SSC CHSL cut-offs and normalisation, explained

Suppose two candidates each work out a raw Tier 1 score of 150 from the answer key. When results come, one sees a normalised score a few marks lower and the other a few marks higher. Nothing has gone wrong. They sat different shifts, and SSC adjusted every score so that the difficulty of each paper is taken into account.

Normalisation and cut-offs cause more anxiety than almost anything else in the CHSL cycle, mostly because they are misunderstood. Once you see how they work, two things become clear: you cannot control them, and you don't need to. What you can control is a target score, and this guide ends with a method for setting one.

Why normalisation is needed

SSC CHSL runs across many shifts on different days, and each shift has its own question paper. However carefully the papers are set, some turn out easier than others. If raw scores were compared directly, a candidate's rank would depend partly on luck: which shift they happened to be given.

Normalisation removes most of that luck. It converts raw scores from every shift onto a common scale before candidates are ranked.

How it works, in plain terms

SSC has described its method as percentile-based. The idea runs like this:

1. Within each shift, every candidate gets a percentile: the percentage of candidates in that shift who scored at or below them.
2. Each shift is large, and candidates are spread across shifts without regard to ability. So SSC assumes the spread of ability is similar in every shift.
3. If that holds, a candidate at the 90th percentile of a hard shift is as able as one at the 90th percentile of an easy shift, even if their raw scores differ.
4. Raw scores are then mapped to normalised scores so that equal percentiles get equal normalised marks.

An illustration (numbers invented for the example). In Shift A, a harder paper, the 90th-percentile candidate scored 138 raw. In Shift B, an easier paper, the 90th-percentile candidate scored 152 raw. Normalisation puts both on the same normalised score, which will lie between the two. The Shift A candidate gains marks and the Shift B candidate loses some. Neither did anything different. The adjustment simply corrects for the paper each one got.

The cut-offs you will meet

Tier 1 is qualifying. SSC shortlists candidates for Tier 2 category by category, usually in some multiple of the vacancies. The Tier 1 cut-off is simply the normalised score of the last candidate shortlisted in each category.

The final cut-off works differently. Final merit comes from Tier 2 Sections I and II (mathematics and reasoning, then English and GA). Posts are then filled in rank order according to each candidate's preferences, eligibility (including passing the right skill test) and the vacancies in each category. That is why the final cut-off for DEO posts can differ from that for LDC/JSA posts, and why it differs by category. Our guide to post preferences explains allocation in detail.

Reading past cut-offs sensibly

SSC publishes cut-off marks with its results. They are useful, but only if you read them carefully:

• Compare like with like. Only cut-offs from the same pattern mean anything. A cut-off from a year with a different Tier 2 or a different marking scheme tells you little.
• Read them as a range, not a number. Cut-offs move with vacancies, paper difficulty and the number of candidates. A year with many more vacancies usually has a lower cut-off, and the reverse.
• Use your own category and your target post. A general figure for the whole exam is not your cut-off.
• Remember that Tier 1 is only a gate. Scraping past the Tier 1 cut-off still leaves you to build rank in Tier 2.

Setting your own target

Rather than chasing one number, turn past cut-offs into section targets.

1. Collect the official cut-offs for your category over the last few cycles of the current pattern.
2. Take the highest of them and add a margin. That is your comfortable target.
3. Split it across sections according to your strengths.
4. Translate each section target into attempts and accuracy, because that is what you actually control on the day.
5. Check your mocks against it. You want to be above the target consistently, not once.

Why accuracy matters more than attempts

Tier 1 gives +2 for a right answer and −0.50 for a wrong one. Compare two candidates:

Candidate B attempted ten more questions and got two more right, yet finished level. The two extra right answers earned 4 marks, and the eight extra wrong answers took exactly 4 marks away. All that extra time bought nothing. In Tier 2, with +3 and −1, the penalty bites harder:

Same knowledge, six marks apart. The candidate who stopped at 24 lost nothing by leaving six questions they couldn't solve.

What you can control

Practice

1. Why does SSC normalise CHSL scores?
2. Your shift was harder than most. What is likely to happen to your normalised score compared with your raw score?
3. Which parts of Tier 2 make up the merit score, and out of how many marks?
4. In Tier 1, a candidate gets 70 right and 10 wrong. What is the raw score?
5. In a Tier 2 module, a candidate gets 20 right and 6 wrong. What is the score?
6. Why can two candidates with the same final score end up in different posts?

Answers:

1. Different shifts have papers of different difficulty, and normalisation puts all shifts on a common scale.
2. It is likely to be higher than the raw score.
3. Session I Sections I and II: mathematics, reasoning, English and GA, out of 360.
4. 70 × 2 − 10 × 0.5 = 140 − 5 = 135.
5. 20 × 3 − 6 × 1 = 60 − 6 = 54.
6. Allocation also depends on their preferences, eligibility (including the skill test) and the vacancies left when their rank is reached.

What to do next

• Download the official cut-off notices for your category from the last few cycles of the current pattern.
• Write your comfortable target and split it into section targets.
• Convert each section target into a number of attempts at your mock accuracy.
• Track your last five mocks against the target, not against friends. The Tier 2 strategy guide shows how to plan attempts section by section.