How to Use Mock Tests for UPSC Prelims Preparation: Elimination, Confidence and Half-Knowledge

The short answer: use UPSC mock tests to measure and improve your judgement, not just your knowledge. Tag every answer as sure, eliminated or guess while you attempt it. After scoring, find your half-knowledge errors, compute your net marks by tag and set an attempt policy from your own numbers. Treat CSAT papers as insurance for the qualifying line.

Why UPSC mocks need a different method

In Prelims General Studies, the options are written to be close, and each wrong answer costs one-third of a question's marks. A candidate who knows 70% of the syllabus can still score badly by answering too many questions on partial knowledge, or by leaving many questions that could have been answered by elimination. The score therefore depends on two things that reading cannot train: how well you can tell what you know from what you only half-know, and how well you use that knowledge when you attempt. A mock is the only place to practise both. The pattern is described in the UPSC Prelims exam pattern guide.

Step 1: tag every answer while you attempt

Next to each question, write a code: S if you are sure, E if you have eliminated at least one option and are choosing among the rest, and G if you are guessing. It takes a second and generates the most valuable data a mock can give. A "sure" tag should mean that you could explain the answer to someone else.

Step 2: build the tag table

After scoring, count how many answers in each tag were right and wrong. An illustrative table from one paper (2 marks for a right answer, about 0.67 for a wrong one):

The reading: sure answers are 90% accurate and worth 82.7 marks. Eliminated answers are 53% accurate and add 22.6 marks net: elimination pays, and this candidate should keep doing it. Pure guesses are 17% accurate and lose 2.7 marks: below the 25% break-even, they should stop. The decision: keep answering after elimination; stop guessing blind. A candidate who has no tag data would apply a slogan ("never guess" or "always guess"), and either slogan would cost marks for this candidate.

Step 3: hunt half-knowledge errors

A half-knowledge error is a wrong answer where you knew part of the answer: two of four statements right, or the right concept applied to the wrong case. These are the costliest errors, because the confidence they produce makes you attempt, and the penalty applies. For each one:

1. Go to the source (a textbook, a government document, an official website) and read the whole passage, not just the line that settles the question.
2. Write, in one line, the rule or fact that you had wrong.
3. Note what made you confident: a similar term, an old note, a half-remembered line.

After a month you will have a personal list of 50 to 100 half-knowledge traps. Review it weekly with the schedule in spaced repetition, and you will find that the same traps stop appearing.

Step 4: classify other errors

• Did not know: read the source and add to notes.
• Misread: underline "not", "only", "always", "incorrect" before reading options.
• Over-elimination: you eliminated a correct option on a false belief; correct the belief at its source.
• Time pressure: practise the three-round plan so that you do not rush the last 15 questions.

Step 5: set your attempt policy from your data

Do not borrow a policy. After five full papers, add up your tag tables. If E-tag accuracy is comfortably above 25% and G-tag accuracy is below, your policy is: answer S and E, skip G. If your E-tag accuracy is near 25%, tighten: answer only when two options are gone. If it is high, as in the example, you can be bolder. Repeat the check every five papers, because your skill changes as you prepare. The arithmetic is explained in our negative marking guide and accuracy versus attempts.

Step 6: use the three-round plan

Within the two hours for 100 questions, work in rounds. Round 1 (about 60 minutes): answer the sure questions and mark the rest E or N. Round 2 (about 35 to 40 minutes): reason through the E questions. Round 3: apply your policy to what is left. Keep ten minutes for transfer and checking. Practise this in every full-length mock; it is the habit that most often changes a candidate's score without any extra reading. The GS approach is on the GS Prelims page.

CSAT: use papers as insurance

Paper II is a qualifying paper with a minimum of 33%, or 66 marks of 200. Treat it as insurance. Take a full CSAT paper every fortnight and two short sets a week, and aim for a margin: scores of 90 to 100 in mocks leave room for a bad day. If you score near the line, increase the practice before it becomes an emergency. Apply confidence tagging here too, since each wrong answer costs about 0.83 marks. See the CSAT page and a daily reading comprehension routine.

A six-week cycle for the full-test phase

Mistakes to avoid

• Looking only at the total score. The tag table is where decisions come from.
• Reading a solution and moving on. Go to the source for half-knowledge errors.
• Using a slogan policy. Build it from your own data.
• Skipping CSAT practice until a poor score forces it.
• Taking too many papers in the last fortnight, leaving no time to revise.

A checklist for each mock

1. Tag answers as S, E or G as you attempt.
2. Within an hour, note how the paper felt and where you lost time.
3. Build the tag table and compute net marks by tag.
4. Review half-knowledge errors against the source first.
5. Classify other errors and update the error log.
6. Choose one change for the next paper.

To begin, try a free UPSC mock test for a first tag table, read the UPSC mock test strategy for the stage-by-stage plan and browse the UPSC hub for tests. Please confirm the exam rules in the UPSC notice; tests on TestMint are independent practice papers.