Why You Get UPSC Prelims Questions Wrong on Topics You Actually Knew

Question 47. You read it and you knew the topic — you’d revised it three weeks ago. Two options looked plausible. You read the stem again. You marked B, moved on, came back later, changed it to C, and moved on again.

It was B.

That evening the review shows a red mark against question 47 and the correct option beside it. You read the explanation, recognise every part of it, and think: I knew that. Then you go to question 48.

What the review doesn’t show you is that question 47 took nearly four minutes across two visits — and that near the end of the paper there were two questions you never properly read.

Your score records which answers were wrong. It doesn't record which questions were expensive — and some of those expensive questions cost you twice.

01 · The half-review

You review the outcome. You never review the cost.

Watch how a Prelims mock actually gets reviewed. You check the score. You open the solutions. You go through the questions marked wrong, read each explanation, and sort them into rough categories in your head — didn’t know this one, misread that one, silly mistake on this one. Perhaps you note the topics down for revision.

Every part of that is about outcome. Right or wrong, and why.

None of it is about cost. Not one step in that review asks how long any question took, or which questions absorbed time out of proportion to what they returned. And in an exam where the constraint is a fixed hundred and twenty minutes, cost is not a footnote to performance. It’s half of it.

This is why the topic you knew can still show up in the wrong column. Knowing something and executing on it inside a time budget are two different variables, and most mock reviews barely capture the second.

02 · Two dimensions

Every question has two results, not one

Every question you attempt produces two outcomes at the same time.

The first is the one you already track: you got it right, you got it wrong, or you left it.

The second is the one nothing in your review records: what it cost you. Twenty seconds, or ninety, or four minutes spread across three visits with an answer change in the middle.

Those two results are independent. A question can be right and cheap, right and expensive, wrong and cheap, or wrong and expensive. Your score sheet collapses all four into two columns, which means two very different questions — one you got wrong in fifteen seconds, one you got wrong in three and a half minutes — look identical in your review and get the same amount of your attention afterwards.

They are not the same question. One cost you a mark. The other cost you a mark and a chunk of the paper.

03 · The budget

120 minutes is a budget, not a rate

The arithmetic is simple enough. A hundred questions, a hundred and twenty minutes: about seventy-two seconds a question on average.

That number is useful and it is also easy to misuse. The moment most aspirants see it, they start trying to force every question into seventy-two seconds — which is the wrong lesson, and a stressful way to sit an exam.

Seventy-two seconds is a budget, not a rate. It’s the average you get if you spend evenly, and spending evenly is not the goal. Some questions deserve twenty seconds, because you either know them or you don’t and staring won’t change that. Some deserve ninety, because there’s a chain of reasoning to work through and it will get you there. Some genuinely deserve a hundred and fifty.

The question is never “am I inside seventy-two seconds?” It’s “is the time I’m spending here buying me anything?” Two minutes on a hard question you then answer correctly is two minutes well spent. Two minutes of re-reading the same stem while nothing new arrives is two minutes converted into nothing.

04 · Four kinds of question

The four things that can happen on a question

Put the two dimensions together and every question you attempt lands in one of four places.

Fast + right

Clean execution. You knew it, you recognised it, you moved. These questions are also what funds the rest of the paper — every one of them returns time to the budget.

Slow + right

Expensive, but productive. You worked and the work paid. Worth watching rather than fixing — a paper with too many of these can run out of time, but each one individually was a reasonable trade.

Fast + wrong

A knowledge gap, or a misread stem. Cheap, in exam terms — it cost you the mark and almost nothing else. This is the quadrant your existing revision process is already built to handle.

Slow + wrong

The most expensive combination. You lost the mark and you spent time that could have gone somewhere it would have returned something. Question 47 lives here — and it's invisible in a review that only sorts by right and wrong.

Notice that the fix is different in each quadrant. Fast-and-wrong is a content problem — revise the topic. Slow-and-wrong is often more than a content problem. It’s a decision problem: you stayed in a question that had stopped giving you anything.

Reviewing your mock by outcome alone means treating both as the same thing, and reaching for the same remedy — read more — when only one of them is short of reading.

05 · First answer, final answer

Is your first instinct right? That’s a question about you.

Every aspirant has been told to trust their first instinct. It’s one of the most repeated pieces of advice in Prelims preparation, and few aspirants actually check whether it’s true of them.

It’s checkable. On any question where you changed your answer, four things can have happened:

First answer → final answer → result

First answer right → you changed it → wrong. The change cost you.
First answer wrong → you changed it → right. The change saved you.
First answer wrong → you changed it → still wrong. The time went, the mark was never there.
First answer right → you left it → right. Nothing to learn, and worth counting anyway.

Count those across several mocks and you have something the advice can’t give you: a pattern that is actually about your own attempts. For some people, second-guessing reliably destroys marks. For others, the revisit is where they catch their own misreadings, and changing answers is a net gain.

Whether your first instinct is right is a fact about you, not a rule about everyone.

The point isn’t which conclusion you reach. It’s that you look at your own record before adopting somebody else’s rule about how your mind behaves under pressure.

06 · The move-on decision

Moving on is time allocation, not surrender

The reason aspirants stay too long in a question is rarely miscalculation. It’s that skipping feels like losing. You’ve read it, you’ve half-worked it, you can nearly see it — walking away now feels like wasting the ninety seconds already spent rather than protecting the minutes still left.

That instinct is understandable and it is backwards. The ninety seconds are gone either way. The only thing still in your control is what happens to the next ninety.

Know · Work · Move

Know it? Answer it and go.
Can you see a short chain of reasoning to the answer? Work it.
Still circling with no new information arriving? Move on.

The third line is the one that matters, and notice what it’s keyed to. Not a stopwatch — a condition. Is anything new still arriving? When you’ve read the stem twice and the options twice and nothing has shifted, further reading is unlikely to shift it either.

Skipping isn’t abandoning the question. You’re protecting the remaining budget and keeping the option to return if time allows. And with a wrong answer carrying a penalty, the option to come back later with a clearer head is worth more than the option to force it now.

07 · The counter-question

“Doesn’t this turn Prelims into a stopwatch exercise?”

No. The timer is not the objective. It’s evidence.

Speed by itself doesn’t clear Prelims. People clear it by getting enough questions right, and speed matters only because time is the resource those right answers are bought with. If you walk into the hall watching a clock instead of reading the stem, you’ve taken a diagnostic tool and turned it into a distraction.

A question that took two minutes isn’t automatically a failure. If it was genuinely difficult and you got it right, those two minutes may have been the best spend of the paper. The signal worth acting on isn’t any single slow question — it’s the repeated pattern of spending disproportionate time without a corresponding return.

That pattern is invisible in a single paper and much clearer across several. Which is the argument for looking at the cost dimension at all: not to race, but to find out where your time has been going and whether it has been buying anything.

08 · Where Swādhyāya fits

The second layer

Everything above can be done with a stopwatch and a notebook, and if you’re willing to maintain that, it works. The difficulty is that recording per-question time by hand during a timed mock is precisely the kind of overhead that gets abandoned by the third paper. So the cost dimension usually goes unrecorded — not because aspirants don’t want it, but because capturing it manually competes with the exam itself.

Swādhyāya records it while you attempt, so the second layer is there when you review:

Know which questions were expensive. Time per question, accumulated across every revisit rather than just the last one, with the questions that ran long flagged and the number of times you returned to each.
Know your own instinct. Your first answer is kept alongside your final one, so the four outcomes above stop being folklore and become a count of your own attempts.
Know where your time starts costing you. See how your accuracy changes with the time you spend, so you can develop a move-on threshold from your own attempts rather than someone else’s advice.

None of this replaces knowing the syllabus. It tells you something the score alone can’t: not just which questions you got wrong, but which ones you paid too much for.

The goal isn’t to answer every question quickly. It’s to spend your limited time where it has the best chance of producing marks.

One decision at a time.