How to Handle Negative Marking in Government Exams: A Complete Strategy Guide

Negative marking is the single most misunderstood element of government exam strategy. Most candidates either ignore it entirely — attempting every question regardless of confidence — or become so afraid of it that they under-attempt and leave marks on the table unnecessarily. Both extremes cost marks.

The correct approach is neither reckless nor overcautious. It is calculated — based on probability, accuracy data from your mock tests, and a clear decision framework for every question type.

---

Understanding What Negative Marking Actually Means

Before building a strategy, understand the mathematics.

Standard negative marking (SSC, JKSSB, most state PSCs): 0.25 marks deducted per wrong answer.

This means: 4 wrong answers cancel 1 correct answer.

If you attempt a question you are completely uncertain about (random guess, 25% chance of being right):

• Expected marks from 4 attempts = 1 correct × 1 mark = 1 mark
• Expected penalty from 3 wrong = 3 × 0.25 = 0.75 marks
• Net expected gain = 0.25 marks per attempt

Pure random guessing at 0.25 negative marking is marginally positive in expectation — but only in theory. In practice, your accuracy on uncertain questions is usually lower than 25% because you are drawn to wrong options for specific reasons.

SBI PO, IBPS PO Mains negative marking: 0.25 marks per wrong answer — same calculation applies.

RRB NTPC, Group D negative marking: 1/3 mark (0.33) per wrong answer.

This means: 3 wrong answers cancel 1 correct answer.

At 1/3 negative marking, random guessing (25% accuracy) is net negative:

• Expected gain from 4 attempts = 1 mark
• Expected penalty = 3 × 0.33 = 1 mark
• Net = 0 (break even in theory, negative in practice)

SSC CHSL, CGL Tier 1: 0.50 marks per wrong answer.

This means: 2 wrong answers cancel 1 correct answer.

At 0.5 negative marking, random guessing is clearly net negative. Be more selective here than at 0.25 negative marking.

IBPS PO Mains: 0.25 marks per wrong answer — but sectional cutoffs apply. This changes the calculus for borderline sections — sometimes you must attempt to clear the section minimum, even at some risk.

---

The Core Decision Framework

For every question in a government exam, you face one of four situations:

Situation 1 — You Know the Answer with Confidence

Action: Attempt immediately. No calculation needed. Mark the answer and move on. Every second spent second-guessing a question you know is wasted time.

Situation 2 — You Can Eliminate 2 of 4 Options

Action: Attempt. With 2 options remaining, your probability of being right is 50%. At 0.25 negative marking:

• Expected value of attempting = (0.5 × 1) – (0.5 × 0.25) = 0.375 marks This is clearly positive — always attempt when you can eliminate two options.

At 0.5 negative marking:

• Expected value = (0.5 × 1) – (0.5 × 0.5) = 0.25 marks Still positive — attempt.

Situation 3 — You Can Eliminate 1 of 4 Options

Action: Generally attempt at 0.25 negative marking, use judgment at 0.5. With 3 options remaining, your probability is approximately 33%.

At 0.25 negative marking:

• Expected value = (0.33 × 1) – (0.67 × 0.25) = 0.165 marks Positive — worth attempting if you have eliminated one option logically (not randomly).

At 0.5 negative marking:

• Expected value = (0.33 × 1) – (0.67 × 0.5) = -0.005 marks Near zero — use judgment. If the eliminated option was based on genuine knowledge, attempt. If it was a guess, skip.

Situation 4 — Complete Uncertainty (No Elimination Possible)

Action: Skip. At 0.25 negative marking, a pure random guess has marginal positive expected value but in practice your instinct draws you to wrong answers more than probability suggests. The safer policy is to skip completely uncertain questions.

At 0.5 negative marking: always skip if completely uncertain.

At 1/3 negative marking (Railway exams): always skip if completely uncertain.

---

Building Your Personal Accuracy Benchmark

The framework above is based on probability theory. Your actual accuracy on uncertain questions may be higher or lower than theoretical probability. The only way to know is data from your mock tests.

After every mock test, track this:

After 5–10 mock tests, you will have real data on your accuracy at each confidence level. This data, not theoretical probability, should drive your exam strategy.

Common findings from tracking:

• Most candidates have 85–95% accuracy on high-confidence questions
• Accuracy drops to 55–70% when 2 options are eliminated
• Accuracy drops to 35–50% when 1 option is eliminated
• Accuracy is often below 30% on complete guesses (worse than random, because wrong options are attractive for wrong reasons)

Use your actual numbers to calibrate how aggressively to attempt at each confidence level.

---

Subject-wise Negative Marking Strategy

Reasoning

Reasoning questions are either solvable or not — there is rarely a halfway position. If you cannot solve a puzzle or series within 90 seconds, skip it and come back if time allows. Do not attempt reasoning questions you cannot actually solve — guessing on reasoning questions is lower accuracy than guessing on knowledge questions.

Exception: Simple analogies, classifications, and direction questions can often be answered from partial knowledge. Apply the elimination framework here.

Quantitative Aptitude / Mathematics

Two types of Maths questions require different approaches:

Calculation questions (profit/loss, percentages, SI/CI): If you start the calculation and reach a reasonable-looking number, attempt. If your calculation is going wrong or taking too long, skip — a wrong calculation gives a wrong answer with certainty.

Approximation questions: These are designed for quick elimination — use approximation to narrow to one or two options, then apply the framework.

Data Interpretation: Complete the full DI set before marking answers. Often, solving one question within a DI set helps you solve others faster. Never mark a DI answer mid-set and move on.

General Awareness / General Knowledge

This is the section where negative marking strategy differs most from other sections.

GK questions are binary — you either know the answer or you do not. If you know it, mark it immediately. If you partially know it (recognise the correct answer when you see it among the options), you are essentially in Situation 2 — attempt.

For GK questions where you have no relevant knowledge at all: skip. Unlike Reasoning or Maths where you might logically derive the answer, GK without knowledge is pure guessing.

One useful technique: If a GK option seems completely implausible (a date that is clearly wrong, a name associated with a different domain entirely), eliminating it puts you in Situation 3 or 2 — recalculate accordingly.

English

Error spotting and fill-in-the-blanks often allow elimination based on grammatical rules you partially know. Apply the framework — if a rule eliminates even one option, your accuracy improves significantly.

Comprehension questions should almost always be attempted — the passage is there, the answer is derivable, and skipping comprehension questions is leaving easy marks behind.

---

Time Management and Negative Marking

Negative marking strategy and time management are connected.

Do not spend time calculating whether to attempt a borderline question. The calculation takes 15–20 seconds. At that level of uncertainty, the expected value difference between attempting and skipping is less than 0.2 marks. Spending 20 seconds on this decision is almost never worth it.

The better rule: Make your decision policy before the exam, not during it. Decide: “I attempt if I can eliminate at least one option with genuine knowledge.” Apply this consistently in the exam without recalculating each time.

Two-pass strategy for negative marking:

First pass (80% of time): Go through all questions. Answer everything you are confident about. Mark borderline questions (those you can eliminate at least one option on) for review. Skip complete unknowns.

Second pass (20% of time): Return to marked questions. With fresh eyes, decide — eliminate, attempt, or final skip. Often a question that seemed unclear on first pass becomes clearer on second look.

---

Common Negative Marking Mistakes

Mistake 1: Changing correct answers during review. Research consistently shows that first instincts in exams are more often correct than second thoughts. Only change an answer if you have a specific reason — you noticed a word you missed, you recalled a relevant fact. Do not change answers based on vague unease.

Mistake 2: Attempting questions at the end of the exam from panic. With 5 minutes left and 15 unattempted questions, the temptation is to guess everything. Resist this. Random guessing in the final minutes at 0.5 negative marking will cost you more marks than leaving questions blank.

Mistake 3: Not tracking your accuracy in mock tests. Negative marking strategy without data is guesswork. Track your accuracy at each confidence level across at least 5 mock tests before your real exam.

Mistake 4: Applying the same strategy across all exam types. 0.25 negative marking (SSC, JKSSB) allows more liberal attempting than 0.5 (SSC CGL Tier 2) or 1/3 (Railways). Adjust your strategy for each exam’s specific penalty.

Mistake 5: Being too conservative in sections with sectional cutoffs. In exams with sectional cutoffs (IBPS, SBI), under-attempting to avoid negative marking can mean failing the sectional minimum even if your overall score is fine. Monitor your attempted count per section and adjust.

---

Exam-wise Negative Marking Quick Reference

---

Practice Drill — Building the Habit

In your next 5 mock tests, do the following:

1. Before starting, commit to your decision policy (e.g. “attempt if I can eliminate at least 1 option”)
2. During the test, mark each question as: Confident / Eliminated 1 / Eliminated 2 / Skipped
3. After the test, review: how accurate were you in each category?
4. Adjust your policy based on real data

Within 5 mock tests this process becomes automatic — you develop an instinct for when to attempt and when to skip that is calibrated to your actual accuracy rather than theoretical probability.

---

Published by ExamzPrep — free government exam preparation for serious aspirants. Last updated June 2026.