Some time after the Tier-1 window closes on 30 October 2026, the Commission will put up a tentative answer key together with your individual response sheet, and for a few days you will be able to see exactly what you marked against exactly what was correct. That is the moment this page is written for - and also, honestly, the moment many candidates get the arithmetic wrong and spend a fortnight believing a number that was never their number. The formula itself is simple: two marks for each correct answer, half a mark deducted for each wrong one, nothing for a question left blank. The complications come afterwards. Your raw score is not your final score, because the Commission normalises marks across shifts before anything is decided. And a raw score only means something when you hold it against what cut-offs have actually done in past cycles, which is a wider band than most people assume. This page gives you the calculation, a worked example, and an honest way to interpret the answer. It does not guess at this year's cut-off, and you should be suspicious of anything that does.
Maximum marks are 200 across 100 questions. Each correct answer is worth +2 and each wrong answer costs -0.5; a question you did not attempt neither earns nor costs anything. So your raw score is simply (correct x 2) minus (wrong x 0.5). Worked example: you attempted 82 questions, of which 64 were correct and 18 wrong. That is (64 x 2) = 128, minus (18 x 0.5) = 9, giving a raw score of 119. Note that with the 15-minute sectional locks introduced in the 2026 cycle, questions left behind when a section closes count as unattempted and cost you nothing - they simply earn nothing either. The raw figure you compute is not the mark that appears in the result: SSC normalises marks across shifts, and the scorecard shows both. Always verify the marking scheme printed in this cycle's notice on ssc.gov.in.
You do not have to wait until November to learn what your number looks like. Our free full mocks are 100 questions in four sections of 25 with the 2026 sectional locks, scored at +2 and -0.5 exactly as Tier-1 is, so the score you see is computed on the Commission's own arithmetic. Candidates who have watched that number move across several mocks read the official answer key far more calmly than those seeing their first real score on results week.
A raw mark on its own is meaningless until you know where it sits. After a mock, the AI estimates your All-India rank against real 2020-24 cut-off data and tells you what it is comparing you against. It is an estimate and the app says so plainly - cut-offs move with vacancies, applicants and paper difficulty - but it is anchored in published figures rather than in a number someone posted on a forum.
Negative marking hurts most when the wrong answers were avoidable. After each mock the AI sorts your mistakes into conceptual, calculation, careless and time-pressure errors, so you can see how many marks went to genuine gaps and how many to misreading the question. In the fortnight before the exam, that split tells you what is actually worth fixing and what is simply a matter of slowing down by three seconds.
Our bank holds around 31,000 questions, of which roughly 29,000 are genuine previous-year SSC questions carried with the Commission's official answer keys. That matters for score calculation specifically: you learn the Commission's conventions for what it treats as correct, which is exactly the judgement you will need when you read the tentative key and decide whether a question is worth challenging.
Raw score = (number of correct answers x 2) - (number of wrong answers x 0.5). Questions you did not attempt do not enter the calculation at all, so leaving a question blank is neutral rather than harmful. The maximum is 200 across 100 questions. Count from your official response sheet rather than from memory, since the sheet records exactly which option was saved for each question.
You can try, but the result is usually wrong and always unsettling. Candidates consistently misremember how many questions they attempted, and in a 60-minute paper with four locked sections the tail end of each section is precisely where recall is weakest. A rough self-estimate right after the exam typically lands several marks either side of the real figure, which is enough to move you from calm to panicked for no good reason. Wait for the response sheet, which is exact, and download it the day the login window opens.
No, and this is the most common misunderstanding. Because Tier-1 runs across many shifts over a month-long window, SSC normalises marks so candidates who sat different papers can be compared fairly - and since its notice dated 2 June 2025 it does this by the equipercentile method, which works from your percentile rank within your own shift. Your normalised mark can therefore land above or below the raw number you calculated. The official scorecard published on ssc.gov.in shows the marks the Commission is actually using.
Across the recent cycles for which SSC has published Tier-1 results, the unreserved cut-off for the general pool of posts has moved within a band of roughly 114 to 153 out of 200, and it has shifted by more than fifteen marks between consecutive cycles more than once. OBC cut-offs have typically sat a few marks below UR, with SC and ST lower again, and the posts carrying an extra Tier-2 paper - Assistant Audit Officer and Junior Statistical Officer among them - have had their own, noticeably higher cut-offs, above 165 for UR in at least one cycle. Treat that band as context, not as a target line for 2026: this year has 12,256 vacancies and over 28 lakh applicants, and only the Commission's own result notice will give the real figure. Read the official result PDFs on ssc.gov.in for exact past numbers.
The tentative key is exactly that - tentative. Candidates may challenge questions during the objection window for a per-question fee stated in the answer-key notice, and after reviewing the challenges the Commission publishes a final answer key. Results are computed on that final key, and any revision applies to every candidate who was given that question, not only to those who objected. So if you are confident a key is wrong, raise the objection with a proper justification, but do not assume a single disputed question changes your standing - recompute your score on the final key when it appears, and treat that as your raw figure.
SSCIntel — free AI-powered SSC mock tests