves an intermediate step rather than the final question. These patterns are individual, they are yours, and the only way to fix them is to name them from your own work, which the careless-mistakes elimination method treats in depth and which the countdown converts into a single dedicated self-audit day. With the machine and its behaviors clear, the plan itself can be specific, and the rest of this guide is that plan, fourteen days, one job each, ending with the morning routine.

One last structural fact shapes how the final-week simulations should be run: the quantitative section comes second on the digital exam, after the reading and writing portion, with a short scheduled break between them. This matters for your rehearsal because it means your math performance happens when you are already an hour or more into the sitting, with the freshest part of your concentration spent on the verbal section. A student who only ever practices math cold, first thing, rehearses a condition that will not exist on the day. When you run your timed module simulation around five days out, do it after some other cognitive work rather than as your first act of the morning, so the rehearsal matches the fatigue state you will actually face, and so the warm-up routine you build for test morning is calibrated to wake the right part of your mind at the right moment rather than too early.

The fourteen-day countdown, one job per day

What follows is a calendar, not a list of suggestions. Each day carries a single primary job, a clear output, and a tie to the companion article that teaches the underlying skill in full. The structure front-loads diagnosis and repair across the first week, draws a hard line against new content at the one-week mark, and then tapers deliberately into rest and rehearsal across the final stretch. The daily study time is meant to rise gently through the first half and fall through the second, never exceeding a couple of focused hours, because a tired brain in this window subtracts more than it adds. If your test date does not leave you a clean fourteen-day runway, compress the first week and protect the last three days untouched, since the taper is the part of the plan you can least afford to skip.

Day 14: the diagnostic and the error map

The window opens with a full, timed, screen-based practice assessment taken under conditions as close to the real thing as you can manage. Sit it in one sitting, on the Bluebook application if you can, with the clock running and no pauses, because a leisurely untimed run measures something other than what you need to measure. The point of this day is not the resulting number, which you should glance at and then set aside. The point is the error map you build afterward, and building it well is the most consequential hour of the entire fortnight.

Here is how that map gets built, walked through as a tutor would narrate it. Take every item you missed and every item you got right but felt unsure of, and write each one on its own line. For each, record three things: the topic, the tier, and the failure mode. The tier comes from the analysis of how the quantitative section distributes its content across difficulty bands, and assigning it is the move that turns a pile of mistakes into a plan. Suppose you missed a percentage-increase item, a system-of-equations item, a quadratic-vertex item, and a rare combinatorics item. The percentage miss is Tier 1, high frequency and low difficulty, which means it is bleeding you points across every practice run and it goes to the top of the repair list. The system and the quadratic are Tier 2, moderate frequency and moderate difficulty, the band where most score growth actually lives, so they go in the middle. The combinatorics item is Tier 3, rare and hard, and with only fourteen days you may rationally decide to leave it alone, because the hour you would spend learning it returns less than the same hour spent making the percentage and system items automatic.

Now record the failure mode for each, because the tier tells you how much a fix is worth and the failure mode tells you what the fix actually is. The percentage miss, on inspection, turns out not to be ignorance of percentages at all: you set up the multiplier correctly but computed a 20 percent increase as multiplication by 0.20 rather than by 1.20, a classic execution slip rather than a knowledge gap. That goes in your careless-error log, not your study list. The system miss is genuine: you tried to solve it algebraically, made a substitution error, and never thought to graph both lines in Desmos and read the intersection. That is a method gap, and it points at the calculator-rehearsal day. The quadratic miss is conceptual: you did not remember that the vertex sits at x equal to negative b over 2a, which is a formula the reference sheet does not supply, so it goes to the formula-review day. By the end of this exercise you have not a vague sense of weakness but a sorted, prioritized map: which misses are Tier 1 fires, which are Tier 2 growth, which are Tier 3 you will skip, and for each whether the cure is concept, method, or care. That map drives the next six days. Once you have built it, a session of targeted practice questions with worked solutions lets you confirm the map against fresh items rather than trusting a single test, and the immediate feedback turns the diagnosis into rehearsal.

Day 13: Tier 1 repair, the points you should never lose

The second day belongs entirely to Tier 1, the high-frequency low-difficulty content that gates your Module 1 accuracy and therefore your route to the harder, higher-ceiling second module. These are the items you cannot afford to miss, and the cruel truth is that strong students lose more points here than they expect, precisely because they spend their preparation chasing hard problems and treat the easy ones as beneath attention. Work through fresh examples of every Tier 1 topic your diagnostic flagged: linear equations and their graphs, slope and intercept in every form, percentages and percent change, ratios and proportions, and straightforward data reading from tables and bar charts. The goal is not to learn these, you already know them, but to make execution clean and fast so that the early minutes of a module bank certain points without a wobble.

Spend the day’s energy on the failure modes you logged, not on volume for its own sake. If your percentage misses traced to the multiplier confusion, drill nothing but percentage problems until the 1.20 reflex for a 20 percent increase and the 0.80 reflex for a 20 percent decrease are automatic, and until you can run a percent-change chain, an increase followed by a decrease, without resetting the base incorrectly. If your linear misses traced to confusing slope-intercept with standard form, rehearse converting between them until it is mechanical. The standard you are aiming for is not “I can do this” but “I cannot get this wrong even while tired,” because tired is the state you will partly be in on the day, and Tier 1 reliability is what holds when concentration frays. End the day by re-solving the Tier 1 items you missed on the diagnostic, with the failure mode named aloud before each, so the repair is anchored to the specific slip rather than to the topic in general.

Day 12: Tier 2 repair, where the score actually grows

The third day moves to Tier 2, the moderate-frequency moderate-difficulty band that the content analysis identifies as the region where most realistic score improvement lives. These are systems of equations, quadratics in their several forms, functions and function notation, exponential and linear modeling, the common geometry of triangles and circles, and the data-analysis items that go a step beyond simple reading. A student who has Tier 1 locked and who converts even half of their Tier 2 misses into reliable points moves a meaningful distance up the scale, far more than the same student would gain by finally cracking a single Tier 3 curiosity.

This is the day to drill a weak Tier 2 area to the point of fluency, and it deserves a worked walkthrough because the method matters. Suppose your diagnostic showed repeated misses on systems of equations, and your map flagged the failure mode as method rather than concept: you understand what a solution to a system is, but you default to algebra, make errors in the manipulation, and never reach for the graph. The repair is to rebuild the topic around the most reliable method for the digital format. Take a representative item: a question gives you two linear equations and asks for the value of x at the point where they meet. The slow, error-prone path is substitution by hand. The fast, reliable path is to type both equations into Desmos exactly as written, let the application plot both lines, and read the intersection point directly off the graph, then confirm the x-value matches an answer choice. Walk through three or four such items this way until the move is automatic, then escalate: a system where one equation is a parabola and the other a line, which has two intersection points, where the graph instantly shows you both and saves you from a sign error in the algebra. Then a system presented as a word problem, where the real skill is translating the words into two equations before the graph does the rest. The generalizable principle to carry away is that on the digital exam a system question is usually a graphing question in disguise, and recognizing that converts a method gap into a reliable point. Do the same fluency drill for whichever single Tier 2 area your map ranked highest, and leave the rest for the second half of the week if time allows.

Day 11: the Desmos rehearsal day

The fourth day is reserved for the calculator, because the embedded Desmos graphing tool is the single largest source of recoverable speed and accuracy on the digital quantitative section, and because its power is entirely contingent on rehearsal. A technique you have to reconstruct under time pressure is a technique you will skip when the clock is loud, so the job today is to make the core moves reflexive, the way a touch typist does not think about the keys. The full method set lives in the complete Desmos strategy for the digital exam, and today you practice it until your hands move without consulting your memory.

Rehearse the moves that recur most. Solving an equation by graphing both sides and reading the intersection, which turns many algebra items into a single plotted picture. Finding the zeros of a function by graphing it and reading where it crosses the horizontal axis, which dissolves most quadratic-root questions. Fitting a line or a curve to a table using a regression, which handles the modeling items that ask for the equation of best fit. Using a slider to test how a parameter changes a graph, which makes the “which value of k” questions visual rather than algebraic. Graphing a system to find an intersection, the move you rehearsed yesterday on the systems repair. Spend the day taking ordinary Tier 1 and Tier 2 items and deliberately solving them the Desmos way even when algebra would also work, because the point is not to choose the calculator every time on the day, it is to have the choice available without hesitation. By evening the test should be whether you can open a fresh modeling item and have the regression set up within a few seconds, hands ahead of thought.

Day 10: the formula and concept pass

The fifth day is a final, deliberate pass through every formula and rule that the on-screen reference sheet does not provide, because the sheet is generous with geometry and silent on much of the algebra and statistics you will actually use. The complete inventory belongs to the formula and concept reference sheet, and the job today is not to learn these for the first time, which would violate the no-new-content discipline arriving in two days, but to confirm that each is instantly available and to flag any that still feel shaky.

Run the pass by domain so nothing slips through. In algebra, confirm the slope formula, the three forms of a line and how to convert among them, the quadratic formula, the vertex location at negative b over 2a, the discriminant and what its sign tells you about the number of real roots, the exponent rules, and the conversion between radical and fractional-exponent notation. In geometry, note which formulas the sheet supplies so you do not waste a second hunting for them, then drill the ones it omits: the distance and midpoint formulas, the equation of a circle and how completing the square recovers its center and radius, the angle relationships created by a transversal, the similar-triangle ratios, and the common Pythagorean triples that let you skip the theorem entirely when you recognize a 3-4-5 or a 5-12-13. In statistics and probability, confirm the mean, the basic probability ratio, the counting principle, and conditional probability from a two-way table. For each item, the standard is recall within a second or two and correct application on one quick example. Anything that fails that standard gets a small flag and a five-minute fix on the spot, because after the hard stop you will not be allowed to open it again.

Day 9: the careless-error self-audit

The sixth day is the one most students skip and the one that often returns the most points, because the slips that survive into the final two weeks are rarely random and almost always patterned. The full method belongs to the careless-mistakes elimination guide, and today you apply it to your own work by reviewing every error log you have built this week, the diagnostic misses and the slips you caught during the Tier 1 and Tier 2 days, and naming the recurring patterns.

The patterns are individual, but they fall into a recognizable set. Misreading what the item asks for, where you solve correctly for x when the question wanted x plus three, or you find the value when it wanted the percent. Solving for the wrong quantity, where you stop at an intermediate result that happens to appear among the answer choices as a trap. Sign errors, where a dropped negative flips your answer. Unit errors, where you answer in minutes when the question wanted hours, or in the wrong dimension entirely. And premature selection, where you grab the first choice that matches a number you computed without checking whether it answers the actual question. For each pattern you find in your own logs, write a one-line cure phrased as a behavior, not a wish: not “be more careful with signs” but “before bubbling, reread the final line of the question and confirm I solved for that exact quantity.” Three or four such cures, internalized as habits, close a category of loss that no amount of additional topic study would touch. Spend the rest of the day doing a short mixed set with the cures taped beside you, deliberately executing each one, so the behavior is rehearsed and not merely resolved.

Day 8: the second full assessment

The seventh day closes the diagnostic-and-repair week with a second full, timed, screen-based assessment, taken under the same conditions as the first. Its purpose is comparison and confirmation: you want to see the Tier 1 fires extinguished, the targeted Tier 2 area converted, and the careless patterns reduced, and you want fresh data on what still leaks. Take it in one sitting, then build the error map exactly as you did on Day 14, sorting misses by tier and failure mode.

Read the comparison honestly. If the Tier 1 misses are gone, the repair worked and you can trust those points on the day. If a Tier 1 miss persists, it jumps to the very top of your remaining priority, because a fire that survived a week of attention is the single biggest threat to your route through Module 1. If a Tier 2 area you drilled has firmed up, note it and move on, and if a different Tier 2 area now stands out, it becomes the target for your mid-week drilling day. Resist the urge to read too much into the headline number, because a single point swing between two practice runs is noise, not signal, and chasing it will pull you toward panic studying exactly when the plan calls for the opposite. The output of today is a short, final priority list, no more than three items, that you will address in the few remaining active days before the taper takes over. With that list in hand, the diagnostic phase is complete, and the next day draws the line that protects the rest of the plan.

Day 7: the hard stop on new content

One week out, you draw a line, and on the far side of it lies every topic you have not yet learned. From this day forward you open no new material. Not the combinatorics item you skipped, not the obscure function transformation, not the rare statistics concept that showed up once. The reasoning is not motivational, it is mechanical: a topic introduced inside the final week cannot be moved from fragile to reliable in time, and fragile knowledge is a liability, because it produces hesitation and false confidence on exactly the hard items where a clean “I will flag this and move on” would have served you better. The students who arrive frazzled are almost always the ones who broke this rule, who spent the last week cramming new content and consequently let their reliable skills go stale while gaining nothing solid in return.

Spend today instead on the short priority list from yesterday’s assessment, confined entirely to content you already know. If a Tier 1 miss survived, this is the day to extinguish it for good, with focused repetition until it cannot recur. If a Tier 2 area needs one more pass, give it a measured session, not a marathon. The mood of the day shifts here, from acquisition to consolidation, and your study time should begin its gentle descent. You are no longer building, you are polishing, and the difference shows in how the work feels: lighter, more confident, more like rehearsal than like learning. Mark the hard stop somewhere you will see it, because the temptation to break it grows as the date nears, and the plan only works if the line holds.

Day 6: rhythm with easy and medium sets

With new content sealed off, the sixth day builds rhythm on Tier 1 and Tier 2 items, the band that gates Module 1 and grows the score. The aim is flow: a steady, unhurried pace that banks certain points cleanly and keeps the careless cures running in the background. Work a set of mixed easy and medium items at a comfortable speed, not racing the clock yet, attending to clean execution and to the three-pass instinct that the pacing strategy describes, where you clear the certain points first rather than getting stuck early. A continued run through a targeted practice set with worked solutions suits today well, because the immediate feedback keeps the rhythm honest and surfaces any cure that has not yet become automatic.

The purpose of a rhythm day is partly psychological and partly mechanical. Mechanically, it keeps the reliable skills warm without the strain of timed pressure, so they stay sharp through the taper. Psychologically, it rebuilds the confidence that a week of error-hunting can erode, because spending six days finding everything wrong with your performance is demoralizing if it is not balanced by days that remind you how much you can do well. End the session on a run of items you solve cleanly, deliberately, so the last impression of the day is competence rather than struggle, and so your relationship with the section going into the final stretch is calm rather than anxious.

Day 5: the timed module simulation

The fifth-to-last day is for one timed module simulation, a single thirty-five-minute block taken under the clock, to rehearse pace rather than to learn anything. This is not a full assessment, which would be too taxing this close to the date, but a focused dress rehearsal of the rhythm you will need: the three-pass sweep, the flag-and-return discipline, the decision of when to stop wrestling a hard item and bank the time elsewhere. The adaptive logic means Module 1 accuracy gates your route, so run the simulation with accuracy as the priority and speed as the servant, exactly as the module-routing analysis recommends.

Treat the simulation as a behavior rehearsal, not a score check. Notice whether your pacing instinct fires correctly: do you move on from a stuck item before it eats two minutes, or do you sink into it out of stubbornness? Do you flag and return cleanly, or do you forget the flags? Does the Desmos move come automatically on the items that call for it, or do you still hesitate? The answers tell you what the final two days of light review should reinforce. Keep the analysis afterward short and behavioral, focused on pacing and execution rather than on which topics you missed, because at this point topic gaps are mostly closed and the remaining gains are in how you run the clock. The taper begins in earnest tomorrow, and from here the plan does less, not more, on purpose.

Day 4: light targeted review

Four days out, the work goes light and targeted. A short session on the one or two items from the priority list that still feel less than automatic, a quick re-pass of the careless cures, a brief Desmos warm-up to keep the hands fluent. Nothing today should last long or feel heavy. If the priority list is genuinely clear, this becomes a confidence day: a small set of items you solve cleanly, chosen to remind you what reliable execution feels like. The instinct to do more, to squeeze in one more topic or one more full module, is exactly the instinct the taper is designed to override, because the marginal point you might gain from extra cramming is smaller than the point you will lose from arriving depleted.

This is also the day to begin attending to the non-academic preparation, because logistics handled early are logistics that do not generate anxiety later. Confirm the test center location and the route, check that the Bluebook application is installed and updated on the device you will use, gather the admission ticket and an acceptable photo identification, and locate an approved calculator as a backup even though Desmos is embedded. Handling these now, while you still have days of margin, means the final forty-eight hours can be spent resting rather than scrambling, and a calm logistical runway is part of how you arrive ready.

Day 3: the taper deepens

Three days out, the taper deepens, and the daily work shrinks to a light touch. A brief review of formulas, a short read-through of your careless cures, perhaps a handful of easy items to keep the machine warm, and then you stop. The logic of the final-three-days taper deserves its own explanation, because it is the part of the plan students most distrust and most often sabotage. The reasoning is that quantitative performance under timed pressure depends not only on what you know but on the state of the system executing that knowledge, and that system, your attention, your working memory, your error discipline, degrades with fatigue and recovers with rest. The slips that the careless-error day was built to fight are fatigue-sensitive: a rested brain rereads the question and catches the wrong-variable trap, a tired one does not. So the taper is not a reward for finishing the work, it is the final phase of the work, the part that converts everything you built into something that will actually fire on the day.

Concretely, the final three days descend from a light review on Day 3, to a very brief warm-up on Day 2, to no studying at all on Day 1. Each step down is deliberate. The aim is to arrive on test morning with your skills warm but your mind fresh, the way an athlete tapers training before a competition rather than peaking exhaustion the day before. Trust the descent even though it will feel, to a student conditioned by months of daily study, uncomfortably like doing nothing. Doing nearly nothing, on purpose, in these three days, is doing exactly the right thing.

Day 2: the brief warm-up

Two days out, keep contact light. A short warm-up of a few easy items, solved cleanly, exists only to keep the rhythm from going cold, not to teach or to test. Reread your careless cures one final time, glance at the formula flags, and then close the books. Spend the rest of the day on ordinary life: rest, normal meals, normal sleep, time away from the exam. The work is done, and the most productive thing you can do now is protect the rest that will let the work show.

This is the day to lock the logistics you confirmed earlier into a simple, written plan for the morning: what time you will wake, what you will eat, when you will leave, what you will carry. Lay out the admission ticket, the identification, the backup calculator, a watch if your center allows one, a snack and water for the break, and a light layer in case the room runs cold. Having all of this assembled and a plan written removes a whole category of morning friction, and a frictionless morning is a calmer mind, which is itself worth points.

Day 1: rest and logistics, no studying

The day before the exam, you do not study the quantitative section at all. Not a problem, not a formula, not a flashcard. The temptation to do “just a little” is strong and it is wrong, because a last-minute cram cannot add reliable skill at this point and can easily disturb the rest you need. The job today is recovery and readiness: a normal, restful day, light activity if it relaxes you, ordinary meals, and an early enough night to bank real sleep, since sleep is the single most performance-relevant variable left in your control.

Run through the morning plan once so it is fresh, confirm everything is packed, and then deliberately set the exam aside. Anxiety in the final hours is normal and does not predict a poor result, but feeding it with frantic review makes it worse, while a calm evening lets it settle. If you need something to do with nervous energy, a short walk and an early bedtime serve you far better than a problem set. You have spent two weeks building exactly the readiness this day is meant to preserve, and preserving it is now the entire assignment.

Test morning: the routine

On the morning itself, the goal is to arrive warm, calm, fed, and unhurried, and the routine that produces that state is simple enough to execute on autopilot, which is the point, because autopilot is what you want running while your decision-making stays reserved for the items. Wake with enough margin that nothing is rushed, and eat a real breakfast with protein and something slow to digest rather than only sugar, because a stable blood-sugar curve sustains attention across two sections far better than a spike that crashes mid-module. Arrive early enough that traffic or a parking problem cannot rattle you, with everything packed the night before so the morning is only execution.

A light cognitive warm-up helps and over-preparation hurts. Solving two or three easy items on the way, the kind you can do cleanly, wakes the quantitative part of your mind the way a runner does a few strides before the gun, so that the first real item of Module 1 does not catch you cold. The emphasis is on easy and clean, never on hard or new, because the warm-up exists to switch the machine on, not to test it, and a hard item that you stumble on would do the opposite of what you want, planting doubt at the worst moment. Once the section begins, fall into the rhythm you rehearsed: clear the certain points first, flag and return, let Desmos do the work it is fast at, and reread the final line of each question before you commit, executing the careless cures you spent a whole day building. You arrive having engineered your own peak, and the only job left is to run the plan you already know.

Worked examples: turning fragile points into reliable ones

The repair days are abstract until you see what the work actually looks like on a single item, so here are several walkthroughs of the kind you will run during the first week, narrated the way a tutor would talk you through them. Each one ends with the principle that generalizes, because the goal is never to memorize one solution but to carry away a move you can reuse on the next item of the same shape. Read these as models for how to drill your own flagged topics, not as a substitute for drilling them.

Start with a quadratic-vertex item, the kind that traces to a concept gap in your error map and gets fixed on the formula day. A function is given as f of x equal to x squared minus six x plus one, and the prompt asks for the minimum value of the function. The slow, anxious response is to start plugging in values and hope a pattern appears. The reliable response begins with a fact the on-screen reference sheet does not supply: the vertex of a parabola sits at x equal to negative b over 2a, and for a parabola opening upward that vertex is the minimum. Here a is one and b is negative six, so the vertex is at x equal to six over two, which is three. Substitute three back into the function: nine minus eighteen plus one, which is negative eight. The minimum value is negative eight. Then confirm it the digital way by typing the function into the graphing tool and reading the lowest point of the curve, which lands at the same place. The principle to carry away is that any “minimum or maximum value of a quadratic” question is a vertex question, the vertex lives at negative b over 2a, and the calculator confirms in seconds what the formula produces, so the formula and the graph check each other rather than competing. A topic that felt fragile on the diagnostic becomes a two-step reflex.

Move to an exponential-versus-linear modeling item, the sort of Tier 2 content where the deeper analysis of linear and exponential models and when each applies does the underlying teaching, and which makes a good target for a mid-window drilling session if your map flagged it. A problem describes a population that grows by the same percentage each year and asks which kind of model fits. The trap is to assume that anything increasing is linear because linear feels simpler. The discriminating question is whether the quantity changes by a constant amount each step, which is linear, or by a constant factor, which is exponential. “The same percentage each year” is a constant factor, so the model is exponential, and the equation takes the form of a starting value times a growth factor raised to the number of years, where a five percent annual increase makes the factor 1.05. Drill three or four of these, deliberately asking each time whether the change is by amount or by factor, until the distinction is automatic, then escalate to a decay version where a quantity loses a fixed percentage and the factor drops below one. The generalizable principle is that the words “constant amount” point to linear and the words “constant percentage” or “constant factor” point to exponential, and naming which one the problem describes resolves the model before any computation begins.

Now a Tier 1 percentage chain, the high-frequency content you cannot afford to miss and the place where the careless multiplier slip lives. A price rises by twenty percent and then the higher price is reduced by twenty percent, and the question asks how the final price compares to the original. The intuitive wrong answer is that the two changes cancel and the price returns to the start, and the trap is built precisely to catch that intuition. The reliable method is the multiplier: a twenty percent increase multiplies by 1.20, a twenty percent decrease multiplies by 0.80, and applying both means multiplying by 1.20 and then by 0.80, which gives 0.96. The final price is ninety-six percent of the original, four percent lower, not equal. The reason the changes do not cancel is that the decrease is taken on the larger amount, so it removes more than the increase added. The principle to carry away is that percent changes compound through multiplication rather than addition, that an increase followed by an equal-percentage decrease always lands below the start, and that the multiplier method protects you from the additive intuition the test exploits. This is exactly the kind of item to drill on the Tier 1 day until the multipliers are reflexive.

Then a wrong-variable trap, the failure mode the careless-error audit is built to catch. A problem sets up a relationship, asks you to find a quantity, and offers among its choices both the value of x and the value of the thing the question actually wanted, which might be x plus five, or two x, or the price before tax rather than after. You solve correctly, find x equal to seven, see seven sitting right there in the choices, and select it, except the question asked for x plus five, which is twelve, also sitting in the choices as the intended answer. Nothing about your algebra was wrong; the loss came entirely from not rereading the final line. The cure, written as a behavior rather than a wish, is to reread the last sentence of the question after solving and before bubbling, confirming that the number you are about to select answers that exact sentence. Rehearse this on a mixed set with the cure taped beside you, and the wrong-variable loss, which is one of the most common and most maddening on the section, closes.

Finally a conditional-probability item from a two-way table, a statistics concept the formula pass keeps available and one students often overcomplicate. A table breaks a group down by two categories, and the question asks for the probability that a member has one property given that they already have another. The word “given” is the signal, and it changes the denominator. Instead of dividing by the whole group, you divide only by the subgroup named after “given,” because that subgroup is now your entire population. If the question asks for the probability that a student plays a sport given that they are a senior, the denominator is the number of seniors, not the number of students, and the numerator is the number of seniors who play a sport. Read the relevant row or column, take the two numbers, and divide. The principle to carry away is that “given” narrows the denominator to the conditioning group, and that conditional-probability items are reading exercises on the table far more than they are calculation, so the whole skill is identifying which subgroup the condition selects.

These five span the failure modes your map will sort: a concept fix, a method-and-modeling fix, a high-frequency execution fix, a careless-behavior fix, and a reading-precision fix. Run your own flagged topics through the same narration, always ending by stating the move that generalizes, and you convert a pile of misses into a set of reusable reflexes, which is the entire job of the repair days. A further practice set with worked solutions lets you test each reflex against fresh items, and the immediate feedback confirms whether the fix held or whether the failure mode still lurks.

Reading your two assessments against each other

The diagnostic on Day 14 and the second assessment on Day 8 are not two separate scores to compare for reassurance, they are a before-and-after pair whose difference, read carefully, tells you exactly where the remaining gains are. The headline numbers matter least, because a single-point swing between two practice runs is noise produced by sleep, item luck, and ordinary variance, and treating it as signal is how students panic or relax at the wrong moment. What matters is the change in the error map, tier by tier and failure mode by failure mode.

Read the Tier 1 line first, because it gates everything downstream. If the high-frequency easy misses that appeared on the diagnostic are gone from the second assessment, your repair worked and you can trust those points on the day, which frees your final active days for polishing rather than firefighting. If even one Tier 1 miss survived a full week of targeted attention, it does not get demoted to “still working on it,” it gets promoted to the single highest priority for your remaining days, because a fire that survived a week of water is the biggest threat to your route through the first module. A surviving Tier 1 miss almost always means the failure mode was misdiagnosed: what you logged as a concept gap was really an execution slip, or what you treated as care was really a genuine hole. Relabel it honestly and attack the real cause.

Read the Tier 2 line next, looking for conversion. The area you drilled to fluency on Day 12 should show fewer misses or none, confirming that the method or concept fix took. If a different Tier 2 area now stands out that the diagnostic did not flag, that is not a failure, it is the natural result of fixing one leak and revealing the next, and it becomes the target for your light targeted-review day. Do not try to drill every Tier 2 area that appears; pick the one with the highest frequency or the clearest, most fixable failure mode, because in the days remaining you are choosing the highest-return single target, not attempting comprehensive coverage.

Read the failure-mode column across both assessments last, because the careless patterns are where the quietest gains hide. If the wrong-variable trap caught you on the diagnostic and your cure was rehearsed in between, the second assessment should show it gone, and seeing it gone is the proof your behavioral cure works under timed conditions rather than only in calm practice. If a careless pattern persists, the cure was either not specific enough or not rehearsed enough, and the fix is to sharpen the cure into a concrete behavior and run it on a short mixed set with the cure visible beside you. The output of this whole comparison is a short final priority list, no more than three items, that your last active days address before the taper takes over the schedule entirely.

Deciding which weak spots to abandon

A two-week window forces a discipline most students resist: deciding, on purpose, not to fix some of what is broken. Time is the binding constraint, and an hour spent on a rare, hard topic is an hour not spent making a common, easy one reliable, so the question is never “can I improve this” but “is this the highest-return use of the hour.” The tier system answers it. A Tier 3 miss, the rare and hard content that shows up sparingly and mostly in the harder second module, frequently fails the return test, because learning it from a fragile start in the final fortnight buys you an occasional, uncertain point at the cost of the warmth and reliability of skills that earn points on every run.

The decision rule is concrete. If a topic is high-frequency, fix it regardless of difficulty, because it pays on every assessment. If a topic is moderate-frequency and the failure mode is a clear, fixable method or concept gap, fix it, because that is where realistic growth lives. If a topic is low-frequency and hard, and especially if your error map shows it as a genuine concept hole rather than a slip, the rational move in a two-week window is to abandon it as a planned flag-and-skip on the day, freeing the time for consolidation. Abandoning is not giving up; it is a deliberate allocation, and on the day it converts into a calm decision to flag the rare hard item and move on rather than sinking three minutes you needed elsewhere. The students who refuse to abandon anything spread their final days too thin and arrive with everything half-warm, while the students who triage honestly arrive with the high-value content reliable and a clear plan for the rest. The full logic of how content concentrates by frequency, which underwrites this whole decision, lives in the question-pattern analysis, and leaning on it lets you abandon the right things without guilt.

The countdown at a glance

The plan compresses into a single calendar you can print and follow without rethinking it. Each row is one day, its single job, and the companion article that teaches the skill behind it. This is the InsightCrunch math final-review countdown, and it is built so that the decisions are already made, leaving your judgment free for the items themselves on the day.

Day	Single job	Output	Companion
14	Full timed assessment, then build the error map	Misses sorted by tier and failure mode	Question-pattern analysis
13	Tier 1 repair	The points you should never lose, made reliable	Careless-mistakes method
12	Tier 2 repair, drill the weakest area to fluency	One growth area converted	Question-pattern analysis
11	Desmos rehearsal until the moves are reflexive	Core calculator techniques automatic	Desmos strategy
10	Formula and concept pass	Every off-sheet formula instantly available	Formula reference sheet
9	Careless-error self-audit	Three or four behavioral cures	Careless-mistakes method
8	Second full timed assessment	Short final priority list of at most three items	Question-pattern analysis
7	Hard stop on new content; clear surviving fires	Line drawn, consolidation begins	Pacing strategy
6	Rhythm sets at a comfortable pace	Reliable skills warm, confidence rebuilt	Pacing strategy
5	One timed module simulation	Pacing and flag-and-return rehearsed	Module 1 vs Module 2
4	Light targeted review; confirm logistics	Loose ends closed, route and ticket ready	Test day complete guide
3	Taper deepens, light review only	Skills warm, fatigue clearing	Test day complete guide
2	Brief warm-up, then stop; pack the bag	Morning plan written, kit assembled	Test day complete guide
1	No studying, rest and sleep	Recovery banked	Test day complete guide
Morning	Eat, arrive early, light warm-up, run the plan	A calm, warm, unhurried start	Test day complete guide

The shape of the calendar is the whole argument in one picture: diagnosis and repair stacked in the first week, a hard line at the seven-day mark, and a deliberate descent through the final stretch. If you remember nothing else, remember that the work bends downward at the end on purpose, and that the bend is preparation, not laziness. For the full logistical side of test morning, the dedicated test day complete guide covers the kit, the timing, and the break strategy in detail.

Adjusting the countdown to your score target

The fourteen-day skeleton is the same for everyone, but where you spend the repair days should bend toward where your points actually are, and that depends on the band you are reaching for. The frequency logic does not change, since Tier 1 reliability gates the first module for every test-taker, but the marginal point sits in different places for a student consolidating a middle result and a student reaching for the top of the scale, and tailoring the emphasis is part of spending the fortnight well.

For a student whose practice work lands in the middle band and who wants to firm it up, the highest-return emphasis is Tier 1 and the easier half of Tier 2, because that is where the leaks are and where reliable execution converts directly into a steadier first module. A middle-band student who still drops the occasional percentage chain or linear-form item is bleeding the exact points the routing punishes most, so the Day 13 repair and the rhythm day matter more than any pursuit of hard content. The decision to abandon Tier 3 is easy here: a rare, hard item is not where a middle-band score grows, and the time it would consume is far better spent making the common content unmissable. For this student the taper is, if anything, even more important, because a steadier first module under rested conditions is precisely what nudges a middle result upward, and fatigue-driven slips are the main thing standing between the student and the next band.

For a student reaching for the top of the scale, the calculus shifts, because reliable Tier 1 is assumed rather than earned and the marginal point lives in the harder second module the routing unlocks. This student should still verify Tier 1 reliability on the diagnostic, since a single careless slip there can cap the route before the hard module is ever reached, but once that is confirmed the repair days lean toward the trickier Tier 2 content and the more challenging items that the analysis of the hardest question types treats in depth. The abandonment decision is harder for this student, because some Tier 3 content does appear in the route to the top and a genuine concept hole there may be worth a measured fix if the failure mode is clean. Even so, the discipline holds: a top-band student who breaks the hard stop to cram a rare topic in the final week risks the same fragility-and-fatigue trap as anyone else, and the points lost to a frazzled, slip-prone first module would cost more than the exotic item could ever gain. The timed module simulation matters most for this student, because at the top of the scale the difference between bands is often pace and flag-and-return discipline on the hard module rather than raw knowledge.

For a student who has improved quickly and whose practice scores are still climbing, the temptation is to assume the climb continues automatically and to coast through the final two weeks. The countdown protects against this by insisting on the second assessment, which tells you honestly whether the climb has plateaued and where, so you spend your last active days on the real current weak spot rather than on a stale picture of it. A fast-improving student often finds that the failure modes have shifted: the concept gaps that defined the early climb have closed, and what remains is execution and pace, which means the careless-error audit and the rhythm day carry more weight than further topic study. Whatever the band, the rule that does not bend is the hard stop on new content and the taper into rest, because those protect the points everyone has already earned, and protecting earned points is the surest math in the entire window.

The single most damaging error in the last two weeks is cramming new material, and it is worth naming the misconception precisely because it feels so productive. The belief is that one more topic, learned at the eleventh hour, will be the topic that appears and saves you. The reality is that a topic learned inside the final week stays fragile, and fragile knowledge does not behave like reliable knowledge on the day. It produces hesitation, it consumes time you cannot spare, and worst of all it generates a false confidence that lures you into attempting an item you should have flagged and skipped. Meanwhile the hours spent on the new topic are hours not spent keeping your reliable skills warm, so the net effect of late cramming is usually negative: you gain a fragile maybe and lose a reliable certainty. The hard stop at the one-week mark exists to make this mistake impossible to commit, and holding the line is one of the highest-return decisions in the entire plan.

The second mistake is skipping the taper, or worse, inverting it by studying hardest in the final days. Students conditioned by months of daily effort distrust rest, reading a light final week as slacking, and so they push through to the night before and arrive depleted. This is precisely backward. The quantitative section punishes fatigue through exactly the careless-error channel that a whole day of the plan was built to defend, because the wrong-variable slip and the dropped-negative slip are fatigue-sensitive, caught by a rested mind and missed by a tired one. A student who tapers well arrives with the same knowledge as a student who crammed, but with a sharper instrument to execute it, and on a timed test of execution that difference is points. Trust the descent.

The third mistake is letting a single practice score, high or low, hijack the plan. Practice numbers wobble for reasons that have nothing to do with your real ability: a bad night’s sleep, an unfamiliar item set, simple variance. A test-taker who panics at one low practice result and responds by abandoning the taper to cram, or who relaxes at one high result and stops the repair work early, has let noise overrule the plan. The countdown is built to be robust to a single data point, which is why the second assessment on Day 8 is used to confirm patterns rather than to chase a number, and why the analysis after it stays short and behavioral. Read trends across your week of work, not the headline of any one run.

The fourth mistake is neglecting logistics until they become a crisis. A student who has not confirmed the route, checked the Bluebook installation, or located identification until test morning has manufactured a category of stress that the plan deliberately eliminates by handling all of it on Day 4 and Day 2. A frictionless morning is a calmer mind, and a calmer mind catches the traps a frazzled one walks into, so the logistical preparation that looks like busywork is in fact part of the score.

The verdict on the last two weeks

If the question is what to do in the final fortnight before the quantitative section, the answer is unambiguous: diagnose hard in the first week, draw a firm line against new content at the one-week mark, and taper deliberately into rest across the final stretch. Do not learn new material in the last seven days, do not skip the taper, and do not let a single practice number rewrite the plan. The points that are still available to you this late are not in new topics, they are in the reliable execution of topics you already know, in the careless slips you can name and cure, in the calculator moves you can make reflexive, and in the rested clarity that lets all of it fire cleanly on the day. A student who accepts that the final two weeks are for consolidation and tapering rather than for one last push will, all else equal, outscore the student who studies harder right up to the night before, because peaking on a single date is something you engineer, and the engineering is mostly about knowing when to stop.

The plan above is that engineering written out, fourteen days with one job each, ending with a morning routine simple enough to run on autopilot. Print it, follow it, and reserve your judgment for the items rather than for second-guessing the schedule. For the verbal side of the same final stretch, the companion reading and writing final two weeks countdown mirrors this logic for the other section, and the two together cover the whole runway to the date.

Frequently Asked Questions

What should I do in the last two weeks before the SAT math?

In the final fortnight, shift from learning to consolidation and tapering. Open the window with a full timed assessment and sort every miss by difficulty tier and failure mode, then spend the first week repairing the highest-frequency leaks first, rehearsing your calculator moves, confirming off-sheet formulas, and auditing your careless-error patterns. At the one-week mark, stop opening any new content entirely, because a topic learned this late stays fragile and helps less than the reliable skills it crowds out. Across the final stretch, taper deliberately: lighter review each day, a single timed module simulation around five days out, confirmed logistics, and no studying at all the day before. The goal is to arrive with skills warm and mind fresh. You are not adding knowledge in these days, you are protecting what you already have and arranging it to fire cleanly under timed pressure on the morning that counts.

Should I study new math material the week before the SAT?

No. Draw a hard line at the one-week mark and open no new content past it. The reasoning is mechanical rather than motivational: a topic introduced inside the final week cannot move from fragile to reliable in time, and fragile knowledge is a liability on the day, because it breeds hesitation and a false confidence that lures you into attempting items you should have flagged and skipped. Worse, the hours spent on new material are hours not spent keeping your reliable skills warm, so late cramming usually nets out negative, trading a shaky maybe for a sure thing you let go stale. Spend the final week instead on consolidating content you already know: extinguishing any high-frequency miss that survived your repair work, polishing your strongest growth area, and rehearsing pace. If a rare topic is still a gap one week out, the rational move is to accept it as a flag-and-skip on the day rather than to chase it.

How do I structure a 14-day SAT math countdown?

Front-load diagnosis and repair, draw a hard line against new content at the one-week mark, then taper. The first day is a full timed assessment whose real output is an error map sorting misses by tier and failure mode. The next days repair the highest-frequency leaks first, then the growth-band topics, then rehearse the embedded calculator, confirm the formulas the on-screen sheet omits, and audit your careless-error patterns. A second full assessment around the halfway point confirms what the repair fixed and what still leaks. At seven days out, stop all new content and switch to consolidation: rhythm sets, one timed module simulation, light targeted review, and confirmed logistics. The final three days descend from light review to a brief warm-up to no studying at all, ending with a simple test-morning routine. The structure works because the decisions are made in advance, leaving your judgment free for the items rather than for the schedule.

When should I take my final practice test before the SAT?

Take your last full timed assessment about a week out, not in the final days. The plan places a second full assessment around the seven-to-eight-day mark, both to confirm that your week of repair fixed the leaks it targeted and to produce a short final priority list. A single timed module simulation, not a full assessment, fits around five days out as a pace rehearsal. Taking a full test inside the last seventy-two hours is a mistake, because it taxes you when the taper calls for rest and it tempts you to chase a single noisy number into panic cramming. The closer you get to the date, the less you should test and the more you should rest. If your runway is short, protect the final three days as taper days regardless, and compress the assessments into the earlier part of the window rather than pushing them late.

What should I do the day before the SAT math?

Do not study the quantitative section at all. Not a problem, not a formula, not a flashcard. A last-minute cram cannot add reliable skill this late and can easily disturb the rest you need, so the day before is for recovery and readiness rather than review. Keep the day normal and restful, with ordinary meals, light activity if it relaxes you, and an early enough night to bank real sleep, which is the most performance-relevant variable left in your control. Run through your written morning plan once, confirm the admission ticket, identification, and backup calculator are packed, and then set the exam aside deliberately. Anxiety in the final hours is normal and does not predict a poor result, but feeding it with frantic review makes it worse while a calm evening lets it settle. You spent two weeks building exactly the readiness this day is meant to preserve, and preserving it is the entire assignment.

How do I taper my studying before test day?

Tapering means scaling the work down on purpose across the final stretch so you arrive warm but fresh. Concretely, the last three days descend in steps: a light review of formulas and cures three days out, a brief warm-up of a few clean easy items two days out, and no studying at all the day before. Each step down is deliberate. The logic is that timed performance depends not only on what you know but on the state of the system executing it, and that system degrades with fatigue and recovers with rest. The careless slips that cost the most points are fatigue-sensitive, caught by a rested mind and missed by a tired one, so the taper directly protects your accuracy. It will feel, after months of daily study, uncomfortably like doing nothing, but doing nearly nothing on purpose in these final days is doing exactly the right thing. Trust the descent rather than fighting it.

How many practice tests should I take in the final two weeks?

Two full timed assessments, plus one shorter timed module simulation, is the right volume. Place a full assessment at the very start of the window to build your initial error map, and a second around the one-week mark to confirm what your repair fixed and to set a final priority list. Around five days out, a single thirty-five-minute timed module rehearses pace without the strain of a full test. More than this is counterproductive, because additional full tests in a two-week window eat the energy the taper is meant to preserve and tempt you to chase noisy score swings. Practice numbers wobble for reasons unrelated to ability, so piling on tests to watch the number move is a trap. The assessments exist to diagnose and to rehearse rhythm, not to generate a stream of scores, and once they have done that job, more testing subtracts from your readiness rather than adding to it.

What should test-day morning look like for SAT math?

Aim to arrive warm, calm, fed, and unhurried, running a routine simple enough to execute on autopilot. Wake with enough margin that nothing is rushed and eat a real breakfast with protein and something slow to digest rather than only sugar, because a stable blood-sugar curve sustains attention across two sections far better than a spike that crashes mid-module. Arrive early enough that traffic or parking cannot rattle you, with everything packed the night before. A light cognitive warm-up helps and over-preparation hurts: solving two or three easy items on the way wakes the quantitative part of your mind the way a runner does a few strides before the gun, so the first real item does not catch you cold. Keep the warm-up easy and clean, never hard or new, because its job is to switch the machine on, not to test it. Once the section starts, fall into the rehearsed rhythm: certain points first, flag and return, let the calculator do its fast work, and reread the final line before committing.

How do I use my error analysis in the final review?

Your error analysis is the engine that drives the whole plan, so treat it as a structured map rather than a vague sense of weakness. After each timed assessment, write every miss on its own line and record three things: the topic, the difficulty tier, and the failure mode. The tier tells you how much a fix is worth, since a high-frequency easy miss bleeds points constantly while a rare hard miss may be worth skipping in a two-week window. The failure mode tells you what the fix actually is: a conceptual gap points to formula or topic review, a method gap points to calculator rehearsal, and an execution slip points to your careless-error cures rather than to any study list. Sort the whole pile this way and you get a prioritized to-do list: which fires to fight first, which growth areas to drill, which curiosities to abandon, and for each whether the cure is concept, method, or care. That sorted map, not the raw score, is what your repair days act on.

Should I rest completely the day before the SAT?

Yes, rest from studying the quantitative section completely, while keeping the day otherwise normal. Complete rest from review does not mean lying in bed all day, which can leave you restless and anxious, but it does mean closing the books on the exam entirely: no problems, no formulas, no flashcards. The work that mattered is already done, and the most productive thing you can do now is protect the rest that lets it show. Keep ordinary routines, eat normally, do something light and relaxing, and get to bed early enough to bank real sleep, since sleep is the single most performance-relevant variable still in your control. If nervous energy needs an outlet, a short walk serves you far better than a problem set. The temptation to do “just a little” review is strong and it is wrong, because it cannot add reliable skill this late and can easily disturb the recovery you need for clean execution on the day.

How do I build rhythm with easy problems before the test?

Rhythm days work mixed easy and medium items at a comfortable, unhurried pace, with the goal of flow rather than speed. The point is to keep your reliable skills warm without the strain of full timed pressure, so they stay sharp through the taper, and to rebuild the confidence that a week of error-hunting can erode. Work a steady set, attending to clean execution and to the instinct to clear certain points first before getting stuck on anything hard. Run your careless-error cures in the background, rereading the final line of each question before committing, so the protective habits stay rehearsed. End the session on a run of items you solve cleanly and deliberately, so the last impression of the day is competence rather than struggle. Doing this in the days after your hard stop on new content keeps the machine warm and your relationship with the section calm, which matters because a calm, confident state on the day catches traps that an anxious one misses.

When should I do a timed module simulation?

Place a single timed module simulation around five days out, after your second full assessment and before the deepest part of the taper. It is one thirty-five-minute block under the clock, not a full test, and its purpose is to rehearse pace rather than to learn or to measure. Use it to check that your pacing instinct fires correctly: do you move on from a stuck item before it eats two minutes, do you flag and return cleanly, does the calculator move come automatically when an item calls for it? Because the module routing means first-module accuracy gates your score ceiling, run the simulation with accuracy as the priority and speed as its servant. Keep the analysis afterward short and behavioral, focused on how you ran the clock rather than on which topics you missed, since topic gaps should mostly be closed by this point. The simulation tells you what the final light-review days should reinforce, then the taper takes over.

How do I review formulas efficiently in the last week?

Run a single deliberate pass through every formula and rule the on-screen reference sheet does not provide, confirming instant recall rather than learning anything new. The sheet supplies geometry area and volume formulas and the special-right-triangle ratios, but it omits much of the algebra and statistics you will actually use. Go domain by domain so nothing slips through: in algebra confirm the slope formula, the line forms and their conversions, the quadratic formula, the vertex location, the discriminant, and the exponent rules; in geometry drill the distance and midpoint formulas, the circle equation, the transversal angle rules, the similar-triangle ratios, and the common Pythagorean triples; in statistics confirm the mean, the basic probability ratio, the counting principle, and conditional probability from a two-way table. For each, the standard is recall within a second or two and correct application on one quick example. Anything that fails that standard gets a small flag and a five-minute fix on the spot, since after the hard stop you should not reopen it.

What is the goal of the final Desmos drill before the SAT?

The goal is to make the core calculator moves reflexive, so you use them without hesitation under time pressure rather than reconstructing them mid-module. The embedded graphing tool is the largest source of recoverable speed and accuracy on the digital quantitative section, but only if your hands already know the moves, because a technique you have to think about is a technique you will skip when the clock is loud. Rehearse the moves that recur most: solving an equation by graphing both sides and reading the intersection, finding zeros by reading where a function crosses the axis, fitting a line or curve to a table with a regression, using a slider to test how a parameter changes a graph, and graphing a system to find where two equations meet. Practice by deliberately solving ordinary items the calculator way even when algebra would also work, so the choice is available instantly on the day. By the end, opening a fresh modeling item and setting up the regression should take only a few seconds, hands ahead of thought.

How is the two-week plan different from a longer study schedule?

A longer schedule is built for acquisition, and the two-week plan is built for consolidation and tapering, which inverts almost every property. During the long middle of preparation the goal is learning, which is slow, tolerates mistakes, and rewards volume. The final fortnight makes the goal retrieval under timed pressure, drops the tolerance for new mistakes to near zero, and makes volume actively harmful past a modest ceiling. Carrying the acquisition mindset into the last two weeks is the central error, because it keeps you opening new topics you can only half-learn while your reliable skills go stale. The two-week plan therefore does less, not more: it diagnoses precisely, repairs the highest-value leaks, draws a hard line against new content, and tapers into rest. Where a longer schedule asks how much you can learn, the two-week plan asks how cleanly you can execute what you already know, and that difference reshapes every day of the window.

Why does the taper matter so much for math performance?

Because timed quantitative performance depends not only on what you know but on the state of the system executing that knowledge, and that system recovers with rest and degrades with fatigue. The slips that cost the most points late in preparation are not knowledge gaps but execution errors, and they are fatigue-sensitive: a rested mind rereads the question and catches the wrong-variable trap or the dropped negative, while a tired mind walks into it. So the taper is not a reward for finishing the work, it is the final phase of the work, the part that converts everything you built into something that will actually fire on the day. A student who tapers well arrives with the same knowledge as a student who crammed but with a sharper instrument to use it, and on a timed test of execution that difference shows up as points. Skipping or inverting the taper by studying hardest in the final days is the most common way strong students underperform their real ability.

What if my SAT date does not leave a full fourteen days?

Compress the diagnostic-and-repair phase and protect the taper. The taper is the part of the plan you can least afford to cut, so if your runway is short, keep the final three days light no matter what: a descent from light review, to a brief warm-up, to no studying the day before. Fold the diagnosis and repair into whatever time remains before that: take one timed assessment as early as you can to build your error map, spend your active days on the highest-frequency leaks and your single weakest growth area, do one focused calculator rehearsal and one formula pass, and run your careless-error audit. Skip the second full assessment if you must, and trust trends from your earlier work rather than chasing a fresh number. The principle holds at any length: diagnose what leaks, fix the high-value leaks first, stop new content before the date, and arrive rested. A shorter window simply means tighter triage on the front half, with the taper held sacred on the back half.

Should I focus on Module 1 or Module 2 in my final review?

Focus your final review on the accuracy that gates Module 1, because the adaptive structure makes first-module performance the precondition for reaching the harder, higher-ceiling second module. The early items of a module are where certain points should be banked cleanly, and dropping a high-frequency easy item there can route you away from the top of the scale entirely. So in the last two weeks, rehearse unhurried, error-free execution on the easier and medium content, run your careless cures, and let accuracy beat speed every time on the first module. This does not mean ignoring harder content, but it does mean that a student still dropping an occasional easy percentage or linear item should fix that before chasing exotic difficulty, because the easy miss costs more through the routing than the hard miss gains. The timed module simulation around five days out is the place to rehearse this priority, treating first-module accuracy as the gate and pacing as its servant.

What is the biggest mistake students make in the last two weeks?

Cramming new material, with skipping the taper a close second, and the two often travel together. The cramming mistake feels productive because it looks like effort, but a topic learned inside the final week stays fragile, and fragile knowledge breeds hesitation and false confidence on exactly the hard items where a clean flag-and-skip would have served better. Meanwhile the cramming hours starve your reliable skills of the warmth they need, so the trade is usually negative: a shaky maybe gained, a sure thing let go stale. The taper mistake compounds it, because a student who studies hardest in the final days arrives depleted, and fatigue attacks accuracy through the careless-error channel that costs the most points. The cure for both is the same discipline: draw a hard line against new content at the one-week mark, then taper into rest on purpose. Peaking on a single date is engineered, and the engineering is mostly about knowing when to stop pushing.

Can I really improve my math score in only two weeks?

Yes, but understand what kind of improvement two weeks can buy. You will not build new conceptual range from scratch in this window, because moving a topic from fragile to reliable takes longer than fourteen days. What you can move, and often by a meaningful margin, are the recoverable points: the high-frequency easy items you keep slipping on, the careless patterns you can name and cure, the calculator moves you can make reflexive, and the timing decisions you make badly under pressure. These are usually more numerous than students expect, and they respond fast because the underlying skill already exists and only the execution is leaking. A student who diagnoses precisely, repairs the highest-value leaks first, stops new content before it can do harm, and arrives rested will typically out-perform their recent practice runs, not because they learned more but because they finally executed cleanly what they already knew. The realistic gain comes from consolidation and tapering, not from cramming, which is exactly why the plan does less near the end rather than more.