How to Guess on a Multiple Choice Question and Actually Be Right

There is a specific feeling, about two seconds into a multiple choice question, where you realize you do not know the answer and you are going to have to do something about it anyway. Most people treat what happens next as luck. It is not entirely luck. There is a real amount of skill in guessing, and the gap between people who have it and people who do not is wider than the gap in what they actually know.

This is not about knowing more things. It is about converting the partial, sideways, half-buried knowledge you already have into a correct answer more often than a coin would.

The math is more encouraging than you think

Four options, no information: you are at twenty five percent. Kill one option and you are at thirty three. Kill two and you are at fifty.

Written out like that it seems trivial, but notice the shape of it. The first elimination buys you eight points. The second buys you seventeen. Elimination accelerates, and the last one you manage is always worth more than the first. This matters because people give up on elimination too early. They knock out the obviously silly option, feel like they have done the work, and then pick between three.

Across a fifty question quiz, the difference between guessing blind and eliminating one option every time is about four extra correct answers. That is usually the difference between second and first.

Your gut is often a real memory wearing a disguise

You look at four options and one of them feels right. No reason attached. Just a pull.

That pull is worth taking seriously, because memory has two settings and only one of them comes with a receipt. There is recollection, where you retrieve the fact along with where you learned it, and there is familiarity, where the thing simply registers as previously encountered without any context at all. Familiarity is the sensation of recognizing someone at the supermarket and having no idea how you know them. It is real information. It is just information stripped of its source.

So when an option feels warm and you cannot say why, that is frequently because you have encountered it before, in the right context, and the context did not survive. Trust it more than you trust an option you have reasoned your way toward from general principles. Reasoning about what sounds plausible is how you get talked into the distractor that was designed to sound plausible.

The exception is when the familiar option is a famous name that has nothing to do with the question. If you are asked who discovered something and Einstein is on the list, Einstein feels warm because Einstein always feels warm. That is not familiarity with this question. That is fame.

Read the question before you read the options

This sounds like advice for eleven year olds and it is the single most common failure among adults who are otherwise good at this.

What happens is that your eyes reach the options before your brain has finished processing the stem, and from that moment you are no longer answering the question. You are evaluating four candidates against a half-formed impression of what was asked. Distractors are built precisely to be attractive under that condition. A well-made wrong answer is the correct answer to a slightly different question, and if you never fully loaded the actual question, you cannot tell the difference.

The habit worth building is to read the stem, look away, and say what you think the answer is before your eyes touch the options. Even a vague version helps. If you can produce "some kind of French treaty, seventeen hundreds" from memory before you look, you will spot the correct option instantly and you will be immune to the one that says sixteen hundreds.

Pay particular attention to negation and to qualifiers. Which of these is not, which was the first, which is the largest by area rather than by population. Those words carry the entire question and they are exactly the words that get skipped when you are reading fast. A question you got wrong because you missed a "not" is not a gap in your knowledge, and it counts the same as one that is.

Tells in a badly written question

Question writing is hard and a lot of it is done quickly. When it is done quickly, the correct answer leaks. There is a whole research literature on this, going back to the 1960s, filed under the wonderful term test-wiseness, and the findings have been stable for sixty years.

The long one. Writers make the correct answer precise, because they want it to be defensible. Precision takes words. The three wrong ones get written fast and stay short. If one option is noticeably longer or more qualified than the others, it is worth a second look.

Absolutes. Options containing always, never, all, none, or only are wrong more often than they are right, because reality has exceptions and question writers know it. An option hedged with usually or generally is a safer bet than one that claims totality.

Grammar that does not fit. If the stem ends in "an" and only one option starts with a vowel, the writer has told you. Same with singular and plural agreement, and with tense. This is pure carelessness and it happens constantly.

The odd one out, inverted. If two options are near-identical and differ in one detail, the answer is usually one of those two. Writers build a distractor by taking the real answer and changing something. The two options that are clearly unrelated to everything else are frequently filler.

All of the above. In lazily written tests it is right more often than chance. It also has a useful property: if you are confident that two of the individual options are correct, it is the answer, and you do not need to evaluate the third.

Now the warning

Every tell above is a symptom of a writer who was not paying attention. Against a writer who is, they invert.

Anyone who has written questions seriously knows this list, and the good ones deliberately break it: they pad a wrong option to be the longest, they put the absolute in the correct answer where it happens to be true, they check that the article agrees with all four options. We have written about this from the other side in how to write trivia questions, and the core discipline is that the difficulty has to live in the wrong answers rather than in the right one.

So treat the tells as a tiebreaker, never as a strategy. If you have genuinely narrowed it to two and have nothing left, let the tell decide. If you are using tells to skip past thinking about the actual question, you will get taken apart by anyone competent.

Change your answer

Almost everyone believes that your first instinct is right and that second-guessing is how you talk yourself out of a correct answer. Almost everyone is wrong, and this is one of the better documented findings in the whole area.

When people change answers on a multiple choice test, the changes run roughly two to one from wrong to right. The belief persists anyway, and the reason it persists is nicely human: changing from right to wrong is memorable and stings, while changing from wrong to right feels like nothing at all. You remember the times it cost you. You do not encode the times it saved you.

The practical version is narrow, so do not overapply it. If you come back to a question and something specific has surfaced, a reason, a detail, a reconsidered reading of what was actually being asked, change it. If you are just anxious and circling, leave it alone. The finding is about new information, not about doubt.

When there is a clock

Timed formats change the calculation completely, and this is where most people bleed points without noticing.

If a question is scored the same whether you take four seconds or fourteen, and there are more questions waiting, then time spent agonizing over one you do not know is time stolen from ones you do. Set yourself a rule and follow it mechanically: eliminate what you can in the first few seconds, pick the best remaining option, move. An unanswered question scores zero with certainty, and a fast guess scores twenty five percent, which is infinitely better.

The harder discipline is walking away from a question you almost have. That feeling of being close is often accurate and it is still usually not worth the twenty seconds, because the next three questions are unread. Trivia is a rate game more than people treat it as one.

What none of this fixes

Guessing well raises the floor. It does not raise the ceiling. Against a field that knows the material, a great guesser loses to a mediocre one who happens to know forty more facts, every single time.

The thing that actually moves you is exposure, repeated, to questions you have not seen before, with the answer revealed immediately after you commit. That is retrieval practice, and it is a much better use of your time than memorizing lists, for reasons we went into in why you blank on answers you definitely know. If you want the broader version, nine habits that actually stick covers the parts that are not about guessing at all.

And it is worth knowing which of your instincts are actively lying to you. A surprising number of the things that feel most certain are the ones that turn out to be wrong at trivia night, and confident wrongness is not something better guessing can save you from.

The short version

Eliminate ruthlessly, past the point where it feels like you have done enough. Trust a warm feeling over a plausible argument, unless the warmth is just fame. Use the writing tells only to break a tie you cannot break otherwise. Change your answer when you have a reason and not when you are merely nervous. On a clock, pick and move.

None of that will make you good at trivia. All of it will make you better at trivia than you currently are, which is a different and much faster thing to achieve.

← Back to the blog