GATE's paper structure is stable and public — yet every year candidates lose real marks to misunderstanding the three question formats and the negative-marking arithmetic. Here is the pattern, precisely, and the attempt policy that falls out of it.
The frame
- 65 questions, 100 marks, 180 minutes, computer-based; you sit one assigned session, though high-volume papers run multiple sessions (which is what normalization then equates).
- General Aptitude — 10 questions, 15 marks (5 one-mark + 5 two-mark): verbal reasoning, quantitative aptitude, basic data interpretation.
- Paper-specific component — 55 questions, 85 marks. CS, EC, EE, ME and CE include a separately identified Engineering Mathematics share within those 85 marks; DA and several other papers place all 85 marks in the subject component without that separate 13-mark label. Use your paper's official syllabus/pattern.
The three question formats
MCQ — one correct option, negative marking applies.
- 1-mark MCQ: wrong answer costs -1/3
- 2-mark MCQ: wrong answer costs -2/3
MSQ — one OR MORE correct options, NO negative marking, NO partial credit. You must select ALL correct options and nothing extra to score. The trap is psychological: no negative marking invites guessing, but the all-or-nothing rule means near-misses score zero. Treat MSQs as precision questions: they test whether you know each statement's truth value independently.
NAT — type the number, NO options, NO negative marking. The answer is a signed real number entered through the virtual numeric keypad; the question tells you the required decimal places. Carry sensible precision through your working and round only at the end. With no options to eliminate, method fluency is everything — NAT strategy and the virtual calculator.
How to treat each format
- MCQ: solve, then decide whether the evidence is strong enough to accept negative-marking risk. Elimination changes the decision; a vague feeling does not.
- MSQ: test every option independently as true or false. Do not stop after finding one correct statement, and do not add an option merely because there is no negative marking.
- NAT: write the target unit and expected sign, complete the method, then enter the signed value at the precision stated in the question. There is no option list to expose an order-of-magnitude mistake.
In review, record format as well as subject. If Thermodynamics MCQs are strong but Thermodynamics NATs are weak, "revise Thermodynamics" is too broad; the actual gap is numerical execution.
The attempt policy the maths dictates
- 1-mark MCQ, two options eliminated: guess. With two options left, EV = 0.5 x (+1) + 0.5 x (-1/3) = +1/3 per guess — positive, so take it.
- 2-mark MCQ, pure blind guess: at four options, EV = 0.25 x (+2) + 0.75 x (-2/3) = 0.5 - 0.5 = exactly 0. Blind guessing 2-mark MCQs is EV-zero at best — with any elimination it turns positive, so eliminate first or leave it.
- MSQ and NAT: no marking penalty for a wrong answer, but time still has value. Attempt them when you can make a considered judgement without sacrificing more solvable questions.
- A NAT blank scores zero; a reasoned estimate can be useful when time permits, but an endless calculation still carries opportunity cost.
A worked attempt decision
Consider a 2-mark MCQ with four choices. A blind choice has zero expected value under the official penalty. If you can eliminate one option, the chance of being correct becomes one in three while the wrong-answer penalty stays two-thirds of a mark; the expected value becomes positive. That does not make random guessing a strategy — it shows why elimination work has measurable value.
For MSQ, the calculation is different because scoring is all-or-nothing and there is no negative mark. Selecting an extra doubtful option can turn an otherwise correct set into zero. The rational approach is not "select more because there is no penalty"; it is "judge every statement, then submit the exact set supported by the concept."
What this means for practice
Your mocks must mirror the real format mix — a practice diet of MCQs only leaves you untrained for exactly the formats (MSQ precision, NAT method-fluency) where GATE hides its discrimination. The full preparation system: GATE 2027 guide; how marks become your final score: normalization explained. Compute your paper from an answer key with the free GATE marks calculator.
After a mock, create a format ledger with four columns: subject, format, failure reason and fix. Typical fixes are concept revision, option-by-option MSQ practice, NAT unit/precision checks, or a stricter MCQ risk rule. Review the ledger before the next mock so the correction is tested, not merely written down.
Quick answers
Where does negative marking apply? Under the official GATE 2027 rules, it applies to wrong MCQ answers. MSQ and NAT questions have no negative marking.
Can an MSQ receive partial marks? The current brochure describes MSQ scoring as all-or-nothing; confirm later official notices if rules change.
Should every NAT be attempted? There is no negative marking, but use remaining time intelligently. A considered estimate may be worthwhile; an endless calculation can still cost other questions.
Why practise with the virtual calculator? NAT performance depends partly on entering a signed, precision-controlled number efficiently. Familiarity reduces avoidable input and transcription errors.
