Cutoff, normalisation and tie-breaking, explained properly

A candidate scores 138 in an exam where the cutoff is announced as 134, and does not make the list. Another scores 131 in the same exam and does. Both results are correct, and neither is an error, because the number that decided them was not the raw score either candidate calculated after the exam.

Three separate mechanisms sit between the marks you count and the result you get: normalisation, the cutoff itself, and tie-breaking. Most of the confusion after a result comes from treating them as one thing.

The cutoff is an outcome, not a target

The most persistent misunderstanding is that a cutoff is a pass mark set in advance. It is not. On almost every large recruitment exam, the cutoff is whatever score the last selected candidate got, discovered after the fact, once the ranked list is cut at the number of candidates the next stage needs.

That makes it a function of three things, none of which you control: how many vacancies were notified, how many candidates appeared, and how hard the paper turned out to be. A cutoff falling year on year does not mean the exam got easier; often it means the paper got harder or the vacancy count changed. A cutoff rising does not mean the standard rose.

Two consequences worth internalising. First, last year’s cutoff is a reference point, not a target, and preparing to clear it exactly is preparing to miss. Second, cutoffs are published separately per category and often per post and per region, so the single number people quote in conversation is usually the general-category, all-India one and may have nothing to do with the list you are on.

Normalisation, and why it exists

When an exam cannot seat every candidate at once, it runs in shifts across several days, and each shift gets a different question paper. Those papers are built to the same blueprint, but they are never exactly equal in difficulty. Without a correction, a candidate who happened to draw the harder shift would be penalised for the draw rather than for their performance.

Normalisation is that correction. It converts raw scores from each shift onto a common scale, using the distribution of scores within each shift as the evidence for how hard that shift actually was. Broadly: a shift where scores across the whole cohort came out low is inferred to have been harder, so scores from it are adjusted up relative to a shift where scores came out high.

The important properties, in plain terms:

• It is comparative, not absolute. Your normalised score depends on how everyone in your shift performed, not only on how you performed.
• It can move a score in either direction. A raw score can normalise down if your shift was, statistically, an easy one.
• It applies to multi-shift stages only. A stage everybody sits at the same time, from the same paper, has nothing to normalise.
• The formula is published by the conducting body, and differs between bodies. SSC, the railway boards and the banking recruitments do not all use the same method.

This is why the score on the official scorecard can differ from the score you calculated from the answer key, and why that difference is not a mistake to be appealed.

What you can actually calculate yourself

You can calculate your raw score accurately, and that is worth doing, both after mock tests and after the real exam once the provisional answer key is out. Raw score is questions correct times marks per question, minus questions wrong times the negative marking penalty. Unattempted questions score nothing and cost nothing.

The negative marking calculator does this arithmetic with your exam’s specific penalty fraction, which is the part people get wrong from memory. What you cannot calculate is the normalised score, because the input to that is the whole cohort’s performance, which is not available to you. Treat any tool or forum post claiming to predict a normalised score or a cutoff as a guess.

Tie-breaking

Large exams produce genuine ties, and every recruitment publishes an order of tie-breaking criteria applied in sequence until the tie resolves. Typical criteria, in the order they commonly appear:

• Marks in a specified section or paper, so that a tie is broken by performance in the part of the exam considered most relevant to the post.
• Date of birth, with the older candidate ranked higher.
• Alphabetical order of name, as a final resort.

The exact list and its order are set per recruitment and stated in the notification, not inferred from another exam. Where an exam includes a skill or physical stage, the tie-break rules may sit in a different place in the sequence entirely.

Date of birth as a tie-break criterion is worth knowing about, because it is the one people are surprised by: two candidates on identical marks can be separated by years of birth alone, and no amount of preparation changes that input.

What to do with all this

Prepare against the paper, not against a number. Aim comfortably above the range recent cutoffs have sat in rather than at a specific figure, because the figure will move and you do not control what moves it.

After the exam, calculate your raw score honestly from the answer key, including the negative marking, and use that to decide whether to keep preparing for the next stage or the next cycle. Do not spend the waiting period on cutoff prediction; it is the one part of the process where effort produces nothing.

Confirm the rules for your exam

The normalisation method, the tie-breaking sequence, whether a stage runs in multiple shifts at all, and how cutoffs are split by category and region are all set per recruitment and published by the conducting body. Each exam page here lists the pattern and links the official notification. Read the result and normalisation policy for your specific exam before drawing any conclusion from a score, including anything in this article.