One Question, Two Different Answers
The attempt-count question cannot be answered for "the paper", because the paper is two papers wearing one admit card. Session I — forty questions of Maths and Reasoning — carries no negative marking and counts only as a qualifying gate. Session II — fifty questions of General Awareness and English — deducts one mark per wrong answer and is, after normalisation, your entire merit score. The exam pattern guide lays out the structure; this article draws out the one consequence aspirants ask about most.
Any answer that quotes a single number for the whole exam — and coaching folklore is full of them — has already made the merged-score mistake. The sessions reward opposite instincts, and your attempt strategy has to hold both at once.
Session I: Attempt All Forty, Without Exception
With no penalty for a wrong answer, the arithmetic of Session I is closed: an unattempted question scores zero with certainty, while a blind guess scores three marks a quarter of the time and costs nothing when wrong. There is no accuracy level, no confidence level, no shift difficulty at which leaving a Session I question blank is correct. The answer to "how many should I attempt in Session I?" is forty — asked and answered.
The practical work is making that happen under the clock. The rehearsed shape: clear reasoning first (its families are shallow and fast), give arithmetic the larger remaining share with a hard rule for abandoning any question that runs long, and reserve the final two minutes for the sweep — filling every remaining blank, mechanically, without deliberation. The Session I strategy covers the full routine; the sweep is its non-negotiable line.
Session II: Attempt Count Is an Output, Not a Target
Here the folklore inverts reality. Aspirants set themselves a target — "I will attempt 45 of 50" — and then mark enough guesses to hit it, as if the count itself earned marks. It is exactly backwards: in Session II the attempt count is what falls out of applying a per-question decision rule honestly, and forcing the count higher than the rule allows converts marks into deductions. Two candidates with identical knowledge can differ by ten-plus net marks purely on this discipline.
The decision rule is expected value, and it takes one minute to internalise for life. A question carries three marks; a wrong answer costs one.
- Blind guess, four live options: a quarter chance of +3 against three quarters of −1 — break-even on average, which under exam variance and near a cutoff is a trade not worth making.
- One option eliminated: a third chance of +3 against two thirds of −1 — modestly positive; acceptable when elimination is genuine rather than hopeful.
- Two options eliminated: half of +3 against half of −1 — clearly profitable; this guess should always be taken.
- Sure or near-sure answers: attempt, obviously. The rule exists for everything in between.
Note what the rule does not reference: how many you have attempted so far. Each question is its own trade. The complete treatment, including the psychology of "feeling lucky" late in a paper, is in the negative marking strategy.
Why Chasing a Number Backfires
The deeper reason target-counts fail is that Session II marks are ranked after normalisation, against other candidates — and what a quoted "safe attempts" figure hides is the accuracy attached to it. Forty-five attempts at seventy per cent accuracy nets fewer marks than thirty-eight attempts at ninety per cent, and the difference widens once deductions are counted. A good score in this exam is an accuracy story, not an attempt story, and every previous cycle's selected candidates include modest attempt counts with excellent accuracy.
There is also a clock argument. Question selection — the skill of spending your forty-five minutes on the questions you can actually convert — is worth more than raw attempt volume, and it is precisely what a forced target destroys. The candidate racing to attempt fifty spends time on their worst questions; the candidate applying the rule spends it on their best. The speed and accuracy guide covers how the two trade off in practice.
Building the Discipline in Mocks
The elimination rule is easy to recite and hard to obey at question thirty-one with the clock falling — which is why it must be drilled where failure is free. In every timed paper, apply the rule explicitly: mark guesses in your head as one-elimination or two-elimination as you make them, then audit them in review. Count the blind guesses that slipped through and what they netted after deductions. Most aspirants discover their "instinct" guesses run a loss, and the discovery — made in a mock — is what makes the discipline stick.
Real previous year shifts on the SSC MTS previous year papers page apply the exact split marking — no deduction in Session I, minus one in Session II — so your practice score is the score the real exam would have given you, and your guess audit runs on honest numbers. A mock with flattering marking rehearses precisely the wrong instincts.
Does a Hard Shift Change the Rule?
The commonest mid-exam temptation is recalibrating on perceived difficulty: the shift feels brutal, everyone must be struggling, so surely looser guessing is justified. Resist it, for two reasons. First, your impression of a shift's difficulty is formed from a sample of one nervous candidate — you — and shifts routinely feel harder than they mark. Second, normalisation already does the adjusting: if your shift genuinely ran harder, the normalisation that produces the merit list compensates for it across all candidates, which means the per-question expected-value arithmetic is untouched. A blind guess is the same losing trade on a hard shift as on an easy one; the only thing difficulty changes is how many questions qualify as genuinely attempt-worthy, and the rule handles that by itself.
The same logic disposes of shift-to-shift strategy rumours — the idea that an evening shift wants more attempts, or that a particular day's paper rewarded aggression. Every such claim is reconstructed from anecdote after the fact, and none of it survives contact with how the scoring actually works. The rule is difficulty-proof and rumour-proof, which is precisely what makes it worth drilling: it removes an entire category of exam-hall decisions by having made them in advance.
The Answer, Stated Once
So: how many questions should you attempt in SSC MTS? Session I — all forty, with the final-sweep habit making sure of it. Session II — every question you know, every question with two options eliminated, one-elimination guesses when the elimination is real, and nothing blind; the count that results is your correct count for that shift, whether it lands at thirty-six or forty-six.
If you want a self-test of whether the rule has truly settled, watch your last five minutes of Session II in your next mock. The undisciplined candidate spends them marking hurried guesses on everything still blank, converting a decent paper into a deducted one. The disciplined candidate spends them re-checking marked-for-review questions and leaves the genuine unknowns untouched. Which of the two you are at minute forty-one is not decided at minute forty-one — it was decided in the mocks you took this month.
Aspirants who internalise this stop asking the question, because it stops having a number for an answer. It has a rule — and the rule, unlike a number, works on every shift, at every difficulty, all the way to the merit list.