Skip to content

SAT · technique

How to guess well

On this page · 5 sections

Guessing well means turning questions you can't fully solve into points anyway. On the digital SAT this is a real skill, because the scoring rewards it. There is no penalty for a wrong answer. A blank and a wrong answer both earn zero, so any blank is a point you gave away for free.

The test has 98 questions in 134 minutes. Reading and Writing comes first: two modules of 27 questions, 32 minutes each. Math comes second: two modules of 22 questions, 35 minutes each. Most questions have four choices, so a blind guess is right 25% of the time. Cross out two wrong choices and your odds jump to 50%. Cross out three and you have the answer. About a quarter of the Math questions are student-produced responses with no choices at all, which need a different kind of guess.

The test is adaptive by module. How you do on the first module of a section decides whether your second module is harder or easier, and the harder route opens up the higher scores. That changes how you should guess. In module 1, careful work on the questions you can solve matters more than speed, because sloppy errors on easy questions can route you down. Each module also hides a few unscored trial questions you cannot identify. So you can never tell whether a strange question even counts, which is another reason to answer it and move on.

Good guessing is not luck. It is a fixed procedure: budget your time, eliminate choices by rule, make a decision, and never leave a box empty when the clock runs out.

The steps

  1. 1

    1. Know your per-question budget before you start

    Reading and Writing gives you 32 minutes for 27 questions, which is about 70 seconds each. Math gives you 35 minutes for 22 questions, which is about 95 seconds each. These are averages, not limits, because many questions take far less. Use them as a bail-out signal. If you pass roughly double the budget on one question (about 2 minutes in R&W, about 3 in Math) and still have no clear path, stop. That question is now a guessing question, not a solving question.

  2. 2

    2. Eliminate by rule, using the strike-through tool

    Bluebook has an answer eliminator that lets you cross out choices on screen. Use it every time, because it stops you from rereading choices you already rejected. In Reading and Writing, cross out a choice for a specific reason. Common reasons are that it goes beyond the text, it is true but answers a different question, only half of it fits, or it uses words that are too strong, like "always" or "proves". In grammar questions, look at what changes between the choices, since that tells you the rule being tested. If two choices are the same in meaning and differ only in a way that doesn't matter, neither can be the single right answer.

  3. 3

    3. In Math, attack from the answers

    When you can't set up the algebra, the four choices are a shortcut. Plug each one back into the equation or condition. Start with one of the two middle choices (SAT numeric choices are usually listed in order), so you learn whether to go higher or lower. Use the built-in Desmos calculator to graph both sides of an equation and read off where they meet. Estimate with figures, because the Math directions say figures are drawn to scale unless the question states otherwise. Also check signs and size. If a choice would make a length negative or a probability greater than 1, cross it out without solving anything.

  4. 4

    4. Down to two? Test both against the exact task

    Many points are lost here, with two choices left that both feel right. Reread the question stem, not the passage, and name exactly what it asks for. For example, it might ask what would weaken a claim, what logically completes the text, or what the value of y is rather than x. Then ask which choice does the whole job. The wrong survivor usually does part of it, or does a different job very well. Pick one, mark the question for review if time allows, and move on. Going back and forth between two choices for another minute rarely changes the result.

  5. 5

    5. Mark, move, and return

    Within a module you can jump back to any question. Use the "Mark for Review" flag on anything you guessed with doubt or skipped. Always put in a provisional answer before you leave, even if you plan to return. Then if time runs out, your best guess is already there. On the second pass, start with the flagged questions where you got down to two, because those have the best payoff for the time.

  6. 6

    6. Run a final sweep before time expires

    Watch the timer. With about two minutes left, check the question navigator for any unanswered question and fill each one in. For multiple-choice questions you never read, pick one letter and use it for every blind guess, which saves you deciding. For a student-produced response, enter a reasonable number, such as a value you estimated or a number from the setup, since an empty box can't score. Once a module closes you cannot return to it, so this sweep has to happen inside each module.

Worked through

Here is one Reading and Writing question, worked from first read to final click. This is an original practice item.

Text: Ecologist Mara Ostrander studied crows living in cities. She proposed that crows in noisy neighborhoods rely more on visual signals, and less on calls, to warn each other about predators.

Question: Which finding, if true, would most directly support Ostrander's hypothesis?

A) Crows in quiet and noisy neighborhoods made warning calls at similar volumes. B) When shown a model hawk, crows in noisy neighborhoods flicked their wings more often and called less often than crows in quiet neighborhoods did. C) Crows in noisy neighborhoods lived in larger groups than crows in quiet neighborhoods. D) Crows in every neighborhood stopped calling once the model hawk was removed.

Step 1: name the task. The stem asks for support of a hypothesis. So first break the hypothesis into its parts. It compares two places, noisy and quiet. It also makes a two-sided claim: noisy-area crows use more visual signals and fewer calls.

Step 2: eliminate by rule. Choice C is about group size, which the hypothesis never mentions, so it is off-task. Cross it out. Choice D describes all crows the same way, with no noisy-versus-quiet comparison, so it can't support a claim about a difference. Cross it out.

Step 3: compare the two survivors. Suppose you are torn between A and B. Choice A is about calls, which sounds related. But it only mentions call volume, and says nothing about visual signals. Similar volume in both places doesn't even show that noisy crows call less. It does half the job at best. Choice B covers both halves: more wing-flicking, which is a visual signal, and fewer calls, measured against quiet-area crows. It does the whole job, so B is the answer.

Step 4: decide and move. If you reached step 3 after about a minute and still felt unsure, you would still pick B. You got there by a rule, not a hunch, and choosing between two left you at 50% odds instead of 25%. Mark the question for review and go on to the next one.

Now a quick Math version of the same habit. Say a question asks for k if f(x) = x² − 6x + k has exactly one x-intercept, with choices −9, 0, 6 and 9. You don't remember the discriminant rule, so plug in the choices, starting with a middle one. With k = 6, graphing x² − 6x + 6 in Desmos shows two crossings. A bigger k lifts the parabola, so try a higher choice. With k = 9, the function becomes (x − 3)², which touches the x-axis only at x = 3. The answer is 9, and you never needed the formula.

Common mistakes

  • Leaving questions blank. There is no penalty for wrong answers, so a blank scores the same as a wrong answer. Every blank multiple-choice question throws away a free 25% chance or better. A blank student-produced response still gives up a free chance at the point, though a guess there is less likely to be right.

  • Rushing module 1 to save time. Module 1 results decide whether your second module is the harder one, which is the route to high scores, so careless errors on easy early questions cost more than one point.

  • Crossing out choices because they "sound wrong" instead of for a named reason. Then you end up eliminating the right answer, often because it uses plain wording while a wrong choice sounds more impressive.

  • Picking a choice that does part of the job. The most common trap is a choice that matches the topic, or half of a two-part claim, but not the exact task the stem names.

  • Spending three or four minutes stuck between two choices. That time could solve two easier questions, while more staring rarely changes your pick.

  • Skipping a student-produced response without entering anything, or letting a module time out with unanswered questions. Once the module closes, those questions are gone for good.

Checklist

  • I know my average budget: about 70 seconds per Reading and Writing question and about 95 seconds per Math question.
  • I have a set bail-out point, and when I hit it I switch from solving to eliminating and guessing.
  • I can give a specific reason for every choice I cross out, such as off-task, too strong, half-right, or not in the text.
  • When I'm down to two choices, I reread the question stem and pick the choice that does the whole task.
  • I can backsolve a Math question by plugging in answer choices, and I use Desmos to check equations by graphing them.
  • I put a provisional answer on every question before I mark it and move on.
  • I check the question navigator in the last two minutes of every module and fill in every blank, including student-produced responses.
  • I take module 1 carefully, because it decides which second module I get.

More SAT technique